The role of ischemic preconditioning in the recruitment of rolling and adherent leukocytes in hepatic venules after ischemia/reperfusion

Role of ischemic preconditioning in liver surgery and hepatic transplantation

INTRODUCTION

The purpose of this review is to summarize intraoperative surgical strategies available to decrease ischemia-reperfusion injury associated with liver resection and liver transplantation.

MATERIAL AND METHOD

We conducted a critical review of the literature evaluating the potential applications of hepatic ischemic preconditioning (IPC) for hepatic resection surgery and liver transplantation. In addition, we provide a basic bench-to-bedside summary of the liver physiology and cell signaling mechanisms that account for the protective effects seen with hepatic IPC.

El preacondicionamiento isquémico del hígado: de las bases moleculares a la aplicación clínica

Ischemia-reperfusion injury is produced when an organ is deprived of blood flow (ischemia), which is then restored (reperfusion). In certain circumstances, this injury leads to irreversible organ damage. Several therapeutic strategies have been used to reduce the severity of this injury. One of these strategies is the application of brief and repetitive episodes of ischemia-reperfusion before prolonged ischemia-reperfusion (ischemic preconditioning). In the present article we review the molecular mechanisms through which ischemic preconditioning confers protection against ischemia-reperfusion injury. The application of ischemic preconditioning during liver surgery is discussed, both in normothermic situations such as liver resection and in situations of low temperature such as liver transplantation.

Dynamic changes of post-ischemic hepatic microcirculation improved by a pre-treatment of phosphodiesterase-3 inhibitor, milrinone

BACKGROUND

Phosphodiesterase-3 inhibition has been shown to attenuate hepatic warm ischemia-reperfusion injury. The aim of this study was to investigate the effect of milrinone, phosphodiesterase-3 inhibitor, on post-ischemic microcirculation of rat livers by intravital microscopy.

MATERIALS AND METHODS

Male Wistar rats were randomly assigned to three groups; group A, milrinone pre-treatment; group B, ischemic pre-conditioning; and group C, no pre-treatment. All animals underwent a 60-min warm ischemia of the left lateral liver lobe. Microvascular perfusion and leukocyte-endothelial interaction were observed by intravital videomicroscopy. Hepatocellular viability and cellular damage were quantified by adenosine triphosphate tissue concentration as well as alanine aminotransferase and lactate dehydrogenase blood levels, respectively.

RESULTS

In groups A and B, cyclic AMP hepatic tissue concentration was elevated significantly. After reperfusion, microvascular perfusion in hepatic sinusoids was significantly better maintained, and the number of adherent leukocytes was reduced in sinusoids and in post-sinusoidal venules in these rats. Serum transaminase blood levels were suppressed significantly in these groups compared with controls.

CONCLUSION

The demonstrated improvement of hepatic microcirculation is certainly derived from milrinone induced cell protection in ischemia reperfusion of the liver. This effect is outlined by improved energy status and reduced liver enzyme liberation and mimics the effect of ischemic pre-conditioning.

Effects of ischemic preconditioning on cyclinD1 expression during early ischemic reperfusion in rats

AIM: To observe the effect of ischemic preconditioning on cyclinD₁ expression in rat liver cells during early ischemic reperfusion.

METHODS: Fifty-four SD rats were randomly divided into ischemic preconditioning group (IP), ischemia/reperfusion group (IR) and sham operation group (SO). The IP and IR groups were further divided into four sub-groups (n = 6). Sham operation group (SO) served as the control group (n = 6). A model of partial liver ischemia/reperfusion was used, in which rats were subjected to liver ischemia for 60 min prior to reperfusion. The animals in the IP group underwent ischemic preconditioning twice for 5 min each time prior to the ischemia/reperfusion challenge. After 0, 1, 2, and 4 h of reperfusion, serum and liver tissue in each group were collected to detect the level of serum ALT, liver histopathology and expression of cyclinD₁ mRNA and protein. Flow cytometry was used to detect cell cycle as the quantity indicator of cell regeneration.

