Improvement of rat liver graft quality by pifithrin-α–mediated inhibition of hepatocyte necrapoptosis



Early graft dysfunction due to ischemia reperfusion injury remains a major clinical challenge in liver transplantation. Because apoptosis may contribute to graft dysfunction, we studied whether transient inhibition of p53 is capable of improving graft quality by reducing apoptotic cell death. Rat livers were harvested and stored for 24 hours or 48 hours in a 4°C solution containing either pifithrin-α (PFT-α), a specific p53-inhibitor, or the vehicle dimethyl-sulfoxide. Storage was followed by 2-hour reperfusion with 37°C Krebs-Henseleit buffer in an isolated liver perfusion system. Besides caspase-3 activation, apoptosis was quantified using fluorescence microscopy and hematoxylin-eosin histology. Trypan blue allowed for assessment of cell membrane damage, indicating both secondary apoptosis and primary necrosis. Bile flow, oxygen consumption, K+-excretion and enzyme release served as indicators of overall graft quality. Upon 2-hour reperfusion, livers developed procaspase activation as well as a mixture of apoptotic and necrotic cell death, representing necrapoptosis. In livers that had been stored for 48 hours, necrapoptotic injury was more pronounced compared with that after 24-hour storage. PFT-α effectively attenuated caspase activation as well as hepatocellular apoptosis and necrosis. Attenuation of both modes of cell death by PFT-α was associated with improved liver function, metabolism, and integrity. Experiments with the caspase inhibitor z-VAD-fmk confirmed that apoptosis is one mode of cell death in cold ischemia reperfusion. In conclusion, inhibition of p53-dependent apoptosis by PFT-α reduces hepatic preservation-reperfusion injury and improves primary organ function and metabolism. Fortification of the preservation solution with PFT-α may represent a promising and easily applicable approach to mitigate reperfusion injury in liver transplants. (HEPATOLOGY 2004;39:1553–1562.)

The mounting number of patients awaiting liver transplantation and the still-limited pool of donor organs underline the expanding necessity of using all available organs for transplantation. However, cold storage at 4°C as a standard technique for organ preservation is quite limited, with irreversible injury occuring after prolonged periods beyond 16 hours to 24 hours. Hence current principles of organ preservation chiefly strive to both confer optimal primary graft function and prolong organ ischemic tolerance.

Preservation injury is mainly due to cold ischemia reperfusion injury and represents the major cause of primary graft nonfunction following liver transplantation.1 The cellular compartments in liver preservation injury include Kupffer cells, which become primed and activated2, 3; endothelial lining cells, which become rounded and detached4; and hepatocytes, which although minimally affected by hypothermia4, 5 encounter necrotic cell death upon reperfusion.6 A growing body of literature suggests that besides hepatocellular necrosis, apoptosis may occur, significantly contributing to organ damage caused by ischemia reperfusion and transplantation.7–9

As a morphologically distinct form of cell death, apoptosis is characterized by cell shrinkage and rounding up, nuclear and cytoplasmic condensation, nuclear fragmentation, and DNA cleavage. Apoptosis can be induced by different initiating mechanisms such as oxidative stress, physical injury, mitochondrial dysfunction, and various ligand/receptor interactions (Fas, Fas-ligand, TRAIL, TNFRp55).10–12 Because the genes for these membrane death receptors are, at least in part, under the transcriptional control of the p53 tumor suppressor and have been shown to be upregulated under conditions of stress in a p53-dependent manner,13, 14 it is tempting to speculate that pifithrin-α (PFT-α), a p53-blocking agent,15 may protect against apoptotic cell death upon cold ischemia reperfusion and thus ameliorate graft injury. We therefore procured livers with PFT-α–supplemented histidine-tryptophan-ketoglutarate (HTK) solution and studied hepatocellular apoptosis and necrosis, proapoptotic protein expression, and liver dysfunction during poststorage reperfusion.


PFT-α, pifithrin-α; HTK, histidine-tryptophan-ketoglutarate; ALT, alanine transaminase; AST, aspartate transaminase; KHB, Krebs-Henseleit bicarbonate; PAS, periodic acid Schiff; DMSO, dimethyl sulfoxide.

Materials and Methods


Sprague-Dawley rats of either sex weighing 300–350 g were used. Animals were housed one per cage at 22°C–24°C with a 12-hour dark–light cycle, and were kept on water and standard chow ad libitum. After approval by the local animal care committee, the experiments were conducted according to the National Institutes of Health Guide for Care and Use of Laboratory Animals (NIH publication 86-23, revised 1985).

Liver Procurement.

