|Home | About | Journals | Submit | Contact Us | Français|
Liver injury induced by ischemia/reperfusion (I/R) is the prime factor in delayed or loss graft function following transplantation. CD4+ T lymphocytes are key cellular mediators of antigen-independent inflammatory response triggered by I/R. We attempted to modulate rat liver I/R injury by targeted gene therapy with CD40Ig, which blocks the CD40–CD154 costimulation pathway. One hundred percent of Ad-CD40Ig-pretreated orthotopic liver transplants (OLTs) subjected to 24 h of cold (4°C) ischemia survived >14 days (vs 50% in untreated/Ad-β-gal groups). Ad-CD40Ig treatment decreased sGOT levels and depressed neutrophil infiltration, compared with controls. These functional data correlated with histological Suzuki’s grading of hepatic injury, which in untreated/Ad-β-gal groups showed severe necrosis (>60%) and moderate to severe sinusoidal congestion; the Ad-CD40Ig-pretreated group revealed minimal sinusoidal congestion/necrosis. Unlike in controls, OLT expression of mRNA coding for IL-2/IFN-γ remained depressed, whereas that of IL-4/IL-13 reciprocally increased in the Ad-CD40Ig group. Ad-CD40Ig reduced frequency of TUNEL+ cells and proapoptotic Caspase-3, but enhanced antioxidant HO-1 and antiapoptotic Bcl-2/Bcl-xl expression. Thus, prolonged blockade of CD40–CD154 by CD40Ig exerts potent cytoprotection against hepatic I/R injury. These results provide the rationale for a novel gene therapy approach to maximize the organ donor pool through the safer use of liver transplants exposed to prolonged cold ischemia.
The damage to the liver caused by ischemia and reperfusion (I/R) represents a continuum of processes that culminate in hepatocellular injury. These processes are triggered when the liver is transiently deprived of oxygen and reoxygenated and can occur in a number of clinical settings, such as those associated with low flow states and with diverse surgical procedures or during organ procurement for transplantation. In fact, I/R injury to the liver, an Ag-independent component of “harvesting” insult, represents an important problem affecting transplantation. It causes up to 10% of early organ failure and can lead to the higher incidence of acute and chronic liver graft rejection . The primary pathophysiological events of this injury include initial recruitment and interaction of neutrophils to the endothelium, mediated by the selectin family of adhesion molecules ; Kupffer cell release of proinflammatory cytokines/chemokines, resulting in microcirculatory disturbances ; sinusoidal endothelial cell death; and loss of parenchymal (hepatocytes) and nonparenchymal (Kupffer/Ito cells) viability . Although numerous studies have given us insight into the inflammatory response to I/R injury, the exact mechanisms that lead to liver failure remain to be elucidated.
The demonstration that systemic immunosuppression (CsA, FK506) attenuates hepatocellular injury implies the involvement of T lymphocytes in I/R [5,6], data supported by our own findings in T-cell-deficient (nude) mice , as well as in rats in which treatment with FTY720 ameliorated the insult in parallel with redistribution of recirculating T cells from blood into lymph nodes . The CD4+ Th1 proinflammatory cytokine producers represent the prime cellular mediators in hepatic I/R injury . Consequently, anti-inflammatory effects of IL-10 relate not only to inhibition of Kupffer cells but also to resident T cell inhibition , consistent with our own findings on cytoprotection rendered by Ad-IL-13 in rat models of hepatic I/R injury  or in recipients with disrupted STAT4 transcription pathway . All these results are consistent with an emerging paradigm on the pivotal role of T cells in the mechanism of I/R injury.
