Donor graft adenoviral iNOS gene transfer ameliorates rat liver transplant preservation injury and improves survival

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

The exact role of inducible NOS (iNOS) in liver ischemia/reperfusion (I/R) injury is controversial. This study was designed to investigate whether donor liver pretreatment with adenovirus encoding iNOS (AdiNOS) ameliorates I/R injury associated with liver transplantation. Orthotopic syngeneic LEW rat liver transplantation (OLT) was performed after 18 or 24 hours' preservation in cold UW. AdiNOS or control gene vector (AdLacZ) was delivered to the liver by donor intravenous pretreatment 4 days before graft harvesting. Uninfected grafts also served as control. Recipients were sacrificed 1 to 48 hours posttransplantation. An abundant hepatic iNOS protein expression and marked serum NO elevation was observed in the AdiNOS-treated group, without affecting endothelial nitric oxide synthase (eNOS) expression, before harvesting and after OLT. AdiNOS pretreatment markedly improved liver function assessed by serum aspartate aminotransferase/alanine aminotransferase levels and reduced liver necrosis formation. AdiNOS treatment also was associated with reduced ICAM-1 mRNA expression and neutrophil accumulation in the liver graft after OLT compared with untransfected or AdLacZ-treated group. Furthermore, AdiNOS delivery significantly improved transplant survival, compared with AdLacZ or saline controls. AdiNOS pretreatment did not attenuate I/R-induced apoptotic cell death in the liver graft. Administration of a selective inhibitor for iNOS abrogated the protection afforded by AdiNOS pretreatment. In conclusion, donor pretreatment with AdiNOS led to improved liver graft injury and posttransplantation survival. Downregulation of ICAM-1 mRNA and neutrophil infiltration may be associated with the mechanisms by which AdiNOS pretreatment confer the protection against transplant-associated hepatic I/R injury. (HEPATOLOGY 2006;43:464–473.)

Liver ischemia/reperfusion (I/R) injury is observed in a variety of clinical situations, such as hemorrhagic shock, septic shock, cardiac arrest, and liver surgery including tumor resection and transplantation. Severe I/R injury has the potential to contribute to significant liver failure and mortality. Although many experimental studies have suggested protective agents against liver I/R injury, very few have yet reached clinical practice.1 Therefore, identification of specific therapeutic targets may lead to strategies to combat I/R injury.

During warm and cold liver I/R, expression of inducible nitric oxide synthase (iNOS) is upregulated in reperfused liver, which results in relatively higher levels of NO production than that constitutively produced by endothelial NOS (eNOS) under normal conditions. eNOS-derived NO, which is necessary to maintain microvascular blood flow, is generally accepted to have a beneficial role in liver I/R injury.2–4 However, the role of iNOS in liver I/R injury is much more controversial.5, 6 We have recently shown that blockade of the L-arginine/NOS pathway by the non-selective NOS inhibitor N-nitro-L-arginine methyl ester, which inhibits both eNOS and iNOS, worsened liver preservation injury.7 Others have reported that a selective iNOS inhibitor increased liver injury during hepatic warm I/R.8, 9 Moreover, enhanced liver damage was reported in the warm I/R model of iNOS-deficient mice.10, 11 In contrast, other studies reported no effect or even a protective effect of iNOS inhibitors on warm hepatic I/R injury12, 13 and also demonstrated improvement of liver function after warm I/R in iNOS-deficient mice.4 Therefore, the exact role of endogenous iNOS expression during liver I/R remains to be defined.

Similar to iNOS, heme oxygenase (HO)-1 is also induced after hepatic I/R and is considered to be a cytoprotective response.14, 15 Upregulation of endogenous HO-1 system by HO-1 inducer or gene therapy successfully protects liver against warm and cold I/R injury.16, 17In vitro, NO behaves as an potent HO-1 inducer,18–20 and iNOS-derived NO has been shown to induce HO-1 expression in mouse hepatocytes.21

To provide local NO synthesis during transplant preservation injury, we transduced donor liver grafts with adenovirus encoding iNOS (AdiNOS) and observed a beneficial effect of AdiNOS pretreatment on hepatic cold I/R injury. Transplant survival was also significantly improved when the AdiNOS was delivered in a model of severe I/R injury. Furthermore, we found that decreased ICAM-1 expression and neutrophil infiltration was involved in the mechanism of AdiNOS pretreatment-mediated protection.