Pentobarbital-anesthetized animals (50 mg/kg intraperitoneally) were placed in the supine position and tracheotomized to facilitate spontaneous respiration. A polyethylene catheter in the left carotid artery allowed for injection of the fluorescent dye bisbenzimide (H33342, 2 μmol/kg; Sigma, Deisenhofen, Germany) and blood sampling. After transverse laparotomy and cannulation of the common bile duct, bile was collected for 20 minutes to assess baseline values. Livers were then flushed via the abdominal aorta with 100 mL of 4°C HTK solution (Köhler Chemie, Alsbach-Hähnlein, Germany) by gravity of 100 cm H2O. Livers were immediately excised, weighed, and stored in 4°C HTK solution.

Isolated Liver Reperfusion.

After 24- or 48-hour storage, livers were flushed with 40 mL of 37°C Ringer's lactate. Aliquots of the effluent flush were sampled for analysis of electrolyte concentrations, pH, and alanine transaminase (ALT) and aspartate transaminase (AST) activities.

Livers were then reperfused for 2 hours through the portal vein in a nonrecirculating fashion with freshly prepared Krebs-Henseleit bicarbonate (KHB) buffer saturated with 95% oxygen and 5% carbon dioxide at a flow rate of 2 mL/min × g liver tissue using a pulsatile perfusion pump (beta/4, ProMinent, Heidelberg, Germany). Portal venous pressure was assessed continuously throughout the reperfusion period. Portal venous resistance was calculated from portal venous pressure and flow rate and expressed in mm Hg × min/mL.16

To assess liver excretory function and tissue integrity, bile flow was collected at 30-minute intervals. Aliquots of effluent fluid were collected for liver enzyme analysis. Moreover, perfusate samples were simultaneously withdrawn from the portal inflow and the venous outflow after 5, 30, 60, 90, and 120 minutes for direct analysis of pO2, pCO2, and pH, as well as for electrolyte concentrations. Liver oxygen consumption was calculated by: oxygen consumption (μmol/min × g liver tissue) = (pO2 inflow − pO2 outflow) (mm Hg) × 0.00136 (μmol/mL × mm Hg)/flow rate (mL/min × g liver tissue).17

Fluorescence Microscopy.

Using a modified fluorescence microscope (Axiotech, Zeiss, Jena, Germany) attached to an ultraviolet filter system (330–380/>415 nm), the microscopic images were recorded by a charge-coupled device video camera and stored on videotape for off-line analysis.18 Using a water immersion objective (W63x/0.90, Zeiss), hepatocellular apoptosis was assessed during the isolated perfusion procedure by visualizing bisbenzimide-stained hepatocytes.19 Quantitative analysis was performed in 10 fields per liver by counting the number of cells showing apoptosis-associated condensation, fragmentation, and/or crescent-shaped formation of nuclear chromatin, and is given in percent of all cells visible.19

Trypan Blue Uptake.

Parenchymal cell membrane damage was assessed by trypan blue perfusion.20 After 2-hour reperfusion, trypan blue (Merck, Darmstadt, Germany) was added to the circuit at a concentration of 200 μM and was perfused via the portal vein for additional 10 minutes. After flushing with KHB buffer to remove excess dye, whole livers were fixed by flushing with 1% paraformaldehyde, stored in 10% formalin for 2–3 days, embedded in paraffin and processed for light microscopy. Within 20 fields per section, trypan –blue positive tissue areas were planimetrically assessed (CapImage, Zeintl, Heidelberg, Germany) and given in percent of the total area of observation.

Sampling and Assays.

Bile samples were weighed and standardized per gram of liver wet-weight (μL/min × g liver tissue), assuming a specific weight of 1g/mL.21 ALT and AST activities were analyzed in perfusate samples by means of standard spectrophotometric techniques. At the end of both cold preservation and 2-hour reperfusion, liver tissue was sampled for histology and Western blot analysis.


Liver tissue was fixed for 2–3 days in 4% formalin and embedded in paraffin. From paraffin-embedded tissue blocks, 5-μm sections were cut and stained with hematoxylin-eosin for analysis of hepatocellular apoptosis and vacuolation as well as venular endothelial detachment. Apoptotic cells were morphologically identified through cell shrinkage, chromatin condensation, chromatin fragmentation, and apoptotic bodies; they were then counted and expressed in percent of all cells within 15 consecutive high-power fields. Cytoplasmic vacuolation of hepatocytes was scored in 20 high-power fields from 0 to 4 according to Calabrese et al.,22 where none = 0; minimal (<10% hepatocytes) = 1; mild (10%–40% hepatocytes) = 2; moderate (40%–70% hepatocytes) = 3; and severe (>70% hepatocytes) = 4. Endothelial detachment was assessed as the number of postsinusoidal venules showing detachment in percent of the total number of vessels analyzed. Sections were further stained for glycogen content using the periodic acid Schiff (PAS) method. Semiquantitative assessment of PAS-positive tissue was performed as follows: 1 = <30% hepatocytes; 2 = 30%–70% hepatocytes; and 3 = >70% hepatocytes.22

Cleaved Caspase-3 Immunohistochemistry.

