Partial liver graft transplantation is a surgical advance developed to overcome severe donor shortage. Survival of these grafts involves recovery from cold ischemia and reperfusion (CIR) injury, immediate regeneration and maintenance of function. Here we examined the outcome of partial liver grafts in comparison to whole grafts following CIR injury.
Lewis rats subjected to orthotopic liver transplantation (OLT) with whole grafts preserved in Viaspan® were compared to rats receiving 50% and 30% grafts. Outcome was analyzed by survival and regeneration.
Transplantation was associated with 100% survival for all grafts, whereas 16 h preservation resulted in 100%, 20% and 0% survival in animals receiving whole, 50% and 30% grafts, respectively. CIR induced increased IL-6 levels in 50% and 30% grafts, and activation of STAT3. Cell cycle progression (cyclin D1) and regeneration (BrdU) was initiated in all livers preserved for 1 or 8 h, but not in partial grafts preserved for 16 h.
In conclusion, partial grafts recover from CIR injury through similar molecular pathways to whole grafts. Partial grafts with severe injury fail to achieve cellular proliferation despite the early initiating signals. This failure could be attributed to the impaired ability of the parenchyma to respond to initiating signals for regeneration.
Liver transplantation has become the only successful therapy for end stage liver disease, creating a demand for organs that overwhelmingly exceeds the available supply. In an effort to develop strategies to overcome the severe shortage of cadaveric organs, the liver transplant community has developed surgical techniques in the transplantation of partial grafts, either by the splitting of one cadaveric liver for two recipients, or the use of a right or left lobe from an adult living donor as a living donor liver transplant (LDLT). Except in pediatric recipients, these grafts are typically significantly smaller than a calculated standard liver volume, with a minimal volume of liver mass necessary for successful recovery (1–3). Both animal studies and clinical experience have demonstrated decreased function and poor survival of grafts that do not make these minimal requirements, a condition termed ‘small-for-size’, manifested by prolonged cholestasis and appearance of significant ischemic-type injury on pathologic review (4,5). However, the mechanisms of this type of injury, the reasons for graft failure and particularly the ability of the partial liver graft to regenerate after transplantation are not well defined.
Liver grafts, whole or partial, are affected by cold ischemia and reperfusion (CIR) injury. In the LDLT setting, this injury may be minimal, but in cadaveric donors, it can be quite extensive. We have previously shown in whole liver grafts that the pathways of recovery from this injury resemble regeneration post-partial hepatectomy (PH) (6), with the amount of regeneration correlating with the degree of injury. Studies using knockout animals with targeted disruption of the tumor necrosis factor receptor or interleukin 6 (IL-6) genes have demonstrated that ‘the pro-inflammatory’ cytokines, TNF-α and IL-6, appear to act as growth factors following PH and are necessary for initiation of normal hepatic regeneration (7–9). These studies have also demonstrated that cytokine-responsive transcription factors, such as nuclear factor kappa B (NF-κB) and signal transduction and activator of transcription 3 (STAT3) are activated during the first few hours after PH coinciding with activation of the JAK-STAT transcriptional pathway and the induction of immediate-early growth response genes. These investigators have defined a pathway of regeneration that involves TNF-α-induced activation of NF-κB that upregulates IL-6 expression, leading to activation of STAT3 followed by expression of selected target immediate early genes associated with the cell cycle (10). STAT3 is critical for cells to progress from G1 to S phase, and is crucial for cell cycle progression (11). Cyclin D1 is then up-regulated and translocates to the nucleus, completing the cycle by moving the cell into S phase, and autonomous cellular replication is achieved (12). We and others have shown that the IL-6/STAT3 pathway is an important component in the recovery of whole liver grafts following CIR injury (6,13).
In the transplant setting, it is important for the graft to regenerate to replace lost cells damaged by CIR. In the case of partial grafts the need for immediate rapid regeneration is even more profound. There is little experimental evidence examining the regeneration response of partial grafts or how well they tolerate cold preservation (13). In this study we wished to determine the effect of CIR injury on the survival and regeneration of small-for-size grafts. We explored the role of the IL-6/STAT3 pathways in the functional recovery and initiation of regeneration of partial liver grafts in comparison to whole grafts injured by CIR.