RESULTS: Compared with IR group, IP group showed a significantly lower ALT level in 1 h to 4 h sub-groups (P < 0.05). Proliferation index(PI) indicated by the S-phase and G₂/M-phase ratio [(S+G₂/M)/(G₀/G₁+S+G₂/M)] was significantly increased in IP group at 0 and 1 h (26.44± 7.60% vs 18.56 ± 6.40%,41.87 ± 7.27% vs 20.25 ± 6.70%, P < 0.05). Meanwhile, cyclinD₁ protein expression could be detected in IP group. But in IR group, cyclinD₁ protein expression occurred 2 h after reperfusion. The expression of cyclinD₁ mRNA increased significantly in IP group at 0 and 1 h (0.568 ± 0.112 vs 0.274 ± 0.069, 0.762 ± 0.164 vs 0.348 ± 0.093, P < 0.05).

CONCLUSION: Ischemic preconditioning can protect liver cells against ischemia/reperfusion injury, which may be related to cell proliferation and expression of cyclinD₁ during early ischemic reperfusion.

Role of Heme Oxygenase in the Protection Afforded Skeletal Muscle During Ischemic Tolerance

OBJECTIVE

Ischemic tolerance (IT) is known to improve resistance to ischemia/reperfusion (I/R)-induced injury; however, the mechanisms remain unknown. The authors hypothesized that induction of heme oxygenase (HO), a heat shock protein, would provide anti-inflammatory benefits during IT, thereby preventing leukocyte-derived I/R injury.

METHODS

Male Wistar rats were randomly assigned to sham (n = 4), I/R (n = 9), preconditioning (PC)+I/R (n = 7), chromium mesoporphyrin, to inhibit HO (CrMP; n = 4), or PC+I/R+CrMP (n = 6) groups. PC consisted of 5 cycles of I/R, each lasting 10 min, induced by tightening a tourniquet placed above the greater trochantor of the hindlimb. Twenty-four hours later, the hindlimb underwent 2 h of no-flow ischemia followed by intravital microscopy during 90 min reperfusion to assess capillary perfusion (#/mm), tissue injury (ratio of ethidium bromide to bisbenzimide labeled cells/100 microm2), leukocyte rolling (Lr, #/1000 microm2), and adhesion (La, #/1000 microm2) in postcapillary venules of the extensor digitorum longus (EDL) muscle.

RESULTS

In the I/R group, Lr was significantly increased (7.1 +/- 0.4) compared to sham (3.1 +/- 0.4). PC+I/R increased Lr (10.8 +/- 0.72), which was further exacerbated by the removal of HO (14.2 +/- 1.3). La (7.8 +/- 2.0) was significantly increased compared to sham (2.4 +/- 0.9), while PC returned La back to sham levels (1.9 +/- 0.7). Removal of HO activity, via CrMP, had no significant effect on La (3.9 +/- 0.7). However, CrMP removed the protection to microvascular perfusion (I/R = 9.4 +/- 1.1, PC = 16.6 +/- 1.8, sham = 20.5 +/- 2.8, PC+CrMP+I/R = 12.3 +/- 2.3) and prevented protection from ischemia-induced tissue injury.

CONCLUSION

The data suggest that HO is an important protective mechanism during IT in skeletal muscle, but such protection was by mechanisms other than altered leukocyte-endothelial cell interaction.

Ischemic preconditioning and liver tolerance to warm or cold ischemia: experimental studies in large animals

BACKGROUND

In the rodent, ischemic preconditioning (IPC) has been shown to improve the tolerance of the liver to ischemia-reperfusion under normothermic or hypothermic conditions. The aim of the present study was to test this hypothesis in a dog model, which may be more relevant to the human.

METHODS

Beagle dogs were used in two distinct animal models of hepatic warm ischemia and orthotopic liver transplantation (hypothermic ischemia). IPC consisted of 10 minutes of ischemia followed by 10 minutes of reperfusion. In the first model, livers were exposed to 55 minutes prolonged warm ischemia and reperfused for 3 days (n = 6). In the second model, livers were retrieved and preserved for 48 hours at 4 degrees C in University of Wisconsin solution, transplanted, and reperfused without immunosuppression for 7 days (n = 5). In each model, nonpreconditioned animals served as controls (n = 5 in each group). Also, isolated dog hepatocytes were subjected to warm and cold storage ischemia-reperfusion to model the animal transplant studies using IPC.