Abbreviations

I/R, ischemia/reperfusion; iNOS, inducible nitric oxide synthase; eNOS, endothelial nitric oxide synthase; HO-1, heme oxygenase-1; AdiNOS, adenovirus encoding inducible NOS; pfu, plaque forming unit; AST, aspartate aminotransferase; ALT, alanine aminotransferase; HE, hematoxylin-eosin; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling; LDU, Laser Doppler Unit; PARP, poly ADP-ribose polymerase.

Materials and Methods

Preparation of Adenoviral Vectors.

An E1- and E3-deleted adenoviral vector carrying the human iNOS cDNA was constructed as previously described.22, 23 Concentrations of AdiNOS and the control adenovirus AdLacZ were determined by plaque-forming assay, and expressed as plaque-forming units (pfu). The titers of AdiNOS and AdLacZ were 1010 pfu/mL, and both vectors were diluted to 2 × 109 pfu/mL with saline for intravenous injection to the donor rat.

Orthotopic Liver Transplantation (OLT).

Male Lewis (LEW, RT1l) rats weighing 200 to 300 g (Harlan Sprague Dawley, Inc., Indianapolis, IN) were maintained in a laminar-flow, specific-pathogen–free atmosphere at the University of Pittsburgh. All of the initial transplants were performed by a single surgeon who has been performing rat liver transplantation for more than 5 years (T.K.). The anhepatic and warm ischemia times were well controlled for with warm ischemic time of 14.3 ± 0.8 minutes. Basic techniques of liver harvesting and orthotopic transplantation without hepatic arterial reconstruction were according to the method previously described by Kamada and Calne.24

Experimental Design.

Donor liver was transduced in vivo with AdiNOS or AdLacZ and then the transduced grafts were transplanted into syngeneic rat recipients. Endpoints examined were iNOS expression, liver enzyme release, histology, ICAM-1 mRNA expression and neutrophil infiltration, HO-1 expression, apoptosis, and animal survival. In the first series of experiments, AdiNOS, AdLacZ (2 × 109 pfu/mL) or 1 mL 0.9% saline was injected via the penile vein to prospective donor animals under brief isoflurane anesthesia. Four days later, the donor liver was harvested. All liver grafts were kept in a bath of UW solution at 4°C for 18 or 24 hours of cold preservation period and orthotopically transplanted into syngeneic LEW recipients. Rats were killed at 1, 3, 6, 24, and 48 hours after transplantation for serum and liver grafts sample (n = 4-6 for each time point). In the second series of experiment, we examined the effects of iNOS or HO-1 inhibitors on AdiNOS-mediated protection against preservation injury. After donor pretreatment with AdiNOS, the selective iNOS inhibitor, L-N6-(1-imino-ethyl)lysine (L-NIL; 30 mg/kg, Alexis Biochem., San Diego, CA), was given to the recipients intravenously at 0 and 3 hours after reperfusion.7 The HO-1 inhibitor, tin-protoporphyrin IX (SnPP; 10 μmol/L/kg, Frontier Scientific, Logan, UT), was given to AdiNOS-treated donor subcutaneously at 24 hours before graft harvest and to the recipient intravenously just after reperfusion. As a negative control, the same volume of saline (0.25 mL) was injected to the AdiNOS-treated donor and recipient at the same time points as each inhibitor. The preparation and doses of L-NIL and SnPP were determined according to previous reports.7, 25–27

All procedures in this experiment were performed according to the guidelines of the Council on Animal Care at the University of Pittsburgh and the National Research Council's Guide for the Care and Use of Laboratory Animals.

Liver Function Tests.

Hepatic function and injury after rat liver transplantation was assessed by serum aspartate aminotransferase (AST) and serum alanine aminotransferase (ALT) levels using the Opera Clinical Chemistry System (Bayer Co., Tary Town, NY).

Histopathology.

Liver graft tissues were fixed in 10% formalin, embedded in paraffin, sectioned into 6-μm thickness, and stained with hematoxylin-eosin (HE). The percent necrotic area was quantified by random evaluation of 10 low-power fields (×40) of each HE-stained section using NIH image analysis software by an independent researcher in a blind manner (T.K.). The neutrophils were stained using a naphthol AS-D chloroacetate esterase staining kit (Sigma Diagnostics, St. Louis, MO). Neutrophils, identified by nuclear morphology and bright red positive staining, were counted in 20 high-power fields (×200) per each section and expressed as the number of cells per field of 1 mm2 by an independent researcher in a blind manner (T.K.). Apoptotic cells were detected by HE stain and were evaluated by a liver pathologist (M.A.N.) for features such as cellular contraction with hyperchromatic or fragmented nuclei.

Terminal Deoxynucleotidyl Transferase-Mediated Deoxyuridine Triphosphate Nick-End Labeling Assay.

