During liver transplantation, hemodynamics change considerably when the inferior vena cava and hepatic portal vein are clamped (the start of anhepatic stage, the liver is dissected from the body) and opened (the start of neohepatic stage, the donor liver is reperfused in the body), resulting in vasoplegic syndrome, which is characterized by high cardiac output but low systemic resistance and severe hypotension.1–3 The mechanisms of vasoplegic syndrome during liver transplantation remain unclear, but the inflammatory cytokines induced by gastrointestinal tract congestion may play roles in these mechanisms.3,4 Meanwhile postreperfusion syndrome is defined as a mean arterial pressure decrease of more than 30% over 1 min within 5 min of reperfusion5,6; vasoplegic syndrome differs from postreperfusion syndrome in terms of etiology, symptoms, and treatment.4
In recent years, segments of intact arteries removed from patients undergoing surgery were used to study the vascular responses induced by different factors. In one study, ring segments of isolated pulmonary arteries and radial arteries obtained from patients undergoing lung resection or coronary artery bypass graft surgery were used to examine the contractile responses to different vasoactive drugs.7 In another study, the internal mammary arteries obtained from patients undergoing coronary artery bypass graft surgery were used to measure the vascular responses to severe acidosis.8
Hepatic portal clamping and opening cause major pathophysiological changes during liver transplantation. For example, the inflammatory cytokines induced by gastrointestinal tract congestion during liver transplantation may have an impact on the vascular endothelial cells and may affect vascular tension.3,4 P38 mitogen-activated protein kinase (P38 MAPK) is a very important branch of MAPK pathway, as part of an important intracellular signal transduction system plays a key role in a variety of cellular responses involved in inflammation, cell stress, apoptosis, cell cycle and growth in physiological, and pathological processes.9,10
In this study, we used an isolated hepatic artery ring segment model (obtained from dissected livers derived from orthotopic liver transplantation) to explore whether the inflammatory cytokines induced by gastrointestinal tract congestion exert an effect on vascular endothelial cells and isolated hepatic artery ring tension. In addition, we explored the associated signal transduction pathway.
Materials and Methods
Reagents and Drugs
N-nitro-L-arginine (L-NAME), indomethacin, glibenclamide, BAY 11-7082, wortmannin, SB203580, SB 600125, and PD 98059 were obtained from Sigma-Aldrich Co. (St. Louis, MO). Tumor necrosis factor α (TNFα), interleukin-1 (IL-1), interleukin-6 (IL-6), and the corresponding monoclonal antibodies were obtained from Sigma-Aldrich Co. Primers for P38 MAPK and β-actin were obtained from Promega (Madison, WI). Antibodies against P38 MAPK and phospho-P38 MAPK were obtained from Sigma-Aldrich Co.
Hepatic Artery Ring Segments and Plasma Preparation
Ethical approval for this study was provided by the Ethics Committee (Ethical Committee AFH-108) of the Logistics University of Chinese People Armed Police Force Hospital, Tianjin, China (Chairperson Prof. Sai Zhang) on February 15, 2013, and all patients involved in the study signed an informed consent form. Blood from the superior vena cava was obtained from the jugular vein just before the neohepatic stage, and gastrointestinal congestion samples were obtained from the portal vein at the moment that hepatic portal vein was unclamped. The blood samples were centrifuged for 10 min at 3,000 rpm to obtain the plasma, which was kept at 4°C until experimental use. The hepatic artery ring (internal diameter approximately 1 mm) was collected as soon as the liver was dissected and was placed in ice-cold, oxygenated Krebs solution composed of the following fractions (in mM): NaCl, 119; KCl, 4.7; NaHCO3, 25; CaCl2, 2.5; MgCl2, 1; KH2PO4, 1.2; and D-glucose, 11. Each artery was cut into 3 mm-wide ring segments after removing the adhering fatty tissues. These rings were suspended between two stainless steel hooks in a 5 ml chamber in a Multi Myograph system (Danish Myo Technology A/S, Skejby Science Center Skejbyparken 152 DK-8200 Aarhus N Denmark.). The Krebs solution in the chamber was kept at 37°C and oxygenated with a 95% O2/5% CO2 gas mixture. Each ring was initially stretched to 2 g as the optimal resting tone and was then stabilized for 90 min before the start of each experiment. Rings containing endothelium were confirmed to exhibit more than 85% vasorelaxation when treated with acetylcholine (1 μM) after first being contracted with phenylephrine (1 μM). The rings were then rinsed three times with prewarmed, oxygenated Krebs solution until the stable resting tone returned to 2 g; the resting tone was readjusted to 2 g if necessary. Each experiment was performed on rings from different patients.
