Akt is expected to be an effective target for the treatment of ischemia-reperfusion injury (I/R) due to its anti-apoptotic properties and its ability to activate the endothelial nitric oxide synthase (eNOS) enzyme. Therefore, this study was aimed to determine the efficacy of an active mutant of Akt (myr-Akt) to decrease I/R injury in a model of orthotopic liver transplantation in pigs. In addition, we analyzed the contribution of nitric oxide in the Akt-mediated effects by using an eNOS mutant (S1179DeNOS) that mimics the phosphorylation promoted by Akt in the eNOS sequence. Donors were treated with adenoviruses codifying for myr-Akt, S1179DeNOS or β-galactosidase 24 h before liver harvesting. Then, liver grafts were orthotopically transplanted into their corresponding recipients. Levels of transaminases and lactate dehydrogenase (LDH) increased in all recipients after 24 h of transplant. However, transaminases and LDH levels were significantly lower in the myr-Akt group compared with vehicle. The percentage of apoptotic cells and the amount of activated-caspase 3 protein were also markedly reduced in myr-Akt-treated grafts after 4 days of liver transplant compared with vehicle and S1179DeNOS groups. In conclusion, myr-Akt gene therapy effectively exerts cytoprotection against hepatic I/R injury regardless of the Akt-dependent eNOS activation.
Liver transplantation has become the mainstay of therapy for patients with end-stage liver disease. However, organ shortage significantly reduces the applicability of this procedure. The longer waiting time and the increasing number of patients enlisted for transplantation further stress the importance of reducing the incidence of graft failure. I/R injury is one of the major contributors to post-transplant complications, including primary graft failure and acute rejection (1–5). The pathophysiology of hepatic ischemia reperfusion involves activation of Kupffer cells, which leads to reactive oxygen species formation (6–8). The resulting oxidative stress promotes apoptosis and necrosis of hepatic cells, activates transcription regulators, such as nuclear factor-κB and activator protein-1 (AP-1) (9,10) and generates pro-inflammatory cytokines and chemokines such as interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) (11). This is aggravated by the activation of neutrophils, macrophages and lymphocytes further enhancing reperfusion injury (12,13). All these mechanisms contribute to various degrees in the overall injury. Therefore, the combination of all these theoretical concepts in the design of new therapeutic approaches will produce superior results to those achieved by targeting only individual mechanisms.
An interesting therapeutic candidate to be used as a modulator of some of these pathophysiological mechanisms is the serine/threonine kinase Akt (or protein kinase B), which is a well-studied anti-apoptotic protein activated by the phosphoinositide 3-kinase (PI3-kinase) (14,15). The eNOS enzyme, which synthesizes NO, is one of the targets of Akt. This molecule causes vasodilation, inhibits inflammatory cell adhesion and acts as a scavenger of free radicals. All these functions are considered to be beneficial for graft function, as confirmed by studies showing that eNOS inhibition aggravates liver injury in experimental models of I/R (16). In this context, previous studies have demonstrated that Akt can phosphorylate eNOS on serine 1179 (bovine sequence) or serine 1177 (human sequence), resulting in eNOS activation and NO production (17,18). In addition, Akt is a major regulator of cell survival and may also act as a beneficial cytoprotective agent during I/R injury (14,19).
These data indicate that Akt may elicit biochemical responses leading to graft cytoprotection. Therefore, our aim in the present study was to stimulate Akt activity in donor livers before graft recovery by using an orthotopic model of liver transplantation in pigs. For this purpose, our strategy was to transduce liver grafts with a constitutively active mutant of Akt (myr-Akt) and to discriminate the contribution of NO in the effects resulting from Akt gene transduction by using an eNOS mutant (S1179DeNOS) that mimics the phosphorylation promoted by Akt in the eNOS sequence.
