Hepatic warm ischemia and reperfusion (IR) injury occurs in many clinical situations and has an important link to subsequent hepatic failure. The pathogenesis of this injury involves numerous pathways, including mitochondrial-associated apoptosis. We studied the effect of mitochondrial calcium uptake inhibition on hepatic IR injury using the specific mitochondrial calcium uptake inhibitor, ruthenium red (RR). Rats were subjected to 1 hour of 70% warm hepatic ischemia following RR pretreatment or vehicle injection. Sham-operated animals served as controls. Analysis was performed at 15 minutes, 1 hour, 3 hours, or 6 hours after reperfusion. Serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) concentrations were determined. Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) staining was performed to assess apoptosis, and hepatocellular necrosis was semiquantitated from hematoxylin and eosin–stained tissue sections. RR pretreatment significantly decreased both AST and ALT serum levels after 6 hours of reperfusion (AST: 1,556 ± 181 U/L vs. 597 ± 121 U/L, P = 0.005; ALT: 1,118 ± 187 U/L vs. 294 ± 39 U/L, P = 0.005). Apoptosis was observed within 15 minutes of reperfusion in vehicle-pretreated animals and peaked after 3 hours of reperfusion (98 ± 21 cells/high-power field [hpf]). Apoptosis was inhibited at all time points by RR pretreatment. Histologic evidence of necrosis was not observed prior to 3 hours of reperfusion (23% ± 4%), and maximal necrosis was observed after 6 hours of reperfusion (26% ± 1% percent area). RR pretreatment significantly decreased the necrotic percent area at both the 3-hour and the 6-hour time points (4.2% ± 2%; 3.7% ± 1%, respectively). Hepatic IR injury resulted in both apoptotic and necrotic cell death, which were attenuated by RR pretreatment. In conclusion, these observations implicate mitochondrial calcium uptake in the pathogenesis of hepatic IR injury. (Liver Transpl 2005;11:663–668.)
Hepatic warm ischemia and reperfusion (IR) injury occur in many clinical situations including transplantation, resection, trauma, shock, hemorrhage, and thermal injury. Evidence exists linking the periods of ischemia associated with these events to subsequent liver failure.1 In addition, there is evidence that the duration of the ischemic periods during hepatic resection directly correlates with postoperative liver dysfunction.2–5 The pathogenesis of ischemic hepatocellular injury is complex and involves numerous mediators.6–9 Although many authors have demonstrated apoptosis following hepatic ischemia, the role of mitochondrial-associated apoptotic mechanisms in hepatic IR injury remains unclear. A better understanding of the cellular pathophysiology of IR and identification of methods to attenuate this injury has important clinical implications.
Within the last decade, a series of findings have shown that mitochondria play a central role in intracellular calcium dynamics.10, 11 Mitochondria can take up calcium rapidly via the ruthenium red (RR)-sensitive uniporter that spans the inner mitochondrial membrane. This uptake is driven by the negative potential difference across the mitochondrial membrane (ΔΨ), which is maintained by proton extrusion from the electron transport chain.12 The mitochondrial matrix has a large calcium binding ratio, which allows mitochondria to accumulate extraordinary amounts of calcium, yet retain a negative membrane potential and avoid permeability transition pore formation. These properties allow mitochondria to function as buffers against cytosolic calcium overload; however, this ability to be an effective calcium buffer requires active electron transport.12, 13
Hepatic ischemia increases cytosolic calcium concentrations as a result of both an increased release from the endoplasmic reticulum and other calcium stores as well as decreased calcium extrusion via membrane pumps.14 As a result, rapid mitochondrial calcium uptake occurs in an effort to buffer the cytosolic calcium incease.15 However, as ischemia persists electron transport is inhibited, and ATP levels are depleted. The continued accumulation of calcium under these circumstances compromises ΔΨ. Ultimately if ischemia is not rectified, mitochondrial calcium overload occurs, resulting in complete collapse of ΔΨ, and mitochondrial-initiated cell death.13, 16, 17
RR is a glycoprotein stain that has a high affinity for the mitochondrial calcium uniporter and inhibits uniporter-dependent calcium uptake in a noncompetitive manner.18 RR is a well-known and widely studied tool for modulation of mitochondrial calcium uptake. The uniporter is a gated ion channel and can be completely inhibited by micromolar concentrations of RR.18–24 RR can prevent mitochondrial calcium uptake in isolated mitochondria, cell culture, tissue culture, and in vivo.22, 24–27 We have previously demonstrated that prevention of mitochondrial calcium uptake using RR prevents apoptosis in cultured HepG2 cells during the early rewarming period following an ischemic insult.28 In this article, we investigate the role of pathologic mitochondrial calcium uptake on warm hepatic IR injury. We report here that prevention of mitochondrial calcium uptake using RR pretreatment attenuates reperfusion injury in a rat warm ischemia model.
