Surg Today Jpn J Surg (1994) 24:994-1002 SURC RvToDAv

© Springer-Verlag 1994

The Effect of FK506 on Warm Ischemia and Reperfusion Injury in the Rat Liver

HISAO WAKABAYASHI, YUKIHIKO KARASAWA, SATOSHI TANAKA, YASUTAKA KOKUDO,

First Department of Surgery, Kagawa Medical School, Miki-cho, Kita-gun, Kagawa, 761-07 Japan

and TAKASHI MAEBA

Abstract: The protective effect of FK506 on hepatocytes against ischemia and reperfusion injury was examined by evaluating the following: the high energy phosphorus met- abolism obtained using 31p magnetic resonance spectros- copy (31p-MRS) and the tissue blood flow of the liver in ischemia and the reperfusion process, mitochondrial glutamic oxaloacetic transaminase (m-GOT) and glutamic pyruvic transaminase (GPT), the survival rates of the animals, a histological study and immunohistological staining for intercellular adhesion molecule-1 (ICAM-1) in the liver after ischemia. The rats were treated with FK506 l mg/kg/day i.m. for 4 days before testing. Ischemia was induced by clamping the hepatoduodenal ligament for 30min. In 31p. MRS, the recovery of the hepatic energy status after ischemia, evaluated by ~-ATP/inorganic phosphate (Pi), was signi- ficantly better in the FK506 group. It also coincided with the recovery of tissue blood flow monitored with a laser Doppler flowmeter. In the histological examination, the congestion observed in the periportal region of the control group was mild, while there was less induction of ICAM-1 in the endo- thelial cells of the portal veins and hepatic veins in the FK506 group. From these findings, we concluded that FK506 had a protective effect on hepatocytes against warm ischemia and reperfusion injury, and the mechanism for this could partially be attributed to improved tissue blood flow after ischemia by the modulation of immunological events.

Key Words: liver, ischemia and reperfusion injury, FK506, 31P-MRS, ICAM-1

Introduction

It has recently been reported that FK506 (FK), a powerful immunosuppressive agent, may enhance hepatocyte regeneration, and may also ameliorate

Reprint requests to: H. Wakabayashi (Received for publication on June 22, 1993; accepted on May 6, 1994)

hepatic injury associated with ischemia in animal models. 1-7 This effect is of great interest and is considered to be worthy of thorough investigation, because the augmentation of hepatic regeneration and protection from ischemic injury may be crucial not only in liver transplantation, but also in other types of liver surgery.

We previously reported that FK506 and cyclo- sporine A have a protective effect on the liver against warm ischemia and reperfusion injury. 1 In this study, we focused on the protective effect of FK506 in an attempt to better understand its efficacy and mech- anism. In addition, in order to evaluate the hepatic energy status using 31p magnetic resonance spectro- scopy (31P-MRS), the liver tissue blood flow was continuously monitored in the process of ischemia and reperfusion by using a laser Doppler flow meter. The serum mitochondrial glutamic oxaloacetic transaminase (m-GOT) and glutamic pyruvic transaminase (GPT), the survival rate of the rats and histological changes of the liver were also evaluated. Furthermore, the induc- tion of intercellular adhesion molecule-1 (ICAM-1) after reperfusion, and the effect of FK506 on it were also examined.

Materials and Methods

Adult male Sprague-Dawley rats weighing 200-300g were used in this study. The animals were assigned to either the FK506 group, or the control group. The rats in the FK506 group were treated with FK506 (Fujisawa Pharmaceutical, Osaka, Japan) lmg/kg/day i.m. for 4 days just prior to the examinations. The dose was determined according to the method of Kawano et al. 13 The rats in the control group were treated with saline.

The animals were fed ad libitum for 12 h, and anes- thetized by an intraperitoneal injection of 50mg/kg

H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver 995

sodium pentobarbital (Nembutal, Abbott Labor- atories, North Chicago, IL, USA). Ischemic injury was induced by cross-clamping the hepatoduodenal ligament with a silicone vascular occluder for 30 min. Using this occluder, the vessels could be clamped even while the animals were in the NMR machine.

