Impact of venous systemic oxygen persufflation supplemented with nitric oxide gas on cold-stored, warm ischemia–damaged experimental liver grafts


  • Pramod Kadaba Srinivasan,

    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    Search for more papers by this author
  • Shintaro Yagi,

    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    2. Department of Hepatobiliary, Pancreas, and Transplant Surgery, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Benedict Doorschodt,

    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    Search for more papers by this author
  • Kazuyuki Nagai,

    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    2. Department of Hepatobiliary, Pancreas, and Transplant Surgery, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Mamdouh Afify,

    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    Search for more papers by this author
  • Shinji Uemoto,

    1. Department of Hepatobiliary, Pancreas, and Transplant Surgery, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Rene Tolba

    Corresponding author
    1. Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany
    • Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Pauwels Strasse 30, Aachen, Germany D-52074
    Search for more papers by this author
    • Telephone: +49-241-8080472; FAX: +49-241-8082462


The increasing shortage of donor organs has led to the increasing use of organs from non–heart-beating donors. We aimed to assess the impact of venous systemic oxygen persufflation (VSOP) supplemented with nitric oxide (NO) gas during the cold storage (CS) of warm ischemia (WI)–damaged experimental liver grafts. Rat livers (n = 5 per group) were retrieved after 30 minutes of WI induced by cardiac arrest (the WI group) and were thereafter preserved for 24 hours by CS in histidine tryptophan ketoglutarate solution. During CS, gaseous oxygen was insufflated via the caval vein with 40 ppm NO (the VSOP-NO group) or without NO (the VSOP group). Cold-stored livers without WI served as controls. Liver viability was assessed after the preservation period by normothermic isolated reperfusion for 45 minutes with oxygenated Krebs-Henseleit buffer. After 45 minutes of reperfusion, the VSOP-NO–treated livers showed significantly lower alanine aminotransferase values than the WI-damaged livers (10.2 ± 0.2 versus 78.2 ± 14.6 IU/L), whereas the control livers showed no differences from the VSOP-NO–treated livers. The mitochondrial enzyme release was lower in the VSOP-NO group (4.0 ± 0.7 IU/L) versus the WI group (18.2 ± 4.9 IU/L). An increased portal vein pressure was observed throughout reperfusion (45 minutes) in the WI group (21.7 ± 0.2 mm Hg) versus the VSOP-NO group (12.2 ± 0.8 mm Hg) and the control group (19.9 ± 0.4 mm Hg). Furthermore, the NO concentration in the perfusate after 5 minutes of reperfusion was highest in the VSOP-NO group. The release of malondialdehyde into the perfusate was significantly reduced in the VSOP-NO group (0.9 ± 0.1 nmol/mL) versus the WI group (31.3 ± 5.3 nmol/mL). In conclusion, the resuscitation of livers after 30 minutes of WI to a level comparable to that of nonischemically damaged livers is possible with VSOP supplemented with NO gas. Moreover, the application of VSOP with NO minimizes the extent of injuries caused by oxygen free radicals during preservation. Liver Transpl 18:219–225, 2012. © 2011 AASLD.

The global shortage of livers for transplantation has severely limited the treatment options for patients with end-stage liver disease. To expand the donor pool, donation after cardiac death livers are increasingly being used for transplantation.1-3 However, primary dysfunction (PDF) and primary nonfunction (PNF) after transplantation occur at a higher frequency with donation after cardiac death livers.4-6 These grafts are subjected to warm ischemia (WI) before the cold ischemia preservation period and to reperfusion after transplantation. These events are deleterious to graft viability through various pathways; hepatocellular injury is the common denominator.7, 8 Moreover, the duration of the ischemic periods is directly related to the extent of apoptosis.9-11 Since the early 1970s, it has been demonstrated that venous systemic oxygen persufflation (VSOP) during cold storage (CS) has a protective effect against the apoptotic changes in WI-damaged livers.4

During preservation, the application of nitric oxide (NO), which was declared the Molecule of the Year in 1992 by Science magazine,12 has shown promising results for protecting graft viability.13 The antioxidative and vasodilative properties of NO, particularly during periods of WI, have been demonstrated previously.14-16 Moreover, a decrease in NO levels during reperfusion can lead to the aggravation of reperfusion injury.13 In a clinical study, it has been shown that the inhalation of gaseous NO at a concentration of 80 ppm can accelerate the restoration of liver function in adults after orthotopic liver transplantation.17 Furthermore, Murakami et al.18 showed that NO at a concentration of 30 ppm, when it was administered immediately or in a delayed manner during reperfusion, reduced ischemia/reperfusion injury in lungs. A similar study by Dong et al.19 showed that the administration of NO to non–heart-beating donor (NHBD) lungs before and after retrieval improved posttransplant function.