To study active caspase-3 using immunohistochemistry, 5-μm sections of paraffin-embedded liver specimens were incubated overnight at room temperature with a rabbit polyclonal anticleaved–caspase-3 antibody (1:50, Cell Signaling Technology, Frankfurt, Germany). This antibody detects endogenous levels of the large fragment (17/19 kDa) of activated caspase-3, but not full length caspase-3. A biotinylated anti–mouse/rabbit Ig antibody was used as a secondary antibody for streptavidine-biotin complex peroxidase staining (Link, LSAB-HRP, DakoCytomotion, Hamburg, Germany). 3,3′ diaminobenzidine was used as the chromogen. The sections were counterstained with hemalaun.

Western Blot Analysis.

For whole protein extracts and Western blot analysis of caspase-3, liver tissue was homogenized in lysis buffer (10 mM Tris, pH 7.5, 10 mM NaCl, 0.1 mM ethylenediaminetetraacetic acid, 0.5% Triton-X 100, 0.02% NaN3, 0.2 mM phenylmethyl sulfonyl fluoride), incubated for 30 minutes on ice, and centrifuged for 30 minutes at 16,000g. The supernatant was saved as whole protein fraction. Prior to use, the buffer received a protease inhibitor cocktail (1:100 v/v, Sigma). Protein concentrations were determined using the Lowry assay with bovine serum as the standard.23

Equal amounts of protein per lane (60 μg of whole liver lysate) were separated discontinuously on 12% sodium dodecyl sulfate polyacrylamide gels and transferred to a polyvinyldifluoride membrane (BioRad, Munich, Germany). After blockade of nonspecific binding sites, membranes were incubated for 2 hours at room temperature with a rabbit polyclonal anticleaved–caspase-3 antibody (1:800, Cell Signaling Technology) followed by a secondary peroxidase-conjugated donkey anti–rabbit Ig antibody (1:5,000, Amersham Pharmacia Biotech, Freiburg, Germany). Equal protein loading was proven by Coomassie blue staining of the gels and by Ponceau S staining of the immunoblot membranes.

Protein expression was visualized using luminol-enhanced chemiluminescence and exposure of membrane to blue light–sensitive autoradiography film (Hyperfilm ECL, Amersham Pharmacia Biotech). Signals were assessed densitometrically.

Experimental Protocol.

Livers were subjected to four groups (n = 6 livers each) according to the perfusion solution and the cold storage time. The reversible p53 inhibitor PFT-α (Alexis, Grünberg, Germany) was added to the HTK solution at a concentration of 20 μM15 dissolved in 99.5% dimethyl-sulfoxide (DMSO; Sigma). Control livers were perfused with HTK solution containing an equivalent volume of the vehicle DMSO. Cold preservation of livers was performed for either 24 hours or 48 hours.

In an additional set of experiments, the pan-caspase inhibitor z-VAD-fmk (Alexis)11 was added to the HTK solution at a concentration of 50 μM dissolved in saline. Cold preservation of livers was performed for either 24 hours (n = 3) or 48 hours (n = 4). Because of inactivation of z-VAD-fmk within periods of more than 24 hours, z-VAD-fmk (50 μM) was added a second time at 24 hours in the 48-hour cold-stored livers.

Livers that were harvested as described above but immediately reperfused with an ischemic period of less than 1 hour served as sham-operated controls (n = 5).

Statistical Analysis.

All data are expressed as mean ± SEM. After testing for normal distribution using the Kolmogrov-Smirnov test, differences between the PFT-α and the control groups were assessed using the Student's t test. Pearson product moment correlation was performed to evaluate significant correlations between the parameters studied. Overall statistical significance was set at P < .05.


Isolated Liver Reperfusion.

Over the 2-hour period of reperfusion, portal venous pressure and portal venous resistance progressively decreased from 10–12 mm Hg and 0.44–0.53 mm Hg × min/mL to 6 mm Hg and 0.25 mm Hg × min/mL without significant differences between PFT-α and control groups (Tables 1 and 2). O2 consumption and K+ excretion did not differ between the groups after 24-hour storage (see Table 1). However, after 48-hour storage, O2 consumption was markedly higher in PFT-α–treated livers throughout the 2-hour reperfusion period (see Table 2). In parallel, PFT-α–treated livers showed less K+ excretion upon flushing (0.91 ± 0.04 mmol/L vs. DMSO: 1.45 ± 0.09 mmol/L; P < .05) as well as during the first 5 minutes of reperfusion (see Table 2). Sham-operated controls revealed a portal venous pressure between 4.8 and 3.6 mm Hg, an oxygen consumption constantly above 1 μmol/min × g of liver tissue, and low K+ excretion upon flushing (0.31 ± 0.02 mmol/L).