Material and Methods
Bovine aortic endothelial cells (BAEC) were isolated and cultured as previously described (20). HEK 293 cells (American Type Culture Collection) were maintained in DMEM with 10% heat-inactivated FCS, penicillin/streptomycin (50 U/mL and μg/mL), and l-glutamine (2 mmol/L). Hepatocytes and liver endothelial cells (LECs) were isolated from pigs after 24 h of infection with 1 × 1011 plaque forming units (pfu). of myr-Akt or S1179DeNOS adenoviruses. Briefly, liver samples were resuspended in Hanks' buffer supplemented with 15 mM HEPES, 0.001% DNase and 0.05% collagenase A (Roche Diagnostics, S.L. Mannheim, Germany) for 30 min at 37°C. The resulting liver digestion was filtered and parenchymatic cells were then isolated by centrifugation at 50g. More than 95% of this cell fraction was microscopically identified as hepatocytes. Hepatocytes were collected and subsequently processed for Western blot experiments. Next, the resulting nonparenchymatic cell suspension was layered over a discontinuous 25–50% Percoll gradient (Pharmacia, Uppsala, Sweden). After centrifugation, the middle region between layers was removed and washed with Hanks' solution. This step resulted in a substantial enrichment of LECs but was also contaminated with Kupffer cells, as previously described (21). The contaminating Kupffer cell population was eliminated from the LECs-enriched fraction by allowing the cell suspension to attach on the plastic surface for 15 min under serum-free conditions. The resulting LECs population, which exhibited a typical endothelial-like morphology including cobblestone growth and CD31 expression, were subsequently processed for Western blot experiments.
Three adenoviral constructs expressing β-galactosidase (β-gal), a phosphomimetic form of eNOS (S1179DeNOS) targeted to GFP and a constitutively active mutant of Akt (myr-Akt) targeted to hemaglutinin (HA). were generated as described previously (18,22–24). All vectors were propagated in the HEK 293 cell line and titters were determined by standard plaque assay.
In vitro and in vivo adenoviral transduction
BAECs were infected with 100 multiplicity of infection (m.o.i) of adenovirus containing β-gal, myr-Akt or S1179DeNOS for 3 h. The virus was removed and cells were left to recover for 24 h in complete medium. These conditions resulted in uniform expression of the transgenes in close to 100% of the cells.
In vivo transduction of adenoviral constructs was performed by diluting a dose of 1 × 1011 pfu in 5 mL of saline solution. This concentration of adenovirus was intravenously injected into pigs 24 h before liver harvesting.
Twenty-eight outbred weanling pigs, weighing 25–30 kg, were transplanted with an allograft from a 24 h previously treated donor belonging to one of the following experimental groups: (a) Vehicle group (n = 7), (b) animals infected with β-gal adenovirus (n = 7), (c) animals infected with S1179DeNOS adenovirus (n = 7) and (d) animals infected with myr-Akt adenovirus (n = 7).
Donor procedures: Anesthetic and monitoring procedures were performed as previously described (25,26). After opening the abdomen, the hepatic hilium was exposed. A noninvasive flowmeter (Transonic Systems Inc. HT207. Ithaca, New York) was placed around the hepatic artery and portal vein. Baseline flow values were determined. After cannulation of the portal vein and aorta, the liver was perfused with University of Wisconsin solution in all cases. Liver harvesting was then performed in a standard manner (26) and maintained during 6 h at 4°C.
Recipient procedures: After anesthetic management, a standard hepatectomy was performed as previously described (26). Immediately after the hepatectomy, supra-hepatic vena cava and portal vein anastomoses were performed allowing graft reperfusion. Thereafter, the infra-hepatic vena cava and the hepatic artery anastomoses were performed. Finally, the biliary tract anastomoses was achieved by using an intra-luminal stent. The anhepatic phase lasted no longer than 20 min in any case. Portal and hepatic artery blood flows were registered 1 h after reperfusion.
Postoperative care: Analgesia was given by intra-muscular injection of ketoprofen (1–3 mg/kg) every 8 h after tracheal extubation. The immunosuppressive regimen consisted of methylprednisolone 0.5 mg and tacrolimus 0.04 mg/kg before liver reperfusion. In the recipient, blood samples were taken after 1 h of reperfusion, 24 h after transplantation and at the end of the study. Animals were sacrificed on the fourth postoperative day (96 h) by an i.v. overdose of sodium pentobarbital. The study was approved by the Investigation and Ethics Committee of the Hospital Clinic.