All animal studies were performed in accordance with National Institutes of Health animal care procedures, and they were approved by and carried out under the guidelines of the Vanderbilt University Institutional Animal Care and Use Committee. Male Sprague-Dawley rats weighing between 300 and 400 g were used for these experiments. Animals were housed in cages, kept on a 12:12 hour light/dark cycle, fed daily with rat chow, and given water ad libitum.
Model of Hepatic Ischemia
Following an overnight fast, animals were anesthetized using isoflurane (3%, 2 L O2). Via a midline incision, the portal vein, hepatic artery, and bile duct were identified and isolated. Approximately 70% hepatic ischemia was created by using an atraumatic vessel clamp (S&T Microclamp B-3; Fine Science Tools, Foster City, CA) to occlude the portal vein and hepatic artery branches to the median and left hepatic lobes. This method is widely reported and avoids mesenteric congestion by allowing portal venous drainage via the right lobe and caudate lobe.
At 30 minutes prior to the initiation of hepatic ischemia, animals received RR (50 μg/kg) or vehicle (0.9% NaCl) intravenously. Ischemia was carried out for 1 hour. The total anesthesia time for the animals was 90 minutes. Sham-operated animals received laparotomy, mobilization of the liver, and underwent equivalent anesthesia.
Following vascular clamp removal, animals were allowed to reperfuse for periods of 15 minutes, 1 hour, 3 hours, or 6 hours. At the time of animal sacrifice, serum was collected, and hepatic tissue was preserved for histological analysis. A total of 40 animals were used for this experiment. A total of 10 animals were used for each time point: 4 hepatic ischemia, 4 hepatic ischemia with RR pretreatment, and 2 sham-operated animals.
Serum Transaminase Determinations
Serum collected from each animal at the time of euthanasia was analyzed for aspartate aminotransferase (AST) and alanine aminotransferase (ALT) concentrations. These measurements were carried out by Mid South Vet Lab (Murfreesboro, TN).
Assessment of Apoptosis
In order to quantitate the rate of apoptosis in a sensitive manner, we performed terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) assay on paraffin embedded tissue sections from each animal. This assay was performed following the manufacturer's recommendations for a commercially available kit (ApopTag Plus Fluorescein In Situ Apoptosis Detection Kit; Chemicon International, Temecula, CA). Five randomly selected fields from each tissue section were examined at an original magnification of 200×. The number of positively staining cells per field was recorded. Only hepatocytes were included in this analysis. The average number of positively staining nuclei per field for each animal was used for analysis.
Quantitation of Hepatocellular Necrosis
Paraffin embedded tissue sections were stained with hematoxylin and eosin and examined at an original magnification of 200× (Axio-Plan 2; Zeiss, Hallbegmoos, Germany). Digital photomicrographs were made from 5 randomly selected fields of each tissue section (Axiocam HRc; Zeiss). Necrosis was defined as areas demonstrating one or more of the following characteristics: nuclear pyknosis, cytoplasmic hypereosinophilia, loss of distinct cellular borders, hemorrhage, and neutrophil infiltration. These traits correspond to grade 2 and 3 hepatic injury as defined by Camargo et al.29 ImagePro software (ImagePro Plus v4.5; Media Cybernetics, Silver Spring, MD) was used to create specific overlays of any region of necrosis in each of the photomicrographs. Using these overlays, percent area necrosis was calculated for each photomicrograph. The average percent area necrosis for each animal was then used for analysis.
Statistical significance between study groups was determined using a 2-tailed homoscedastic Student's t test. A P value of 0.05 or below was considered significant.
