SHOCK, Vol. 33, No. 5, pp.

532Y540, 2010

HYPERTONIC SODIUM PYRUVATE SOLUTION IS MORE EFFECTIVE THAN RINGERS ETHYL PYRUVATE IN THE TREATMENT OF HEMORRHAGIC SHOCK
Pushpa Sharma and Paul D. Mongan
Hemorrhagic Shock and Trauma Laboratory, Department of Anesthesiology, The Uniformed Services University, Bethesda, Maryland
Received 19 Feb 2009; first review completed 9 Mar 2009; accepted in final form 21 Aug 2009 ABSTRACTHypertonic sodium pyruvate (HSP), as well as ethyl pyruvate solutions, has been proposed as resuscitative fluids in the treatment of hemorrhagic shock (HS) because of their anti-inflammatory and antioxidant properties. The effectiveness of one pyruvate preparation over the other in the treatment of HS has not been evaluated. The authors aimed to compare two pyruvate solutions for resuscitation and their mechanisms of action in rats during HS. The effects of infusion of low-volume HSP were compared against high-volume Ringers ethyl pyruvate on hemodynamic parameters, inflammatory cascade, and regulation of stress and apoptosis-related proteins in the liver. Sprague-Dawley rats were either treated as sham animals or subjected to computer-controlled arterial hemorrhage (40 mmHg) for 60 min followed by resuscitation with isotonic sodium chloride solution, hypertonic saline, Ringers lactate solution, Ringers ethyl pyruvate, or HSP for 60 min. Animals were continuously monitored for hemodynamic and biochemical parameters in blood. At the end of the experiment, animals were killed, and liver samples were taken for the evaluation of inflammatory and antiinflammatory markers and mediators of oxidative stress, liver injury, and expression of apoptotic signaling proteins. In comparison with Ringers ethyl pyruvate, HSP administration after hemorrhage reduced liver injury, which was associated with increased levels of serum and tissue inflammatory cytokines, inflammatory mediators such as NOS and cyclooxygenase 2, lipid peroxidation, and higher hepatocellular adenosine triphosphate. Cellular apoptotic events related to the activation of caspase-3 and poly(ADP-ribose)polymerase cleavage were also decreased by sodium pyruvate. Resuscitation with small-volume HSP offers significant protection against inflammatory and oxidative stress and in preventing liver injury compared with large-volume Ringers ethyl pyruvate. KEYWORDSRats, apoptosis, hepatic injury, cytokines, oxidative stress

INTRODUCTION One of the main causes of post-traumatic deaths is the development of severe decompensatory hemorrhagic shock (HS) and multiple organ failure (MOF). Although there are numerous factors that influence the development of MOF, there is increasing evidence that hepatic dysfunction plays a central role (1). The increased production of oxygen and nitrogen species, stimulation of prostaglandins and thromboxanes by cyclooxygenase 2 (COX-2) activation, and the resultant production of proinflammatory cytokines from the activated macrophages and Kupffer cells of damaged liver have been identified as key events in the depletion of hepatocellular adenosine triphosphate (ATP) and propagation of liver damage and MOF in HS (2, 3). The goal of HS treatment is to restore tissue perfusion, suppress the initial inflammatory response, and improve cellular energy reserves. However, resuscitation strategies can modulate the inflammatory response and organ-specific injury. Ringers lactate solution infusion can restore the peripheral tissue perfusion but may cause rebleeding and early mortality because of reintroduction of oxygen into the previously ischemic tissue and formation of interstitial edema

Address reprint requests to Pushpa Sharma, PhD, Department of Anesthesiology, 4301 Jones Bridge Rd, Bethesda, MD 20814. E-mail: psharma@usuhs.mil. This study was supported by ONR and Exploratory grant. DOI: 10.1097/SHK.0b013e3181cc02b3 Copyright 2010 by the Shock Society

(4, 5). On the other hand, hypertonic saline solution has been shown to reduce fluid needs, restore hemodynamic stability, and prevent interstitial edema, but may be detrimental in terms of cytokine production and histopathologic change (6). This suggests that despite the restoration of global oxygen delivery, additional pharmacological therapies are needed to prevent or reverse organ damage. We have previously shown that infusion of hypertonic sodium pyruvate (HSP) solution can increase animal survival after HS, and its actions are mainly mediated through the improvement in cellular energetic status of liver and brain, increased redox status, and decreased apoptotic signaling proteins (7Y11). Despite its protective actions, pyruvate in neutral or alkaline aqueous solutions has been questioned because of the potential degradation of pyruvate to parapyruvate, an effect that would impede mitochondrial ATP production. Fink (12) formulated ethyl pyruvate in Ringers lactate (LR) solution and called it Ringers ethyl pyruvate (REP) solution. However, the potentially detrimental effects of ethanol released by ester cleavage may limit the concentrations of REP that can be administered by continuous i.v. infusions. However, there are no published reports to suggest the advantage of using low-volume resuscitation with HSP versus high-volume resuscitation with REP in the treatment of HS. This study compared the efficacy of HSP and REP with a focus on volume sparing, hemodynamics, inflammatory cascade, oxidative damage, metabolic effects, and regulation of stress-related and apoptotic proteins on liver damage in a rat model of HS.
532

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK MAY 2010

MATERIALS AND METHODS
The protocol and experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the Uniformed Services University of the Health Sciences at Bethesda, Md, with adherence to the Guide for Care and Use of Laboratory Animals. All animals were maintained in accordance with the recommendations of the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.

HYPERTONIC SODIUM PYRUVATE

IN

SHOCK

533

was determined by measuring NADH formation at 340 nm using hexokinase and glucose-6-phosphate dehydrogenase as substrates (11, 14).

Measurement of lipid peroxidation

Liver malondialdehyde (MDA) contents were measured using a colorimetric thiobarbituric acid assay kit from Calbiochem, La Jolla, Calif. The malondialdehyde levels were normalized to protein content and expressed as nanomoles per milligram liver protein.