Apoptosis was also determined using the in situ end-labeling technique. Formalin-fixed paraffin-embedded sections (6 μm) of liver were investigated using the ApopTag Peroxidase Kit (Intergen Co., Purchase, NY). The peroxidase activity was visualized with AEC substrate, yielding a brown-red oxidation product. Hematoxylin was used as counterstain. The number of terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL)-positive cells in the liver were counted and expressed as the number of cells per field of 1 mm2 by an independent researcher in a blind manner (T.K.).

Blood Flow.

Graft tissue blood flow was measured using a laser Doppler flowmeter (BLF 21D, Transonic Systems, Ithaca, NY) at 1, 3, 6, and 24 hours after reperfusion. Probe was placed on the right, middle, and left lobes of the liver graft, and 3 separate measurements were obtained in each location by one of the authors (A.I.) without knowledge of the experimental groups. Data were obtained by Laser Doppler Unit (LDU) and expressed the percent increase of naïve unoperated animals. Blood flow of the recipient abdominal aorta and portal vein were also measured.28

Real-Time RT-PCR.

mRNA expression for ICAM-1 and GAPDH was quantified by SYBR Green real-time reverse transcription polymerase chain reaction (RT-PCR) by using an ABI PRISM 7000 Sequence Detection System (PE Applied Biosystems) as previously described.28 The expression of each gene was normalized to GAPDH mRNA content and calculated relative to normal control.

Western Blot Analysis.

Western blotting assay was performed using cytoplasmic proteins (20-50 μg) as previously described.29 Nuclear protein (20 μg) was used for cleaved poly adenosine diphosphate (ADP)-ribose polymerase (PARP) detection. Membranes were incubated with primary rabbit polyclonal antibody for iNOS, eNOS (Transduction Laboratories, Lexington, KY), HO-1 (StressGen Biotech, Victoria, BC, Canada), cleaved caspase-3, cleaved PARP (Cell Signaling Technology, Beverly, MA), or actin (Sigma-Aldrich, St. Louis, MO). After incubation with secondary goat anti-rabbit antibody (Pierce Chemical, Rockford, IL), membranes were developed with the SuperSignal detection systems (Pierce Chemical) and exposed to film. The band densities were measured by NIH Image analysis software and presented in comparison to relative actin band.

Immunohistochemistry for iNOS.

Formalin-fixed paraffin-embedded sections (6 μm) of liver were deparaffinized and incubated with anti-iNOS antibody (Transduction Laboratories) and subsequently incubated with biotinylated goat anti-rabbit immuoglobrins (DAKO A/S., Denmark). After the incubation with a pre-formed avidin and biotinylated horseradish peroxidase complex reagent, the immune complex was visualized with AEC substrate. The sections were counterstained with hematoxylin.

Serum NO Measurement.

Serum NO was determined by measuring serum nitrite and nitrate levels by using a commercially available kit (Cayman Chemical, Ann Arbor, MI).

Statistical Analysis.

Data are represented as the mean ± SEM. Comparisons between the groups at different time points were performed by using the Student t test or analysis of variance (ANOVA) using the Statview program (Abacus Concepts, Inc., Berkeley, CA). Graft survivals were plotted with the Kaplan-Meier method, and the differences among groups were analyzed using the log-rank test. Differences were considered significant at a P value less than .05.

Results

AdiNOS Pretreatment Leads to iNOS Protein Expression in Donor Liver.

Hepatic iNOS protein expression was initially observed at 3 days and was prominent by 4 days after AdiNOS (2 × 109 pfu) injection (Fig. 1A). The dose of 2 × 109 or 3 × 109pfu, but not 1 × 109 pfu, mediated strong iNOS protein expression in the donor liver 4 days after intravenous injection (Fig. 1A). Therefore, we selected the dose of 2 × 109 pfu and the harvest time of 4 days in the subsequent liver transplant experiments. The iNOS protein staining was predominantly localized to hepatocytes in AdiNOS (2 × 109 pfu)-treated donor livers 4 days after intravenous injection, and there was no detectable iNOS staining in saline-treated donor livers (Fig. 1B).

Figure 1.

(A) Time- and dose-dependent iNOS protein expression in donor liver after AdiNOS injection. Western blot analysis for iNOS was performed. iNOS protein expression was initially observed at 3 days, and was prominent at 4 days after AdiNOS (2 × 109 pfu) intravenous injection. Different doses of AdiNOS (1-3 × 109 pfu) or 1 mL saline were injected to the donor rat via penile vein at 4 days before graft harvest. The dose of 2 × or 3 × 109 pfu, but not 1 × 109 pfu, mediated strong iNOS protein expression in the donor liver at 4 days after iv injection. Blot shown is representative of 3 similar experiments. (B) Immunohistochemistry for iNOS. iNOS staining (brown-red) was localized in hepatocytes in AdiNOS-treated donor liver at 4 days after intravenous injection. There was no positive staining of iNOS in saline-treated donor liver. iNOS, inducible nitric oxide synthase.