31p-NMR Examination

The 31p-NMR spectra were obtained on an NMR sys- tem at 6.34 tesla (GSX-270, JEOL, Tokyo, Japan). A 10-mm surface coil was used both for 31P-NMR spectroscopy and for skimming on the water signal. The 31p-NMR spectra were obtained in 4-min blocks of 15 transients before, during, and after the 30-min warm ischemia. The following spectral parameters were used: a 15-msec pulse, 200 scans, and a repeti- tion time of 1.102s. During the NMR examination, the animals were given lactated Ringer's solution through an intravenous catheter to prevent dehydra- tion (10ml/kg/h). The animals used in this examina- tion included six rats from the FK506 group, and eight from the control group.

Tissue Blood Flow in the Liver

The liver tissue blood flow was measured using a laser Doppler flowmeter (Laser Flowmeter ALF21R, Advance, Tokyo, Japan) in ischemia and the reper- fusion process using the same animal model. The measurement was performed on the lower surface of the left lobe where the 31p-NMR spectra was obtained. The blood flow was expressed as a ratio of the preclamp level (%). Six rats were assigned for each group.

Mitochondrial-GOT and GPT

The serum levels of m-GOT and GPT for each rat were measured at 24 h after reperfusion using standard laboratory methods. A blood sample was collected from the inferior vena cava. Each group used six rats.

Histological Studies

The left lobes of the livers were harvested from five rats in each group at 30min, 60min, and 24h after reperfusion in each group, and were fixed and stained for histological examination with hematoxylin and eosin (H&E). Immunohistological staining for ICAM- 1 was performed using monoclonal anti-rat ICAM-1 antibody (Seikagaku Kogyo, Tokyo, Japan) at 60rain and 12 h after reperfusion.

Survival Rate of the Rats

The survival rate of the rats (30 rats for each group) was determined at 24h after reperfusion in each group.

Statistical Analysis

All the data in the figures are given as the mean +_ S.D. Unpaired Student's t-test was used to test the differences between the groups for each examination except for the survival rate where the ~2 test was used. A P value of less than 0.05 was considered to be statistically significant.

Results

~I P-MRS Examination

There were six distinct peaks observed in the 31p_ NMR spectrum in both the FK and control groups. Each peak represented phosphomonoesters (PME), inorganic phosphate (Pi), phosphodiesters (PDE), 7- ATP + [3-ADP, a-ATP + a-ADP, and ~-ATP, res- pectively. There was no difference in the preischemic 31p-NMR spectra between the two groups. The ratio of the area of the 13-ATP peak to the area of the Pi peak was measured to obtain the relative content of ATP. Figure 1 shows the changes in the [3-ATP/Pi ratio in the ischemia and reperfusion process. The recovery of the 13-ATP/Pi ratio was significantly better in the FK506 group, compared with the control. In the FK506 group, the [3-ATP/Pi ratio decreased to 1.16% of the preclamp level at the end of 30-min ischemia, and then recovered to 30.3% at 10min, 61.2% at 30min, and 76.7% at 120min after unclamping. In the control group, it recovered from 1.98% of the preclamp level to 41.9% even at 120min after unclamping.

Tissue Blood Flow

Clamping the hepatic artery and the portal vein for 30min resulted in a significant decrease in the erythrocyte flux. As illustrated in Fig. 2, the ratio of the preclamping level to the erythrocyte flux decreased rapidly as a result of the clamping. At the end of 30 min of ischemia, the ratio was 7.66 + 1.72% in the control group and 5.58 _+ 0.76% in the FK506 group. There was no significant difference observed during the ischemic period. However, after unclamping the ratio recovered rapidly in the FK506 group to 31.22 + 12.48% at 5min, 56.18 + 8.86% at 10min, 70.48 + 6.56% at 15min, and 84.38 + 12.86% at 90rain. In the control group, this ratio recovered to 14.66 +