Animal preparation and surgical procedures

Male Sprague-Dawley rats weighing 300 to 350 g were purchased from Taconic Farms, German Town, NY. Spontaneously ventilating rats were anesthetized using 5% isoflurane and oxygen through a fitted nose cone. The femoral artery and femoral vein were cannulated with polyethylene catheters (PE 50; Clay Adams, Piscataway, NJ). The incision was closed with interrupted sutures. Core temperature (rectal) was maintained at 37-C T 0.2-C throughout the experiment with a heat lamp. The isoflurane concentration was reduced to 2.0%. The MAP was monitored by connecting the femoral arterial catheter to a pressure transducer and a computerized physiograph (Labview 5; National Instruments, Austin, Tex). The resuscitative fluid or shed blood was infused through the femoral vein. Animals were allowed to acclimate to the induction of sedation and stress of surgery for about 10 min before the induction of HS. Induction of HS The computer-controlled rat model of HS was created by the withdrawal of blood at a rate of 0.5 mL/min for 10 to 15 min from femoral arterial catheter using an Instech P720 peristaltic pump (Instech Laboratories, Inc, Plymouth Meeting, Pa) to induce shock (MAP, 40 mmHg). This MAP was maintained at 40 mmHg either by removal or reinfusion of shed blood for 60 min (T60) at the shock phase. Experimental groups At the end of the shock phase, the animals were randomly assigned to the following six groups according to the solution to be infused (n = 8 animals per group). Group 1. Sham: instrumented time-control rats were anesthetized and instrumented in the same manner as rats in the other shock groups but did not undergo arterial hemorrhage or resuscitation, except the withdrawal of blood for laboratory investigation. Group 2. NS: resuscitated with 0.9% isotonic sodium chloride solution equal to 3 times the volume of shed blood. Group 3. HTS: resuscitated with 7.5% hypertonic saline (HTS) at the rate of 5 mL/kg as a control for the HSP group. Group 4. LR: resuscitated with LR solution equal to 3 times the volume of shed blood as a control for the REP group. The LR solution contained 109 mM of NaCl, 4 mM of KCl, 2.7 mM of CaCl2, and 28 mM of sodium L-lactate. Group 5. REP: resuscitated with REP solution equal to 3 times the volume of shed blood. The REP contained 130 mM of NaCl, 4 mM of KCl, 2.7 mM of CaCl2, and 28 mM of ethyl pyruvate (pH 7.4). Group 6. HSP: resuscitated with 2 M HSP infused at a rate of 5 mL/kg. The doses of REP and pyruvate were selected on the basis of previously published investigations (11, 13). All fluids were used at room temperature and infused within 45 min after shock.

Protein assay
Protein content in liver tissue samples was determined according to Bradford assay using bovine serum albumin as standard (15).

Determination of cytokine levels in serum and liver

Protein levels of proinflammatory cytokines (TNF-!, IL-6) and the antiinflammatory cytokine IL-10 were quantified in duplicate in serum and in liver tissue homogenates using an enzyme-linked immunosorbent assay kit specific for rat cytokines (Invitrogen-BioSource International, Inc, Camarillo, Calif). The optical density of each well was measured at 450 nm with a microplate reader (Dynatech Laboratories, Billingshurst, UK). The serum cytokine concentrations were calculated from the corresponding standard curve, normalized to serum volume or liver protein concentration as picogram per milliliter serum or picogram per milligram liver protein.

Western blotting
Liver cell homogenates from HS rats were used for Western blotting according to our published methods (7, 9, 11, 16). In brief, approximately 100 mg of the pulverized frozen liver tissue samples was homogenized in ice-cold lysis buffer (1:10 [wt/vol] consisted of phosphate buffer saline, pH 7.4, 230 mg/mL, phenylmethylsulphonyl fluoride, 1 2g/mL leupeptin, and 1 2g/mL aprotinin in 1% Triton X-100. After brief sonication and centrifugation (1,000g for 10 min), these samples were denatured in Laemmli sample buffer. Approximately 15 2g of protein was resolved on 10% sodium dodecyl sulfate gel and transferred to the nitrocellulose membrane of 0.45-2m pore size. Membranes were blocked in 5% fat-free milk in PBS plus 0.1% Tween-20 for 1 h at room temperature. Membranes were probed with primary antibody to antiYCOX-2 or anti-iNOS (dilution 1:500; Transduction Laboratories, Lexington, Ky), antiYcaspase-3, or antiYpoly(ADP-ribose)polymerase ([PARP] dilution 1:1,000; Biomol Research Laboratories Inc, Plymouth Meeting, Pa). Rabbit monoclonal "-actin was used as the loading control. Detection was accomplished using horseradish peroxidaseYconjugated antibody diluted 1:20,000 for 1 h. The reaction was developed with Super Signal West Pico Chemiluminescence Substrate (Pierce,

Sampling and assays

Blood sampling Arterial blood (500 2L) was collected before hemorrhage (T0) and every 60 min in EDTA tubes. Approximately 200 2L of blood was used for the measurement of pH, base excess, and blood gases (IL 1610 Blood Gas Analyzer; Instrumentation Laboratories, Lexington, Ky), and serum lactate, pyruvate, and glucose with a CMA 600 Analyzer (CMA/Microdialysis, Acton, Mass). The remaining blood was used for the determination of cytokines and liver enzymes. Tissue sampling Before death, the liver was immediately excised, quickly frozen in liquid nitrogen, pulverized, and stored in cryovials at j80-C for the measurement of lipid peroxidation, ATP, and Western blotting procedures. At the end of the 2-h shock and resuscitation experiment (T120), the animals were euthanized with 4% to 5% isoflurane and air through the nose cone.

Assessment of liver injury

The serum contents of liver enzymes, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were determined in our clinical chemistry laboratory at Uniformed Services University of the Health Sciences.