Examination of Plasma from Gastrointestinal Congestion
Plasma from the superior vena cava and gastrointestinal congestion was obtained just before the neohepatic stage of liver transplantation. The hepatic artery was taken from the dissected liver. The temperature, the electrolyte levels, and the acid-base balance in the plasma were detected by blood gas analysis, and the TNFα, PGI2, IL-1, IL-6, and endotoxin (lipopolysaccharides, LPS) levels in the plasma were examined by an enzyme-linked immunosorbent assay. Disturbance of the acid-base balance was rectified based on the blood gas analysis. Monoclonal antibodies against TNFα, IL-1, and IL-6 were then used to antagonize different inflammatory cytokines in the samples from gastrointestinal congestion.
Hepatic Artery Ring Isometric Tension Measurement
The tension in the rings when contracted with 60 mM KCl in Krebs solution was used as the control. After rebalancing, the rings were contracted with 60 mM KCl in the superior vena cava plasma or intestinal congestion plasma; this plasma was incubated for 30 min at 37°C with different inhibitors or signal blockers (100 μmol/L of the nitric oxide [NO] inhibitor L-NAME, 2.5 mmol/L of the PGI2 inhibitor indomethacin, 25 μg/ml of the KATP channel inhibitor glibenclamide, 10 μmol/L of the nuclear factor-κB [NF-κB] blocker BAY 11-7082, 2.0 μmol/L of the phosphatidylinositol-3 kinase/protein kinase B [PI3K/Akt] pathway blocker wortmannin, 20 μmol/L of the P38 MAPK pathway blocker SB 203580, 50 μmol/L of the c-Jun N-terminal protein kinase [JNK] pathway blocker SB 600125, or 60 μmol/L of the extracellular signal-regulated kinase [ERK] pathway blocker PD 98059). The tension of the rings in the plasma was recorded and expressed as a percentage of the tension in Krebs solution.
After incubation with plasma or different drugs for 30 min, the hepatic artery rings were frozen in liquid nitrogen and then homogenized in ice-cold radioimmunoprecipitation assay lysis buffer. Proteins were obtained by centrifugation, and 50 μg of total protein was electrophoresed and subsequently electroblotted on an Immobilon-P (Sigma-Aldrich Co, St. Louis, MO) polyvinylidene difluoride membrane. P38 MAPK and phospho-P38 MAPK levels were detected with primary antibodies against P38 MAPK (1:500; Sigma-Aldrich Co, St. Louis, MO) and phospho-P38 MAPK (1:500; Sigma-Aldrich Co) and with horseradish peroxidase-conjugated secondary antibodies (DakoCytomation, Carpinteria, CA). The signal intensities were analyzed using a gel documentation system (FluorChem; Alpha Innotech Corp. San Jose, CA).