RESULTS

In the first model (warm ischemia), IPC significantly decreased serum aminotransferase activity at 6 and 24 hours post-reperfusion. After 1 hour of reperfusion, preconditioned livers contained more adenosine triphosphate and produced more bile and less myeloperoxidase activity (neutrophils) relative to controls. In the second model (hypothermic preservation), IPC was not protective. Finally, IPC significantly attenuated hepatocyte cell death after cold storage and warm reperfusion in vitro.

CONCLUSIONS

IPC is effective in large animals for protecting the liver against warm ischemia-reperfusion injury but not injury associated with cold ischemia and reperfusion (preservation injury). However, the IPC effect observed in isolated hepatocytes suggests that preconditioning for preservation is theoretically possible.

Ischaemic preconditioning in transplantation and major resection of the liver

BACKGROUND

Ischaemia-reperfusion injury (IRI) contributes significantly to the morbidity and mortality of transplantation and major resection of the liver. Its severity is reduced by ischaemic preconditioning (IP), the precise mechanisms of which are not completely understood. This review discusses the pathophysiology and role of IP in this clinical setting.

METHODS

A Medline search was performed using the keywords 'ischaemic preconditioning', 'ischaemia-reperfusion injury', 'transplantation' and 'hepatic resection'. Additional articles were obtained from references within the papers identified by the Medline search.

RESULTS AND CONCLUSION

The mechanisms underlying hepatic IRI are complex, but IP reduces the severity of such injury in several animal models and in recent human trials. Increased understanding of the cellular processes involved in IP is of importance in the development of treatment strategies aimed at improving outcome after liver transplantation and major hepatic resection.

Modulation of liver oxidant-antioxidant system by ischemic preconditioning during ischemia/reperfusion injury in rats

AIM: To investigate effects of ischemic pre-conditioning on the liver endogenous oxidant-antioxidant system during ischemia/reperfusion injury.

METHODS: Twenty-four male Sprague-Dawley rats were randomly divided into sham-operated (Sham), ischemia/reperfusion (I/R), ischemic pre-conditioning plus ischemia/reperfusion (IPC) groups. Serum ALT, AST and hyaluronic acid levels were assayed and pathologic alterations observed. Liver malondialdehyde (MDA) contents, endogenous antioxidant enzymes, superoxidase dismutase (SOD), catalase (CAT), gultathionine peroxidase (GSH-Px) activities, neutrophils accumulation marker, myeloperoxidase (MPO) activities were measured respectively.

RESULTS: Compared with I/R group, sinusoidal endothelial cells as well as hepatocytes damages, as assessed biochemically and histochemically, were improved significantly in IPC group; neutrophils infiltration was also markedly reduced. In IPC group, liver peroxidation, as measured by MDA contents, was significantly decreased when compared with I/R group; endogenous antioxidant enzymes, SOD, CAT and GSH-Px activities were markedly higher than that in I/R group.

CONCLUSION: Ischemic pre-conditioning exerts protective effects on both hepatic sinusoidal endothelial cells and hepatocytes during liver I/R injury. Its mechanisms may involve dimunition of neutrophils infiltration and modulation of the imbalance of endogenous oxidant-antioxidant system in the organism.

Nitric oxide attenuates ischaemia-reperfusion (I/R) injury in the diabetic liver

BACKGROUND

Liver ischaemia-reperfusion (I/R) occurs during resuscitation from haemorrhagic shock, hepatic transplantation and anatomic resection of the liver. This injury is associated with hepatocellular enzyme release and hepatocyte necrosis. The impact of chronic illnesses such as diabetes mellitus (DM) on hepatic I/R is unknown. This study determines the effect of DM on liver I/R using a murine model of type II DM in which the leptin receptor is defective. Preliminary studies suggest that animal models of DM have impaired endothelial nitric oxide (NO) release. Other studies suggest that NO attenuates hepatic I/R in phenotypically normal animals. We postulated that DM exacerbates hepatic I/R and that exogenous NO administration will attenuate hepatocellular injury.

METHODS

Non-diabetic and diabetic (db/db) mice were anaesthetized and underwent laparotomy with the placement of a microvascular clip on the hepatic artery and portal vein supplying the medial and left lateral lobes of the liver rendering about 70% of the liver ischaemic. Hepatic ischaemia was maintained for 45 min after which time the clip was removed and the liver segments reperfused. The abdomen was closed and the animals maintained for 5 h of reperfusion. Hepatic injury was then assessed by measuring serum alanine and aspartate transaminases (ALT, AST) spectrophotometrically. Sections of liver reperfused for 24 h were stained with haematoxylin and eosin and the percentage of hepatocyte necrosis evaluated using morphometric techniques. Other animals undergoing hepatic I/R received the NO donor (DETA 100 micro g/kg, i.v. 5 min prior to reperfusion). Time-matched, sham-operated animals served as controls. The data are expressed as mean +/- SEM and analysed by ANOVA.