Determination of ATP content in liver

The ATP content in the liver after resuscitation (T120) was determined by deproteinating the frozen pulverized liver tissue in 1.5 M perchloric acid (3 mL/mg protein) for 10 min and then neutralizing with 2 M KOH. The ATP

FIG. 1. Effects of resuscitation on MAP. Data are presented as mean SE from eight animals per group. The initial hemodynamic and laboratory measurements were made immediately before the start of the controlled arterial hemorrhage (T0) and at 60 min after the hemorrhage or sham (T60). The NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, REP, and HSP treatments were started 60 min after the start of hemorrhage for the next 45 min until the end of the experiment (T120). *P G 0.05 ANOVA for repeated measures comparing the within-group MAP at the end of the shock period (T60) with baseline (T0). +P G 0.05 ANOVA for repeated measures comparing the within-group MAP at the end of the recovery period (T120) with baseline (T0).

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

534

SHOCK VOL. 33, NO. 5

SHARMA

AND

MONGAN

Rockford, Ill). Quantitative determinations of signal intensity of the protein bands were performed by using Scion image program from Scion Corporation, Frederick, Md, and density values obtained from three animals per group were presented as mean T SEM. The percentage of PARP fragmentation was calculated by using the formula (optical density of 85-kd fragmented PARP)/optical density of total PARP (116-kd intact PARP + 85-kd fragmented PARP) 100.

Statistical analysis
Data were analyzed using Sigma-Stat 3.0 Software and expressed as mean T SE of eight animals in each group. Statistically significant differences between groups for nonrecurring measurements were assessed using a one-way ANOVA technique. Analysis for differences between and within groups for repeated measurements was performed using ANOVA between groups for the dependent variable (e.g., MAP, etc) with repeated measures over time. Within and between group testing was followed by Tukey or Dunnett multiple comparison tests against controls. The level of statistical significance was accepted at P G 0.05 after correction for multiple comparisons.

period (Fig. 1). Because of similar weights of the animals, the volume of shed blood for the induction of shock did not differ significantly among the various hemorrhaged groups (P 9 0.05) and was in the range previously reported for the same level (40 mmHg) of HS (11, 17). No resuscitative strategy returned the MAP to baseline. In sham animals, MAP remained stable throughout the experiment. In comparison with baseline values, all shocked animals had a significant decrease in PCO2 at T60. All treatments effectively increased PCO2 (P G 0.05) at T120 (Table 1).
Effects of pyruvate formulations on serum osmolality, hemoglobin, electrolytes, and acid-base status

RESULTS
Hemodynamic and physiological response to HS and resuscitation

Data depicted in Figure 1 show that the MAP was similar in all animals at the start of the hemorrhage and decreased to approximately 40 mmHg in all groups during a 10- to 15-min

Table 1 shows the increases in osmolality and decreases in hemoglobin throughout the protocol in all treatment groups. Decline in hemoglobin after hemorrhage suggests dilution from transcapillary refill and fluid infusion as seen after resuscitation. Hemorrhagic shock resulted in severe hyperkalemia and increases in Na+ and Clj concentrations in all groups except HSP. Hyperkalemia did not occur after HSP infusion,

TABLE 1. Effect of resuscitation on hemodynamic parameters, blood gas, and electrolyte analysis Parameters Shed blood, mL/kg Fluid infused, mL P O2 Time, min T0YT60 T60YT120 T0 T60 T120 PCO2 T0 T60 T120 Osmolality, mosmol/kg T0 T60 T120 Hemoglobin, g/dL T0 T60 T120 Na, mmol/L T0 T60 T120 K, mmol/L T0 T60 T120 Cl, mmol/L T0 T60 T120 Sham (n = 8) 12.4 T 0.2 1.5 T 0.4 78.0 T 1.1 80.7 T 3.3 82.5 T 2.3 41.8 T 0.4 42.8 T 0.8 48.8 T 0.4 286 T 15.5 302 T 13.1 270 T 13.6 13.6 T 0.2 10.0 T 0.3 12.0 T 0.1 135.5 T 1.2 131.0 T 1.5 136 T 2.3 4.7 T 0.2 4.1 T 0.4 4.5 T 0.5 103 T 1.8 101 T 1.6 96.9 T 2.2 Control (n = 8) 11.1 T 1.6 34.5 T 2.9 73.0 T 1.2 60.0 T 1.6 88.0 T 0.4* 43.7 T 0.6 34.9 T 1.7 40.5 T 0.5 270.5 T 15.3 278 T 14.1 290 T 13.5* 10.8 T 0.1 6.8 T 0.3 131.5 T 1.4 142.3 T 2.4 143.5 T 1.2 4.2 T 0.3 5.8 T 0.4 6.9 T 0.4 105.2 T 2.1 112 T 2.1

HTS (n = 8) 13.5 T 1.9 3.4 T 1.6 86.0 T 1.6 84.2 T 9.0 108.0 T 3.4* 45.1 T 0.7 28.1 T 1.0 33.9 T 3.4 284 T 12.5 290 T 13.9 318 T 9.6* 11.9 T 0.3 6.8 T 0.5 133.4 T 0.8 156.5 T 2.3 145 T 1.1 4.3 T 0.2 5.5 T 0.4

LR (n = 8) 10.3 T 1.3 32.6 T 2.5 76.0 T 4.13 71.1 T 12.0 94.9 T 7.3* 42.5 T 0.3 30.4 T 0.5 32.2 T 1.1 279 T 16.5 281 T 14.0 305 T 11.9* 12.0 T 0.4 7.1 T 0.8 135.0 T 0.8 150 T 1.8 136.4 T 1.4 4.2 T 0.1 5.8 T 0.6 6.9 T 0.2 102 T 2.9 118.0 T 2.5

REP (n = 8) 10.0 T 2.1 30.2 T 1.8 59.0 T 0.5 79.3 T 3.5 85.5 T 1.2* 46.8 T 0.9 32.5 T 2.5 38.6 T 0.6*

HSP (n = 8) 12.5 T 1.0 3.7 T 2.1 83.8 T 0.8 76.8 T 2.8 92.0 T 1.5* 42.5 T 1.6 33.9 T 1.0 45.4 T 1.8* 287 T 13.5 289 T 16.0 307.6 T 8.2* 14.5 T 0.4 6.1 T 0.4 5.5 T 0.13* 132.8 T 0.7 149.3 T 1.5 152.9 T 0.6 4.5 T 0.2 4.9 T 0.4 4.3 T 0.2k 98.9 T 1.1 114.1 T 1.2 87.7 T 1.9k