Table 1. Hemodynamic and Metabolic Parameters During Reperfusion of Livers With KHB Buffer for a Total of 2 Hours
 5 Minutes30 Minutes60 Minutes90 Minutes120 Minutes
  1. NOTE. Livers were preserved for 24 hours with 4°C HTK solution, substituted with either DMSO (control) or PFT-α. Data are given as mean ± SEM.

Portal venous pressure (mm Hg)     
 DMSO11.7 ± 0.67.3 ± 0.46.3 ± 0.36.0 ± 0.45.8 ± 0.5
 PFT-α11.8 ± 0.77.0 ± 0.46.3 ± 0.46.2 ± 0.75.7 ± 0.6
Portal venous resistance (mm Hg × min/mL)     
 DMSO0.52 ± 0.030.33 ± 0.020.28 ± 0.020.27 ± 0.020.26 ± 0.02
 PFT-α0.53 ± 0.030.31 ± 0.020.28 ± 0.020.27 ± 0.030.25 ± 0.03
O2 consumption (μmol/min × g liver tissue)     
 DMSO0.77 ± 0.040.80 ± 0.040.83 ± 0.040.84 ± 0.040.83 ± 0.04
 PFT-α0.78 ± 0.060.78 ± 0.050.82 ± 0.050.86 ± 0.040.85 ± 0.03
K+ efflux (mmol/L)     
 DMSO0.02 ± 0.06−0.05 ± 0.040.18 ± 0.060.16 ± 0.040.07 ± 0.04
 PFT-α−0.01 ± 0.06−0.11 ± 0.040.07 ± 0.030.07 ± 0.070.16 ± 0.04
Table 2. Hemodynamic and Metabolic Parameters During Reperfusion of Livers With KHB Buffer for a Total of 2 Hours
 5 Minutes30 Minutes60 Minutes90 Minutes120 Minutes
  • NOTE. Livers were preserved for 48 hours with 4°C HTK solution, substituted with either DMSO (control) or PFT-α. Data are given as mean ± SEM.

  • *

    P < .05 vs. DMSO.

Portal venous pressure (mm Hg)     
 DMSO10.3 ± 0.68.3 ± 0.97.3 ± 0.66.7 ± 0.76.5 ± 0.8
 PFT-α9.8 ± 0.96.3 ± 0.96.5 ± 0.86.2 ± 0.96.0 ± 0.9
Portal venous resistance (mm Hg × min/mL)     
 DMSO0.46 ± 0.030.37 ± 0.040.33 ± 0.030.30 ± 0.030.29 ± 0.03
 PFT-α0.44 ± 0.040.28 ± 0.040.29 ± 0.030.27 ± 0.040.27 ± 0.04
O2 consumption (μmol/min × g liver tissue)     
 DMSO0.74 ± 0.110.65 ± 0.110.71 ± 0.080.72 ± 0.080.72 ± 0.08
 PFT-α0.94 ± 0.080.94 ± 0.060.96 ± 0.06*0.97 ± 0.06*0.96 ± 0.08*
K+ efflux (mmol/L)     
 DMSO0.14 ± 0.05−0.02 ± 0.050.16 ± 0.050.17 ± 0.040.10 ± 0.03
 PFT-α−0.02 ± 0.04*−0.10 ± 0.050.14 ± 0.040.12 ± 0.030.12 ± 0.03

Fluorescence Microscopy.

Cold storage of DMSO-treated livers for 24 hours showed 7% apoptotic cells. Subsequent reperfusion caused a progressive increase of the number of apoptotic cells with 17% at 120 minutes of reperfusion (Fig. 1A). Storage for 48 hours resulted in 10% apoptotic cell death, while additional 2-hour reperfusion induced a dramatic increase to almost 60% (Figs. 1B and 2). PFT-α treatment afforded a significant reduction of apoptotic cell injury after both preservation and reperfusion. At 120 minutes of reperfusion, the number of apoptotic cells decreased by 54% and 67% after 24- and 48-hour storage, respectively (see Fig. 1). Sham-operated control livers revealed less than 2% of apoptotic hepatocytes throughout the 2-hour reperfusion.

Figure 1.

Apoptotic cell death, as determined by fluorescence microscopy, in cold-stored livers during reperfusion with Krebs-Henseleit bicarbonate buffer for a total of 2 hours. Livers were preserved for (A) 24 hours or (B) 48 hours with 4°C HTK solution, substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. *P < .05 vs. DMSO.

Figure 2.