RESULTS

Serum AST and ALT levels were significantly higher in db/db animals vs. non-diabetics, even in the absence of hepatic I/R (P < 0.01). Serum AST and ALT levels in db/db mice undergoing hepatic I/R were nearly five times greater than that of non-diabetic animals (P < 0.01). Histologic examination of the livers of the diabetic animals undergoing I/R demonstrated significantly greater hepatocellular necrosis (zone III; 30-40%) when compared with non-diabetic animals sustaining the same injury (zone III; 3-10%). The NO donor DETA totally prevented the increase in serum ALT and AST release associated with I/R in both the diabetic and non-diabetic mice when compared with animals not receiving this agent (P < 0.01).

CONCLUSION

This is the first study suggesting that DM exacerbates hepatic I/R and that NO donors will prevent this hepatocellular injury in the diabetic. Sixteen million Americans have DM. Understanding the effect of this chronic illness on the inflammatory response to injury is essential to improving clinical outcomes in these medically compromised patients.

Ischemic preconditioning and intermittent clamping improve murine hepatic microcirculation and Kupffer cell function after ischemic injury

The aim of this study was to evaluate whether the protective effect of intermittent clamping and ischemic preconditioning is related to an improved hepatic microcirculation after ischemia/reperfusion injury. Male C57BL/6 mice were subjected to 75 or 120 min of hepatic ischemia and 1 or 3 hours of reperfusion. The effects of continuous ischemia, intermittent clamping, and ischemic preconditioning before prolonged ischemia on sinusoidal perfusion, leukocyte-endothelial interactions, and Kupffer cell phagocytic activity were analyzed by intravital fluorescence microscopy. Kupffer cell activation was measured by tissue levels of tumor necrosis factor (TNF)-alpha, and the integrity of sinusoidal endothelial cells and Kupffer cells were evaluated by electron microscopy. Continuous ischemia resulted in decreased sinusoidal perfusion rate and phagocytic activity of Kupffer cell, increased leukocyte-endothelial interactions and TNF-alpha levels. Both protective strategies improved sinusoidal perfusion, leukocyte-endothelial interactions and phagocytic activity of Kupffer cells after 75-minutes of ischemia, and intermittent clamping also after 120 minutes ischemia. TNF-alpha release was significantly reduced and sinusoidal wall integrity was preserved by both protective procedures. In conclusion, both strategies are protective against ischemia/reperfusion injury by maintaining hepatic microcirculation and decreasing Kupffer cell activation for clinically relevant ischemic periods, and intermittent clamping appears superior for prolonged ischemia.

Hepatic ischemia reperfusion injury: pathogenic mechanisms and basis for hepatoprotection

This review highlights recent advances in our understanding of mechanisms underlying reperfusion injury to the liver after warm hepatic ischemia. Sinusoidal endothelial cells and hepatocytes are targets of injury in the early 'cytotoxic' phase, although participation of apoptosis in the cell-death process remains contentious. Kupffer cells may play an important role as the initial cytotoxic cell type and are likely a source of reactive oxygen species and proinflammatory mediators, particularly tumor necrosis factor (TNF)-alpha. The latter are involved with subsequent neutrophil activation and recruitment. Microcirculatory disruption results from an imbalance between the actions of vasoconstrictors and vasodilators, such as nitric oxide, and also has a major impact on reperfusion injury. There is growing evidence that a brief prior ischemia-reperfusion period, termed 'ischemic preconditioning', is hepatoprotective. This can be mimicked by drugs that produce oxidative stress, and by interleukin-6 and TNF-alpha; both these cytokines are involved with priming hepatocytes to enter the cell cycle. Several mechanisms have been implicated including mobilization of adenosine and activation of adenosine type 2 receptors, nitric oxide, abrogation of TNF synthesis, preservation of energy metabolism, protection of the microcirculation and accelerated cell-cycle entry. A better understanding of preconditioning mechanisms will lead to novel approaches to improve outcomes of liver surgery.

