Cytoprotective function oftetrahydrobiopterin in rat liverischemia/reperfusion injuryYuzuru Hara, MD, Kenichi Teramoto, MD, PhD, Kozo Ishidate, PhD, and Shigeki Arii, MD, PhD,Tokyo, Japan

Background. Tetrahydrobiopterin (BH4) is a key coenzyme of nitric oxide synthase (NOS), which isassociated with a cytoprotective function in various ischemia-reperfusion (I/R) injury models. Therehave been a few reports on the efficacy of BH4 in the treatment of I/R injury in other organs; therefore,the aim of this study was to investigate the effect of BH4 related with NOS reaction in hepatic I/R injury.Methods. A model of 70% liver I/R injury with a 100-minute ischemic time was created in rats, and thenon--ischemic lobes were then resected. The rats were given BH4 (BH4 group) or saline solution (salinegroup) before reperfusion. The specific inducible NOS blocker 1400W was used to evaluate the effect ofendogenous inducible NOS in the I/R hepatic injury. Survival, nitric oxide products (nitrate and ni-trite), NOS expression, and nitrotyrosine (ie, the peroxynitrite product) were measured after reperfusion.Results. On day 7, the survival rate was 62.5% in the BH4 group, as opposed to 14.3% in the salinegroup (P = .0004); 1400W administration to the BH4 group decreased the survival rate to 0% (P =.003). BH4 prevented the significant increase in total bilirubin levels (P < .01) after 12-hour reper-fusion. The increases in serum alanine transaminase levels (after 3 hours and 12 hours of reperfusion)were significantly (P < .01) attenuated in the BH4 group. BH4 increased the nitrate/nitrite concen-trations in liver tissue (P < .05) and reduced nitrotyrosine production, and the protein assay showedthat BH4 increased inducible NOS and endothelial NOS expression. Histologic examination of the liverrevealed that BH4 mitigated the damage that was caused by liver I/R.Conclusion. Exogenous BH4 increased nitric oxide production, which attenuated the hepatic I/R injury.(Surgery 2006;139:377-84.)

From the Department of Surgery, School of Medicine, and Department of Molecular Cell Biology,Medical Research Institute, Tokyo Medical and Dental University, Tokyo, Japan

SURGERY 377

ISCHEMIA/REPERFUSION (I/R) INJURY of the liver is aserious postoperative complication of liver surgery,especially in patients with liver cirrhosis and/orfatty liver. Reperfusion has been shown sometimesto affect hepatic function adversely because it in-duces the release of reactive oxygen species from en-dothelium and neutrophils, and activation ofKupffer cells. Both of these induce inflammatorymediators, such as tumor necrosis factor-alpha andinterleukin 1, which cause increased vascular perme-ability and accumulation of activated neutrophils.1-3

Accepted for publication August 19, 2005.

Supported in part by a Grant-in-Aid from the Ministry of Educa-tion, Science, Sports and Culture of Japan.

Reprint requests: Kenichi Teramoto, MD, PhD, Department ofHepato-Biliary-Pancreatic Surgery, Tokyo Medical and DentalUniversity, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan.E-mail: teraken.srg1@tmd.ac.jp.

0039-6060/$ - see front matter

� 2006 Mosby, Inc. All rights reserved.

doi:10.1016/j.surg.2005.08.019

Nitric oxide (NO) has been the focus of a greatdeal of attention because it is a substance that isrelated to I/R injury, and 3 isoforms of NO synthase(NOS) have been identified as responsible for thegeneration of NO. Two are constitutive isoforms(ie, neuronal NOS and endothelial NOS [eNOS]),and the other is an inducible form (iNOS). NO thatis produced in reactions that are catalyzed by eNOShas been thought to have a cytoprotective effect,because selective eNOS inhibitors exacerbate I/Rhepatic injury.4-6 In I/R hepatic injury, NO hasbeen reported to have a vasodilator effect, an anti-platelet effect, and a neutrophil-activation-inhibit-ing effect.7-11 On the other hand, NO also has acytotoxic effect that causes lipid peroxidation,peroxynitrite formation, and inhibition of mito-chondrial respiration, especially when present inexcessive amounts.12,13 Because it has these oppos-ing effects, the contribution of NO to liver damageis still a matter of controversy.