996 H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver

fl -ATP/Pi 1.5 "~

1.0

0.5

0.0 • . . . . . A o 3'o" ' " ' ' go" ' " ' "9'o" ' " ' ' 1 1 o '

Occlusion A Time (min) Reperfusion

Fig. 1. [3-adenosine triphosphate/inorganic phosphate ([3- 31 ATP/Pi) obtained by in vivo P magnetic resonance spec-

troscopy (31P-MRS) in the ischemia and reperfusion process in the control group (open circles, n = 8) and the FK506 group (closed circles, n = 6). The values are the mean + S.D. Asterisks denote P < 0.05

3.15% at 5min, 22.16 + 7.15% at 10rain, 28.22 _+ 10.26% at 20rain, and 70.52 + 8.57% at 120rain after reperfusion. This shows that statistically significant differences did exist between the two groups in the recovery of tissue blood flow after reperfusion.

Serum m-GOT and GPT Levels

As seen in Fig. 3, the serum m-GOT and GPT values in the two groups of animals studied increased 24h after reperfusion. The value of m-GOT (IU/1) was 4.4 + 1.8 in the control group, and 2.8 + 1.3 in the FK506 group before ischemia. Twenty-four hours after ischemia these values increased to 37.4 _+ 16.7 in the control, and 8.6 + 5.2 in the FK506 group. The GPT value (IU/1) was 20 + 6.0 before ischemia, and 438.2 + 269.8 after ischemia in the control group, while it was 17.6 _+ 3.8 before ischemia and 58.2 + 39.6 after ischemia in the FK506 group. Statistically significant differences were found between the groups regarding the increasse in both the serum m-GOT and GPT values (P < 0.05).

Survival Rate

The 24-h survival rate for the FK506 group was 93%, compared to only 77% in the control group. The dif- ference was statistically significant (P < 0.05).

Tissue Blood Flow (%) 1001 6O

40

° " ~ ' I ' I " ~ " I ' I • 3 0 ' i , i • ; 0 " I ' I " 9 i o " [ " I " 1 ~ 0 ' I

Occlusion ~p ~ Time (min) Reperfusion

Fig. 2. Changes in the liver tissue blood flow monitored by a laser Doppler flowmeter in the ischemia and reperfusion process in the control group (open circles, n = 6) and the FK506 group (closed circles, n = 6). The values are the mean _+ S.D. Asterisks denote P < 0.05

m-GOT(IU/I) 60] 50

4O

30.

20.

10.

0' i

Preischemia 24-hrs a~fter Reperfusion

GPT(IU/1) 800

0 i I Preischemia 24-hrs after

Reperfusion

Fig. 3. Changes in the serum mitochondrial glutamic oxa- loacetic transaminase (m-GOT) and glutamic pyruvic transaminase (GPT) values before ischemia and 24h after reperfusion in the control group (open circles, n = 6) and FK506 group (closed circles, n = 6). All values are the mean _+ S.D. Asterisks denote P < 0.05

H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver 997

Fig. 4A,B. At 30 rain after reper- fusion, severe congestion was ob- served around the portal area in the control group (A), while it was much milder in the FK506 group (B). No leukocyte plug- ging sinusoids were observed; furthermore, no changes were seen in the hepatocytes. (H&E, x40)

Histological Study

In the histological examination by H&E staining at 30 rain and 60 rain after reperfusion, severe congestion was observed around the portal area in the control group. However this congestion was mild in the FK506 group. There were no leukocytes plugging the sinu- soidal blood flow or infiltrating the lobules found in either group. No remarkable changes in the hepa- tocytes of either group were observed (Fig. 4).