The CD40–CD154 costimulation pathway provides the essential second signal in the initiation and maintenance of T-cell-dependent immune responses . Although the efficacy of CD154-targeted therapy to prevent acute rejection, and to induce tolerance in some transplantation models, has been well established , little is known about how CD40–CD154 signaling may affect I/R injury in transplant recipients. By utilizing a nontransplant murine knockout system, we have recently documented the requirement for CD154 in the mechanism of hepatic reperfusion injury following “warm” ischemia . The present study was designed to analyze the role of the CD40–CD154 pathway in a more clinically relevant rat model of ex vivo “cold” ischemia followed by orthotopic liver transplantation (OLT). Our results show that prolonged in vivo blockade of CD40–CD154 interactions following pretreatment of liver isografts with replication-deficient adenovirus encoding CD40Ig (Ad-CD40Ig) exerted potent cytoprotection against I/R injury, as evidenced by 100% OLT survival, prevention of apoptosis, depression of Th1-type cytokines, and triggering of local expression of antioxidant/antiapoptotic genes. Thus, by modulating inflammatory pathways that are initiated prior to I/R injury, our results provide the rationale for novel therapeutic approaches to maximize the organ donor pool through the safer use of liver transplants exposed to a prolonged period of cold ischemia.
We analyzed the effects of Ad-CD40Ig gene transfer in our well-established model of 24-h ex vivo hepatic cold (4°C) ischemia, followed by syngeneic OLT [11,15–17]. As shown in Fig. 1, only 50% untreated or Ad-β-gal-pretreated OLTs were alive at day 14, with the majority of death occurring within the first 3 posttransplant days. In contrast, 100% of recipients of OLTs infected with Ad-CD40Ig survived 14 days. To document Ad vector expression, we analyzed OLTs for 5-bromo-4-chloro-indolyl–β-d-galactopyranoside (X-gal) staining. As previously reported , we have now detected ca. 90, 70. and 50% of X-gal+ cells at days 3, 7, and 14 posttransplant, respectively (mean; two or three rats/group; not shown).
To detect CD40Ig expression, we performed serial ELISA. As shown in Fig. 2, serum CD40Ig protein levels (µg/ml), which were the highest in the Ad-CD40Ig group at 24 h posttransplant (194–221 µg/ml), declined after day 7, but still remained consistently elevated, compared with untreated or Ad-β-gal controls.
As shown in Fig. 3, serum glutamic oxaloacetic transaminase (sGOT) levels in OLT recipients were markedly decreased in the Ad-CD40Ig gene transfer group, compared with untreated and Ad-β-gal controls (at day 1, 84 ± 18 vs 281 ± 40, P < 0.005, and 551 ± 184, P < 0.05, respectively; at day 3, 168 ± 66 vs 918 ± 85, P < 0.001, and 1032 ± 215, P < 0.01, respectively; at day 7, 262 ± 18 vs 1258 ± 190, P < 0.05, and 1650 ± 89, P < 0.005, respectively; and at day 14, 366 ± 55 vs 1773 ± 280, P < 0.05, and 2409 ± 120, P < 0.001, respectively).
We assessed hepatic I/R injury using the Suzuki classification . By 24 h, untreated and Ad-β-gal control OLT showed moderate to severe hepatocyte necrosis and sinusoidal/vascular congestion (Figs. 4A and B; Suzuki score 3.0 ± 0.9 and 3.3 ± 0.8, respectively). In contrast, the Ad-CD40Ig group revealed minimal necrosis/sinusoidal congestion and almost complete preservation of lobular architecture (Fig. 4C; Suzuki score 1.1 ± 0.8, P < 0.0005).
To determine whether Ad-CD40Ig affected neutrophil infiltration, we assessed MPO activity in OLT tissue samples. As shown in Fig. 5, Ad-CD40Ig significantly decreased MPO activity (U/g) at 24 h, from 3.9 ± 0.2 in untreated and 4.1 ± 0.3 in Ad-β-gal controls to 1.6 ± 0.4 in the Ad-CD40Ig group ( P < 0.05).
We then used competitive template RT-PCR to analyze cytokine gene expression patterns. Fig. 6 shows results of a time-course study in which OLTs infected with Ad-CD40Ig or Ad-β-gal were screened at days 1, 7, and 14 for mRNA coding for Th1 (IL-2, IFN-γ) and Th2 (IL-4, IL-10) cytokines. In untreated and Ad-β-gal livers, the expression of IL-2 and IFN-γ remained consistently and significantly increased throughout, compared with the Ad-CD40Ig group ( P < 0.05). However, Ad-CD40Ig resulted in a significant ( P < 0.05) and progressive increase in gene transcript levels for IL-4 and IL-13, whereas those of IL-2 and IFN-γ were reciprocally diminished.