Ischemic preconditioning protects from hepatic ischemia/reperfusion-injury by preservation of microcirculation and mitochondrial redox-state

BACKGROUND/AIMS

Ischemic preconditioning (IP) is known to protect hepatic tissue from ischemia-reperfusion injury. However, the mechanisms involved are not fully understood yet.

METHODS

Using intravital multifluorescence microscopy in the rat liver, we studied whether IP exerts its beneficial effect by modulating postischemic Kupffer cell activation, leukocyte-endothelial cell interaction, microvascular no-reflow, mitochondrial redox state, and, thus, tissue oxygenation.

RESULTS

Portal triad cross-clamping (45 min) followed by reperfusion induced Kupffer cell activation, microvascular leukocyte adherence, sinusoidal perfusion failure (no-reflow) and alteration of mitochondrial redox state (tissue hypoxia) (P<0.05). This resulted in liver dysfunction and parenchymal injury, as indicated by decreased bile flow and increased serum glutamate dehydrogenase (GLDH) levels (P<0.05). IP (5 min ischemia and 30 min intermittent reperfusion) was capable to significantly reduce Kupffer cell activation (P<0.05), which was associated with a slight attenuation of leukocyte adherence. Further, IP markedly ameliorated sinusoidal perfusion failure (P<0.05), and, thereby, preserved adequate mitochondrial redox state (P<0.05). As a consequence, IP prevented the decrease of bile flow (P<0.05) and the increase in serum GLDH levels (P<0.05).

CONCLUSIONS

IP may exert its beneficial effects on hepatic ischemia-reperfusion injury by preserving mitochondrial redox state, which is guaranteed by the prevention of reperfusion-associated Kupffer cell activation and sinusoidal perfusion failure.

Tumor necrosis factor suppression and microcirculatory disturbance amelioration in ischemia/reperfusion injury of rat liver after ischemic preconditioning

BACKGROUND

Brief periods of hepatic ischemia produce immediate tolerance for subsequent prolonged ischemia. Although the beneficial effect of this ischemic preconditioning is recognized, the mechanism itself remains poorly understood.

METHODS

Male Wistar rats were divided into two groups: a control group that was subjected to 30 min of ischemia + following reperfusion, and an ischemic preconditioning group that was subjected to 5 min of ischemia + 5 min of reperfusion + 30 min of ischemia + following reperfusion. By using this model, hepatic damage, microcirculatory disturbances, and tumor necrosis factor-alpha protein production and mRNA expression were analyzed during the course of reperfusion in both groups. For the hepatic damage evaluations, hepatic enzyme levels, histology, apoptosis analysis, and intravital microfluorography for dead cells were examined. For the microcirculatory disturbance analysis, an adhesion molecule and intravital microfluorography for endothelial-adherent leukocytes were examined.

RESULTS

In the ischemic preconditioning group, ischemia/reperfusion injuries (shown by hepatic enzymes elevation, histological degeneration, and increases in the number of apoptotic cells and microfluorographic dead cells) were markedly reduced. Moreover, microcirculatory disturbances represented by intercellular adhesion molecule-1 expression and leukocyte adhesion on the endothelium were ameliorated. Tumor necrosis factor-alpha protein production and mRNA expression were also suppressed in the ischemic preconditioning group.

CONCLUSION

The suppression of tumor necrosis factor-alpha and the subsequent amelioration of microcirculatory disturbances were observed, suggesting that the mechanism underlying the protective effect of ischemic preconditioning in hepatic ischemia/reperfusion injuries may involve tumor necrosis factor-alpha and microcirculatory regulation.

Ischemic preconditioning protects against cold ischemic injury through an oxidative stress dependent mechanism

BACKGROUND/AIMS

Ischemic injury in cold preserved livers is characterized by sinusoidal endothelial cell (SEC) detachment and matrix metalloproteinase activity. Upon reperfusion reversible ischemic injury becomes permanent with SEC rapidly undergoing apoptosis. Ischemic preconditioning prevents reperfusion injury after normothermic ischemia. We hypothesized that ischemic preconditioning, through an oxygen free radical burst, protects against injury during cold preservation and reperfusion.