The reduced form of nicotinamide adeninedinucleotide phosphate and the bound cofactors

SurgeryMarch 2006

378 Hara et al

flavin mononucleotide, flavin adenine dinucleotide,tetrahydrobiopterin (BH4), heme, and iron ions areall known to be essential electron carriers in the re-action that is catalyzed by NOS, and the depletion ofBH4 results in a dramatic decrease in NO synthesisin reactions that are catalyzed by both the eNOSand the iNOS isoforms in endothelial cells andblood vessels.14 Sapropterin hydrochloride is ahydrochloride salt of L-erythro-tetrahydrobiop-terin, an active analogue of BH4, and has beenused clinically to treat patients with hereditary hy-perphenylalaninemia who have BH4 deficiency.15

BH4 has been reported to be the rate-limitingfactor in NO synthesis in several different cellsystems, which include murine macrophages,16

human umbilical vein endothelial cells,17 and ratvascular smooth muscle cells.18 The relationship be-tween NO synthesis and BH4 has also been charac-terized,19 and BH4 has been shown to be requirednot only for the activity of NOS itself, but for the ex-pression of the NOS gene in human mesangial cells.Exogenous BH4 has also been reported to restorethe impaired endothelial NO synthesis in acute en-dothelial dysfunction.20 In addition to these find-ings, reactions that are catalyzed by BH4-free NOShave been reported to form peroxynitrite, a strongbiologic oxidant.21-24 In some studies, BH4 hasbeen used to treat I/R injury in several organs, butnot the liver, and it was been shown to be effec-tive.25-28 However, the role of BH4 in I/R injury,especially in relation to NO, is still unclear.

In the present study, we investigated the role ofBH4 in hepatic I/R injury in a rat model of liver ische-mia/reperfusion by measuring survival and changesin NOS and NO metabolites after reperfusion.

MATERIAL AND METHODS

Animals. This study was performed on maleSprague-Dawley rats that weighed 200 to 250 g(Sankyo Labo Service, Tokyo, Japan). The ratswere housed individually in cages with wood chipbedding and were given free access to standard ratfood and water. The animal room was illuminatedon a 12-hour light-dark cycle. Room temperaturewas maintained at 18 to 22�C, and humidity wasmaintained at 60% to 70%. All procedures werecarried out in accordance with the guidelines of theAnimal Experimentation Ethics Committee of theTokyo Medical and Dental University for animalexperiments.

Chemicals and reagents. All chemicals and re-gents were purchased from Sigma Chemical Com-pany (St. Louis, Mo), unless otherwise noted. The1400W (a specific iNOS blocker) was obtained

from Calbiochem (San Diego, Calif). L-erythro-tetrahydrobiopterin dihydrochloride was donated bySuntory Bio-Pharma Tech Center (Tokyo, Japan).The protein assay reagents were from Bio-Rad(Hercules, Calif). Enzyme-linked immunosorbentkits for nitrate/nitrite and monoclonal nitrotyro-sine antibody were from Cayman Chemical (AnnArbor, Mich). Polyclonal anti-iNOS antibody, pol-yclonal anti-eNOS antibody, and secondary anti-bodies to each of them were purchased from SantaCruz Biotechnology (Santa Cruz, Calif).

Surgical procedures. All animals (includingcontrols) were anesthetized with ether, ketaminehydrochloride (10 mg/kg, intraperitoneally), andphenobarbital (10 mg/kg, intraperitoneally) andwere placed in the supine position on a heatingpad to maintain physiologic body temperature. Toinduce 70% hepatic ischemia, laparotomy wasperformed, and the blood supply to the right upperlobe and left lobe of the liver was interrupted byplacing a bulldog clamp on the peripheral branchof the hepatic artery and portal vein to each of thelobes. The vessels to the other 3 lobes were leftintact to prevent blockage of intestinal outflow.Reperfusion was achieved by removing the clamps.Then the shunting lobes of the caudate lobe andthe right lower lobes were resected to eliminate theinfluence of the non-I/R lobes. All animals wererandomized for group assignment as described.

Study 1: Survival analysis. To study the effect ofBH4 on postischemic liver injury, a 7-day survivalcurve was plotted for 4 experimental groups (Fig1): group 1 (saline group; n = 21): the animalswere subjected to anesthesia, laparotomy, 100 min-utes of 70% hepatic ischemia, and an intravenousinjection of 0.4 mL of saline solution 30 minutesbefore reperfusion; group 2 (BH4 group; n = 24):the animals were subjected to anesthesia, laparot-omy, 100 minutes of 70% hepatic ischemia, andan intravenous injection with BH4, 20 mg/kg/0.4 mL, 30 minutes before reperfusion; group 3(I/R + 1400W group; n = 24): the animals were sub-jected to anesthesia, laparotomy, 100 minutes of 70%hepatic ischemia, subcutaneous injection with1400W, 5 mg/rat, just before ischemia, and intrave-nous saline solution injection 30 minutes beforereperfusion; group 4 (I/R + 1400W + BH4 group;n = 24): the animals were subjected to anesthesia,laparotomy, 100 minutes of 70% hepatic ischemia,injection with 1400W just before ischemia, and in-jection with BH4 30 minutes before reperfusion.