Quantitative Reverse-Transcription Polymerase Chain Reaction
Total RNA was extracted with TRIzol (Promega) from the hepatic artery rings frozen in liquid nitrogen after treatment with plasma or different drugs. The reverse-transcription reaction and PCR were performed according to the manufacturers’ protocols. The primers for P38 MAPK were 5′-TGTTGACCGGAAGAACGTTGT-3′ (sense) and 5′-CAAAGTACGCATGCGCAAGAG-3′ (antisense). The primers for β-actin were 5′-GAAATCGTGCGTGACATTAAG-3′ (sense) and 5′-CTAGAAGCATGCGGTGGA-3′ (antisense) (Promega). The PCR products were then electrophoresed on 1.5% agarose gels. The results were analyzed using a digital imaging system (UVItec, Cambridge, UK).
Transmission Electron Microscopy
The structure of the endothelium and vascular smooth muscle of the rings was observed by transmission electron microscopy after the rings were incubated with the superior vena cava plasma or intestinal congestion plasma for 30 min.
Data were recorded as the mean ± SD. The hepatic artery rings were obtained from 30 patients. A two-tailed Student’s t-test or one-way analysis of variance followed by the Newman-Keuls test that was used in the statistical analysis, and p < 0.01 was recognized as statistically significant. The reverse-transcription polymerase chain reaction (RT-PCR) and Western blotting results are representative of three independent experiments. The hepatic artery ring contractile forces are expressed as a percentage of the mean value of two consecutive responses to 60 mM KCl in plasma compared with the tension in Krebs solution.
Acid-Base Balance and Inflammatory Cytokines in Gastrointestinal Congestion Plasma
The base excess (BE) value of the superior vena cava plasma was normal, whereas the gastrointestinal congestion showed a state of metabolic acidosis (Figure 1A, shown as the absolute value of the BE). Therefore, in the next experiment, the acid-base imbalance was rectified with 5% NaHCO3. In contrast, there were no significant differences among the Ca2+, K+, pH, and temperature values (data not shown). Compared with those of the superior vena cava plasma, the PGI2, IL-6, TNFα, and IL-1 levels in the gastrointestinal congestion plasma were increased significantly (Figure 1B–E); however, no difference was found for LPS (data not shown). This result demonstrated that the gastrointestinal ischemia-reperfusion injury during orthotopic liver transplantation resulted in the production and release of significant levels of inflammatory cytokines.
The Effect of the Acid-Base Balance and Inflammatory Cytokines on Hepatic Artery Tension
The effect of the acid-base balance and inflammatory cytokines on hepatic artery tension is expressed as a percentage of the vascular tension caused by 60 mM KCl in plasma compared with the tension in Krebs solution. Compared with the superior vena cava plasma, the gastrointestinal congestion plasma caused significant vasorelaxation (observed as decreased ring tension), and this vasorelaxation could be partially corrected by the addition of monoclonal antibodies against TNFα, IL-1, and IL-6 (1, 2, and 2 μg/ml, respectively) but not by acid-base balance rectification (Figures 2 and 3A). When the superior vena cava plasma was mixed with the inflammatory cytokines TNFα, IL-1, and IL-6 (100, 20, and 20 ng/ml, respectively), the observed vasorelaxation effect was the same as that induced by the gastrointestinal congestion plasma (Figure 3B).
The Mechanisms Underlying the Effect of Inflammatory Cytokines on Hepatic Artery Tension
To explore the mechanism underlying the effect of inflammatory cytokines on hepatic artery tension, the artery rings immersed in gastrointestinal congestion plasma incubated with different drugs (100 μmol/L of the NO inhibitor L-NAME, 2.5 mmol/L of the PGI2 inhibitor indomethacin, 25 μg/ml of the KATP channel inhibitor glibenclamide, 10 μmol/L of the NF-κB blocker BAY 11-7082, 2.0 μmol/L of the PI3K/Akt pathway blocker wortmannin, 20 μmol/L of the P38 MAPK pathway blocker SB203580, 50 μmol/L of the JNK pathway blocker SB 600125, or 60 μmol/L of the ERK pathway blocker PD 98059) were treated with 60 mM KCl. Compared with untreated gastrointestinal congestion plasma, gastrointestinal congestion plasma incubated with L-NAME, indomethacin, glibenclamide, BAY 11-7082, or SB 203580 significantly restored vasocontraction (Figure 4). This outcome suggested that NO, PGI2, the KATP channel, and NF-κB play a role in the vasorelaxation induced by plasma from congestion and that the signal transduction pathway involved in this process is P38 MAPK. This result was also confirmed by Western blotting and RT-PCR (Figure 5). After incubation with plasma and different drugs for 30 min, the hepatic artery rings in the gastrointestinal congestion group also expressed higher levels of phospho-P38 MAPK protein and P38 MAPK mRNA compared with those in the superior vena cava group.