Histology and immunohistochemistry
For hematoxylin-eosin (HE) staining, liver tissue was fixed in formalin and embedded in paraffin. Next, 4-μm sections were made and stained. Histopathology was interpreted by one independent pathologist who was blinded to the study. The histological severity of the liver injury was graded using modified Suzuki's criteria (27). In this classification, sinusoidal congestion, hepatocyte necrosis and ballooning degeneration are graded from 0 to 4. No necrosis, congestion or centrilobular ballooning is given a score of 0, while severe congestion and ballooning degeneration, as well as > 60% lobular necrosis is given a value of 4.
For immunohistochemistry, all livers of recipient pigs that received an orthotopic liver transplant from donor animals infected with myr-Akt and S1179DeNOS adenoviruses were fixed with 10% neutral-buffered formalin (pH 7.2) and embedded in paraffin. Tissue sections were incubated 1 h at room temperature with mouse anti-HA (Roche Applied Science, Mannheim, Germany) or mouse anti-GFP (BD Clontech) antibodies at a dilution of 1:1200 and 1:300, respectively, with PBS + 0.5% BSA and revealed with Dako LSAB2 System, HRP (Dako, Glostrup, Denmark). Immunoreactivity was visualized by light microscopy (Olympus Bx51, Tokyo, Japan).
Eighty micrograms (for activated-caspase-3 detection), 40 μg (for HA and GFP detection) or 30 μg (for CD31 detection) of proteins from liver homogenates or liver cells were prepared in lysis buffer (50 mM Tris–HCl, 0.1 mM EDTA, 0.1 mM EGTA, 1% (v/v) Nonidet P-40, 0.1% SDS, 0.1% deoxycholic acid, 20 mM sodium fluoride (NaF), 1 mM sodium pyrophosphate (Na4P2O7), 1 mM sodium vanadate, 1 mM Pefabloc, 10 μg/mL aprotinin, and 10 μg/mL leupeptin). Nitrocellulose membranes were probed with rabbit anti-activated-caspase-3 (Stressgen Bioreagents, Victoria, BC, Canada), mouse anti-HA (Roche Applied Science, Mannheim, Germany), mouse anti-GFP (BD Clontech) or mouse anti-CD31 (Chemicon, Temecula, CA) antibodies followed by incubation with goat anti-rabbit or goat anti-mouse horseradish peroxidase-conjugated secondary antibodies (Cell Signaling, Beverly, MA).
Liver tissue was fixed with 10% neutral-buffered formalin (pH 7.2) and embedded in paraffin. After rehydratation and permeabilization, tissue samples were incubated with terminal deoxynucleotidyltransferase and labelled deoxynucleotides for 1 h according to the Fluorescein-FragEL DNA fragmentation detection kit manufacturer's protocol (Oncogen Research Products; Cambridge, MA). Samples were then mounted on glass microscope slides and analyzed by fluorescence microscopy (Nikon Eclipse E600 fluorescence microscope) using a standard fluorescein filter. For each sample, the number of TUNEL-positive cells was counted per 400× high-power field. At least four representative fields were evaluated by two blinded observers for each treatment group, from which an average value was calculated.
Lactate dehydrogenase (LDH), alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured by the Advia 1650 automatic analyzer (Bayer, Tarrytown, NY). LacZ gene expression was detected by a chemiluminescent β-gal reporter gene assay (Roche Diagnostics, S.L. Mannheim, Germany). For NO quantification, the release of nitrates and nitrites (NOx) were measured in cell culture media and in serum obtained from the hepatic artery by using a chemiluminescence detector (NO analyzer, Sievers Instruments, Inc. Boulder, CO), as previously described (28,29).
Data are expressed as mean ± SEM. Statistical differences were measured by the unpaired Student t test, one-way analysis of variance with Bonferroni post hoc test, and the Kruskal-Wallis test with the Dunns post hoc test when appropriate. Differences were considered to be significant at p < 0.05.
Characterization of adenoviral constructs
Previous studies have shown that phosphorylation of eNOS by the serine/threonine kinase Akt at serine 1179 (bovine eNOS sequence) or 1177 (human eNOS sequence) significantly increase NO production (17,18). We therefore hypothesized that adenoviral transduction of a constitutively active mutant of Akt may be used as a strategy to increase NO availability in liver grafts. However, Akt also plays fundamental roles in other cellular processes such as survival and metabolic pathways. Therefore, to differentiate the contribution of NO and Akt in our experimental model of orthotopic liver transplantation, we also used a phosphomimetic form of eNOS (S1179DeNOS), which was constructed by mutating serine 1179 to an aspartate residue (Figure 1A).