Study 2: Biochemical determinations. For plasmasampling, other groups of rats were used for thebiochemical analysis (Fig 1). The animals were sub-jected to 100 minutes of 70% hepatic ischemia and

SurgeryVolume 139, Number 3

Hara et al 379

intravenous injection with 0.4 mL of saline solutionor BH4, 20 mg/kg/0.4 mL, 30 minutes before re-perfusion. Blood and liver samples were collectedjust before reperfusion (control group, n = 9),1 hour after I/R (BH4 group and saline group,both n = 5), 3 hours (n = 5, each), and after 12hours (n = 5, each). The blood was collectedfrom the inferior vena cava, centrifuged to obtainplasma, and used to determine alanine transami-nase (ALT) and total bilirubin levels.

Study 3: Assay for nitrate/nitrite in I/R liver tissue.As shown in Fig 1, liver samples were collected at 0,1, 3, or 12 hours after reperfusion according to thegroup (n = 4). The samples were homogenizedimmediately in phosphate-buffered saline solution(pH 7.4) and centrifuged at 10,000g for 20 minutes.The supernatant was ultracentrifuged at 100,000gfor 15 minutes, and the supernatant was passedthrough a 0.45-lm filter. After ultrafiltration of thefiltrate through a 30-kd molecular weight cut-offfilter (Millipore, Billerica, Mass), the sample thatwas collected in the filter was assayed by enzyme-linked immunosorbent assay for nitrate and nitritelevels (Cayman Chemical, Ann Arbor, Mich).

Study 4: Western blotting analysis for NOS expressionand nitrotyrosine. As shown in Fig 1, 6 hours after re-perfusion, liver samples were collected from eachgroup (n = 4) and frozen immediately in liquid

Fig 1. Experimental protocol: I/R injury. The rats weresubjected to 100 minutes of ischemia (I) followed byreperfusion (R). Control, Blood was sampled 70 minutesafter 70% hepatic ischemia (pre-reperfusion); Group 1(saline group), saline solution, 0.4 mL, was administeredintravenously 30 minutes before reperfusion; Group 2,(BH4 group), BH4, 20 mg/kg in 0.4 mL, was administeredintravenously 30 minutes before reperfusion; Group 3,ischemia with the administration of 1400W, and a sa-line injection was given 30 minutes before reperfusion;Group 4, ischemia with the administration of 1400W,and the BH4 was administered 30 minutes beforereperfusion.

nitrogen. The samples were homogenized in 3volumes of a 50-mmol/L Tris-HCl extraction buffer(pH 7.4) that contained 0.5 mmol/L EDTA(pH 7.4), 0.5 mmol/L EGTA (pH 7.4), 1 mmol/L1,4-dithiothreitol plus various protease inhibitors(1 lmol/L leupeptin) and 100 lmol/L p-amidino-phenyl-methanesulfonyl fluoride hydrochloride.After centrifugation at 15,000 rpm for 10 minutesat 4�C, the supernatant was centrifuged at 15,000rpm for 60 minutes at 4�C. The supernatant wascollected, and the protein concentration was mea-sured by the Bradford protein assay (Bio Rad).The liver homogenated samples were aliquotedinto 100-lg portions in 62.5 mmol/L Tris-HClbuffer (pH 6.8) contained 25% glycerol, 2%sodium dodecylsulfate, and 5% b-mercaptoetha-nol, then heat denatured at 100�C for 1 minute. Af-ter loading the samples onto 7.5% to 20% gradientsodium dodecylsulfate–polyacrylamide gels (BioRad), the samples were electrophoresed for 3 hoursat 80 V. The loaded proteins were then electroblot-ted onto nitrocellulose membranes (AmershamBiosciences, Piscataway, NJ). After the membraneswere blocked, rabbit polyclonal antiserum (1:500dilution) directed against iNOS or eNOS was in-cubated with the membranes for 1 hour at roomtemperature. The membranes were incubated for1 hour at room temperature with a horseradish per-oxidase–conjugated goat anti-rabbit secondary an-tibody (1:1,000 dilution) and then with luminolreagent (Santa Cruz Biotechnology). The mem-brane signal was detected on photographic film af-ter approximately a 10-minute exposure. To extractof eNOS protein, 20 mmol/L of 3-[(3-chloramido-propyl) dimethylammonio] propanesulfonic acidwas added to the extraction buffer. The Westernblot analysis for the iNOS and eNOS protein signalswere measured with gel analysis software (NIHImage; National Institutes of Health, Bethesda,Md). The Western blot analysis for nitrotyrosinewas performed with a nitrotyrosine monoclonalantibody (1:250 dilution) and an horseradish per-oxidase–conjugated goat anti-mouse secondaryantibody (1:1,000 dilution; Bio-Rad). All of the rel-ative intensities of the protein were corrected forprotein expression of beta-actin.