By 24 h posttransplant, liver grafts in the untreated and Ad-β-gal control groups showed a large number of apoptotic cells (29 ± 16 and 31 ± 15 cells/field, respectively, Figs. 7A and B). In contrast, the number of apoptotic cells in OLTs that underwent Ad-CD40Ig gene transfer was markedly decreased (7 ± 5; P < 0.005; Fig. 7C).
We used Western blots to analyze the expression of antioxidant (heme oxygenase-1; HO-1), antiapoptotic (Bcl-2/Bcl-xl), and pro-apoptotic (Caspase-3) gene products at days 1, 7, and 14 post-OLT. Their relative expression levels were determined by densitometry. As shown in Fig. 8, the expression of HO-1 and Bcl-2/Bcl-xl was strongly up-regulated after Ad-CD40Ig (1.6–1.8 AU and 2.0–2.2, 1.7–2.3 AU, respectively), compared with untreated or Ad-β-gal controls (0.05–0.2 and 0.1–0.3 AU, respectively). In contrast, Caspase-3 expression was inhibited after CD40Ig (0.4–0.5 AU), compared with the untreated or Ad-β-gal group (1.6–1.8 and 1.7–2.0 AU, respectively).
We report here the results of our studies on the protective effects of a targeted CD40Ig gene therapy in a stringent rat model of hepatic cold ischemia followed by syngeneic OLT. The prolonged blockade of CD40–CD154 in vivo interactions via gene transfer-mediated overexpression of CD40Ig (1) prevented I/R injury in liver grafts, as evidenced by long-term OLT survival; (2) ameliorated hepatocellular damage (sGOT levels), preserved hepatocyte integrity (Suzuki’s criteria), and decreased neutrophil infiltration (MPO activity); (3) prevented prolonged intragraft Th1-type (IFN-γ/IL-2) cytokine up-regulation, seen otherwise in OLTs infected with Ad-β-gal; and (4) prevented apoptosis (TUNEL stains/Caspase 3) and upregulated OLT expression of antioxidant (HO-1)/antiapoptotic (Bcl-2/Bcl-xl) protective molecules. By demonstrating that prolonged blockade of CD40–CD154 interactions prevents I/R injury in a cold model of liver ischemia, our results provide the rationale for novel therapeutic approaches utilizing targeted gene transfer with CD40Ig to maximize donor organ use and function.
These findings complement our recent study in the mouse system in which selective disruption of CD40–CD154 signaling or treatment of wild-type (WT) animals with anti-CD154 mAb prevented hepatic insult in a “warm” 90-min I/R model . However, unlike in the latter “in situ” model , we have now performed orthotopic syngeneic transplants utilizing livers exposed to 24 h of cold (4°C) ischemia, the major component of I/R insult in clinics. Moreover, as systemic anti-CD154 blockade required repeated mAb administration in humans and primates [19,20], resulting in serious side effects such as thromboembolism [21,22], we have now employed a gene therapy approach in which the target organs, i.e., rat livers, were engineered to secrete CD40Ig locally. The CD40Ig, whose affinity to its ligand is predicted to be comparable with that of CD40, could then competitively inhibit APC (CD40)–T cell (CD154) interactions. Indeed, the gene delivery system has an advantage over the conventional therapy as it provides immunosuppression by a single vector administration and eliminates daily treatment. On the other hand, however, Ad-CD40Ig pretreatment a day prior to actual organ procurement may be clinically undesirable. Although the use of soluble CD40Ig may be an alternative, a large amount of the agent is required to block the CD40–CD154 signaling efficiently .
There are just a few other reports in the literature in which CD40Ig gene therapy induced long-term acceptance of rat liver [24,25], cardiac , and composite tissue  allografts. The duration of gene expression and levels of serum CD40Ig are of critical importance for the efficacy of gene therapy. Consistent with other reports [24–27], we have shown that highly elevated CD40Ig levels were detectable systematically by a single donor liver administration of Ad-CD40Ig. These allowed long-term acceptance of OLTs despite a prolonged period of cold ischemia. As the expected duration of Ad transgene expression after systemic infusion is ca. 2–3 weeks, these results may reflect the high degree of stability of CD40Ig protein and/or prolonged hepatocyte expression of the transgene.