Materials and Methods
Male Lewis rats (Harlan, Indianapolis) weighing 240–350 g, were used as donors and recipients. Animals received care according to guidelines of the Institutional Animal Care and Use Committee (IACUC) at the University of Pennsylvania. All animals were kept in a temperature-controlled environment with a 12-h light–dark cycle and were allowed free access to food and water at all times.
All operations were performed under anesthesia by isofluorane inhalation (in O2) using a commercial system (Surgivet, Waukesha, WI). All steps were performed under clean conditions using loupe (2.5×) for IVC, portal vein and common bile duct preparation and anastomosis, using a surgical microscope (GZ6E, Leica, Frankfurt, Germany) for hepatic artery preparation and anastomosis. The surgical procedure was a modification from techniques described by Lee (14) and Kamada (15,16).
After shaving and disinfection of the abdominal wall with Betadine, the abdominal cavity was accessed through a transverse laparotomy and the liver was exposed. All the ligaments of the liver were dissected, a 3–4 mm PE-50 tube stent (outer diameter, 0.023” ID, BD Biosciences) was inserted into the bile duct and secured with a 6-0 suture. The portal vein was dissected free and the pyloric vein and splenic vein were ligated and divided close to the portal vein. The hepatic artery was dissected to the celiac artery. For whole liver transplantation, the liver was perfused and removed at this point. For partial liver graft preparation, partial hepatectomy was performed in the donor animal. To achieve the 50% graft preparation, 3-0 silk sutures were placed around each left lobe (caudal, lateral and central lobes) and all left lobes were ligated and divided, preserving the right lobes. To achieve the 30% liver graft preparation, 3-0 silk sutures were also placed around the central and lateral lobes, which were ligated and then divided. The remaining caudate (left) and triangulated lobe (right) were preserved.
Donor liver perfusion and harvesting
After 100 units of heparin were injected into the penile vein, the hepatic artery was ligated and divided. The liver was perfused with 10 mL of cold Viaspan solution slowly injected into the portal vein. The suprahepatic vena cava was transected close to the diaphragm, allowing outflow of the perfusate. The portal vein and infrahepatic vena cava were divided, allowing for sufficient length for cuff preparation. The liver was freed from all retroperitoneal structures and removed. The graft was placed in Viaspan solution on ice.
Polyethylene cuffs were placed on the portal vein (PE-205, 0.062”ID, 0.082” OD, BD Biosciences) and the infrahepatic vena cava (PE-260, 0.070”, BD Biosciences). The organ remained in cold Viaspan solution for 1, 8 or 16 h, until implantation in the recipient animal.
Following a transverse incision, the recipient liver was freed from all ligaments. The common bile duct was dissected free, ligated and transected close to the liver. The common hepatic artery was dissected allowing 10-mm length for anastomosis. The portal vein, infrahepatic and suprahepatic IVC were dissected, clamped and transected close to the liver in order to preserve the maximum length for anastomosis. The native liver was removed, and the graft liver was placed in an orthotopic position in the recipient animal. The cuff technique was used for both the portal vein and the infrahepatic vena cava anastomosis. The suprahepatic vena cava was anastomosed using a 7-0 nylon suture. The bile duct was re-connected via stent. Finally, the hepatic artery was anastomosed using the sleeve technique. The anhepatic time was consistently kept below 20 min with a mean of 16 min. The two layers of the abdominal wall were closed separately by a running suture (5-0 silk).
There were three major study groups: recipients of syngeneic whole, 50% and 30% grafts. Each graft, whole or partial, was preserved in Viaspan solution at 4°C for 1, 8 or 16 h before transplantation. Three rats from each group were sacrificed at the following times: 90 min, 1, 2, 4 and 7 days following reperfusion of the liver graft. Six animals per group were followed for survival up to 7 days. Two hours before sacrifice at 48-h time point, a single dose of bromodeoxyuridine (BrdU, Sigma, St. Louis, MO, USA) was injected intraperitoneally at a dose of 50 mg/kg. At the time of sacrifice, liver tissue was processed for nuclear extracts and 2 g of tissue was homogenized in guanidine thiocyanate and immediately frozen for further RNA analysis. The remaining tissue was fixed and processed for histology and immunohistochemistry. The formalin fixed liver segments were paraffin-embedded and processed in an automated tissue processor. Five-micron liver sections were analyzed by hematoxylin and eosin (H&E) staining and immunohistochemistry (see below).