Early iNOS Protein Expression Posttransplant Is Attributable to AdiNOS Transgene Expression and Results in NO Synthesis.

Marked expression of iNOS protein was detected in AdiNOS-treated, but not saline- or AdLacZ-treated, donor livers 4 days after injection (before graft storage). At 1 hour after reperfusion, iNOS protein was seen only in AdiNOS-transduced grafts, indicating that iNOS expression was attributable to AdiNOS gene transfer and not to induction of endogenous iNOS mediated by adenovirus or I/R injury. By 3 hours posttransplantation, faint iNOS expression was seen in the saline- and AdLacZ-treated animals, indicating induction of endogenous iNOS mediated by I/R injury (Fig. 2A). AdiNOS-mediated iNOS protein expression was strong at 3 hours after reperfusion and persisted up to 48 hours in the AdiNOS-treated animals (Fig. 2B). Although endogenous iNOS protein expression increased slightly at 3 hours after reperfusion in the AdLacZ group, AdiNOS-mediated iNOS expression was not seen at 48 hours after reperfusion in AdLacZ-treated group (Fig. 2B). Consistent with the iNOS expression, donor animals at 4 days after AdiNOS injection and recipient animals receiving AdiNOS-transduced liver grafts exhibited significant elevations of serum NO levels (nitrite + nitrate) 1 to 48 hours posttransplantation compared with saline or AdLacZ controls (Fig. 2C).

Figure 2.

(A) Comparison of hepatic iNOS and eNOS protein expressions between saline-, AdLacZ- and AdiNOS-treated animals. Representative images of Western blot for iNOS or eNOS from 3 different experiments (upper panel), and band density expressed in comparison with actin (lower panel) were shown. A marked expression of iNOS protein was detected in AdiNOS (2 × 109 pfu)-treated, but not saline- or AdLacZ (2 × 109 pfu)-treated donor liver before graft storage, and 1 or 3 hours after reperfusion. eNOS protein expression was significantly decreased at 1 or 3 hours after reperfusion in all groups as compared with fresh normal liver. There was no significant difference in eNOS protein level at 4 days after injection (before storage) and 1 or 3 hours after reperfusion between each group. Each lane represents a different animal. *P < .01 vs. saline- or AdLacZ-treated group; #P < .05, P < 0.01 vs. fresh normal liver (n = 4 for each group). (B) Time course of iNOS protein expression. Representative image of Western blot for iNOS expression from 3 different experiments is shown (n = 3). AdiNOS-mediated iNOS protein expression was persistent 3 to 48 hours after reperfusion, but not in the AdLacZ-treated group. Saline-treated animals showed same pattern of iNOS expression as AdLacZ-treated animals (data not shown). (C) Serum nitrite + nitrate levels. A marked elevation of serum NO2 and NO3 levels was observed in AdiNOS-treated group at 4 days after intravenous injection (donor) and 1, 3, or 48 hours after reperfusion (recipient). *P < .01 versus saline- or AdLacZ-treated group (n = 4 for each group). iNOS, inducible nitric oxide synthase; eNOS, endothelial nitric oxide synthase.

AdiNOS Pretreatment Does Not Affect eNOS Protein Expression in Liver Graft.

To examine the influence of AdiNOS pretreatment on endogenous eNOS protein expression in liver graft, we performed Western blotting assay for eNOS. There was no significant difference in eNOS protein levels 4 days after injection (before storage) and during early posttransplantation (1-3 hours after reperfusion) between each group, indicating that marked elevation of serum NO in the AdiNOS-treated group was mainly caused by AdiNOS-mediated iNOS protein expression (Fig. 2A). However, hepatic eNOS protein levels were significantly decreased at 1 and 3 hours after reperfusion in each group compared with fresh normal liver or donor liver before cold storage (Fig. 2A). The eNOS protein expression recovered to normal levels by 48 hours after reperfusion (data not shown). These data indicate diminished eNOS protein levels as a result of the I/R injury, and AdiNOS treatment did not directly influence eNOS expression.

AdiNOS Pretreatment Attenuates Hepatic I/R Injury.