280 T 12.5 283 T 13.8 283 T 12.5* 13.2 T 0.1 7.0 T 0.5

5.9 T 0.2*

5.1 T 0.4*

5.6 T 0.3*

4.8 T 0.2* 130.0 T 1.6 151.0 T 1.9 138.2 T 1.7 4.4 T 0.1 5.0 T 0.1

7.3 T 0.3 89.9 T 1.6 111 T 1.1 112 T 1.7

6.4 T 0.3 92.5 T 0.2 112 T 1.5

114 T 1.3

111 T 1.9

103 T 1.1

The initial hemodynamic and laboratory measurements were made immediately before the start of the controlled arterial hemorrhage (T0), at 60 min after hemorrhage or sham (T60), and at the end of experiment at T120. The NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, REP, and HSP treatments were started 60 min after the start of hemorrhage for the next 45 min until the end of the experiment (T120). *P G 0.05, compared with after shock values at T60. P G 0.05, compared with baseline for time-matched data. P G 0.05, comparing the LR group with the REP group. P G 0.05, comparing the HTS group with the HSP group. k P G 0.05, comparing the REP group with the HSP group.

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK MAY 2010

but it resulted in a significant increase in Na+ and decrease in Clj compared with other groups such as NS, HTS, and REP (P G 0.05). In addition, the level of metabolic acidosis (base excess, lactate-pyruvate [l/p] ratios, and pH) induced by HS was similar in all groups at T60. In comparison with the other treatments, HSP was more effective in reversing the metabolic acidosis. In addition, HSP significantly reduced the serum l/p ratio (21.4 T 0.4) in comparison with NS (41.6 T 0.3), HTS (38.9 T 0.3), LR (48.3 T 0.5), and REP (39.5 T 2.6). This suggests that HSP is effective in decreasing the HS-induced chemical and metabolic acidosis caused by the metabolic effects of 2 mM pyruvate.
Effects of pyruvate formulations on serum and liver cytokines

HYPERTONIC SODIUM PYRUVATE

IN

SHOCK

535

One of the significant hallmarks of HS is the release of potent inflammatory mediators into the circulation that are capable of inducing a generalized inflammatory response on remote organ functions. We compared the anti-inflammatory effects of HSP and REP by determining the serum and liver concentrations of proinflammatory cytokines TNF-!, IL-6, and anti-inflammatory molecule IL-10. The capability of

various resuscitative fluids to achieve cytokine balance was evaluated by determining the IL-6/IL-10 ratio because a strong correlation of high IL-6/IL-10 ratios and poor outcomes in infants and adult patients with systemic inflammatory response syndrome has been previously reported (18, 19). As shown in Figures 2 and 3, both the HSP and REP infusion significantly attenuated the cytokine concentrations at the systemic level as well as in the liver compared with other treatments, the values remained higher than the sham group (P G 0.05) at all times after shock. IL-10 levels were not affected by any of the treatment fluids and suggest that the changes in TNF-! and IL-6 by HSP and REP were caused by the reduced production of these cytokines and not caused by the trigger of anti-inflammatory response. In addition, the significantly low ratios of IL-6/IL-10 cytokines in HSP and REP groups compared with NS-, HTS-, and LR-treated animals (P G 0.05) confirmed a reduction in the inflammatory response in both HSP and REP groups. The plasma cytokine levels at baseline (T0) in all groups were similar and significantly lower than the levels in all treatment groups during hemorrhage and resuscitation (datanot shown). The detection range for TNF-! was greater than 5 pg/mL and for IL-6 and IL-10 was greater than 10 pg/mL.

FIG. 2. Effect of HS and resuscitation on cytokine concentrations in serum. Data presented as group means T SE from eight animals per group. Serum samples collected at the end of the experiments at T120 were analyzed for cytokine concentrations. Treatments included NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, REP, and HSP. Hemorrhagic shock increased levels of serum TNF-! and IL-6 (A and B, respectively). Both HSP and REP were equally effective in reducing their levels. IL-10 was not altered by any of the treatment included in this study. bP G 0.05, comparing the treatments with sham for time-matched data. cP G 0.05, comparing the LR group with the REP group. dP G 0.05, comparing the HTS group with the HSP group. e P G 0.05, comparing the REP group with the HSP group.

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

536

SHOCK VOL. 33, NO. 5

SHARMA

AND

MONGAN

FIG. 3. Effect of HS and resuscitation on cytokine concentrations in liver. Data presented as group means T SEM of eight animals per group. Treatments included NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, REP, and HSP. Liver was harvested at the end of the experiments. Measurements were performed in liver tissue homogenates. bP G 0.05, comparing the treatments with sham for time-matched data. cP G 0.05, comparing the LR group with the REP group. dP G 0.05, comparing the HTS group with the HSP group. eP G 0.05, comparing the REP group with the HSP group.