The aims of this study were to evaluate the impact of VSOP in combination with NO at a concentration of 40 ppm on cold-stored, WI-damaged liver grafts and to assess the potentially hepatoprotective effect of NO.


ALT, alanine aminotransferase; CS, cold storage; GLDH, glutamate dehydrogenase; HTK, histidine tryptophan ketoglutarate; IPRL, isolated perfused rat liver; MDA, malondialdehyde; NAC, N-acetylcysteine; NHBD, non–heart-beating donor; NO, nitric oxide; PDF, primary dysfunction; PNF, primary nonfunction; PVP, portal vein pressure; VSOP, venous systemic oxygen persufflation; WI, warm ischemia.


Experimental Protocols

All experiments were conducted in accordance with German federal law regarding the protection of animals. For the care of laboratory animals, Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication 86-23, 1985 revision) was followed. Male Wistar rats (Charles River GmbH, Sulzfeld, Germany) that weighed between 250 and 300 g were used as liver donors. The animals were housed under specific pathogen–free conditions according to the guidelines of the Federation for Laboratory Animal Science Associations in a temperature- and humidity-controlled environment with a 12-hour light/dark cycle, and they were allowed food (a standard rat diet; Sniff-Spezialitäten GmbH, Soest, Germany) and water ad libitum. After the induction of general anesthesia by the inhalation of isoflurane (Abbott GmbH & Co. KG, Wiesbaden, Germany), the abdomen was opened by a midline incision with bilateral subcostal extensions. The liver was carefully mobilized from all ligamentous attachments. In the study groups, cardiac arrest was induced by phrenotomy to induce WI. Thereafter, the hepatic artery was ligated, and the portal vein was cannulated with a 14-gauge polyethylene catheter (B. Braun Melsungen AG, Melsungen, Germany). To wash out the blood, the liver was perfused in vivo with 20 mL of an ice-cold 0.9% saline solution (B. Braun Melsungen). During the initial washout, the abdominal caval vein was incised and bled in order to prevent a hepatic outflow obstruction. After ligation of the infrahepatic caval vein, the phrenic veins, and the right adrenal vein, the liver was explanted and immediately perfused with 60 mL of histidine tryptophan ketoglutarate (HTK) solution (Custodiol, Dr. Franz Köhler Chemie, Alsbach-Hahnlein, Germany) at 4°C. For isolated reperfusion, a similar 14-gauge catheter was inserted into the suprahepatic caval vein.

Each liver was then stored in 125 mL of HTK at 4°C with a cold-water bath (Ministat 125, Peter Huber Kältemaschinenbau GmbH, Offenburg, Germany) for 24 hours. The livers were allocated into 4 groups (5 livers per group): the CS or control group (24 hours of CS), the WI group (30 minutes of WI and 24 hours of CS), the VSOP group (30 minutes of WI and 24 hours of VSOP), and the VSOP-NO group (30 minutes of WI and 24 hours of VSOP supplemented with 40 ppm NO).

In the VSOP and VSOP-NO groups (O2 flow = 0.2 L/minute, NO flow = 22 mL/minute), each liver was persufflated with medical-grade gaseous oxygen via the superior caval vein at a pressure of 18 mm Hg. The HTK solution in which the liver was immerged was supplemented with 20 mM N-acetylcysteine (NAC; Hexal AG, Holzkirchen, Germany). At the margin of each liver lobe, small pinpricks were set into dilated postsinusoidal venules with a fine acupuncture needle (0.18 × 30 mm; Seirin Corp., Shizuoka, Japan); this allowed the gas to escape from the liver microvasculature.

In order to simulate the period of rewarming during reimplantation, each liver was allowed to rewarm in a water bath set at 24°C for 30 minutes. Before reperfusion, each liver was washed out through the portal vein catheter with 10 mL of saline.