Study 5: Histology. Liver samples were also col-lected for histologic examination. After the spec-imens were fixed in 10% neutral buffered formalinand embedded in paraffin, they were cut into 5-lmsections, which were mounted onto Silane-coatedslides and stained with hematoxylin-eosin, accord-ing to standard procedures.

Statistical analysis. All values are expressed asmeans ± SD. The statistical significance of the

SurgeryMarch 2006

380 Hara et al

differences was evaluated by the Peto-Pike test andthe Kaplan-Meier log rank test (study 1) and a fac-torial analysis of variance, followed by the Tukey-Kramer test (studies 2, 3, and 4). The Welch’s t-testwas used to analyze the effects of BH4 on the hepaticlevels of iNOS and eNOS. An associated probabilityof <.05 was considered significant.

RESULTS

Survival after liver I/R. On day 7 the survivalrate was 14.3% in group 1 (saline group), and itwas significantly (P = .0004) higher, 62.5%, ingroup 2 (BH4 group; Fig 2). The effect of 1400W,a specific inhibitor of iNOS, was examined to de-termine whether iNOS-mediated NO played a rolein the increase in survival caused by BH4. As shownin Fig 2, the administration of 1400W before the is-chemia significantly (P = .0003) lowered the highersurvival rate induced by BH4 (group 4 vs 2), whichsuggested that iNOS activity was involved in theeffect of BH4 on survival after liver I/R injury.

Biochemical data. As shown in Fig 3, the ele-vated plasma levels of both ALT (at 3 and 12hours) and total bilirubin (at 12 hours) were atten-uated significantly (P < .01) in group 2 (BH4

group), and the attenuations of both groups sup-ported a protective role of BH4 against liverI/R injury.

Fig 2. Survival rate after hepatic I/R. Male Sprague-Dawley rats were subjected to 100 minutes of 70% hepa-tic ischemia and then injected intravenously with salinesolution (group 1; n = 21) or BH4 (group 2; n = 24)30 minutes before reperfusion. In groups 3 and 4 (n =24), 1400W (5 mg/body) was administered subcutane-ously before the ischemia. Differences among the groupswere evaluated for statistical significance by the Peto-Pike test, followed by the Kaplan-Meier log rank test.*P = .0004 (group 2 vs group 1); *P # .0001 (group 2vs group 3); *P = .0003 (group 2 vs group 4); **P =.0039 (group 3 vs group 4). On day 7, the survival rateof group 1 was 14.3% and was significantly higher (62.5%)in group 2. The administration of 1400W, a specificinhibitor of iNOS, before ischemia caused a significantdecrease in the survival rate prolonged by BH4.

Nitrate and nitrite production. To confirm theup-regulation of NO production by exogenous BH4

administration, the sum of both the nitrate andnitrite levels was measured. The tissue levels ofNO2

� and NO3� at 1 and 3 hours after reperfusion

were increased by the BH4 treatment (P < .05; Fig 4).Western blotting for iNOS and eNOS. Liver

samples that had been collected from all

Fig 3. Effect of BH4 on liver function after liver I/R. Afactorial analysis of variance followed by the Tukey-Kramer test revealed statistically significant differencesin the ALT levels in group 2 (BH4 group), 3 to 12 hoursafter reperfusion (*P < .01). After 12 hours of reperfu-sion, in the BH4 group (group 2; total bilirubin level[T-Bil], 1.6 ± 1.1 mg/dL) prevented the significantincrease in the total bilirubin levels, compared with thesaline group (group 1; total bilirubin level, 4.3 ± 1.8 mg/dL; **P # .01). The values are expressed as means ±SD (n = 5, each in groups 1 and 2 and n = 9 in thecontrol group).