Fluorescence microscopic images of liver tissue (A) at the end of 48 hours of cold storage and (B) after an additional 2 hours of reperfusion, displaying bisbenzimide-stained hepatocytes with condensation (arrow) as well as fragmentation and margination of nuclear chromatin (arrowheads). Note the marked increased of hepatocytes exhibiting apoptotic signs upon 2 hours of reperfusion (ultraviolet epi-illumination; original magnification ×800).

Storage of livers with the pan-caspase inhibitor z-VAD-fmk markedly reduced apoptotic cell death. After 24-hour storage, numbers of apoptotic hepatocytes as assessed by fluorescence microscopy (0 hours, 2.8 ± 0.7%; 1 hour, 7.8 ± 2.9%; 2 hours, 9.2 ± 1.0%) were comparable to those in livers preserved with PFT-α–substituted HTK solution; after 48-hour storage, z-VAD-fmk proved to be slightly less effective in reducing apoptotic cell death (5.0 ± 1.1%, 30.4 ± 5.4%, 29.5 ± 6.5%) when compared with PFT-α.

Bile Flow and Liver Enzyme Activities.

Upon reperfusion after 24-hour storage, bile flow in DMSO-treated livers ranged between 0.14 and 0.18 μL/min × g liver tissue. Hepatocellular excretory function was found better maintained by PFT-α preservation, as indicated by a 1.5- to 2-fold higher bile flow (Fig. 3A). After 48-hour storage, bile flow was markedly lower (<0.08 μL/min × g liver tissue) compared to that after 24-hour storage and did not differ significantly between the two groups (Fig. 3B). Throughout reperfusion, average bile flow of sham-operated livers was 0.55 ± 0.04 μL/min × g liver tissue.

Figure 3.

Bile flow in cold-stored livers during reperfusion with KHB buffer for a total of 2 hours. Livers were preserved for (A) 24 hours or (B) 48 hours with 4°C HTK solution, substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. *P < .05 vs. DMSO.

Analysis of AST and ALT activities in the flushing solution revealed higher levels after 48 hours than after 24 hours (Fig. 4). PFT-α caused a significant reduction of enzyme activities after both 24- and 48-hour cold preservation (see Fig. 4). After 2-hour reperfusion following either 24- or 48-hour storage, no significant differences between PFT-α– and DMSO-preserved livers could be observed (data not shown). In sham-operated control livers, activities of AST and ALT in the perfusate remained below detection limit.

Figure 4.

Liver enzymes (A) AST and (B) ALT in flushing effluent after 24- and 48-hour cold preservation in HTK solution substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. *P < .05 vs. DMSO. Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase.

Liver Histology and Immunohistochemistry.

To confirm apoptotic cell death after liver cold storage and reperfusion, active caspase-3 was studied using immunohistochemistry. These experiments showed positive staining of a considerable number of individual hepatocytes. Interestingly, cells with positive active caspase-3 staining did not necessarily show cell shrinkage but typically showed nuclear condensation, fragmentation, and margination, as well as plasma membrane blebbing (Fig. 5).

Figure 5.

Representative immunohistochemistry of active caspase-3 staining in single hepatocytes of liver specimens, which underwent 24-hour cold storage and 2-hour reperfusion. The cleaved caspase-3 rabbit monoclonal antibody detects endogenous levels of the large fragment (17/19 kDa) of activated caspase-3, but not full length caspase-3. (A) Individual contiguous cells were stained rather than groups of cells. Interestingly, cells with positive active caspase-3-staining did not necessarily show nuclear condensation and cell shrinkage, but frequently showed (B, E) typical nuclear fragmentation, (C, F) nuclear margination, and (D, G) plasma membrane blebbing without cell shrinkage. (Original magnification ×500.)

Quantitative analysis of hematoxylin-eosin–stained tissue sections confirmed the PFT-α–mediated protection against preservation-reperfusion injury (Fig. 6). After both 24- and 48-hour preservation, apoptotic cell death was found significantly reduced in PFT-α–treated livers when compared with DMSO-treated controls. This difference was detected directly after cold storage, but in particular after the 2-hour reperfusion period (see Fig. 6).

Figure 6.

Apoptotic cell death, as determined by analysis of hematoxylin-eosin–stained tissue specimens, in cold-stored livers after reperfusion with KHB buffer for a total of 2 hours. Livers were preserved for (A) 24 hours or (B) 48 hours with 4°C HTK solution, substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. *P < .05 vs. DMSO.

Although cytoplasmic vacuolation of hepatocytes was minimal after storage, it became evident after reperfusion with preferential localization in the pericentral segment of hepatic lobules. In DMSO-treated livers, score values ranged between 3.4 (24-hour storage) and 4.0 (48-hour storage), while vacuolation was less pronounced in PFT-α–treated livers with a score of approximately 3.3 after 48-hour storage (Table 3; Fig. 7). PFT-α conferred protection of endothelial cells as well as hepatocytes as indicated by the amelioration of endothelial cell detachment (see Table 3).