Twenty-four hours after reperfusion, a histological study by H&E staining showed remarkable hepatocyte necrosis in the centrilobular area and leukocyte

infiltration in the control group. These changes were, however, only slight in the FK506 group (Fig. 5)

ICAM-1 stained by monoclonal antibodies was negative in both groups based on immunohistological findings obtained at 60min after reperfusion. How- ever, at 12 h after reperfusion, ICAM-1 was stained on the endothelial cells of the portal veins and hepatic veins in the control group. In the FK506 group, this staining was observed only in a few veins and was much milder than in the control group (Figs. 6, 7). To compare the results of the two groups, the number of portal veins and hepatic veins, which were observed in an entire section, were counted. Consequently,

998 H. Wakabayashi et al. : Effect of FK506 on Warm Ischemia of the Liver

Fig. 5A,B. At 24h after reper- fusion, the presence of both cen- trilobular hepatocyte necrosis and leukocyte infiltration was remarkable in the control group (A), while these changes were slight in the FK506 group (B). (H&E, x40)

the number of positive portal and hepatic veins were evaluated. In the control group, an average of 89 portal veins were observed per section, and 40 of these were positive for ICAM-1 (40/89, 44.9%). In contrast, of the average 97 portal veins observed in the FK506 group, only 15 were positive for ICAM-1 (15/97, 15.7%). An average of 48 hepatic veins were counted in one section of the control group, with 33 of these showing positive staining (33/48, 68.8%), while an average of 24 among 62 hepatic veins were positive in the FK506 group (24/62, 38.7%). There was a statistically significant difference between the two groups. Sinusoid was negative for ICAM-1 even at 24 h after reperfusion in both groups.

Discussion

FK506 is a new immunosuppressive agent which has recently been brought into clinical use. Because of its powerful immunosuppressive action and milder nephrotoxicity, which is one of the major side effects of cyclosporine A, it plays a very important role today in liver transplantation. 8 Recently, it has been repor- ted that FK506 may augment liver regeneration and ameliorate ischemic liver injury in addition to its immunosuppressive action. 1-7 One of the most impor- tant factors influencing allograft and patient survival in liver transplantation is the extension of ischemic injury resulting from harvesting, preservation, and reper-

H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver 999

Fig. 6A,B. Intercellular adhesion molecule-1 (ICAM-1) was stained on the endothelial calls of the portal and hepatic veins. In one whole section, 44.9% of the portal and 68.8% of the hepatic veins were positive for ICAM-1 in the control group (n = 5) (A arrows), while 15.7% of the portal and 38.7% of hepatic veins were positive in the FK506 group (n = 5) (B arrow). (HRP staining, x40)

fusion of the organ. In typical liver surgery, the liver can suffer insult by warm ischemia and reperfusion injury due to clamping of the vessels to control bleed- ing during the operation. To investigate these effects and to better understand the mechanisms involved is therefore important not only for liver transplantation but also for other types of liver surgery.

It has been reported that energy metabolism after ischemic injury can be correlated with subsequent viability in hepatic, renal, and heart ischemia. 9'1° Therefore, by examining the ATP content of the organ in the process of ischernia and reperfusion injury, the degree of organ damage and organ viability

after ischemia can be assessed. The advantages of using 31P-MRS to study liver function and metabolism have been reviewed recently. 11 It allows for a non- destructive study of hepatic phosphorus-containing compounds associated with energy metabolism, such as ATP and Pi, to be conducted simultaneously and repetitively. In this study, using this technique made it possible to examine the changes of the energy metabolism within a very short period of reperfusion (as early as 5 or 10rain). One disadvantage of earlier studies was that the liver tissue had to be excised, while the in vitro assays of metabolites could only be obtained at selected times.

1000 H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver

Fig. 7A,B. The positive staining portal vein for 1CAM-1 in the control group (A), and the nega- tive one in the FK506 group (B). (HRP staining, x200)

In this study, a faster and better recovery of the 13- ATP/Pi ratio was observed in rats treated with FK506, as compared with the control group. Furthermore, as t he increase in hepatic enzymes was restrained, the histological findings associated with ischemic injury were found to be mild in the FK506 group. Conse- quently, a higher survival was observed in these rats. The recovery of the energy metabolism determined by 31P-MRS correlated with the subsequent liver function and the animals' survival rate, which was consistent with previous reports, m These results suggested that FK506 was able to ameliorate ischemia and reper- fusion injury in the rat liver.