The hepatocytoprotective effects of Ad-CD40Ig were reflected by the inhibition of I/R-induced sGOT release and confirmed by histological Suzuki’s criteria . Indeed, unlike in the controls, Ad-CD40Ig-infected OLTs revealed minimal necrosis/sinusoidal congestion and preservation of lobular architecture. Consequently, the 14-day survival of OLTs (a standard measure in this model) subjected to cold ischemia increased from ca. 50% in untreated/Ad-β-gal-treated controls to 100% after Ad-CD40Ig therapy, which matches our recently reported effects of gene therapy-induced Th2-type IL-13 [11,17] or antiapoptotic Bag-1  in this stringent I/R injury rat model. We have previously shown that Ad-mediated gene transfer of antioxidant HO-1 results in ca. 70% animal survival at 14 days post-OLT . Collectively, these results are consistent with the ability of CD40Igmediated CD40–CD154 blockade to prevent Ag-independent OLT injury triggered by cold I/R.
Livers pretreated with Ad-CD40Ig had significantly decreased MPO activity, a marker for neutrophil infiltration. Although MPO assay may also detect some macrophages , our results are consistent with the role of neutrophils in the pathophysiology of hepatic I/R injury . Consistent with the role of CD154 signaling in T cell activation [13,14], and in agreement with our own data , CD40–CD154 blockade markedly diminished the activation of OLT-infiltrating mononuclear cells, as measured by IL-2, and IFN-γ mRNA expression. Locally produced IFN-γ can activate Kupffer cells and lead to a further recruitment and activation of neutrophils . Thus, diminished IFN-γ mRNA may also explain reduced neutrophil infiltration seen in Ad-CD40Ig-transfected OLTs. However, unlike in our Ag-independent I/R model, CD40Ig gene transfer had only a moderate impact on leukocyte infiltration/activation in rat cardiac allograft recipients, in which it failed to prevent chronic rejection .
The question arises as to what other mechanisms may confer Ad-CD40Ig-induced cytoprotection. These may well be complex and interrelated. First, prolonged CD40–CD154 blockade inhibited apoptosis, as assessed by nuclear fragmentation primarily of hepatocytes in OLTs. Considering that the final consequence of I/R injury is the damage to hepatocytes/endothelial cells, decreasing frequency of OLT-infiltrating “cytodestructive” leukocytes should be beneficial. Second, CD40Ig gene transfer triggered increased expression of antiapoptotic (Bcl-2/Bcl-xl) and antioxidant (HO-1) proteins, yet diminished local Caspase-3. Consistent with our studies in allotransplant models [30,31] these protective molecules were associated with a Th2-enriched microenvironment and remained up-regulated selectively in “cytoprotected” OLTs. Indeed, gene therapy-induced antiapoptotic Bcl-2  or Bag-1  induction prevented hepatic I/R injury, whereas apoptosis blockade with a nonspecific caspase inhibitor reduced post-OLT injury . In contrast, overexpression of proapoptotic bax resulted in activation of the apoptosome and Caspase-3 . Moreover, we have shown that HO-1 overexpression by preincubation with cobalt protoporphyrin or Ad gene transfer protects OLTs from otherwise severe I/R injury . Consequently, we detected increased HO-1 protein expression after CD40–CD154 blockade in this study, consistent with the emerging concept that HO-1 represents one of the key adaptive mechanisms against oxidative stress . Thus, CD40Ig may interact with at least three pathways mediating cytotoxic I/R effects: (1) by stabilizing antiapoptotic proteins, such as Bcl-2, and preventing cell death by reactive oxygen species; (2) by inhibiting post-mitochondrial apoptotic events and by affecting apoptosome formation ; and (3) by activating a positive feedback circuit between HO-1 and Th2-type cytokines (e.g., IL-13) to amplify anti-inflammatory and antiapoptotic capacity , possibly via p38 MAPK  or STAT-4-disruption-mediated  signaling transduction pathways.