Fig 4. NO production after exogenous BH4 administra-tion. The sum of nitrate and nitrite levels was measured.At 1 and 3 hours after reperfusion, the tissue levels ofNO2

� and NO3� were increased significantly by BH4 in

group 2 (BH4 group; P < .05). The statistical procedureswere the same as in Fig 3.

SurgeryVolume 139, Number 3

Hara et al 381

Fig 5. Immunoblotting for iNOS and eNOS in groups 1 (saline group) and 2 (BH4 group). Low levels ofiNOS protein were observed in group 1 (saline group), whereas clear expression of iNOS protein wasdetected in group 2 (BH4 group; *P = .00008). eNOS protein was hardly detectable in group 1 (salinegroup) but was present in high levels in group 2 (BH4 group; **P = .0008). The liver samples were col-lected at 6 hours after reperfusion, and the protein concentrations of samples are 20 lg per lane. Thevalues are expressed as means ± SD, with 4 rats in each group. Welch’s t-test was used to analyze the effectsof BH4 on the hepatic levels of iNOS and eNOS.

SurgeryMarch 2006

382 Hara et al

experimental groups 6 hours after reperfusionwere subjected to sodium dodecylsulfate-polyacry-lamide gel electrophoresis followed by Westernblotting analysis for iNOS and eNOS (Fig 5). TheiNOS and eNOS protein levels were very low ingroup 1 (saline group) and significantly higherin group 2 (BH4 group; P = .00008 for iNOS andP = .0008 for eNOS).

Western blot analysis of I/R liver for nitro-tyrosine. To confirm the production of nitrotyro-sine, which is the final reaction product ofperoxynitrite in I/R liver tissue, each sample wassubjected to Western blotting analysis. The tissuelevel of nitrotyrosine at 1 and 3 hours afterreperfusion was increased significantly in theabsence of BH4 administration (P < .05; Fig 6).

Liver histologic condition. Sections of liverspecimens that were collected from group 1 (salinegroup) and group 2 (BH4 group) at 12 hours afterreperfusion were examined histologically; repre-sentative sections are shown in Fig 7. Massivenecrosis with atrophic changes in the remaininghepatic cord was observed in I/R-injured liver ingroup 1 (saline group), and a small number ofhepatocytes remained in the portal area (Fig7, A1). Massive infiltration by granulocytes aroundvessels and large hemorrhages also were observedin the I/R-injured liver in group 1 (saline group;Fig 7, A2). By contrast, hepatocyte swelling and

Fig 6. Western blotting analysis for nitrotyrosine. At1 and 3 hours after reperfusion, the tissue level of nitro-tyrosine (the final reaction product of peroxynitrite in I/R liver tissue) was increased significantly in the absenceof administration BH4 (P < .05). The protein concentra-tions of all samples are 20 lg per lane. The values areexpressed as means ± SD, with 4 rats in each group.The statistical procedures were the same as in Fig 3.

necrosis were observed in a few areas of the liversample from group 2 (BH4 group; Fig 7, B1 and 2).

DISCUSSION

This is first study to show that the administra-tion of BH4, a rate-limiting coenzyme of NOS, is ca-pable of improving hepatic I/R injury. AlthoughBH4 has been reported to be effective against inrenal, gastric, and pulmonary I/R injury25-28, theprecise relationship between BH4 and NOS hadnever been investigated.

BH4 significantly lowered the plasma ALT level(Fig 3), and the histologic examination showedthat massive amounts of necrotic tissue occupiedthe entire liver 12 hours after the I/R injury ingroup 1 (saline group), whereas only focal hepato-cyte necrosis was seen in group 2 (BH4 group;Fig 7). Survival also was improved significantlyby BH4 treatment. These biologic benefits wereobserved in the presence of NOS expression(Fig 5). The beneficial effect of BH4 was abolishedcompletely by a specific functional inhibitor ofiNOS, 1400W,29,30 which suggests that the NOthat is produced by iNOS has a beneficial effecton the outcome of I/R injury (Fig 2).