Table 3. The Effects of PFT-α on Liver Histology After Preservation for Either 24 Hours or 48 Hours and Reperfusion for a Total of 2 Hours With KHB Buffer
 Cytoplasmic VacuolationEndothelial Cell Detachment (%)Trypan Blue Uptake (%)
0 h2 h0 h2 h2 h
  • NOTE. For detailed information of quantitative assessment of liver histology, see Materials and Methods. Data are given as mean ± SEM.

  • *

    P < .05 vs. DMSO.

24 h     
 DMSO0.22 ± 0.093.35 ± 0.1917.6 ± 5.524.5 ± 6.328.1 ± 1.9
 PFT-α0.14 ± 0.092.59 ± 0.19*6.1 ± 3.38.5 ± 1.4*4.8 ± 1.6*
48 h     
 DMSO0.59 ± 0.273.98 ± 0.0225.1 ± 4.562.2 ± 5.261.1 ± 3.8
 PFT-α0.03 ± 0.033.27 ± 0.13*12.5 ± 2.339.5 ± 2.2*23.6 ± 8.1*
Figure 7.

Hematoxylin-eosin–stained tissue sections of cold-stored livers after reperfusion with KHB buffer for a total of 2 hours. Livers were preserved for 24 hours with 4°C HTK solution, substituted with either (A) DMSO or (B) PFT-α. Note the markedly better preserved liver morphology in the PFT-α–treated liver. (Original magnification × 200.)

PAS scores were reduced at 2-hour reperfusion compared with corresponding values at 24-hour storage, denoting use of glycogen as a substrate for energy production upon reperfusion. This reduction was more pronounced in PFT-α–treated livers (Fig. 8A). After 48-hour storage and reperfusion, however, PAS scores did not significantly change in either of the groups, implying hampered ability of liver tissue to use glycogen (Fig. 8B).

Figure 8.

Intensity of hepatocellular PAS staining, as assessed by a semiquantitative scoring index, in cold-stored livers during reperfusion with KHB buffer for a total of 2 hours. Livers were preserved for (A) 24 hours or (B) 48 hours with 4°C HTK solution, substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. Abbreviation: PAS, periodic acid Schiff.

Trypan Blue Uptake.

After cold storage and reperfusion, trypan blue positive cells were mainly located in the midzonal and periportal segments of hepatic lobules. With prolongation of cold storage from 24- to 48-hours, trypan blue–positive areas increased from 28% to 61% in DMSO-treated livers, while PFT-α–treated livers exhibited significantly smaller areas of injury (5% and 24%; see Table 3).

Caspase-3 Protein Levels.

Marked activation of caspase-3 was shown by Western blot analysis of cleaved products of caspase-3 in whole liver lysates after 24-hour storage and 2-hour reperfusion (14 ± 3 optical density × mm2), while caspase activation was lower in lysates sampled directly after the 24-hour storage period (9 ± 2 optical density × mm2). PFT-α was effective to significantly inhibit cold ischemia-induced caspase activation as indicated by a 55% and 33% reduction of cleaved caspase-3 in 24- and 48-hour stored organs (Fig. 9). After 2-hour reperfusion, cleaved products of caspase-3 were only slightly reduced by PFT-α (see Fig. 9).

Figure 9.

Densitometric analysis of cleaved caspase-3 in cold-stored livers after reperfusion with KHB buffer for a total of 2 hours as assessed by Western blot analysis. Densitometric units of cleaved caspase-3 products in the DMSO group were set as 100%. Livers were preserved for (A) 24 hours or (B) 48 hours with 4°C HTK solution, substituted with either DMSO (open bars) or PFT-α (filled bars). Data are given as mean ± SEM. *P < .05 vs. DMSO. The upper panel shows a representative blot of cleaved caspase-3 at 0 hours and 2 hours of reperfusion in livers, which were preserved for 24 hours (left) and 48 hours (right) with 4°C HTK solution, substituted with either DMSO or PFT-α. Caspase cleavage was not detectable in the sham-operated control liver. β-Actin was used to verify equal loading of lanes. Abbreviations: DMSO, dimethyl sulfoxide; PFT-α, pifithrin-α.

Correlation Analysis.

Release of both K+ and AST in the flushing solution, indicating parenchymal cell disintegration, correlated positively with the parameters of subsequent hepatocellular injury at 2-hour reperfusion as determined by apoptosis, cellular vacuolation, endothelial cell detachment, and trypan blue uptake (Table 4). Noteably, both methods of detecting apoptosis (histomorphology and fluorescence microscopy) correlated well with each other (r = 0.81, P < .0001), but also with trypan blue uptake (r = 0.86, P < .0001; r = 0.72, P < .005).