The isolated perfused rat liver (IPRL) system has previously been described in detail.20 In short, the perfusion circuit was prerinsed with 200 mL of a saline solution, and it was subsequently rinsed with 100 mL of Krebs-Henseleit buffer (Sigma-Aldrich Co., St. Louis, MO), which was modified by the addition of calcium chloride (Sigma-Aldrich) and sodium hydrogen carbonate (Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany). Reperfusion was thereafter performed for 45 minutes in a recirculating system at a constant flow of 3 mL/g of liver weight/minute with a roller pump (Masterflex L/S, Cole-Parmer Instrument Co., Vernon Hills, IL) with 220 mL of oxygenated modified Krebs-Henseleit buffer at 37°C. Carbogen (95% oxygen and 5% carbon dioxide) was used for oxygenation, and the perfusate's partial pressure of oxygen, which was measured by blood gas analysis (ABL 5, Radiometer, Copenhagen, Denmark), was maintained at a minimum of 500 mm Hg during the reperfusion period.

Portal Vein Pressure (PVP)

During reperfusion, PVP was continuously measured with a water column connected to the portal vein inflow catheter. The measuring system was calibrated at the start of each procedure.

Enzyme Release

The hepatic effluent was intermittently sampled after 5, 15, 30, and 45 minutes of reperfusion and was analyzed for the release of alanine aminotransferase (ALT) and glutamate dehydrogenase (GLDH). The ALT content of the effluent was measured to assess general hepatic injury by a standard enzymatic method with a Vitros 250 analyzer (Ortho-Clinical Diagnostics, Raritan, NJ). GLDH release was measured with a photometric assay kit (Analyticon Biotechnologies, Lichtenfels, Germany) with photometric reading at 340 nm (EPAC 6140 photometer, Eppendorf-Netheler-Hinz GmbH, Hamburg, Germany).

Lipid Peroxidation

In order to assess oxygen free radical activity after reperfusion, the malondialdehyde (MDA) concentration in the perfusate was determined as described previously.21 In short, 100 μL of the hepatic effluent was mixed with 750 μL of 0.44 M phosphoric acid, 250 μL of a 42 mM aqueous solution of thiobarbituric acid, and 500 μL of water. The samples were incubated in boiling water for 60 minutes and were immediately chilled on ice. The fluorescent lipid peroxidation/thiobarbituric acid adduct was measured with a fluorescence spectrophotometer (Tecan Infinite, Tecan Deutschland GmbH, Crailsheim, Germany). Different dilutions of tetraethoxypropane were used as external standards.

NO Detection by the Griess Reaction

We assessed the NO content in the IPRL perfusate after 5 minutes of reperfusion. First, we added 50 μL of the sample, 50 μL of a blank solution, and 50 μL of a standard solution (0.69 g of sodium nitrite; Sigma-Aldrich Chemie, Munich, Germany) to 1 L of water; 2.5 mL of this solution was further diluted to 250 mL with water, and then the solution was transferred into a 96-well microtiter plate. To each well, 50 μL of a 1% sulfanilamide solution was added, and this was followed by 50 μL of an N-(1-naphthyl)-ethylenediamine solution. Thereafter, the plates were incubated for 30 minutes at room temperature, and the absorbance was measured at 540 nm with a multi-titer plate reader (Tecan Infinite 200, Tecan Group, Ltd., Männersdorf, Switzerland).

A standard curve with a regression coefficient and a slope was set and used for the determination of the perfusate NO concentration. According to the Griess reaction, sulfanilamide and N-(1-naphthyl)-ethylenediamine dihydrochloride were used to detect nitrite (NOmath image), which is 1 of the 2 primary stable and nonvolatile breakdown products of NO.

Hepatic Oxygen Uptake

The functional recovery of the livers was estimated by the assessment of the hepatic oxygen uptake. Perfusate samples were taken from the portal inflow and the venous effluent, and the oxygen content was measured with the ABL 5 blood gas analyzer. The oxygen uptake of the liver was determined as the difference between the portal and venous sites and was expressed as microliters per gram of liver weight per minute.