METHODS

Ischemic preconditioning was achieved in rats by clamping blood supply to the left and median lobes for 10 min followed by 15 min of reperfusion prior to preservation in cold University of Wisconsin solution for 30 h. In a second set of experiments, rats were pretreated with N-acetyl-cysteine (NAC). SEC apoptosis upon reperfusion was assessed in an isolated perfused rat liver (IPRL) model.

RESULTS

SEC detachment and activities of matrix metalloproteinase were significantly reduced in preconditioned livers. A decrease of SEC apoptosis after 1h of reperfusion in the IPRL was noted in preconditioned livers compared to controls. Pretreatment with NAC reversed the beneficial effects of ischemic preconditioning on SEC detachment and apoptosis.

CONCLUSIONS

Ischemic preconditioning is an effective strategy to prevent injury during cold preservation and after reperfusion. The protective effect is possibly mediated by oxygen free radicals.

Ischemic pre-conditioning of 5 minutes but not of 10 minutes improves lung function after warm ischemia in a canine model

BACKGROUND

Protection from reperfusion injury by ischemic pre-conditioning (IPC) before prolonged ischemia has been proven for the heart and the liver. We now assess the efficacy of IPC to protect lungs from reperfusion injury.

METHODS

Eighteen foxhounds (25 to 30 kg) were anesthetized, intubated, and ventilated with a fraction of inspired oxygen of 0.3 at a volume-controlled mode to maintain arterial pCO2 of 30 to 40 mm Hg. After left thoracotomy, we performed warm ischemia for 3 hours by clamping the left hilus, and followed with 8 hours of reperfusion (control, n = 6). In the treated groups, IPC was performed either for 5 minutes followed by 15-minute reperfusion (n = 6, IPC-5), or by 2 successive cycles of 10-minute ischemia, followed by 10-minute reperfusion (n = 6, IPC-10) before prior to the 3-hours warm-ischemia period. Pulmonary compliance and gas exchange were determined separately for each lung, and we recorded pulmonary and systemic hemodynamics. We performed bronchoalveolar lavage (BAL) at the end of the experiment and determined total protein concentration as well as tumor necrosis factor alpha (TNF-alpha) mRNA expression in cell-free supernatant and in BAL cells, respectively. We also assessed the wet/dry ratio of the lung.

RESULTS

In the controls, on reperfusion, we encountered a progressive deterioration of gas exchange, especially of the reperfused left lung, which we could largely avoid using the IPC-5 protocol. Similarly, pulmonary compliance steadily declined but was much better in the ICP-5 group. Parallel to the improvement of gas exchange and lung mechanics, we found less total alveolar protein content and TNF-alpha mRNA expression in BAL cells in the IPC-5 than in the controls. However, we did not find IPC-10 to be paralleled by a significant improvement of lung function. Neither IPC-5 nor IPC-10 influenced the pulmonary vascular resistance index or the fluid accumulation in the lung.

CONCLUSION

The major finding of the present study was that 5 minutes of IPC improved lung function after 3 hours of warm ischemia of the lung.

Persantine Attenuates Hemorrhagic Shock-Induced P-Selectin Expression

Ischemia/reperfusion (I/R), a phenomenon that is associated with conditions such as organ transplantation, trauma, vascular disease, and stroke, involves the recruitment of activated and adherent leukocytes that subsequently mediate tissue injury. Endothelial cell adhesion molecules such as P-selectin mediate I/R-induced leukocyte recruitment and allow the adherent leukocytes to damage the vascular wall and parenchymal cells. This study examines the influence of dypiridamole (persantine) on hemorrhagic shock (H/S)-induced P-selectin expression. H/S was induced in C57BL/6 mice by withdrawing blood to drop the mean arterial blood pressure to 30 to 35 mm Hg for 45 minutes. The mice were resuscitated by infusing the shed blood and Ringer's lactate (50% shed blood volume). In vivo P-selectin expression was determined using a dual monoclonal antibody technique in the heart, lung, liver, kidneys, stomach, small bowel, and colon of a control group, a hemorrhagic shock group, and a hemorrhagic shock group that was pretreated with Persantine (Boehringer, Ingelheim, Ingelheim, Germany). H/S significantly ( P < 0.01) increased P-selectin expression in all regional vascular beds of untreated mice. Persantine treatment largely prevented the H/S-induced P-selectin expression in the same vascular beds. Persantine significantly attenuates the upregulation of P-selectin in the hemorrhagic shock model.