Hepatic transplant preservation injury after 18 hours cold storage, assessed by elevation of serum AST and ALT, was markedly attenuated in the AdiNOS-treated animals 6 to 48 hours after reperfusion compared with saline- or AdLacZ-treated groups (Fig. 3A). Consistent with the improvement in liver function tests, hepatic necrosis was also decreased by AdiNOS pretreatment compared with saline- or AdLacZ-treated groups (Fig. 3B). There was no significant difference in the necrosis between saline- and AdLacZ-treated groups.

Figure 3.

(A) Time course of serum AST (left) and ALT (right) levels after reperfusion. Pretreatment with AdiNOS, but not AdLacZ, significantly attenuated the elevation of serum AST and ALT. *P < .05 versus saline- or AdLacZ-treated group, #P < .05 versus saline-treated group (n = 6 for each group). There was no significant difference between saline- and AdLacZ-treated group in each time point. (B) Hematoxylin-eosin staining and liver necrotic area at 48 hours after reperfusion. Sinusoidal congestion and necrosis (arrows) in liver tissue were markedly reduced by AdiNOS pretreatment as compared with saline- or AdLacZ-treated group. (Original magnification ×200.) *P < .01 versus saline- or AdLacZ-treated group (n = 4 for each group). AST, aspartate aminotransferase; ALT, alanine aminotransferase.

AdiNOS Pretreatment Suppresses Hepatic ICAM-1 mRNA Expression and Neutrophil Accumulation After OLT.

In a time course experiment, we observed that hepatic ICAM-1 mRNA expression peaked at 3 hours after reperfusion in untreated transplant recipients (data not shown). Therefore, we examined the effect of AdiNOS pretreatment on hepatic ICAM-1 mRNA levels at 1 and 3 hours after reperfusion. AdiNOS-transduced grafts exhibited a significant reduction in ICAM-1 mRNA expression at 3 hours compared with saline or AdLacZ controls (Fig. 4A). Neutrophil accumulation, which was prominent in and surrounding areas of liver necrosis, was significantly reduced at 48 hours after reperfusion in AdiNOS-treated group compared with the saline or AdLacZ animals (Fig. 4B). The beneficial effects of AdiNOS pretreatment on ICAM-1 mRNA levels and neutrophil infiltration were significantly reversed by addition of the iNOS-selective inhibitor, L-NIL, indicating that the protective effects are mediated by NO (Fig. 4A-B). Because HO-1 is known to have beneficial effects in some models of liver injury, we injected the HO-1 inhibitor, SnPP, in the AdiNOS-transduced animals and did not see any effect of SnPP on ICAM-1 mRNA or neutrophil infiltration compared with AdiNOS alone. These findings suggest that HO-1 did not play a role in AdiNOS-mediated inhibition of ICAM-1 mRNA and neutrophil infiltration.

Figure 4.

(A) Hepatic ICAM-1 mRNA expression during early reperfusion period analyzed by real-time reverse transcription polymerase chain reaction. Pretreatment with AdiNOS, but not AdLacZ, significantly reduced hepatic ICAM-1 mRNA expression compared with saline control 3 hours after reperfusion. Although SnPP administration did not change the expression of ICAM-1 mRNA, L-NIL completely reversed the effect of AdiNOS. *P < .01 versus saline- or AdLacZ-treated group (n = 4 for each group). (B) Neutrophil infiltration detected by naphthol AS-D chloroacetate esterase staining of the liver tissue at 48 hours after reperfusion. The accumulation of neutrophil was prominent in and surrounding area of necrosis in the saline- or AdLacZ-treated groups, but not the AdiNOS-treated group. (Original magnification ×400.) (C) The number of neutrophils in liver tissue at 48 hours after reperfusion. AdiNOS pretreatment significantly suppressed the neutrophil accumulation at 48 hours after reperfusion. L-NIL treatment of the AdiNOS-transduced recipients resulted in the number of infiltrating neutrophils comparable to levels of saline controls. *P < .01 versus saline- or AdLacZ-treated group (n = 4 for each group).

Effect of AdiNOS Pretreatment on Total Hepatic Blood Flow After OLT.

Because NO is a vasodilator, we assessed total hepatic blood flow detected by laser Doppler flowmeter. Total hepatic blood flow was 31.2 LDU in normal liver and increased by 15% in the donor graft 4 days after AdiNOS injection. Injection of saline as negative control did not influence the total hepatic blood flow. After liver transplantation, total hepatic blood flow was decreased by 40% compared with normal animals at 3 hours after reperfusion to 18.6 LDU. AdiNOS-transduced grafts exhibited an 8% increase in total hepatic blood flow at 3 hours and 28% increase at 24 hours (P < .05) after reperfusion compared with saline controls.