Effects of pyruvate formulations on hepatocellular damage and release of liver enzymes

Data presented in Table 2 suggest that plasma ALT and AST levels were significantly increased in all HS-resuscitated animals compared with sham. Although both REP and HSP caused a significant reduction in ALT and AST levels at T120 compared with NS, HTS, and LR, HSP was significantly more effective than REP in attenuating the hepatocellular injury and the release of these enzymes (P G 0.05).
Effects of pyruvate formulations on lipid peroxidation and hepatocellular ATP

REP, and HSP, respectively (Table 3). Similar to our previous observations, both REP and HSP significantly corrected the cellular ATP reduction in liver, but HSP was more effective and better than REP (P G 0.05).
Effects of pyruvate formulations on stress-related and apoptotic signaling proteins

There was a significant increase in lipid peroxidation (MDA levels; Table 3) of liver tissue after resuscitation (3to 5-fold) compared with sham values. Although REP was significantly more effective than NS, HTS, and LR in minimizing the amount of lipid peroxidation, HSP was still better than all these treatments in reducing lipid peroxidation, which occurred during shock phase. In our rat HS model (hemorrhage followed by subsequent resuscitation with NS, HTS, LR, REP, and HSP), the levels of cellular ATP in the liver were 39% T 1.1%, 36% T 1.0%, 52% T 0.8%, 63% T 1.7%, and 78% T 1.8% of sham in NS, HTS, LR,

To gain insight into the molecular mechanisms of the action of NS, HTS, LR, REP, and HSP, the expression of stressrelated proteins (COX-2 and iNOS) and apoptotic signaling proteomic biomarkers such as caspase -3 activation and PARP cleavage in liver was determined by Western blotting. Figure 4 shows representative Western blots from one animal. The specific bands of stress-related proteins such as COX-2 and iNOS and apoptotic signaling proteins such as caspase-3 and PARP from three individual experiments were subjected to densitometric analysis and normalized with data obtained from sham-operated animals. As indicated by the COX-2/sham ratio in Figure 5A, HSP infusion prevented the increase in COX-2 expression, whereas NS, HTS, LR, and REP failed to inhibit the upregulation of COX-2 induced by HS. In contrast, the effect of HS on iNOS was not affected by any of the resuscitative fluids tested (Fig. 5B).

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK MAY 2010

HYPERTONIC SODIUM PYRUVATE

TABLE 2. Effect of resuscitation on acid-base parameters in HS

IN

SHOCK

537

Parameters Base excess, mEq/L

Time, min T0 T60 T120

Sham (n = 8) 4.1 T 0.1 3.5 T 0.7 3.8 T 0.3 7.41 T 0.01 7.44 T 0.02 7.42 T 0.01 1.2 T 0.01 1.3 T 0.07 1.9 T 0.02 9.3 T 0.4 11.7 T 1.3 12.8 T 0.3

Control (n = 8) 4.0 T 0.2 j9.3 T 1.9* j3.07 T 0.4 7.43 T 0.03 7.26 T 0.03 7.31 T 0.01 1.5 T 0.03 6.2 T 0.9* 12.4 T 0.3 11.4 T 0.5 39.5 T 2.4* 41.6 T 0.3

HTS (n = 8) 5.5 T 0.6 j8.7 T 1.9* j14.5 T 2.4 7.39 T 0.01 7.28 T 0.02 7.37 T 0.04 2.1 T 0.04 6.3 T 1.6* 11.4 T 0.5 11.8 T 0.2 41.5 T 2.2* 38.9 T 0.3

LR (n = 8) 6.4 T 0.6 j9.9 T 1.5* 0.64 T 2.0 7.51 T 0.01 7.31 T 0.03 7.38 T 0.02 1.3 T 0.1 7.1 T 2.4* 12.9 T 1.1 10.6 T 0.9 40.5 T 1.4* 44.3 T 0.5

REP (n = 8) 4.8 T 0.22 j6.5 T 0.4* 2.5 T 0.4 7.45 T 0.03 7.3 T 0.01 7.39 T 0.04 1.3 T 0.2 9.4 T 0.8* 10.5 T 0.6 12.1 T 0.1 37 T 1.7* 39.5 T 2.6

HSP (n = 8) 4.2 T 0.5 j13 T 1.3* 10.0 T 2.4k 7.52 T 0.01 7.29 T 0.02 7.42 T 0.04 1.2 T 0.1 8.4 T 0.5* 14.7 T 0.8k 11.3 T 0.3 40.9 T 1.2* 21.4 T 0.4k

pH

T0 T60 T120

Lactate, mmol/L

T0 T60 T120

l/p ratio

T0 T60 T120

The acid-base measurements were made immediately before the start of the controlled arterial hemorrhage (T0), at 60 min after hemorrhage or sham (T60), and at the end of experiment at T120. The NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, REP, and HSP treatments were started 60 min after the start of hemorrhage for the next 45 min until the end of the experiment (T120). *P G 0.05, compared with baseline for time-matched data. P G 0.05, compared with after shock values at T60. P G 0.05, comparing the LR group with the REP group. P G 0.05, comparing the HTS group with the HSP group. k P G 0.05, comparing the REP group with the HSP group.

Also shown in Figure 5C, the ratio of apoptotic signaling protein and sham treatment indicates that caspase-3 activation after NS, HTS, and LR resuscitation was significantly higher than that of REP and HSP (P G 0.05), and both REP and HSP were equally effective in reducing the caspase-3 activation. The bottom panel of Figure 4 and normalized PARP ratio with sham in Figure 5 show that the intact PARP was abundant in all liver samples as a 116-kd protein, and that the total PARP was significantly increased in all HS groups compared with HSP (P G 0.05), indicating a trigger of DNA repair mechanism by the activation of PARP. Examining the PARP cleavage pattern in these samples shows that a background apoptosis signal was detectable as the faint 85-kd fragments of PARP in the liver of HSP-resuscitated animals, whereas the 85-kd bands had a higher ratio of cleaved PARP in NS-, HTS-, LR-, and REP-infused animals (P G 0.05; n = 3). The densitometric analysis also indicated that the percentage of total PARP fragmentation in the HSP group was significantly less than other treatment groups (PARP fragmentation was

18%, 26%, 22%, 27%, 24%, and 13% in sham, NS, HTS, LR, REP, and HSP groups, respectively). DISCUSSION Hemorrhagic shock and the consequences of its treatment on liver damage are known to be influenced by the volume and composition of resuscitative fluids (6, 7, 9, 20Y23). In this study, using a rat model of HS, we have demonstrated that infusion of low-volume HSP solution is significantly better than standard high-volume isosmotic resuscitation fluids and high-volume REP. The protective effect of HSP on liver damage was related to its metabolic, anti-inflammatory, antioxidant, and antiapoptotic properties. Variation in the intensity of the HS was not a likely cause of differences in the treatment groups because of the similarity of the MAP (40 mmHg), amount of shed blood, base excess, pH, Po2, PCO2, lactate, and l/p ratio throughout the shock period (T60). During the resuscitation and recovery