Although several reports have suggested that this drug could have a hepatotropic action 3'4 and a hepatoprotective effect against ischemic injury, 2'5-7 the mechanism is still unclear, and both effects may be identical. Sakr et al. gave attention to the amelioration of the neutrophil infiltration in a histological examina- tion. 7 Dhar et al. 2 reported similar results. Starzl et al. suggested the possibility that the hepatotropic qualities of FK506 are independent of immunosuppression. 4 Kawano et al. 13 demonstrated that FK506 and cyclos- porine A decreased the level of liver tissue malon- dialdehyde, which can be an index of lipid peroxidative damage, after 60-min ischemia using an ischemically

H. Wakabayashi et al.: Effect of FK506 on Warm Ischemia of the Liver 1001

injured partially hepatectomized model. They also suggested that FK506 could surpress oxidative damage in reperfusion injury. Since Granger et al. la reported the involvement of oxygen free radicals in reperfusion injury, the mechanism of such radicals and their sup- pression have become one of the main areas of investi- gation. The liver has a high content of the enzyme xanthine oxidase, which is known to be a key enzyme in the production of oxygen radicals. Besides resident cells in the liver, such as Kupffer cells, endothelial cells, and hepatocytes, circulating leukocytes are known to be very important radical producers in reperfusion injury. 14A5 In this study, we did not examine the radical production on reperfusion and the effect of FK506 on it. However, it is interesting to note that we observed a clear difference in the recovery of the [3-ATP/Pi ratio within a very short period of reperfusion (as early as 5 or 10min), at which time no leukocyte infiltration was histologically found.

The tissue blood flow as monitored by a laser Dop- pler blood flowmeter showed a faster recovery after reperfusion in the FK506 group, compared with the control group, and it coincided with the recovery of the energy metabolism. During the histological examination performed at 30min and 60min after reperfusion, at which time there already was a statis- tically significant difference observed in the recovery of energy status and tissue blood flow, congestion in the periportal region was mild in the FK506 group compared to the control group. This fact also agrees with the blood flow results. These results suggest that the protective effect observed here may be attributed to the improvement of the recovery of tissue blood flow, particularly in the sinusoidal bloodstream. To clarify this point, the induction of ICAM-1, which is known to be one of the major adherent molecules expressed by vascular endothelial cells, 16 after reper- fusion, and the effect of FK506 on it, were examined. Although there was no induction observed at 60min after reperfusion, it was observed at 12 h after reper- fusion in the control group, while FK506 suppressed the induction.

Microvascular leukocyte accumulation, leukocyte- endothelial cell interaction and their generation of oxygen radicals have been imp]ticated in ischemia- reperfusion injury in a variety of organs. 2°-23 On reperfusion, leukocytes adhere to the endothelial cells, and plug the capillary bloodstream. Recent studies 24 have shown that this capillary plugging may, in fact, be one of the mechanisms responsible for the no- reflow phenomenon, and may ~:hus be a cause of ischemic injury, together with oxygen free radical formation. In a transplanted organ, this phenomenon might sometimes also be considered a cause of pri-

mary graft nonfunction. The finding of this study that the recovery of the tissue blood flow after reperfusion was not satisfactory in the control group could be attributed to this phenomenon. Granger et al. 19 and Marzi et al. 2° independently observed that by using superoxide dismutase, leukocyte-endothelial cell adherence was reduced after either warm or cold ischemia. They also suggested that oxygen-derived free radicals contributed to this adherence and subse- quent microcirculatory failure. FK506 has not been reported to possess any direct scavenging effect. However it is possible that FK506 can suppress the induction of adhesion molecules on both leukocytes and endothelial cells through its immunosuppressive action. Although we did not observe any leukocyte adhesion or the preventive effect of FK506 in the histological studied at 30 min and 60 min after reper- fusion, the induction of ICAM-1 was inhibited by FK506 at 12h after reperfusion. ICAM-1 is a ligand for the CDlla/CD18 (LFA-1) heterodimer. Recent studies have shown that it plays a very important role in the process of inflammation as well as in the rejec- tion response. 22 By suppressing the induction of ICAM-1, FK506 was able to ameliorate the subse- quent inflammatory reaction, and could also protect the graft from a rejection response in transplantation surgery. However, the better recovery of tissue blood flow immediately after reperfusion could not be linked with the interaction between leukocytes and vascular endothelial cells through the adhesion molecules in this study. However, some of the other adhesion pro- teins, such as GMP-140 selectin, are known to be expressed by endothelial cells in the very early phase after stimulation, and promote rapid neutrophil ad- hesion. 23 Therefore, it is possible that FK506 could exert part of its efficacy through improving sinusoidal blood flow after reperfusion by modulating the induc- tion of adherent proteins. However, further study is still needed to better understand this effect.