BAECs were infected with 100 m.o.i. of adenoviruses expressing β-galactosidase, S1179DeNOS and myr-Akt. At this titer, close to 100% of cells were transfected as assessed by cytochemical staining with X-gal (data not shown). As seen in Figure 1B, adenoviral transduction of BAECs with myr-Akt or S1179DeNOS increased NO release (16.49±1.9 and 20.90 ± 0.9 μM of NOx accumulation/1 × 106 cells, respectively), compared with cells infected with β-gal adenoviruses (6.97 ± 1.2 μM NOx accumulation/1 × 106 cells, p < 0.01).
Next, we examined the transduction efficiency of adenoviruses and the tissue distribution of the transgenes in our experimental model. Adenovirus carrying the β-gal gene reporter were infused i.v. into pigs at a dose of 1 × 1011 pfu. Liver, lung, heart, mesentery and kidney were removed from animals 24 h after adenoviral treatment and β-gal activity was analyzed by chemiluminescence. Measurement of β-galactosidase activity demonstrated significant levels of transgene expression in all the tissues analyzed with the exception of the mesenteric tissue were β-galactosidase activity was not detected (Figure 1C).
Next, we studied the efficacy of myr-Akt and S1179DeNOS adenovirus to transduce hepatocytes and LECs. These studies were performed in pigs that were previously infected with 1 × 1011 pfu. of adenoviruses encoding for myr-Akt or S1179DeNOS. We isolated hepatocytes and LECs from the liver of these animals after 24 h of adenoviral treatment to mimic the liver graft situation just before its transplant into recipient pigs. Western blot experiments demonstrated that both proteins were efficiently expressed in these cell populations, although myr-Akt and S1179DeNOS protein abundance, detected using anti-HA and anti-GFP antibodies respectively, were most prominent in hepatocytes. In addition, CD31 was immunodetected in the endothelial cell fraction, indicating the specificity of the cell purification (Figure 2A). The presence and the cellular distribution of these proteins was also confirmed by immunohistochemistry in all the liver samples of recipient pigs that received an orthotopic liver transplant from donor animals infected with S1179DeNOS (Figure 2B) and myr-Akt (data not shown) adenoviruses. Importantly, all animals tolerated the adenoviral treatment at the indicated dose without any adverse effects.
Histological analysis of livers after reperfusion
Histological analysis of hematoxylin/eosin-stained graft biopsies was performed in all transplanted animals that survived until the end of the study and hepatic injury was quantified using Suzuki's classification. Samples revealed significant evidence of injury characterized by the presence of ballooning, vascular congestion, necrotic areas and polymorphonuclear leukocyte infiltration, predominantly around the central veins. These pathophysiological abnormalities were present in the uninfected (Figure 3A) as well as the adenovirus infected animals being much more pronounced in the vehicle (Suzuki score = 4.7 ± 1.7), β-gal (Suzuki score = 4.3 ± 1.2) and S1179DeNOS (Suzuki score = 3.9 ± 1.5) groups (Figure 3A–C, respectively). Importantly, gene transfer of myr-Akt (Figure 3D) attenuated these histological changes in recipient pigs (Suzuki score = 2.7 ± 1.2). However, this histological improvement in the myr-Akt-treated animals was not translated into an increase in the survival rate at the end of the fourth post-operative day compared to the vehicle group (85.7% for vehicle, 42,8% for β-gal, 71.4% for S1179DeNOS and 85.7% for myr-Akt).
Measurement of biochemical and hemodynamic parameters in recipient pigs after orthotopic liver transplantation
To correlate the histological effect of the adenoviral treatment with graft function, we measured AST, ALT and LDH serum concentration in all the surviving animals at 24 h and 96 h after transplantation. Serum levels of ALT, AST and LDH increased in all recipients 24 h after transplant. Among these parameters, AST and LDH levels were significantly lower in the myr-Akt-treated group compared with the vehicle group (793.5 ± 124 vs. 1963.7 ± 506 U/L for AST and 2745.0 ± 245 vs. 5666.5 ± 297 U/L for LDH, respectively). These differences were even more marked 96 h after transplantation with the level of transaminases (ALT and AST) and LDH being reduced by ∼60%, ∼70% and ∼57%, respectively, compared with the vehicle-injected group (Figure 4 A,B).