TABLE 3. Effects of resuscitation on liver damage, lipid peroxidation, and hepatocellular ATP Parameters AST, U/L ALT, U/L MDA, nMol/mg ATP, 2mol/g liver tissue Sham (n = 8) 57 T 3.0 11.0 T 0.9 2.7 T 0.2 2.8 T 0.09 Control (n = 8) 137 T 4.5* 55.2 T 1.4* 12.4 T 0.5* 1.1 T 0.06* HTS (n = 8) 116 T 5.2* 37.5 T 1.3* 7.3 T 0.3* 1.0 T 0.05* LR (n = 8) 158 T 6.0* 41.9 T 2.0* 9.8 T 0.9* 1.4 T 0.07* REP (n = 8) 110 T 3.8* 29.8 T 1.7*

HSP (n = 8) 76 T 4.7* 17.3 T 0.8* 4.2 T 0.2*

6.9 T 0.4* 1.7 T 0.04*

2.2 T 0.03*

Changes in liver functions, accumulation of MDA adducts, and ATP levels caused by HS and resuscitation. Treatments included sham, instrumental control and no hemorrhage, NS, HTS, LR, REP, and HSP. *P G 0.05 comparing the treatments with sham for time-matched data P G 0.05, comparing the LR group with the REP group. P G 0.05, comparing the HTS group with the HSP group. P G 0.05, comparing the REP group with the HSP group.

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

538

SHOCK VOL. 33, NO. 5

SHARMA

AND

MONGAN

FIG. 4. Regulation of stress-related proteins in liver after resuscitation. Representative Western blots of the effects of pyruvate administration on critical indicators of hepatic apoptosis. Liver cell homogenates from HS rats were used for Western blotting according to our published methods (7, 9, 11, 16). Hypertonic sodium pyruvate prevented the activation of HS-mediated COX-2 and caspase-3 and reduced PARP cleavage in comparison to NS (0.9% NaCl, standard volume control), HTS (7.5% NaCl, osmotic control), LR, and REP treatments.

phases, there were no significant differences in the MAP profile between the REP the HSP treatment groups. Although we used 3:1 volume replacement to blood loss regimens for the isotonic fluid treatments and used a treatment volume

(5 mL/kg) and an osmotic control for the hypertonic pyruvate treatment based on results of previous studies (11, 13), we cannot discount potential differences in volume status or cardiovascular performance affecting the results because those

FIG. 5. Normalized protein bands with sham group of animals. Data presented as group means T SE from three animals per group. Using immunoblot analysis, proteins of COX-2, iNOS, casapse-3, and PARP were assessed in lysates of liver of sham or HS-operated rats with resuscitation (n = 3). The specific bands of proteins were quantitated densitometrically and normalized with sham-operated animals to examine the effectiveness of various treatment groups in HS. cP G 0.05, comparing the LR group with the REP group. dP G 0.05, comparing the HTS group with the HSP group. eP G 0.05, comparing the REP group with the HSP group.

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK MAY 2010

parameters were not measured. However, in accordance with previous studies, the replacement of lactate with REP solution was effective in stabilizing the acid-base status associated with HS (13). Furthermore, in this study, infusion of hypertonic HSP was more effective than REP in attenuating the systemic arterial metabolic acidosis as shown by the attenuation of the acidosis by HSP (pH and base excess) at the end of the experiment (T120). There are two potential mechanisms for this effect. First, HSP infusion significantly increased plasma Na+ relative to Clj concentration. This difference between positively and negatively charged electrolytes is associated with a natural consumption of hydrogen ions to preserve electrical neutrality (24). Second, ethyl pyruvate cleaves into ethanol and pyruvic acid, a strong acid with a pK of approximately 3, so that its infusion should lower rather than increase blood pH. Finally, HSP prevented the increase in serum l/p ratios, a marker of cytosolic redox state, which was significantly higher in REP, LR, HTS, and NS groups, suggesting that the effect of the HSP infusion on acidbase status is related to the metabolic consumption of this molecule (7, 9, 11). Consistent with these observations, we found that the degree of HS/R-induced hepatocellular injury (as assessed by leakage of the enzyme ALT and AST into plasma) was significantly attenuated by HSP compared with REP. This protective effect of HSP in reducing liver injury could be attributed to the antioxidant properties of pyruvate because free radical scavengers have shown beneficial effects on liver function after I/R (25). The antioxidant effect of HSP was observed as a significant decrease in lipid peroxidation (accumulation of MDA in the liver) with the HSP treatment. This observation further supports the observation that HSmediated liver injury is associated with oxidative stress. The hepatocellular oxidative stress triggers molecular and cellular processes within the hepatocytes that lead to the activation of Kupffer cells. When activated, Kupffer cells produce signaling molecules (i.e., cytokines) that promote inflammatory reactions as well as more reactive oxygen species, which can damage liver cells (26, 27). Although resuscitation is an obligatory intervention during HS, the therapy may exacerbate inflammation (28). Even blood transfusions contain proinflammatory mediators that both prime and activate neutrophils. Animal models of HS have suggested that different fluids or different rates of resuscitation can have a widely divergent impact on the immune response, neutrophil activation, and tissue injury (29, 30). However, there are conflicting results in the literature regarding the upregulation of circulating cytokines in the blood and their relationships to MOF (2, 31). In this study, we chose TNF-!, IL-10, and IL-6 as the target cytokines for measurement because they have been well correlated with clinical severity and prognosis in animal and human studies of HS (26, 32). We found that the serum levels of proinflammatory cytokines TNF-! and IL-6 were significantly increased after HS, and both HSP and REP were equally effectively in inhibiting their increase in blood as well as in liver. The antiinflammatory effects of REP may not be caused by its pyruvate content because there was only one tenth of pyruvate