On the other hand, Viebahn et al. 24 have suggested that FK506 may have a protective effect on the hepa- tocytes themselves. They examined the viability of hepatocytes after cold hypoxia-reperfusion in vitro, and found that FK506 ameliorated the reoxygenation injury. As the role of mitochondrial damag e in the pathogenesis of ischemic cell injury is also controver- sial, one of the major issues that must be investigated is how FK506 can influence mitochondrial function through its hepatotropic and hepatoprotective actions. Pelekanou et al. 25 tested the effect of FK506, cyclos- porine A, and rapamycine on the state 3 respiration of isolated rat liver mitochondria. They found that cyclosporine A partially inhibited the metabolism of the mitochondria, while FK506 did not. Thus, they speculated that the mitochondria might lack an FK506-

1002 H. Wakabayashi et al. : Effect of FK506 on Warm Ischemia of the Liver

binding protein homologue. However, the protective effect on mitochondrial function may be independent of the FK506-binding protein. Direct evidence, how- ever, is still lacking.

In this study we did not make a bypass from the portal system to the systemic circulation to avoid a pooling of the splanchnic blood during clamping. Marubayashi et al. 1° observed that a portafemoral shunt prepared to relieve portal congestion enhanced the rate and extent of ATP resynthesis by the reflow following ischemia, and that it was accompanied by an increase in the survival rate of the rat. However, they found no significant difference in the recovery of ATP between the groups with and without the shunt, when the duration of ischemia was less than 60rain. In this study, the duration of ischemia was 30min. Thirty minutes of ischemia thus seemed to be short enough to conduct this study without using a shunt. However, it could not be ruled out that FK506 affected the con- gested splanchnic blood during the clamping.

In conclusion, FK506 is able to protect hepatic energy metabolism from ischemia and reperfusion injury. This effect may enhance graft function after transplantation, as well as liver function after other types of liver surgery. However, further study is needed to more fully understand the mechanisms involved.

Acknowledgments. This work was supported by a Grant-in- Aid for general scientific research (no. 16202) from the Ministry of Education, Science and Culture of Japan.

References

1. Wakabayashi H, Karasawa Y, Maeba T, Tanaka S (1992) Effect of FK506 and cyclosporine A on hepatic energy status in tile rat after warm ischemia, as monitored by 31p nuclear magnetic resonance spectroscopy in vivo. Transplant Proc 24:1993-1995

2. Dhar DK, Nagasue N, Kimoto T, Uchida M, Takemoto Y, Nakamura T (1992) The salutary effect of FK506 in ischemia- reperfusion injury of the canine liver. Transplantation 54:583-588

3. Francavilla A, Barone M, Todo S, Zeng Q, Porter KA, Starzl TE (1989) Augmentation of rat liver regeneration by FK506 compared with cyclosporine. Lancet 25:1248-1249

4. Starzl TE, Porter KA, Mazzaferro V, Todo S, Fung J, Francavilla A (1991) Hepatotropic effects of FK506 in dogs. Transplantation 51:67-70

5. Mazzaferro V, Scotti-Foglieni CL, Porter KA, Trejo-Bellido J, Carrieri G, Todo S, Fung JJ, Francavilla A, Starzl TE (1990) Studies of hepatotropic qualities of FK506 and CyA. Transplant Proc 22:93-95