RNA isolation and semiquantitative reverse transcription-polymerase chain reaction
Total RNA was extracted from the liver by homogenization with a high speed rotor (Wayne Industry, Toledo, OH, USA) in GIT buffer (Qiagen, Santa Clarita, CA, USA) and subsequently purified by Qiagen RNAeasy columns. Five hundred nanograms of RNA samples were reverse transcribed and amplified in a 50-μL reaction mixtures containing 0.3 mM dNTPs, 1X of 5XEZ RTth Taq buffer, five units of RTth Taq polymerase (Perkin Elmer, Stamford, CT) and 0.6 μM of 5' and 3' primers. The reaction protocols were as follows: 60°C for 30 min, 95°C for 30 sec, 60°C for 1.5 min, 30 cycles. The primers used were IL-6-5'-GGAGACTTCACAGAGGATACC; IL-6-3'-GCTCTGAATGACTCTGGC. Amplified bands were electrophoresed and analyzed by a gel-doc analyzer for semiquantitation (BioRad, St Louis, MO, USA). The amount of total RNA and the number of cycles in each cytokine were adjusted to obtain a linear correlation between RNA amount and ethidium visible light fluorescence and then normalized to an internal standard RT-PCR control β-actin.
Western blot analysis
STAT3 and P-STAT3 expression was measured by western blot analysis. Such measurement required preparation of whole cellular extracts in 1% NP-40, 20-mM Tris, 150-mM NaCl and 4-mM EDTA with protease inhibitors NaF, Na2MoO4 (1 mM/L) and antipain, aprotinin, bestatin and leupeptin (each at 2 mg/mL) (Boehringer-Mannheim Corp). SDS-polyacrylamide gel electrophoresis (PAGE) of cellular extracts (20 μg) was performed using 10% acrylamide gels, according to Laemmli's method, followed by Coomassie blue staining. Primary antibodies were purchased from Cell Signaling and secondary antibodies from Amersham Biosciences. Positive signals were detected by the enhanced cheluminescence method (Amersham Life Sience, Inc., Arlington Heights, IL, USA). Equal loading was assessed by β-actin.
Hepatocyte nuclear staining for BrdU, a thymidine analogue capable of incorporation into actively replicating DNA, was performed essentially as described (6). Paraffin-embedded liver sections were deparaffinized and rehydrated then quenched for endogenous peroxidases in 0.03% hydrogen peroxide at 60°C for 15 min. Sections were then blocked with Avidin D, Biotin blocking reagent (Vector Laboratories, Burlingame, CA, USA) and rabbit serum (4% in PBT), followed by the primary antibody staining with anti-BrdU monoclonal antibody (Boehringer Mannheim), at 1:500 dilution. After rinsing with phosphate buffered saline, rabbit anti-mouse immunoglobulin (Vector Laboratories, Burlingame, CA, USA), 1:200 dilution, was applied (45 min at room temperature). Sections were incubated for 30 min with the horseradish peroxidase-ABC-conjugated reagent from the Vecta Elite kit and the signal was developed with the DAB Substrate kit for Peroxidase (Vector Laboratories). The sections were counterstained with Gills #2 Hematoxylin (Fisher Scientific, Atlanta, GA, USA). Total BrdU-labeled hepatocytes were determined by counting positively stained hepatocyte nuclei in at least three random low power (10×) microscope fields per liver, by two investigators (6). The mean for each time point was plotted with indicated standard deviations, and statistical analysis was determined with a significance of p < 0.05. Histologic assessment of the liver was also made by hematoxylin and eosin staining with interpretation by a single-blinded pathologist.
Staining for Cyclin D1 was performed mainly as described above. The primary antibody staining was performed with mouse anti-cyclin D1 monoclonal antibody (Pharmingen, San Diego, CA, USA) at 1:500 dilution (overnight at 4°C), and the secondary horse anti-mouse (Vector Laboratories) at 1:200 dilution.