HYPERTONIC SODIUM PYRUVATE

IN

SHOCK

539

concentration in the HSP solution. However, the presence of ethyl alcohol might be a contributing factor to the antiinflammatory properties of REP (33). The inflammatory cytokines work to maintain homeostasis under stressful conditions through a complex chain reaction or cascade that results in the inflammatory response. This response is modulated by a negative feedback system, which consists not only of the self-inhibitory action of TNF-! and IL-6, but also of the production of anti-inflammatory mediators such as IL-10 (34). However, both the REP and HSP solutions did not alter the IL-10 levels in either systemic circulation or in liver. Inflammatory processes are also mediated by multiple molecular mechanisms of which COX-2 is a major inflammatory mediator (35). Prostaglandins, synthesized by COX-2 and released by macrophages after injury, mediate the powerful inflammatory response, augment caspase-3 activation, and increase Bax/Bcl2 ratio after trauma and hemorrhage (36). However, COX-2 inhibition has been shown to have potent anti-inflammatory effects (37). In the present study, our data showed that inhibition of COX-2 by HSP may have also contributed to the significant decrease in oxidative damage and improved liver functions compared with REP. These observations corroborate the findings of Begay and Gandolfi (38), who reported that COX-2 inhibitors can prevent an increase in ALT and necrosis in injured liver. Furthermore, we also found that the altered liver functions and increased accumulation of MDA are associated with the massive proinflammatory response to HS in which TNF-! and IL-6 were released by the liver (10Y15 times the serum levels of these cytokines). Together, these data support an antiinflammatory role for HSP over REP as a resuscitative adjuvant based on these numerous measurements, indicating a reduction in inflammation and cellular damage. In addition, decreased caspase-3 activation and reduced PARP cleavage in HSP-resuscitated animals indicate some form of protection against cellular death signaling. The role of PARP fragmentation in signaling of apoptosis as a result of DNA damage is well defined. The PARP at the mitochondrial level induces release of cytochrome c and apoptosis-inducing factors (39). Because of the conflicting role of PARP in apoptosis and necrosis, it is difficult to define the HS-mediated cell death by apoptosis or necrosis, but it is clear from our study that inhibition of PARP fragmentation is associated with attenuation of liver injury and loss of cellular ATP. In comparison to HSP, the lack of significant benefit to REP in our rat model has many possible explanations, such as 1) the amount of pyruvate in REP was very small in comparison with its concentration in HSP; 2) the pathophysiological mechanism of HS in our model may be different from other model systems in which ethyl pyruvate has been shown to have protective effects (e.g., septic shock); and 3) the presence of ethyl alcohol in REP may have masked the metabolic properties of pyruvate. These issues underscore the need to better understand the pharmacology of HSP as a novel resuscitative fluid. In conclusion, in comparison with REP, HSP treatment provides a significant protection to the liver in an acute controlled model of rat HS through the improvement in hemodynamic parameters and tissue energetic, reduced inflammatory cascade,

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.

540

SHOCK VOL. 33, NO. 5

SHARMA

AND

MONGAN

and inhibition of stress-related and apoptotic signaling proteins. Long-term outcome studies are needed to further evaluate any potential beneficial effects of HSP in severe HS. ACKNOWLEDGMENTS
The authors thank Sean Rotolo for assistance with Western blotting and Michael J. Dymond for technical help.

REFERENCES
1. Wang P, Hauptman JG, Chaudry IH: Hepatocellular dysfunction occurs early after hemorrhage and persists despite fluid resuscitation. J Surg Res 48: 464Y470, 1990. 2. Hierholzer C, Harbrecht B, Menezes JM, Kane J, MacMicking J, Nathan CF, Peitzman AB, Billiar TR, Tweardy DJ: Essential role of induced nitric oxide in the initiation of the inflammatory response after hemorrhagic shock. J Exp Med 187:917Y928, 1998. 3. Basu S: Oxidative injury induced cyclooxygenase activation in experimental hepatotoxicity. Biochem Biophys Res Commun 254:764Y767, 1999. 4. Rizoli SB: Crystalloids and colloids in trauma resuscitation: a brief overview of the current debate. J Trauma 54:82Y88, 2003. 5. Stern SA, Wang X, Mertz M, Chowanski ZP, Remick DG, Kim HM, Dronen SC: Under-resuscitation of near-lethal uncontrolled hemorrhage: effects on mortality and end-organ function at 72 hours. Shock 15:16Y23, 2001. 6. Doucet JJ, Hall RI: Limited resuscitation with hypertonic saline, hypertonic sodium acetate, and lactated Ringers solutions in a model of uncontrolled hemorrhage from a vascular injury. J Trauma 47:956Y963, 1999. 7. Mongan PD, Capacchione J, West S, Karaian J, Dubois D, Keneally R, Sharma P: Pyruvate improves redox status and decreases indicators of hepatic apoptosis during hemorrhagic shock in swine. Am J Physiol Heart Circ Physiol 283:H1634YH1644, 2002. 8. Mongan PD, Fontana JL, Chen R, Bunger R: Intravenous pyruvate prolongs survival during hemorrhagic shock in swine. Am J Physiol 277:H2253YH2263, 1999. 9. Mongan PD, Karaian J, Van Der Schuur BM, Via DK, Sharma P: Pyruvate prevents poly-ADP ribose polymerase (PARP) activation, oxidative damage, and pyruvate dehydrogenase deactivation during hemorrhagic shock in swine. J Surg Res 112:180Y188, 2003. 10. Sharma P, Karian J, Sharma S, Liu S, Mongan PD: Pyruvate ameliorates post ischemic injury of rat astrocytes and protects them against PARP mediated cell death. Brain Res 992:104Y113, 2003. 11. Sharma P, Walsh KT, Kerr-Knott KA, Karaian JE, Mongan PD: Pyruvate modulates hepatic mitochondrial functions and reduces apoptosis indicators during hemorrhagic shock in rats. Anesthesiology 103:65Y73, 2005. 12. Fink MP: Ringers ethyl pyruvate solution: a novel resuscitation fluid. Minerva Anestesiol 67:190Y192, 2001. 13. Fink MP: Ringers ethyl pyruvate solution: a novel resuscitation fluid for the treatment of hemorrhagic shock and sepsis. J Trauma 54:141Y143, 2003. 14. Passonneau JV, Lowery O: Enzymatic Analysis, A Practical Guide. Totowa, NJ: Humana Press, 121Y126, 1993. 15. Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248Y254, 1976. 16. Musser JB, Bentley TB, Griffith S, Sharma P, Karaian JE, Mongan PD: Hemorrhagic shock in swine: nitric oxide and potassium sensitive adenosine triphosphate channel activation. Anesthesiology 101:399Y408, 2004. 17. Majde JA: Animal models for hemorrhage and resuscitation research. J Trauma 54:100Y105, 2003. 18. Ng PC, Li K, Wong RP, Chui K, Wong E, Li G, Fok TF: Proinflammatory and anti-inflammatory cytokine responses in preterm infants with systemic infections. Arch Dis Child Fetal Neonatal Ed 88:F209YF213, 2003.