6. Sakr MF, Zetti GM, Farghali H, Hassanein TH, Gavaler JS, Starzl TE, Van Thiel DH (1991) Protective effect of FK506 against hepatic ischemia in rats. Transplant Proc 23:340-341

7. Sakr MF, Zetti GM, Hassanein TI, Farghali H, Nalesnik MA, Gavaler JS, Starzl TE, Van Thiel DH (1991) FK506 ameliorates the hepatic injury associated with ischemia and reperfusion in rats. Hepatology 13:947-951

8. Starzl TE, Todo S, Fung J, Demetris A J, Venkataramman R, Jain A (1989) FK506 for liver, kidney, and pancreas trans- plantation. Lancet 2:1000-1004

9. Bore PJ, Paratheofanis I, Sells RA (1979) Adenosine triphos- phate regeneration and function in the rat kidney following warm ischemia. Transplantatoin 27:235-237

10. Marubayashi S, Takenaka M, K Dohi, Ezaki H, Kawasaki T (1980) Adenine nucleotide metabolism during hepatic ischemia and subsequent blood reflow periods and its relation to organ viability. Transplantation 30:294-296

11. Cohen SM (1983) Application of nuclear magnetic resonance to the study of liver physiology and disease. Hepatology 3:738-749

12. Granger DN, Rutili G, Maccord JM (1981) Superoxide radicals in feline intestinal ischemia. Gastroenterology 8t:22-29

13. Kawano K, Kim Y, Goto S, Ono M, Kobayashi M (1991) A protective effect of FK506 in ischemically injured rat liver. Transplantation 52:143-145

14. Hernandez LA, Grisham MB, Twohig B, Arfors KE, Marian JM, Granger DN (1987) Role of neutrophils in ischemia- reperfusion-induced microvascular injury. Am J Physiol 253: H699-703

15. Grisham MB, Hernandez LA, Granger DN (1986) Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am J Physiol 251 :G567-574

16. Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA (1986) Induction by IL-1 and interferon-gamma: tissue dis- tribution, biochemistry, and function of a natural adherence molecule (ICAM-1). J Immunol 137:245-254

17. Korthuis RJ, Grisham MB, Granger DN (1988) Leukocyte depletion attenuates vascular injury in postischemic skeletal muscle. Am J Physiol 254:H823

18. Engler RL, Dahlgre MD, Morris DD, Peterson MA, Schmid- Sch6nbein GW (1986) Role of leukocytes in response to acute myocardial ischemia and reflow in dogs. Am J Physiol 251: H314-322

19. Granger DN, Benoit JN, Suzuki M, Grisham MB (1989) Leukocyte adherence to venular endothelium during ischemia. Am J Physiol 257:G683-688

20. Marzi I, Knee J, Bfihren V, Menger M, Trentz O (1992) Reduc- tion by superoxide dismutase of leukocyte-endothelial adherence after liver transplantation. Surgery 111:90- 97

21. Schmid-Sch6nbein GW (1987) Capillary plugging by granulo- cytes and the no reflow phenomenon in the microcirculation. Federation Proc 46:2397-2401

22. Hoffmann MW, Wonigeit K, Steinhoff G, Herzbeck H, Flad HD, Pichlmayr R (1993) Production of cytokines (TNF-alpha, IL-l-beta) and endothelial cell activation in human liver allograft rejection. Transplantation 55:329-335

23. Geng JG, Bevilacqua MP, Moore KL, McIntyre TM, Prescott SM, Kim JM, Bliss GA, Zimmerman GA, McEver RP (1990) Rapid neutrophil adhesion to activated endothelium mediated by GMP-140. Nature 343:757-760

24. Viebahn R, de Groot H, Lauchart W, Becker HD (199l) Hy- poxic and reoxygenation injury in human and rat hepatocytes- Influence of FK506 and cyclosporine. Transplant Proc 23: 3183-3184.

25. Pelekanou V, Trezise AE, Moore AL, Kay JE (1991) FK506 and Rapamycin do not affect platelet aggregation or mito- chondrial function. Transplant Proc 23:3200-3201