Fig 7. Histologic examination of the liver 12 hours afterhepatic I/R. Group 1 (saline group), A1 (3 100) and A2(3 400): massive necrosis and atrophic changes in the ofremaining hepatic cord (black arrows) occupied the en-tire liver bed; large hemorrhages (white arrowheads) andmassive granulocyte infiltration around the vessels werealso observed (black arrowheads). The white arrows pointto the few areas of remaining hepatocytes. Group 2(BH4 group), B1 (3 100) & B2 (3 400): hepatocyte swell-ing (white arrowheads) and necrosis (white arrows) wereobserved in a few areas (hematoxylin and eosin stain).

SurgeryVolume 139, Number 3

Hara et al 383

BH4 is a critical cofactor because its presence isrequired for NOS to dimerize.31 The BH4-satu-rated subunit of NOS generates NO, and theBH4-free subunit generates superoxide and the ni-troxyl anion, which results in the production ofperoxynitrite and is known to have a cytotoxiceffect and induce apoptosis.22,31-34 In the ischemicacute renal failure model, on the other hand,BH4 exhaustion of ischemic kidney was speculatedon the basis of the Western blot results afterlow-temperature sodium dodecylsulfate-polyacryla-mide gel electrophoresis.26 Although more precisestudy of the mechanisms that are responsible forthe effects of BH4 in liver I/R injury is required,the results of the present study suggested thatBH4 may have been exhausted in the I/R-injuredliver in group 1 (saline group). Thus, the superox-ide and nitroxyl anions that were generated in thecells instead of NO may have caused the severenecrotic damage in the liver, and the administra-tion of BH4 may have mitigated the I/R hepaticinjury by dimerizing NOS and generating NO.

We examined Western blots for iNOS and eNOSin our model to assess whether their expressionwas altered during liver I/R injury (Fig 5). The re-sults showed that the expression of both iNOS andeNOS was increased in group 2 (BH4 group) com-pared with group 1 (saline group; Fig 5) and thatnitrate/nitrite production was increased by BH4

administration (Fig 4). Fig 6 shows a decrease innitrotyrosine in group 2 (BH4 group), which sug-gests a decrease in peroxynitrite, which has cyto-toxic effects. Taken together, these findings suggestthat BH4 improved I/R hepatic injury because itresulted in increased NO production and de-creased peroxynitrite production.

The role of NO in I/R hepatic injury is still amatter of controversy. Many studies have investi-gated the role of NO that is produced by iNOS byvarious methods of iNOS inhibition, such as iniNOS knock-out mice or by the use of the iNOS-specific inhibitors ONO-1714, aminoguanidine,and L-N-lysine hydrochloride,5,35-42 but the modelsof I/R injury have differed slightly from study tostudy. Some studies found that the inhibition ofiNOS prevented I/R hepatic injury,5,35-39 but asmaller number of investigators have reportedfinding that iNOS inhibition did not affect orprevent I/R hepatic injury.40-42 However, most ofthe arguments about the data that have beenobtained for NO and iNOS did not take BH4

into account. Instead of NO production, the mon-omer NOS enzyme system in the absence of BH4

catalyzes the production of cytotoxic peroxynitrite.Therefore, further study of BH4 tissue

concentrations appears necessary to elucidate theeffects of NO in the I/R injury model.

In summary, the results of the present studystrongly suggest that BH4 supplementation miti-gated the I/R injury. The effectiveness of BH4

administration depends on NOS production.Exogenous BH4 may be a useful agent for theimprovement of liver I/R injury.

The authors thank Mrs Keiko Gomisawa, Departmentof Surgery, Tokyo Medical and Dental University, forher technical assistance and Suntory Bio-Pharma TechCenter, Tokyo, Japan, for their gift of sapropterinhydrochloride.

REFERENCES

1. Mosher B, Dean R, Harkema J, Remick D, Palma J, CrockettE. Inhibition of Kupffer cells reduced CXC chemokine pro-duction and liver injury. J Surg Res 2001;99:201-10.

2. Peralta C, Prats N, Xaus C, Gelpı E, Rosello-Catafau J.Protective effect of liver ischemic preconditioning on liverand lung injury induced by hepatic ischemia-reperfusionin the rat. Hepatology 1999;30:1481-9.

3. Wanner GA, Ertel W, Muller P, Hofer Y, Leiderer R, MengerMD, et al. Liver ischemia and reperfusion induces a sys-temic inflammatory response through Kupffer cell activa-tion. Shock 1996;5:34-40.

4. Pannen BH, Al-Adili F, Bauer M, Clemens MG, Geiger KK.Role of endothelins and nitric oxide in hepatic reperfusioninjury in the rat. Hepatology 1998;27:755-64.