AdiNOS Pretreatment Does Not Decrease Apoptotic Cell Death After OLT.

To determine the effect of AdiNOS pretreatment on apoptosis, we examined the expression of cleaved caspase-3, an active form of caspase-3, and cleaved PARP, one of the main targets of caspase-3, in liver graft by Western blot analysis. We also counted the number of apoptotic cells by TUNEL stain as well as HE. Because the increase in the number of apoptotic cells peaked at 3 hours after reperfusion in untreated transplant recipients (data not shown), we selected the 3-hour time-point for evaluation of apoptotic cell death. Unexpectedly, AdiNOS pretreatment significantly increased the expression of cleaved caspase-3 compared with saline- or AdLacZ-control group (Fig. 5A). However, no significant difference was seen in expression of cleaved PARP protein expression. The number of apoptotic cells significantly increased 3 hours after reperfusion; however, there was no significant difference among saline-, AdLacZ-, or AdiNOS-treated animals detected by either TUNEL or HE (Fig. 5B-C).

Figure 5.

Assessment of apoptosis at 3 hours after reperfusion. (A) Western blotting assay for cleaved caspase-3 and cleaved PARP using cytoplasmic and nuclear protein isolated from the liver grafts 3 hours after reperfusion. Pictures were representative from 2 independent experiments (4 different animals). Each band density was expressed in comparison with relative actin immunoblot of cytoplasm or nucleus. (*P < .01 vs. saline- or AdLacZ-treated group, n = 4 for each group). (B) The number of TUNEL-positive stainings were counted and expressed as the number of cells per field of 1 mm2. The number of TUNEL-positive cells significantly increased 3 hours after reperfusion; however, there was no significant difference in the number of TUNEL-positive cells among saline-, AdLacZ, or AdiNOS treatment. (P < .01 vs. fresh normal liver, n = 4 for each group). (C) Apoptosis was also evaluated by HE stain. The number of apoptotic cells were similar to those detected by TUNEL stain. (P < .01 vs. fresh normal liver, n = 4 for each group). PARP, poly ADP-ribose polymerase; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling.

AdiNOS Pretreatment Does Not Influence Posttransplantation Hepatic HO-1 Expression.

Because NO has been shown to induce HO-1 expression, we tested for the effect of AdiNOS pretreatment on HO-1 levels. AdiNOS injection increased HO-1 protein expression in donor liver 4 days after intravenous injection (before storage) compared with saline or AdLacZ treatment (Fig. 6). However, after liver transplant, HO-1 protein expression was increased 6 and 48 hours after reperfusion in all groups regardless of treatment (Fig. 6). This indicates that the increase in HO-1 expression after OLT is attributable to the I/R injury itself and not to AdiNOS pretreatment.

Figure 6.

HO-1 protein expression in donor liver. Expression of HO-1 protein in donor liver (before storage) was evaluated by Western blot analysis. AdiNOS pretreatment significantly increased hepatic HO-1 protein expression at 4 days after intravenous injection (before storage) as compared with saline- or AdLacZ-treated animal. There was no difference in HO-1 expression between groups after transplantation. Gels are representative from 2 independent experiments from 4 different animals in each group. *P < .01 vs. saline- or AdLacZ-treated group. HO-1, heme oxygenase-1; NL, normal.

Inhibition of iNOS Abrogates AdiNOS-Mediated Hepatoprotection.

To confirm that hepatic iNOS expression is involved in the mechanism of AdiNOS-mediated protection, animals were treated with saline or the iNOS-selective inhibitor L-NIL after AdiNOS pretreatment. As expected, animals receiving liver grafts transduced with AdiNOS showed an increase in serum nitrite + nitrate levels 6 to 48 hours after reperfusion (Fig. 7A). L-NIL injection significantly reduced serum NO levels in AdiNOS-treated recipients to near normal levels during entire reperfusion period (Fig. 7A). Importantly, iNOS inhibition abrogated the beneficial effect of AdiNOS pretreatment on hepatic I/R injury assessed by serum ALT levels at 24 hours after reperfusion, indicating the protective role of iNOS-derived NO in this setting (Fig. 7B). Administration of the HO-1 inhibitor SnPP did not alter the beneficial effect of AdiNOS pretreatment, indicating that HO-1 is not involved in AdiNOS-mediated protection (Fig. 7B). Use of SnPP alone did not influence the degree of liver injury posttransplantation (ALT 3,071.5 ± 52.5 IU/L in OLT with SnPP vs. 2,546.2 ± 602.1 IU/L in saline control at 24 hours), and is consistent with only basal levels of HO-1 expression in saline or AdLacZ-treated donor grafts (Fig. 6A).