Next, we analyzed the hemodynamic effect of the adenoviral treatments by determining mean arterial pressure, heart rate, portal flow, arterial flow and serum NOx concentration 1 h after graft reperfusion (Table 1). Contrarily to what was expected, these hemodynamic parameters including NOx were not statistically different among β-galactosidase-, vehicle-, myr-Akt- and S1179DeNOS-treated groups. Thus, these results suggest that myr-Akt decrease graft injury through a mechanism not related with NO production.
Table 1. Hemodynamic parameters at 1 h after graft reperfusion
MAP = mean arterial pressure, HA: hepatic arterial. Results are given as mean ± SEM.
73.0 ± 5.8
60.2 ± 4.4
65.5 ± 3.6
61.5 ± 3.2
Heart rate (beats/min)
134.0 ± 19.6
115.5 ± 15.5
136.8 ± 20.5
120.2 ± 13.4
Portal blood flow (mL/min)
684.8 ± 36.8
703.0 ± 87.4
791.7 ± 121.1
612.0 ± 110.7
HA blood flow (mL/min)
99.7 ± 17.0
104.2 ± 21.4
127.6 ± 20.8
130.8 ± 15.3
62.5 ± 24.2
80.3 ± 16.5
85.7 ± 26.7
72.0 ± 16.8
Improvement of graft cell survival by myr-Akt treatmentin vivo
TUNEL assay was used to further examine the possibility of a relationship between myr-Akt treatment and the inhibition of cell death in liver grafts. As a positive control of the TUNEL assay, apoptosis was induced by incubation of liver sections with DNase I. No staining was observed in the negative control in which terminal deoxynucleotidyl transferase enzyme was omitted (Figure 5A). At the time point of 96 hr after liver transplantation, TUNEL staining was observed in the liver of all experimental groups with immunoreactivity localized within the nuclei of hepatocytes. However, cell viability was greatly reduced in animals treated with S1179DeNOS adenoviruses compared with the vehicle group (Figure 5E,C, respectively). By contrast, myr-Akt treatment significantly improved the percentage of viable cells compared with the other experimental groups (Figure 5D).
Additionally, we measured the amount of active caspase-3 in livers of vehicle-, S1179DeNOS- or myr-Akt-treated animals 96 h posttransplant. As shown in Figure 6, the amount of activated-caspase 3 protein was significantly lower in liver grafts transduced with myr-Akt than in the S1179DeNOS or the vehicle groups.
The most important finding of this study was that myr-Akt transduction of liver donors significantly improved graft function after cold ischemia liver transplantation. This was illustrated by reduced AST, ALT and LDH levels in the myr-Akt-pre-treated recipient animals compared to the other three experimental groups. myr Akt over-expression also led to decreased apoptosis, as measured by TUNEL and by activated-caspase-3 abundance, suggesting that cytoprotection of liver cells preserves liver function.
Our results are consistent with previous studies showing that Akt is a major regulator of cell survival. Several Akt substrates including BAD, caspase 9, transcription factors of the Forkhead family, IKK and p38 mitogen-activated protein kinase have been identified to be responsible for the ability of Akt to promote survival (19,30–33). The precise mechanism of how activation of Akt leads to cytoprotection of liver grafts was not investigated in our study. However, it is likely that myr-Akt promotes graft cytoprotection through the interaction with some of the downstream effectors described above and, additionally, by the indirect inhibition of the caspase-3 pathway activation.