RR Pretreatment Decreases Serum Transaminase Levels Following Ischemia
AST and ALT, established surrogate markers of hepatic IR injury, were measured at each reperfusion time point.30 Peak AST and ALT levels following 1 hour of 70% hepatic ischemia occurred after 3 hours of reperfusion (Fig. 1). There was a trend for the animals pretreated with RR to have lower AST and ALT levels after 3 hours of reperfusion, and the maximal effect of RR pretreatment occurred following 6 hours of reperfusion. Inhibition of mitochondrial calcium uptake with RR resulted in a 2.5-fold reduction in AST levels and a 4-fold reduction in ALT levels at the 6-hour reperfusion time point. Pretreatment of sham-operated animals with RR did not alter AST or ALT levels when compared to sham operated animals receiving vehicle alone (data not shown).
Apoptotic Cellular Injury Begins Early in the Reperfusion Period
Recent data has shown that apoptotic mechanisms of hepatocyte death play an important role in hepatic injury following an ischemic insult.31–33 Fragmentation of DNA strands is generally accepted as a specific marker of late apoptosis; therefore, we used TUNEL staining to evaluate the role that apoptotic cell injury played in our model of warm IR. Examples of these methods are demonstrated in Figure 2. The majority of apoptosis was noted in hepatocytes located in zone 3. The rate of apoptosis was quantified by counting the number of positively staining cells per high-power field [hpf], and 5 fields were averaged for each animal. Cellular apoptosis was observed as early as 15 minutes following reperfusion when compared to the apoptotic rates observed in the sham operated animals (6.3 ± 1.2 cells/hpf vs. 0.8 ± 0.2 cells/hpf; P = 0.029). The rate of TUNEL positivity increased after 1 hour and subsequently peaked at the 3-hour time point with 98 ± 21 cells/hpf. The rate of apoptosis decreased by 6 hours of reperfusion; however, the levels remain significantly higher than sham (10.6 ± 1.8 cells/hpf vs. 0.8 ± 0.2 cells/hpf; P = 0.024).
Modulation of Mitochondrial Calcium Uptake During IR Decreases the Rate of Hepatocellular Apoptosis
Our previous work has demonstrated that inhibition of mitochondrial calcium uptake using RR decreased the rate of apoptosis during the early rewarming period.28 To determine whether the attenuation of IR injury seen in the RR pretreated animals was related to a decreased rate of apoptosis, TUNEL staining at each time point compared the RR pretreated animals to vehicle treated animals. Representative sections are illustrated in Figure 2. RR pretreated animals showed a decrease in apoptosis beginning at the 15-minute time point (6.3 ± 1.2 cells/hpf vs. 3.3 ± 0.3 cells/hpf; P = 0.011). As summarized in Figure 3, RR significantly decreased the rate of apoptosis at all reperfusion time points measured in this study. While RR pretreatment did not reduce the rate of apoptosis to sham levels at any time point, the peak rate of apoptosis in pretreated animals is 9.5 ± 0.5 cells/hpf at the 3-hour time point.
RR Pretreatment Attenuates the Percentage of Hepatic Necrosis Resulting From IR Injury
Despite the well-defined role of apoptotic cell injury in hepatic IR injury, debate remains as to whether the majority of the parenchymal injury noted during reperfusion results from apoptotic or necrotic cell death.34, 35 Many groups have reported histological evidence of both apoptosis and necrosis following hepatic IR.34–36 We examined hematoxylin and eosin–stained tissue sections of liver taken at each time point in this study in order to determine the relationship of necrosis and apoptosis in our model and to evaluate any protection afforded by RR pretreatment.
Similar to the observation with apoptosis, necrotic cells were notably concentrated in zone 3. Necrosis was semiquantitated as described in the methods section and expressed as a percentage of the total tissue section. Histologic criteria for grade 2 or grade 3 injury29 was not seen in tissue sections taken at 15 minutes or 1 hour following reperfusion. After 3 hours of reperfusion, the fields examined had an average of 22% ± 4% necrosis. This increased to 26% ± 8% by 6 hours of reperfusion. As summarized in Figure 4, RR pretreatment decreased the observed amount of hepatocellular necrosis to 5% ± 2% and 3% ± 1% after 3 hours and 6 hours of reperfusion (P = 0.006 and 0.026), respectively.