Electron Microscopy

Tissue samples were immersed in a 2% glutaraldehyde and paraformaldehyde solution in phosphate-buffered saline. The samples were sliced into 0.5-mm pieces, which were postfixed with osmium tetroxide (an osmic acid fixative) and embedded in Epon 812 (Serva, Heidelberg, Germany). Semithin sections were then stained as described previously.22 Thin sections were thereafter stained with uranyl acetate and lead citrate and were examined with an electron microscope (EM 400 T/ST, Philips, Amsterdam, the Netherlands) at the electron microscopy core facility (Department of Pathology, University Hospital Aachen, Aachen, Germany).

Statistical Analysis

The results are expressed as means and standard errors of the mean for each group. The statistical analyses of the groups at each time point were tested with a 2-way analysis of variance and Bonferroni's post hoc test. A P value < 0.05 was considered statistically significant. Calculations were made with Prism 5.0b for Macintosh (GraphPad Software, Inc., San Diego, CA).


Enzyme Release

The release of the following enzymes was measured as a general parameter of hepatocellular damage.


The release of ALT after 45 minutes of reperfusion was significantly higher for the WI group (78.2 ± 14.6 U/L) versus the VSOP-NO group (10.2 ± 0.2 U/L; Fig. 1). In comparison with the NHBD group, the VSOP group (15.8 ± 0.4 U/L), the VSOP-NO group, and the control group (10.2 ± 0.2 U/L) showed statistically significant lower levels (P < 0.001)

Figure 1.

Release of ALT as a general parameter of parenchymal injury upon reperfusion. **P < 0.001 versus the WI group (n = 5 per group).


GLDH release (Fig. 2) was measured as a parameter of severe hepatic injury and mitochondrial damage. The release of this enzyme was measured in the effluent, which was collected at different times. The results showed higher values for the WI group (18.2 ± 4.9 U/L) versus the VSOP-NO group (4.0 ± 0.7 U/L) and the control group (3.1 ± 0.6 U/L, P < 0.001). Only the VSOP-NO group did not differ from the control group.

Figure 2.

Release of GLDH as a parameter of major hepatic injury at the mitochondrial level upon reperfusion. *P < 0.001 versus the WI group (n = 5 per group).

Lipid Peroxidation

The MDA levels in tissue were measured to estimate the damage to the liver parenchyma due to oxygen free radical activity (Fig. 3). The MDA level was higher in the WI group (31.3 ± 5.3 nmol/mL) versus all other groups. Statistical significance was observed only for the WI group versus the VSOP-NO group (P < 0.05).

Figure 3.

Release of MDA as a parameter of lipid peroxidation due to oxygen free radicals. *P < 0.05 versus the WI group (n = 5 per group).


NO was measured after 5 minutes of reperfusion to estimate the amounts of endogenous and exogenous NO present in the grafts (Fig. 4). The NO values were highest in the VSOP-NO group (1.5 ± 0.3 μmol/L) and lowest in the control group (0.5 ± 0.0 μmol/L, P < 0.05).

Figure 4.

NO in the venous effluent after 5 minutes of reperfusion. The Griess reaction was used to measure the amount of NO in the effluent. *P < 0.05 versus the control (n = 5 per group).


The WI group had a higher PVP throughout the postischemic reperfusion phase, and this reached significance in comparison with the VSOP-NO group after 45 minutes of reperfusion (21.7 ± 0.2 mm Hg for the WI group versus 12.2 ± 0.8 mm Hg for the VSOP-NO group, P < 0.05; Fig. 5).

Figure 5.

The time course of PVP was measured as an index for total vascular resistance upon reperfusion. *P < 0.05 versus the WI group (n = 5 per group).

Electron Microscopy

The ultrastructure of the liver parenchyma (Fig. 6) was evaluated with electron microscopy. The main focus of the study was the mitochondria of the hepatocytes.

Figure 6.

Ultrastructural analysis with electron microscopy. The liver sections show nuclei (N), mitochondria (M), smooth endoplasmic reticulum (S), rough endoplasmic reticulum (R), vacuole (V) and lipid droplets (L) in the different study groups (magnification ×22.000).

Figure 6 depicts liver sections before and after reperfusion for each group. After CS (before reperfusion) and after reperfusion, the WI group showed a low electron density and extensive damage due to vacuolization throughout the graft. The WI group also showed mitochondrial swelling and other organelle damage, whereas the control group without WI showed intact mitochondria and other cell organelles both before and after reperfusion.