Student's t-test was used to compare hepatocyte replication and cytokine levels. p < 0.05 was considered significant.
Liver mass recovery and graft survival
Orthotopic liver transplantation (OLT) was successfully performed with whole or partial grafts. Survival of animals receiving a whole graft preserved for 1, 8 and 16 h was 100%. Survival of animals receiving a partial graft (50 or 30%) preserved for 1 or 8 h was 100% up to 7 days. Eighty-nine to ninty-seven percent of the original liver mass was recovered by 7 days post-transplantation in both groups. When the partial grafts were preserved for 16 h, by 1 week the survival of 50% partial grafts dropped to 20% and to 0% for the 30% grafts (Figure 1), with death occurring from liver failure, manifested by abdominal ascitis and liver jaundice (n = 6 animals/group).
Increased hepatocyte injury in partial grafts preserved for 16 h
Whole and partial grafts with moderate cold preservation (8 h) demonstrated mild sinusoidal dilatation and inflammation (Figure 2A,C and E). Whole grafts with 16 h of cold ischemia demonstrated mild zone-2 necrosis and rare infiltrating inflammation (Figure 2B). Fifty percent partial grafts preserved for 16 h demonstrated early apoptotic figures, accompanied by sinusoidal inflammation and zone-2 necrosis (Figure 2D). Thirty percent grafts preserved for 16 h demonstrated focal congestion and hepatocyte collapse, with zones 2 and 3 necrosis, and hemorrhagic congestion (Figure 2F).
Recovery from ischemic insult induces early intragraft IL-6 in grafts preserved for extended periods
Ninety minutes following transplantation, a profile of early intrahepatic IL-6 expression was assessed using semiquantitative RT-PCR. Whole grafts preserved for 1 or 8 h demonstrated very low levels of IL-6 (Figure 3A), consistent with our previous results (6). The same low levels were observed with partial grafts preserved for 1 h. However, partial grafts preserved for 8 h demonstrated a significant increase compared to whole grafts with the similar preservation times (p < 0.05) (Figure 3B). All grafts, partial or whole, preserved for 16 h, showed increased levels of IL-6 at this time point, with no statistically significant differences.
Extended CIR injury in partial grafts is associated with early and increased STAT3 activation
To investigate activation and progression of the cell cycle pathways, we studied the levels of the phosphorylated IL-6-dependent transcription factor STAT3 (P-STAT3) in whole cell extracts obtained from the transplanted liver 90 min post-reperfusion. P-STAT3 was increased in both whole and partial grafts with 8- and 16-h cold preservation, demonstrating successful initiation of this pathway by CIR injury (Figure 4).
Cyclin D1 is absent in partial grafts with prolonged cold preservation time
Progression of cell cycle was assessed by cyclin D1 immunohistochemistry. Cyclin D1 expression was present in both the cytoplasm and nucleus at 24 h in all grafts, partial or whole, preserved for 8 h (Figure 5A,C and E). However, a difference was observed between whole and partial grafts with 16 h of cold preservation. By day 1, expression was mainly nuclear in whole grafts (Figure 5B). Partial grafts demonstrated large areas with no cyclin D1 expression (Figure 5D,F). These areas coincided with the necrotic areas seen with H&E staining.
Moderate cold ischemia does not impact cell cycle
We investigated the effect of CIR on hepatic cell cycle progression from G1 to S phase with DNA synthetic activity of partial grafts. Using BrdU immunohistochemistry, we have previously demonstrated the S-phase peak to be at 48 h after whole graft transplantation (6). However, not knowing the timing of cell cycle events in partial grafts, we monitored BrdU uptake starting from day 1, day 2 up to day 7. Fifty percent grafts with minimal ischemia (1 h) demonstrated a steady amount of BrdU uptake throughout the week (Figure 4A). As the ischemic time lengthened (8 h) these grafts demonstrated increased BrdU uptake on day 1, which then drops to average levels (Figure 6A). Thirty percent grafts with minimal ischemia displayed increased uptake compared to the whole and 50% grafts with minimal ischemia, (Figure 6B) similar to the pattern of BrdU uptake seen in the 50% grafts preserved for 8 h. As may be expected due to increased need for restoration of mass, 30% grafts preserved for 8 h demonstrate a more sustained pattern of BrdU uptake, with the number of stained nuclei peaking equally on days 1 and 2 (Figure 6B). The 50% partial grafts with 16 h of preservation were unable to demonstrated significant BrdU uptake as compared to those with lesser CIR injury (Figure 6C). These partial grafts did not survive for long, and none of the 30% grafts with 16-h preservation survived long enough to assess BrdU uptake.