19. Taniguchi T, Koido Y, Aiboshi J, Yamashita T, Suzaki S, Kurokawa A: Change in the ratio of interleukin-6 to interleukin-10 predicts a poor outcome in patients with systemic inflammatory response syndrome. Crit Care Med 27:1262Y1264, 1999. 20. Chaudry IH: Cellular mechanisms in shock and ischemia and their correction. Am J Physiol 245:R117YR134, 1983. 21. Deb S, Martin B, Sun L, Ruff P, Burris D, Rich N, DeBreux S, Austin B, Rhee P: Resuscitation with lactated Ringers solution in rats with hemorrhagic shock induces immediate apoptosis. J Trauma 46:582Y588, 1999. [discussion 588Y589]. 22. Mulier KE, Beilman GJ, Conroy MJ, Taylor JH, Skarda DE, Hammer BE: Ringers ethyl pyruvate in hemorrhagic shock and resuscitation does not improve early hemodynamics or tissue energetics. Shock 23:248Y252, 2005. 23. Fink MP: Ethyl pyruvate: a novel anti-inflammatory agent. Crit Care Med 31:51Y56, 2003. 24. Stewart PA: Modern quantitative acid-base chemistry. Can J Physiol Pharmacol 61:1444Y1461, 1983. 25. Lim SP, Andrews FJ, Christophi C, OBrien PE: Microvascular changes in liver after ischemia-reperfusion injury. Protection with misoprostol. Dig Dis Sci 39:1683Y1690, 1994. 26. Keller SA, Paxian M, Lee SM, Clemens MG, Huynh T: Kupffer cell ablation attenuates cyclooxygenase-2 expression after trauma and sepsis. J Surg Res 124:126Y133, 2005. 27. Koo DJ, Chaudry IH, Wang P: Kupffer cells are responsible for producing inflammatory cytokines and hepatocellular dysfunction during early sepsis. J Surg Res 83:151Y157, 1999. 28. Silliman CC, Clay KL, Thurman GW, Johnson CA, Ambruso DR: Partial characterization of lipids that develop during the routine storage of blood and prime the neutrophil NADPH oxidase. J Lab Clin Med 124:684Y694, 1994. 29. Rhee P, Burris D, Kaufmann C, Pikoulis M, Austin B, Ling G, Harviel D, Waxman K: Lactated Ringers solution resuscitation causes neutrophil activation after hemorrhagic shock. J Trauma 44:313Y319, 1998. 30. Knoferl MW, Angele MK, Ayala A, Cioffi WG, Bland KI, Chaudry IH: Do different rates of fluid resuscitation adversely or beneficially influence immune responses after trauma-hemorrhage? J Trauma 46:23Y33, 1999. 31. Gaddipati JP, Sundar SV, Calemine J, Seth P, Sidhu GS, Maheshwari RK: Differential regulation of cytokines and transcription factors in liver by curcumin following hemorrhage/resuscitation. Shock 19:150Y156, 2003. 32. Roumen RM, Hendriks T, van der Ven-Jongekrijg J, Nieuwenhuijzen GA, Sauerwein RW, van der Meer JW, Goris RJ: Cytokine patterns in patients after major vascular surgery, hemorrhagic shock, and severe blunt trauma. Relation with subsequent adult respiratory distress syndrome and multiple organ failure. Ann Surg 218:769Y776, 1993. 33. Phelan H, Stahls P, Hunt J, Bagby GJ, Molina PE: Impact of alcohol intoxication on hemodynamic, metabolic, and cytokine responses to hemorrhagic shock. J Trauma 52:675Y682, 2002. 34. Karakozis S, Hinds M, Cook JW, Kim D, Provido H, Kirkpatrick JR: The effects of interleukin-10 in hemorrhagic shock. J Surg Res 90:109Y112, 2000. 35. Hierholzer C, Billiar TR: Molecular mechanisms in the early phase of hemorrhagic shock. Langenbecks Arch Surg 386:302Y308, 2001. 36. Zeini M, Hortelano S, Traves PG, Martin-Sanz P, Bosca L: Simultaneous abrogation of NOS-2 and COX-2 activities is lethal in partially hepatectomised mice. J Hepatol 40:926Y933, 2004. 37. Strong VE, Mackrell PJ, Concannon EM, Naama HA, Schaefer PA, Shaftan GW, Stapleton PP, Daly JM: Blocking prostaglandin E2 after trauma attenuates pro-inflammatory cytokines and improves survival. Shock 14: 374Y379, 2000. 38. Begay CK, Gandolfi AJ: Late administration of COX-2 inhibitors minimize hepatic necrosis in chloroform induced liver injury. Toxicology 185:79Y87, 2003. 39. Heeres JT, Hergenrother PJ: Poly(ADP-ribose) makes a date with death. Curr Opin Chem Biol 11:644Y653, 2007.

Copyright 2010 by the Shock Society. Unauthorized reproduction of this article is prohibited.