In conclusion, by modulating inflammatory pathways that are initiated prior to I/R injury, the prolonged blockade of CD40–CD154 interactions by CD40Ig gene transfer protects against severe I/R injury in a rat liver model of ex vivo cold ischemia followed by syngeneic OLT. Our findings may have potential clinical utility to maximize the organ donor pool through the safer use of liver transplants exposed to a prolonged period of cold ischemia.
Generation of replication-defective Ad coding for CD40Ig was described . Briefly, the mouse CD40Ig cDNA was placed under the transcriptional control of a murine CMV promoter. Ad-CD40Ig was constructed by homologous recombination in 293 cells with dl324 viral DNA and an Ad5-derived, E1- and E3-deleted adenovirus. Purified Ad-CD40Ig was then amplified in 293 cells grown in SMEM suspension medium (GIBCO, Grand Island, NY) with 10% newborn calf serum (GIBCO). The titration of the virus was analyzed by plaque assay. Virus stocks of 2.5 × 1010 plaque-forming units (pfu)/ml were stored at −80°C. The generation of Ad encoding Escherichia coli β-galactosidase (Ad-β-gal) was described .
Male Sprague–Dawley rats weighing 250 – 280 g (Harlan Sprague – Dawley, Inc., San Diego, CA) were maintained under conditions approved by the UCLA Chancellor’s Animal Research Committee. All animals were housed in microisolator cages in virus-free facilities and fed laboratory chow ad libitum.
Sprague –Dawley rats were infused iv with Ad-CD40Ig or Ad-β-gal (2.5 × 109 pfu). Twenty-four hour later, livers were harvested and stored for 24 h at 4C in UW solution prior to being transplanted orthotopically into syngeneic Sprague–Dawley recipients [11,15– 18]. Control animals were injected iv with 2.5 × 109 pfu of Ad-β-gal or remained untreated. Animals were followed for survival, and separate groups were sacrificed serially at 1, 3, 7, and 14 days. OLT and blood samples were harvested for future analysis.
OLT injected with Ad-β-gal were embedded in OCT freezing compound and snap-frozen in liquid nitrogen. Cryosections (8 lµm thick) were fixed with 1.25% glutaraldehyde at 4C and then stained with X-gal overnight at 37C, as described . Cells expressing -galactosidase cleave X-gal to yield a blue chromophore in the cytoplasm. Representative areas were scored for percentage of infected (blue-stained) cells with a light microscope.
ELISA to detect CD40Ig levels was performed, as described . Briefly, serum was collected from animal blood, and supernatants were obtained from homogenized liver grafts with 50 mM Tris, 150 mM NaCl, 0.1% SDS, 1% sodium deoxycholate, and 1% Triton X-100, pH 7.2). Flat-bottom 96-well microtiter plates were coated with 50 l/well of diluted anti-mouse CD40 Ab (BD Pharmingen, San Diego, CA). Nonspecific binding sites were blocked with Blocking Buffer (10% fetal bovine serum, 10% newborn calf serum or 1% BSA in PBS). Plates were rinsed, and samples of supernatant (100 µl) were added, followed by addition of 100 µl/well biotinylated rat anti-mouse CD40 Ab (2 µg/ml in Blocking Buffer). After washing, streptavidin–peroxidase conjugate (BD Pharmingen) was added, and the plates were incubated for color development; the reaction was terminated with 50 µl of stopping solution (20% SDS, 50% DMF). Plates were read at 405 nm in an ELISA reader. The linear region of CD40 curves was obtained in a series of eight twofold dilutions of mouse CD40 standard (2000–15 pg/ml).
sGOT levels, as an indicator of hepatocellular damage, were measured by an autoanalyzer (ANTECH Diagnostics, Los Angeles, CA).
Liver grafts were harvested, sliced, preserved in 10% neutral-buffered formalin, cut into 5-µm section, and stained with hematoxylin and eosin by standard methods. The histological severity of I/R injury in the OLTs was graded using Suzuki’s criteria . 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.