5. Isobe M, Katsuramaki T, Hirata K, Kimura H, Nagayama M,Matsuno T. Beneficial effects of inducible nitric oxidesynthase inhibitor on reperfusion injury in the pig liver.Transplantation 1999;68:803-13.

6. Liu P, Xu B, Spokas E, Lai PS, Wong PY. Role of endogenousnitric oxide in TNF-alpha and IL-1beta generation in hepa-tic ischemia-reperfusion. Shock 2000;13:217-23.

7. Palmer RM, Ferrige AG, Moncada S. Nitric oxide releaseaccounts for the biological activity of endothelium-derivedrelaxing factor. Nature 1987;327:524-6.

8. Radomski MW, Palmer RM, Moncada S. Comparative phar-macology of endothelium-derived relaxing factor, nitricoxide and prostacyclin in platelets. Br J Pharmacol 1987;92:181-7.

9. Furlong B, Henderson AH, Lewis MJ, Smith JA. Endothe-lium-derived relaxing factor inhibits in vitro platelet aggre-gation. Br J Pharmacol 1987;90:687-92.

10. Lefer AM, Tsao PS, Lefer DJ, Ma XL. Role of endothelialdysfunction in the pathogenesis of reperfusion injury aftermyocardial ischemia. FASEB J 1991;5:2029-34.

11. Kurose I, Wolf R, Grisham MB, Granger DN. Modulation ofischemia/reperfusion-induced microvascular dysfunctionby nitric oxide. Circ Res 1994;74:376-82.

12. Beckman JS, Beckman TW, Chen J, Marshall PA, FreemanBA. Apparent hydroxyl radical production by peroxyni-trite: implications for endothelial injury from nitricoxide and superoxide. Proc Natl Acad Sci USA 1990;87:1620-4.

13. Haddad IY, Pataki G, Hu P, Galliani C, Beckman JS, MatalonS. Quantitation of nitrotyrosine levels in lung sections ofpatients and animals with acute lung injury. J Clin Invest1994;94:2407-13.

SurgeryMarch 2006

384 Hara et al

14. Kinoshita H, Milstien S, Wambi C, Katusic ZS. Inhibition oftetrahydrobiopterin biosynthesis impairs endothelium-de-pendent relaxations in canine basilar artery. Am J Physiol1997;273:H718-24.

15. Kaufman S. Unsolved problems in diagnosis and therapy ofhyperphenylalaninemia caused by defects in tetrahydro-biopterin metabolism. J Pediatr 1986;109:572-8.

16. Sakai N, Kaufman S, Milstein S. Tetrahydrobiopterin isrequired for cytokine-induced nitric oxide production in amurine macrophage cell line (RAW 264). Mol Pharmacol1993;43:6-10.

17. Werner-Felmayer G, Werner ER, Fuchs D, Hausen A,Reibnegger G, Schmidt K, et al. Pteridine biosynthesisin human endothelial cells: impact on nitric oxide-mediatedformation of cyclic GMP. J Biol Chem 1993;268:1842-6.

18. Gross SS, Levi R. Tetrahydrobiopterin synthesis: anabsolute requirement for cytokine- induced nitric oxidegeneration by vascular smooth muscle. J Biol Chem1992;267:25722-9.

19. Saura M, Perez-Sala D, Canada FJ, Lamas S. Role of tetra-hydrobiopterin availability in the regulation of nitric-oxidesynthase expression in human mesangial cells. J Biol Chem1996;271:14290-5.

20. Tiefenbacher CP, Chilian WM, Mitchell M, DeFily DV.Restoration of endothelium-dependent vasodilation afterreperfusion injury by tetrahydrobiopterin. Circulation1996;94:1423-9.

21. Rusche KM, Spiering MM, Marletta MA. Reactions catalyzedby tetrahydrobiopterin free nitric oxide synthase. Biochem-istry 1998;44:15503-12.

22. Gorren AC, List BM, Schrammel A, Pitters E, Hemmens B,Werner ER, et al. Tetrahydrobiopterin-free neuronal nitricoxide synthase: evidence for two identical highly anti-cooperative pteridine binding sites. Biochemistry 1996;355:16735-45.

23. Mayer B, Pitters E, Pfeiffer S, Kukovetz WR, Schmidt K.A synthetic peptide corresponding to the putative dihydro-folate reductase domain of nitric oxide synthase inhibitsuncoupled NADPH oxidation. Nitric Oxide 1997;1:50-5.