Liver cells can be induced to proliferate in response to stimuli such as toxic injury, liver resection and ischemic insult. The complex process of liver regeneration has been studied intensively in animal models, and many investigators have pursued the molecular mechanisms involved in the regulation of this process (6–10). Immediate regeneration is clearly required following transplantation of partial liver grafts from living and cadaveric donors, a strategy often used for overcoming the shortage of organs. In this study we examined the regenerative response of partial liver grafts after CIR injury, focusing on the IL-6/STAT3 pathway and its role in the recovery of the partial graft. We demonstrate that the early initiating signals for the regenerative pathways are present, but progression of the cell cycle is not possible in the small for size grafts with significant injury from cold preservation.
Recent studies have highlighted TNF-a and IL-6 as major initial growth factors involved in regeneration in vivo (7,8). These factors appear to originate from the Kupffer cells, and these very early signals are necessary to prime the hepatocytes for replication (17,18). In our current model, we found increased early levels of IL-6 mRNA in partial grafts with both moderate and significant CIR injury. Other investigators have reported a blunted regenerative response in this setting, with decreased IL-6 protein levels at earlier time points (19). It is possible that CIR interferes at a translational level and does not allow for the continued expression of the already transcribed IL-6. In this study, markers of regeneration downstream of IL-6, such as STAT3 phosphorylation, cyclin D1 expression and BrdU uptake, were increased in the partial grafts with up to 8 h of cold preservation, indicating the liver segment's attempts at a more rigorous regenerative response.
Although there were no major differences in the activation of the IL-6/STAT3 pathway in whole or partial grafts preserved for 1, 8 and 16 h, there was a significant difference noted in cell cycle progression and survival in partial grafts preserved for 16 h. Despite the activation of initiating signals in the IL-6/STAT3 pathway in the grafts that fail, this alone does not seem to be sufficient to support progression of the cell cycle cascade in order to sustain the regeneration process. This lack of downstream effects of IL-6 suggests impairment not in the activation of non-parenchymal cells but in proliferation at the hepatocyte level. Prolonged cold ischemia in partial grafts might alter the cell cycle machinery at various levels, as has been described in the case of steatotic livers (20).
In hypothesizing why these grafts do not survive, it may be that small for size liver grafts with significant injury simply do not have enough of an energy supply to sustain both regeneration and maintain homeostasis of the recipient. Clinical and experimental experience shows that regeneration after re-section is related to histopathologic condition of the parenchyma. The recovery processes of cirrhotic or steatotic livers are less vigorous after partial hepatectomy compared to normal livers, with a diminished rate and extent of the regeneration response (21–23). In partial grafts with ischemic injury, the remaining hepatocytes need to reach a metabolic compromise between sustained differentiated function and cellular proliferation. These competing demands for energy may overwhelm the segment, resulting in liver failure. In our model of CIR and partial grafts, it is apparent that the maintenance of differentiated hepatic function, recovery from ischemic injury and hepatocyte proliferation, can be overwhelming.
In conclusion, the successful outcome of the cold-preserved liver graft in the transplant setting is dependent upon a robust regenerative response, replacing lost or insufficient liver mass, and maintenance of metabolic homeostasis. The degree of injury correlates with the magnitude of the needed regenerative response. In this study, we demonstrate that the repair mechanisms involving initiation of the cell cycle leading to hepatocyte regeneration are similar in the injured partial livers compared to whole grafts, however progression of the cell cycle, leading to hepatocytes proliferation, is unable to be sustained in partial grafts with extended cold preservation. Further studies delineating the pathways involved in injury and repair of the liver graft may lead to the development of methods aimed at enhancement of these processes within the small for size graft.
This work was supported in part by the National Institute of Health Center Grants P30 DK50306 and DK058315 and the Biesecker Center for Pediatric Liver Disease.