Table 4. Correlation Between Parameters of Liver Morphology at 2 Hours of Reperfusion With K+ and AST Efflux in Flushing Solution
 Regression Coefficient*
  • *

    Pearson product moment correlation.

Apoptosis (bisbenzimide)0.55 (P < .01)0.87 (P < .0001)
Apoptosis (hematoxylin-eosin)0.83 (P < .0001)0.50 (P = .057)
Cytoplasmic vacuolation0.48 (P < .05)0.50 (P < .05)
Endothelial cell detachment0.46 (P < .05)0.71 (P < .005)
Trypan blue uptake0.77 (P < .0005)0.52 (P < .05)


The p53 protein plays a key role in the control of the cell response to various kinds of stress, with the activation of p53 resulting in the arrest of cell proliferation and/or apoptosis.13 PFT-α, a small molecule that reversibly blocks p53-dependent transcriptional activation and apoptosis, has been shown to protect mice from side effects of cancer therapy15 and neurons from death induced by ischemic and excitotoxic insults.24 Herein we report that PFT-α reduces procaspase-3 activation and both hepatocellular apoptosis and necrosis after cold storage and reperfusion, ensuing better liver function, metabolism, and tissue integrity.

Hepatic Cold Preservation Injury.

Although their compositions markedly differ, both University of Wisconsin and HTK solution are used for cold preservation of liver allografts. In the present study, liver preservation was accomplished with HTK solution. Although originally developed as a cardioplegic solution, the use of HTK solution has been extended and covers other organs in addition to the liver.25 HTK solution is thought to be equally as appropriate as University of Wisconsin solution for liver transplantation, even if cold ischemia extends to 15 hours.25 HTK solution imparts certain advantages, in particular the ability of histidine to enter the cells, affording effective intracellular buffering and preventing detrimental pH fall. Hydrogen ion and lactate accumulation from ATP breakdown and anaerobic glycolysis are known to suppress glycolytic enzymes, preventing ATP regeneration.26 Due to continuous removal of acidic products, HTK solution promotes anaerobic glycolysis and ATP preservation.25

Our aim was to investigate the effect of transient inhibition of p53 in reperfusion injury of cold ischemic livers. Ideally, such a study should be performed in a transplantation model, but too many confounding factors are present to conclusively identify specific mechanisms. Therefore, we used an ex vivo isolated liver perfusion model with a blood cell–free perfusate. Being aware that leukocytes and platelets play an enormous role in mediating reperfusion injury,27 it was our intent to simplify the model to exclusively study the individual p53-dependent pathway of hepatocellular apoptosis.

Flushing of livers before reperfusion allowed to collect effluent fluid for determination of K+ levels and transaminase activities. In parallel with other studies,28, 29 these measures portrayed a valuable tool of predicting organ damage, as indicated by significant correlations with morphological characteristics of final tissue injury. AST levels were approximately eightfold higher than corresponding values of ALT, indicating mitochondrial injury as a probable incentive for postischemic liver dysfunction.

The analysis of apoptotic cell injury in our model is primarily based on fluorescence microscopic assessment of nuclear morphology. In a previous study, we could demonstrate that condensation and fragmentation of nuclear chromatin, as visualized by fluorescence microscopy, indeed indicates apoptosis, corresponding with established criteria of apoptosis as assessed by scanning and transmission electron microscopy.30 The fact that z-VAD-fmk was capable of reducing the number of cells that we identified as apoptotic cells—because of their characteristic changes in nuclear chromatin morphology—underlines our finding that apoptosis is one mode of cell death after cold storage and reperfusion.

In contrast to most confirmed apoptosis models in which z-VAD-fmk completely abolished apoptotic cell death, the inhibitor was only partially effective in our preservation-reperfusion model. This may be due to the fact that z-VAD-fmk was given in the preservation solution during 4°C ischemia. Although there is no information on z-VAD-fmk action in 4°C liver storage, experiments analyzing apoptosis in cryopreserved hepatocytes using z-VAD-fmk in the cryopreservation solution also demonstrated an only partial reduction of apoptosis as indicated by a 30% diminution of caspase-3-like protease activity.31

There is an ongoing discussion not only on the predominant mode of cell death (i.e., apoptosis vs. necrosis), but also on the extent of apoptosis in postischemic hepatic reperfusion. In contrast to a recent study demonstrating quantitatively irrelevant numbers of apoptotic cells after partial no-flow warm ischemia reperfusion,32 we herein show 7% apoptotic cells after 24-hour storage and 17% after an additional 2 hours of reperfusion. This supports the view that (1) apoptosis occurs already during cold preservation, as also shown by Rauen et al.33 in in vitro systems, and (2) reperfusion represents its own pathogenic entity, enhancing preservation-induced damage either by aggravating or by unmasking the injury implicated during cold ischemia. In the present model, reperfusion of cold-stored livers was accomplished through machine perfusion forcing oxygenated perfusate into the liver, which to some extent differs from the in vivo situation in which no reflow, vasoconstriction, and plugging of sinusoids might have more impact on cellular ATP depletion, thus favoring necrotic instead of apoptotic cell death.