24. Stuehr DJ. Mammalian nitric oxide synthase. BiochimBiophys Acta 1999;1411:217-30.

25. Schmid RA, Hillinger S, Walter R, Zollinger A, Stamm-berger U, Speich R, et al. The nitric oxide synthase cofactortetrahydrobiopterin reduces allograft ischemia-reperfusioninjury after lung transplantation. J Thorac CardiovascSurg 1999;118:726-32.

26. Kakoki M, Hirata Y, Hayakawa H, Suzuki E, Nagata D,Tojo A, et al. Effects of tetrahydrobiopterin on endothelialdysfunction in rats with ischemic acute renal failure. J Am SocNephrol 2000;11:301-9.

27. Hillinger S, Sandera P, Carboni GL, Stammberger U, Zalu-nardo M, Schoedon G, et al. Survival and graft function in alarge animal lung transplant model after 30 h preservation

and substitution of the nitric oxide pathway. Eur J Cardio-thorac Surg 2000;20:508-13.

28. Ishii M, Shimizu S, Nawata S, Kiuchi Y, Yamamoto T.Involvement of reactive oxygen species and nitric oxide ingastric ischemia-reperfusion injury in rats: protective effectof tetrahydrobiopterin. Dig Dis Sci 2000;45:93-8.

29. Garvey EP, Oplinger JA, Furfine ES, Kiff RJ, Laszlo F,Whittlei BJ, et al. 1400W is a slow, tight binding, and highlyselective inhibitor of inducible nitric-oxide synthase in vitroand in vivo. J Biol Chem 1997;272:4959-63.

30. Boer R, Ulrich WR, Klein T, Mirau B, Haas S, Baur I. Theinhibitory potency and selectivity of arginine substrate sitenitric-oxide syntherase inhibitors is solely determined bytheir affinity toward the different isoenzymes. Mol Pharma-col 2000;58:1026-34.

31. Stuehr DJ. Structure function aspects in the nitric oxidesynthases. Annu Rev Pharmacol Toxicol 1997;37:339-59.

32. Pou S, Pou WS, Bredt DS, Snyder SH, Rosen GM. Genera-tion of superoxide by purified brain nitric oxide synthase.J Biol Chem 1992;267:24173-6.

33. Mayer B, Klatt P, Werner ER, Schmidt K. Kinetics andmechanism of tetrahydrobiopterin-induced oxidation ofnitric oxide. J Biol Chem 1995;270:655-9.

34. Scott-Burden T. Regulation of nitric oxide production bytetrahydrobiopterin. Circulation 1995;91:248-50.

35. Kimura H, Katsuramaki T, Isobe M, Nagayama M, MeguroM, Kukita K. Role of inducible nitric oxide synthase in pigliver transplantation. J Surg Res 2003;111:28-37.

36. Meguro M, Katsuramaki T, Kimura H, Isobe M, NagayamaM, Kukita K. Apoptosis and necrosis after warm ischemia-reperfusion injury of the pig liver and their inhibition byONO-1714. Transplantation 2003;75:703-10.

37. Lee VG, Johnson ML, Baust J, Laubach VE, Watkins SC,Billiar TR. The roles of iNOS in liver ischemia-reperfusioninjury. Shock 2001;16:355-60.

38. Meguro M, Katsuramaki T, Nagayama M, Kimura H, IsobeM, Kimura Y. A novel inhibitor of inducible nitric oxide syn-thase (ONO-1714) prevents critical warm ischemia-reperfu-sion injury in the pig liver. Transplantation 2002;73:1439-46.

39. Liu TH, Robinson EK, Helmer KS, West SD, Castaneda AA,Chang L. Does upregulation of inducible nitric oxide syn-thase play a role in hepatic injury? Shock 2002;18:549-54.

40. Yagnik GP, Takahashi Y, Tsoulfas G, Reid K, Murase N,Geller DA. Blockade of the L-arginine/NO synthase pathwayworsens hepatic apoptosis and liver transplant preservationinjury. Hepatology 2002;36:573-81.

41. Hsu CM, Wang JS, Liu CH, Chen LW. Kupffer cells protectliver from ischemia-reperfusion injury by an inducible nitricoxide synthase-dependent mechanism. Shock 2002;17:280-5.

42. Rivera-Chavez FA, Toledo-Pereyra LH, Dean RE, Crouch L,Ward PA. Exogenous and endogenous nitric oxide but notiNOS inhibition improves function and survival of ischemi-cally injured livers. J Invest Surg 2001;14:267-73.