This study investigated the role of mitochondrial calcium overload in warm hepatic IR injury. By using RR pretreatment to specifically inhibit uniporter dependent mitochondrial calcium uptake, we were able to globally attenuate hepatic reperfusion injury as measured by AST and ALT levels. Further analysis demonstrated a decrease in both hepatocellular necrosis and apoptosis in the pretreated rats. The decrease in apoptosis was immediately evident at the 15-minute time point and continued throughout the study period. While, RR pretreatment did not completely block the occurrence of apoptosis, it was able to reduce the peak rate of apoptosis approximately 10-fold at the 3-hour time point. Similarly, the peak appearance of necrosis in this study which occurred at the 6-hour time point was reduced approximately 10-fold by RR pretreatment.
Although debate over the roles of apoptosis versus necrosis in the etiology of hepatic IR injury remains,34, 35, 37, 38 many authors have demonstrated that apoptosis occurs following an ischemic injury.31, 36, 39, 40 In our study, we observed hepatocellular apoptosis beginning within 15 minutes of reperfusion. The rate of apoptosis subsequently peaked after 3 hours of reperfusion, which also corresponded with the first histological evidence of necrosis. Although apoptosis was still detected, the major histologic injury seen after 6 hours of reperfusion was hepatocellular necrosis. The coexistence of apoptosis and necrosis in hepatic IR injury has been noted by others, and some have proposed that the apoptotic mechanisms and necrotic mechanisms of cell death overlap.37, 38
Mitochondria play a central role in apoptotic cell death.41, 42 In addition, it has become clear that intracellular calcium signaling controls a host of cellular functions including apoptosis.12, 13 Mitochondria are active participants in cellular calcium homeostasis, and they are known to act as buffers against cytosolic calcium overload provided adequate electron transport exists to preserve ΔΨ.12, 13 In fact, mitochondrial calcium uptake serves to stimulate mitochondrial ATP production as part of a poorly understood mechanism by which cytosolic adenosine diphosphate serves to drive mitochondrial ATP production.12, 43 During ischemic conditions, however, the mitochondria are unable to maintain electron transport, which results in complete collapse of ΔΨ and mitochondrial-associated apoptotic cell death.13, 16, 17 While several mechanisms of rewarming apoptosis have been demonstrated, the role of mitochondrial calcium overload in this injury remains unclear. We have previously shown that cold ischemic storage leads to mitochondrial calcium overload that can be blocked by RR.21 In addition, storage of cells in an RR-containing solution during the period of cold ischemia prevents a caspase-independent apoptosis that occurs early in the rewarming period.28
The results of this study implicate mitochondrial calcium uptake and subsequent overload during the period of hepatic ischemia in the pathogenesis of hepatic IR injury. RR pretreatment led to an attenuation of early reperfusion apoptosis that remained throughout the study period. The observed decrease in hepatocellular necrosis was a surprising finding. Mitochondrial calcium overload has a clear role in some etiologies of apoptosis, but the role of mitochondrial calcium regulation in necrotic cell death is unknown. Necrotic cell death is generally believed to result from an acute metabolic disruption leading to rapid adenosine triphosphate (ATP) depletion, cellular swelling, and activation of degradative enzymes, which ends with the rupture of the cell membrane. Previous work by Qian et al.44 demonstrated that hepatocytes experience a rapid onset of permeability transition pore formation following calcium ionophore treatment, which leads to profound ATP depletion and necrotic cell death. However, when ATP levels were maintained by various means during these experiments, permeability transition pore formation still occurred, but instead of rapid necrotic cell death, the cells underwent apoptosis.44 From these experiments, Lemasters et al.37, 38 have proposed the necroapoptosis pathway of cell death in which a common death signal may culminate in either necrotic or apoptotic cell death.
This theory offers a plausible explanation for the observations in our study. In addition, recently RR has been demonstrated to preserve tissue ATP levels following a severe burn injury, and intracellular ATP has been demonstrated as a requirement for mitochondrial associated apoptosis.27, 45 In our study, apoptosis preceded necrosis; however, following 3 hours of reperfusion, the rate of necrosis increased as the rate of apoptosis decreased. In addition, RR decreased the rate of both apoptosis and necrosis by equivalent magnitudes, suggesting that a common cell death trigger was being affected. Given the known specificity of RR for the mitochondrial calcium uniporter, our data indicate that mitochondrial calcium overload can lead to both apoptotic and necrotic cell death.