In the VSOP group, only slight swelling of mitochondria in the postreperfusion phase was observed. The VSOP-NO group showed nearly full, intact mitochondrial structures and cell organelles that were comparable to those of the control group. The electron density was significantly higher in this case, and the structures were better preserved. These findings were in line with the data on enzyme release and lipid peroxidation.


The universal shortage of donor organs has necessitated the expansion of the donor pool through the consideration of organs from less than ideal donors.2, 4, 23 In liver transplantation, steatotic livers account for almost 30% of the marginal livers that are procured,13 and NHBD livers are increasingly being employed for transplantation. However, the incidence of PDF and PNF is more frequent with NHBD grafts versus heart-beating donor grafts.3, 24-26 A possible strategy for reducing PDF and PNF is the improvement of preservation techniques. Currently, CS with University of Wisconsin or HTK solution is the clinical gold standard. It is well established that VSOP, when it is applied to kidney and liver grafts, can improve the preservation quality and thereby increase graft viability.27-32

However, the homogeneous distribution of oxygen in the tissue of WI-damaged grafts has not been reported.33 VSOP alone has demonstrated inferior results for the preservation of ischemically damaged grafts because of the generation of oxygen free radicals.31 The oxygen distribution in WI-damaged grafts can be improved by the addition of antioxidants or vasodilators (eg, NAC) to the rinse solution.31

In this study, NAC was added to the rinse solution and NO was added to persufflated oxygen during the CS of WI-damaged rat livers to expand the VSOP technique. After 24 hours of VSOP supplemented with NO, we assessed the preservation quality with the time-honored IPRL model.

We demonstrated that the exposure of WI-damaged livers to VSOP with 40 ppm NO during CS resulted in lower PVP values, lower levels of ALT and GLDH, reduced lipid peroxidation, and well-preserved structural integrity.

It has previously been reported that the antioxidative effects of NO can protect hepatocytes during periods of WI.13, 15, 16 In our study, the NO levels were higher in the VSOP-NO group, and this could, at least in part, account for the better condition of the VSOP-NO–treated grafts.

WI is known to cause a rapid deterioration of graft viability; this is further aggravated by reperfusion and leads to PDF or PNF.34 Both events lead to an increase in the production of oxygen-derived free radicals, which cause the displacement of electrons from the lipid molecules in the cell membrane and lead to cellular injury and ultimately cell death. Therefore, we assessed the effluent level of MDA as a marker of lipid peroxidation caused by oxygen-derived free radicals. Our results showed a greater than 50% reduction of MDA levels in WI-damaged liver grafts exposed to VSOP-NO versus CS alone. This observation, in conjunction with the lower aminotransferase and mitochondrial enzyme release levels, supports the hypothesis that oxygen-derived free radicals are one of the main sources of WI-related graft injury.

In clinical practice, a high PVP is known to lead to PDF because of the destructive effect on sinusoidal endothelial cells, which leads to an insufficient supply of oxygen to the hepatocytes.35 PVP was reduced by approximately 30% to 40% at the end of reperfusion in the VSOP-NO–treated liver grafts versus the cold-stored livers. The lower PVP of the VSOP-NO–treated livers could have been related to the vasodilative properties of NO or to well-preserved endothelial integrity, which may have prevented vascular occlusion.14, 36

The WI-damaged liver grafts that were cold-stored in HTK also showed higher levels of ALT and GLDH release upon reperfusion. The higher amounts of the mitochondrial enzyme GLDH in the effluent were reflective of mitochondrial damage, which led to the impairment of oxidative phosphorylation (this is responsible for the supply of energy necessary for restoring substrates after the ischemic period).37 The electron microscopy examination confirmed on a visual level the higher release of GLDH in the untreated livers and showed an overall decrease in mitochondrial integrity.

We are aware of the immunological and cytochemical shortcomings of the IPRL model with respect to transplantation (eg, the possible impact of leukocyte or thrombocyte interactions). However, we can conclude from the results that VSOP and NO represent a feasible combination of medical gases for increasing graft viability after preservation, and this may point to a novel and safe approach to more effective preservation of WI-damaged liver grafts. In conclusion, the treatment of WI-damaged livers with VSOP and NO during 24 hours of CS in HTK leads to improved graft quality and structural integrity.


The authors thank Pascal Paschenda, Mareike Schulz, and Martyna Wojcieszak for their skillful technical assistance.