Figure 7.

Effect of iNOS or HO-1 inhibitor on AdiNOS-mediated protection against preservation injury. (A) The iNOS inhibitor, L-NIL, significantly reduced serum NO levels in AdiNOS-treated recipients to near normal levels during entire reperfusion period. (B) After the pretreatment with AdiNOS (2 × 109 pfu), L-NIL (iNOS inhibitor, 30 mg/kg), SnPP (HO-1 inhibitor, 10 μmol/L/kg), or saline was injected to the donor or recipient rats as described in Materials and Methods. Liver function was assessed by serum ALT levels at 6, 24, and 48 hours posttransplantation. A significant decrease in serum ALT level was observed in the AdiNOS+saline group compared with the AdLacZ-treated control. Administration of L-NIL significantly increased serum ALT level compared with AdiNOS + saline group at 24 hours after reperfusion, whereas serum ALT levels were not statistically different between AdiNOS+saline and AdiNOS+SnPP groups *P < .05 vs. AdLacZ-control group; #P < .05 vs. AdiNOS + saline group (n = 5 for each group). iNOS, inducible nitric oxide synthase; HO-1, heme oxygenase-1; ALT, alanine aminotransferase.

AdiNOS Delivery Improves Transplant Survival and Hepatic Injury After Prolonged Cold Storage.

To determine the clinical relevance of AdiNOS transduction, we prolonged cold-storage time from 18 hours up to 24 hours to establish a more severe ischemia/reperfusion injury. Cold preservation for 24 hours resulted in significantly higher serum ALT levels (∼6,000 IU) compared with those in the recipients grafted with 18 hours' cold preservation at 24 hours after reperfusion (Fig. 8A). This level of ALT release would also be designated as a severe transplant preservation injury in human liver grafts and closely mimics the degree of necrosis seen in the clinical setting with severe I/R injury. Furthermore, the 24 hours of cold storage resulted in 70% mortality early posttransplantation, compared with no mortality in the 18 hours' cold storage model. AdiNOS delivery, but not AdLacZ, significantly ameliorated liver injury at both time points and also significantly improved survival, going from 30% to 70% (Fig. 8B). There was no significant difference between AdLacZ-treated animals and saline-injected transplants alone. These findings convincingly demonstrate that AdiNOS transduction of donor grafts can improve both liver injury and survival in a clinically relevant model of severe I/R injury.

Figure 8.

AdiNOS delivery improves transplant survival and hepatic injury after prolonged cold storage. (A) Cold preservation for 24 hours resulted in significantly higher serum ALT levels at 24 hours after reperfusion compared with those recipients receiving grafts with 18 hours' cold preservation. AdiNOS delivery, but not AdLacZ, significantly ameliorated liver injury at both cold storage times. *P < .05 vs. 18 hrs; #P < .05 vs. saline- or AdLacZ-treated group. (B) In correlation with serum ALT levels, AdiNOS induction significantly improved recipient survival with the hepatic grafts preserved for 24 hours, going to 70% from 30% in saline-controls or AdLacZ (log rank test). ALT, alanine aminotransferase.

Discussion

The major and novel findings of this study are: (1) donor pretreatment with a single injection of AdiNOS results in early transgene expression in the graft liver posttransplantation; (2) AdiNOS pretreatment significantly improves liver transplant I/R injury and survival; (3) the protective effects on liver graft damage are mediated in part by decreased ICAM-1 expression and neutrophil infiltration.

Previously, we have shown the endogenous hepatic iNOS protein expression is transiently upregulated after OLT after 18 hours of cold storage.7 The exact role of endogenous iNOS expression remains controversial in this setting. Two opposing hypotheses are that iNOS expression contributes to the liver injury or serves as a defense mechanism of protection. The purpose of this study was not to resolve the controversy regarding the specific role of endogenous iNOS expression but rather to demonstrate a beneficial effect with transduction of donor grafts by AdiNOS. In the current study, donor pretreatment with AdiNOS dramatically induced hepatic iNOS protein expression with marked NO generation and significantly ameliorated hepatic I/R injury associated with cold storage and OLT. Extension of cold ischemia time from 18 to 24 hours produced a severe liver I/R injury that resulted in 70% early recipient mortality. AdiNOS delivery, but not AdLacZ, significantly ameliorated liver injury at both time points, and also significantly improved survival, going from 30% to 70%. These findings highlight the clinical utility of donor graft manipulation to improve posttransplantation outcome. Additionally, iNOS inhibition by L-NIL, which effectively abrogated induced NO synthesis, significantly impaired AdiNOS-mediated protection against hepatic preservation injury. These results strongly suggest that iNOS over-expression in the early posttransplantation period plays a beneficial role against I/R injury. NO has been shown to prevent mitochondrial permeability transition-dependent necrotic cell death after I/R in cultured rat hepatocytes,30 and iNOS-deficient mice have been reported to increase the susceptibility to heart, kidney, and liver I/R injury.10, 31, 32