The presence of MPO, an enzyme specific for neutrophils, was used as an index of liver neutrophil accumulation . Briefly, the frozen tissue was thawed and placed in 4 ml iced 0.5% hexadecyltrimethylammonium bromide and 50 mmol potassium phosphate buffer solution with the pH adjusted to 5. Each sample was homogenized for 30 s and centrifuged at 15,000 rpm for 15 min at 4C. Supernatants were then mixed with hydrogen peroxide–sodium acetate and tetramethylbenzidine solutions. The change in absorbance was measured spectrophotometrically at 655 nm. One unit of MPO activity was defined as the quantity of enzyme degrading 1 mol peroxide per minute at 25C per gram of tissue.
Frozen OLT samples were homogenized, total RNA was extracted, and RNA concentration was determined by a spectrophotometer. A total of 5 µg of RNA was reverse-transcribed using oligo(dT) primers and Superscript reverse transcriptase (GIBCO). Oligonucleotide primer pairs for the 5′ and 3′ rat cytokine regions were selected based on the published sequences . To compare the relative level of each cytokine in different samples, competitors for IL-2, IFN-γ, IL-4, IL-13, and β-actin were constructed, and the competitive template RT-PCR amplification was performed . The samples were subjected to different cycle numbers at the annealing temperature that was optimized empirically for each primer pair: 40 cycles, 60C (IL-2), 45 cycles, 60C (IL-4), 40 cycles, 55C (IL-13), 35 cycles, 60C (IFN-γ), and 32 cycles, 63C (β-actin). PCR products were analyzed in ethidium bromide-stained 2% agarose gels and photographed with Polaroid film under UV light. The PCR results were scanned, and the densities of WT cDNA and gene-specific competitor bands were analyzed using Kodak Digital Science 1D Analysis Software (version 2.0). All samples were normalized against the respective β-actin WT cDNA/CT DNA ratio.
A commercial in situ histochemical assay (Klenow-FragEL; Oncogene Research Products, Cambridge, MA) was performed to detect the DNA fragmentation characteristic of apoptosis in formalin-fixed paraffin-embedded OLT sections, as described [17,18]. The results were scored semiquantitatively by averaging the number of apoptotic cells/field at 200× magnification. Six fields were evaluated per tissue sample.
Protein was extracted from tissue samples with protein lysis buffer (50 mM Tris, 150 mM NaCl, 0.1% SDS, 1% sodium deoxycholate, and 1% Triton X-100, pH 7.2). Proteins (30 µg/sample) in SDS-loading buffer (50 mM Tris, pH 7.6, 10% glycerol, 1% SDS) were subjected to 12% SDS–polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane (Bio-Rad, Hercules, CA). The gel was then stained with Coomassie blue to document protein loading. The membrane was blocked with 3% dry milk and 0.1% Tween 20 (USB, Cleveland, OH) in PBS. Polyclonal rabbit anti-rat HO-1 Ab (Stressgen, Victoria, BC, Canada), polyclonal rabbit anti-rat Bcl-2/Bcl-xl and anti-rat Caspase-3 Abs (Santa Cruz Biotechnology, Santa Cruz, CA) were used. The membranes were incubated with Abs and developed according to the Amersham Enhanced Chemiluminescence protocol. Relative quantities of HO-1, Bcl-2/Bcl-xl, and Caspase-3 proteins were determined by densitometer and expressed in absorbance units (Kodak Digital Science 1D Analysis Software).
For statistical analysis, comparisons between the groups were done using repeated-measures analysis of variance. If differences were established, we then used Student’s t test for judging statistical significance, where P values of less than 0.05 were considered statistically significant. The values are expressed as means ± SD.
This work was supported by NIH Grants RO1 AI23847, AI42223, and DK062357 (J.W.K.W.) and The Dumont Research Foundation. J.W.K.W. is holder of the Ralph and Joan Goldwyn Chair in Transplantation Immunobiology. Work at the GTRI (P.R.L. and M.G.C.) is funded in part by the Board of Governors at Cedars–Sinai Medical Center. P.R.L. is holder of the Bram and Elaine Goldsmith Chair in Gene Therapeutics.