The amount of apoptotic cell death is nicely mirrored by the procaspase-3 activation, being 1.5- to 2-fold higher at the end vs. the beginning of reperfusion. Moreover, there is a marked increase of caspase-3 processing compared with sham-operated controls. The impact of caspase-mediated apoptosis has been emphasized in studies in which liver injury after ischemia reperfusion was prevented by application of caspase inhibitors,8, 34, 35 similarly as in the present study. In parallel to apoptotic cell death, cell membrane damage, as assessed by trypan blue uptake, occurred in dependency to storage time, ranging between 20% and 60%. Thus, in contrast to warm hepatic ischemia reperfusion injury, which is thought to preferentially occur through oncotic necrosis,32 both modes of cell death seem to substantially contribute to liver damage upon cold ischemia reperfusion. This is in line with the view of others, indicating that by inducing mitochondrial permeability transition, ischemia reperfusion causes both apoptosis and necrosis.36 In fact, apoptosis and necrosis may not be unrelated as initially thought, but rather may share common events, resulting in “necrapoptosis” or “aponecrosis.”37, 38 In necrapoptosis, mitochondrial permeability transition initiates a chain reaction that culminates in either apoptosis or necrosis, possibly depending on ATP supply.36 Rapid and complete cellular ATP depletion may direct cells toward necrosis, while apoptotic signaling may proceed if ATP depletion is delayed (cold ischemia) or restored (reperfusion). According to the definition of Lemasters,37 the mixture of apoptotic and necrotic cell death after cold preservation and reperfusion may represent a typical necrapoptotic response.

In addition, it should be kept in mind that cell membrane damage is characteristic not only of primary necrotic cell death, but also of secondary apoptosis. In this case, a considerable number of trypan blue–positive cells may have undergone apoptosis, which is supported by our result that apoptotic cell death, as determined by fluorescence microscopy and hematoxylin-eosin histomorphology, significantly correlated with trypan blue uptake. The observation that the majority of trypan blue uptake was located periportally might be due to the fact that ischemia reperfusion–induced sinusoidal perfusion failure is known to occur preferentially in periportal and midzonal segments,39 resulting in rapid ATP depletion and thus more pronounced cell death.

PFT-α and Hepatic Cold Preservation Injury.

In the present study, the 20-μM dose of PFT-α was chosen because it has demonstrated selective inhibition of p53 transcriptional activity and thus prevention of DNA damage–induced apoptosis.15 We further supplemented the preservation solution with PFT-α because this avoids special pretreatment of organ donors and can easily be done by the transplant surgeon, thus representing an attractive tool in clinical practice. A variety of in vivo studies have shown that PFT-α mediates antiapoptotic properties,24, 40–42 disproving the concern that the herein observed effect of PFT-α is specific for a buffer-perfused isolated tissue.

Being aware that the mechanisms of hepatic cold ischemia reperfusion injury are multifactorial,43 characteristic triggers (e.g., ATP depletion, hypoxia, acidosis, reactive oxygen species, and cytokines) are known to cause p53 activation with execution of p53-dependent cell apoptosis.44 Because apoptotic cell death may precede necrosis6 and aggravate the inflammatory response,45 targeting the apoptotic pathway using PFT-α in the preservation solution may constitute a valid strategy against cold storage–induced organ injury. This view may also explain why PFT-α–treated livers exhibited reductions in both apoptotic and necrotic cell death, finally resulting in improved hepatocellular metabolism and excretory function.

Others have shown that PFT-α does not alter phosphorylation or sequence-specific DNA binding of p53, but slightly lowers the levels of nuclear and not cytoplasmic p53.15 In line with this, we could recently confirm that PFT-α affects the nuclear/cytoplasmic ratio, thereby promoting antiapoptotic signals with proliferation and enhancement of wound healing.41 Assuming that this is not the only mechanism of PFT-α action, reduced activation of procaspase-3 in PFT-α–treated livers implies that antiapoptotic properties of PFT-α include the downstream caspase cascade. However, the connection between p53 and the caspase cascade is only beginning to be understood44 and is beyond the scope of this study.

In conclusion, we show that reduction of hepatocellular apoptosis and necrosis by targeting p53 using PFT-α causes a favorable effect on overall graft quality, as indicated by lower enzyme and K+ release as well as higher O2 consumption and bile flow upon reperfusion. Thus p53-targeting agents such as PFT-α may serve as a novel therapeutic adjuvant to improve liver preservation in hepatic transplantation.