Downregulation of eNOS and subsequent decrease in local NO production by eNOS in the early reperfusion period was reported to be associated with I/R injury after human liver transplantation.33 Killing of sinusoidal endothelial cells and denudation of the sinusoidal lining, which is typically observed after cold I/R,34 seems to account for eNOS-downregulation because there was no significant reduction of eNOS mRNA levels in reperfused liver graft compared with donor liver before preservation.33 In the current study, we also observed significant decrease in eNOS protein levels at 1 and 3 hours after reperfusion in both controls and the AdiNOS-treated group. In addition, we could not see marked iNOS protein expression and elevation of serum NO during early reperfusion in both controls. Therefore, sufficient supply of iNOS-derived NO, which could compensate the decrease in eNOS-derived NO, during early reperfusion may contribute to hepatoprotective effects of AdiNOS pretreatment.

ICAM-1 expression on endothelial cells contributes to neutrophil accumulation at the site of inflammation, and neutrophil recruitment into sinusoids during reperfusion contributes to the development of hepatic I/R injury.35, 36 In the current study, we found that AdiNOS pretreatment significantly reduced hepatic ICAM-1 mRNA levels during early reperfusion periods, and suppressed neutrophil accumulation in liver graft at later reperfusion periods. Although the data do not allow us to conclude a direct relationship between ICAM-1 mRNA expression and liver injury in this model, the diminished ICAM-1 mRNA and neutrophil infiltration in the AdiNOS-treated group are associated with less hepatic injury. Furthermore, the blockade of iNOS by L-NIL completely abrogated the decrease of ICAM-1 mRNA or neutrophil infiltration, which was not altered by administration of HO-1 inhibitor, suggesting that the beneficial effect is NO-dependent. These findings are consistent with previous reports that inhibition of endogenous NO synthesis increased hepatic ICAM-1 mRNA expression and enhanced neutrophil-dependent reperfusion injury in rat hepatic warm I/R model37 and that both endogenous and exogenous NO enhancement attenuated neutrophil infiltration and warm I/R injury of the dog liver.38

We also examined the potential role of AdiNOS treatment on total hepatic blood flow and HO-1 expression. AdiNOS pretreatment did produce a small but significant increase in total hepatic blood flow. However, hepatic HO-1 expression was not different posttransplantation in the AdiNOS group compared with AdLacZ or saline controls, and is consistent with the observation that blockade of endogenous HO-1 by SnPP did not reverse the protective effect of AdiNOS on ICAM-1 mRNA and polymorphonuclear cell infiltration.

Apoptotic cell death is observed in both hepatocytes and sinusoidal endothelial cells during hepatic warm and cold I/R.39, 40 Although the contribution of apoptosis to overall I/R injury remains controversial,41–43 inhibition of apoptotic signals by caspase inhibitors has been shown to improve liver graft injury after cold I/R.44, 45 Anti-apoptotic effect of iNOS-derived NO was demonstrated in cultured hepatocytes and was associated with downregulation of caspase cascade.46, 47 In the current study, AdiNOS pretreatment significantly increased cleaved caspase-3 expression without exacerbating apoptotic cell death, indicating that AdiNOS-mediated hepatoprotection does not appear to be associated with modulation of apoptosis. Upregulation of caspase-3 does not always result in acceleration of apoptosis. Indeed, caspase-3 activation was observed in ischemic preconditioned rat brain tissue, without apoptotic cell death, and that was essential for ischemic preconditioning-mediated neuroprotection.48

In conclusion, donor pretreatment with AdiNOS results in a marked increase in hepatic iNOS protein and NO production, which leads to amelioration of hepatic cold preservation injury. Downregulation of ICAM-1 mRNA and subsequent neutrophil infiltration appear to be an important mechanism by which the overexpression of iNOS plays a beneficial role in cold hepatic I/R injury. Donor graft treatment with AdiNOS not only produced an approximately 60% reduction in graft injury but also significantly improved survival when prolonged cold-storage was used. We acknowledge the potential limitations of this study for clinical application. However, given the expanding use of marginal liver grafts in the United States because of organ shortage, strategies that improve outcome in cases of severe I/R injury are desperately needed.

Ancillary