Other investigators working in the field of ischemia-reperfusion injury have also demonstrated the beneficial impact of Akt activation on liver and heart dysfunction (34,35). These authors demonstrated that gene transfer of myr-Akt decreased apoptotic cell death and subsequent I/R injury in experimental models of warm I/R injury in rats. In agreement with these results, we have also found that hepatic transduction of myr-Akt protects liver grafts. However, we extended these experimental evidences in several significant ways. First, we demonstrated that myr-Akt gene transfer is effective in an experimental model of orthotopic liver transplantation, which is different from the warm I/R injury model in that cold preservation is performed and both innate and adaptive immune responses are present. Second, we used a large animal model providing, thereby, an additional basis for the utility of Akt as a therapeutic tool in an experimental model that more closely resembles the human physiology.
It has been shown that Akt can phosphorylate eNOS resulting in eNOS activation and an increase in NO production of several folds in comparison to basal levels (17,18). We have also reported that the intra-venous administration of myr-Akt adenovirus to cirrhotic rats resulted in enhanced intra-hepatic release of NO and, as a result, normalization of portal pressure (29). Therefore, Akt-dependent eNOS phosphorylation may be an important mechanism in the attenuation of I/R injury. However, and unexpectedly, the present study shows that mimicking Akt-dependent eNOS phosphorylation does not inhibit cell death or improve the hemodynamic parameters, compared with control treatments. Furthermore, transduction of the eNOS mutant, S1179DeNOS, dramatically worsened graft viability. Thus, it could be suggested that augmentation of NO may be deleterious for the liver graft following transplantation. However, this contention should be made with caution considering the numerous data showing a beneficial role of NO in preventing I/R injury obtained in studies aimed at manipulating NO bioavailability in grafts with the use of different strategies including eNOS and iNOS inhibition (36–38), treatment with NO donors or l-arginine supplementation (39–41) and liver graft adenoviral iNOS gene transfer (42). This contradiction most likely reflects the impaired eNOS activity already described in the post-transplant setting. There are several possible explanations that could account for reduced eNOS activity, apart from cofactor availability. For example, release of high amounts of arginase and decreased plasma levels of l-arginine have been reported after graft reperfusion in human and pig liver transplantation (43,44). Under these conditions, the lack of the enzymatic substrate for S1179DeNOS in transduced liver grafts could result in enzyme uncoupling and superoxide production, as described elsewhere (45). Thus, the observed increase in apoptosis induced by the eNOS mutant could be ascribed to the concomitant generation of radical species in the liver graft. This unwanted phenomenon that results from manipulating eNOS enzymatic activity does not occur when NO donors are used to increase NO bioavailability in liver grafts.
In summary, treatment of liver grafts with adenovirus encoding for myr-Akt markedly improves biochemical and cytoprotective parameters after liver transplantation, in comparison to uninfected and S1179DeNOS-treated groups. More complex, however, is the interpretation of the survival rates observed between the infected and noninfected groups at the end of the study. Animals infected with adenovirus encoding for β-gal and S1179DeNOS showed a lower survival index compared with the vehicle and myr-Akt-treated groups. Although the histological examination of transduced grafts did not show additional liver damage attributed to adenoviral infection, we cannot rule out the possibility that this discrepancy between survival rates reflects the ongoing anti-adenoviral adaptative immune response that may accelerate graft injury. Nevertheless, the treatment with myr-Akt seemed to overcome this hypothetical detrimental aspect of adenovirus treatment with the superimposed advantage of improving biochemical and histological parameters compared to vehicle treatment. However, there are clear clinical limitations of our gene therapy approach for increasing Akt activity in liver grafts. High viral titer, such as that used in our study, may induce an immune response in recipients that could be translated into the death of hepatic transduced-cells and the elimination of transgenes several days after treatment. A second limitation is that this immune response may make the expression of new transgenes difficult upon a second administration of virus. These limitations, however, may be overcome by the development of new and more efficient viral vectors.
Our results, therefore, support the important role of Akt in the pathogenesis of graft injury and raise the possibility that the development of new strategies addressed to increase the enzymatic activity of Akt should be explored as a therapeutic tool to protect liver grafts from I/R injury during transplantation.
We are indebted to Dr. Kenneth Walsh for kindly providing the myr-Akt adenovirus. This work was supported by grants from Dirección General de Investigación Científica y Técnica (SAF2006-07053 to WJ), Fondo de Investigación Sanitaria (PI041198 to MM-R), Fundació La Marató de TV3 (000610 to W.J.) and Instituto de Salud Carlos III (C02/03).