Structural Changes in the Endothelium and Vascular Smooth Muscle Induced by Gastrointestinal Congestion
The structure of the endothelium and vascular smooth muscle was observed by transmission electron microscopy after the rings were incubated with the gastrointestinal congestion plasma or the superior vena cava plasma for 30 min. The endothelial cells in the rings incubated with the superior vena cava plasma exhibited a normal structure. In contrast, the endothelial cells in the rings incubated with the gastrointestinal congestion plasma exhibited vacuolation, structural changes, and loose associations between the cells and with other tissues. However, the structure of the vascular smooth muscle was not altered in either group (Figure 6).
All patients in the 30 cases involved in the study underwent successful surgeries. Although one patient died of the original liver dysfunction, the others recovered gradually. A retrospective examination showed that 26 of the 30 patients in the neohepatic stage needed vasoactive drug support. In total, 21, 6, and 3 patients had blood pressure decreases of >30%, >20%, and >10%.
Gastrointestinal congestion plasma exhibited a state of metabolic acidosis because of the ischemia and hypoxia caused by clamping of the inferior vena cava and hepatic portal vein. The ischemia-reperfusion injury of the gastrointestinal tract could also cause inflammatory cell activation and inflammatory cytokine release (TNFα, IL-1, and IL-6).11,12 Whereas metabolic acidosis rectification had no effect on artery ring tension, monoclonal antibodies against TNFα, IL-1, and IL-6 significantly increased artery ring tension. In fact, the addition of TNFα, IL-1, and IL-6 to superior vena cava plasma caused a result that was similar to the results obtained with gastrointestinal congestion plasma, suggesting that inflammatory cytokines, rather than metabolic acidosis, play an important role in the vasorelaxation induced by gastrointestinal congestion. This result is consistent with other experiments demonstrating that inflammatory cytokines can influence ring tension.13,14
Based on the results showing that vasorelaxation induced by gastrointestinal congestion plasma could be partially reversed by indomethacin, glibenclamide, BAY 11-7082, and SB 203580 and the results showing that hepatic artery rings treated with this plasma expressed higher levels of phospho-P38 MAPK protein and P38 MAPK mRNA, we concluded that the NO, PGI2, the KATP channel, NF-κB, and the P38 MAPK signal transduction pathway were involved in the vasorelaxation induced by this type of plasma.
Hepatic artery vascular endothelial cells produce NO because of inducible nitric oxide synthase (iNOS) activation caused by acidosis and inflammatory cytokines. NO activates soluble guanylate cyclase to produce cyclic guanosine monophosphate mononuclear phosphate, which causes decreased Ca2+ levels and increased vasorelaxation.15–18 Acidosis and inflammatory cytokines also activate cyclooxygenase to increase PGI2 in vascular endothelial cells, which results in vasodilation.15–18
Nuclear factor-κB is a nuclear transcription factor that is widely present in many types of cells in vivo and that can specifically bind to the κB locus of various gene promoters and enhancers to promote gene transcription.19,20 Many inflammatory mediator genes contain κB-binding sites; thus, NF-κB can regulate the expression of numerous cytokines and inflammatory mediators and participate in intracellular signal transduction.19–22
The opening of the KATP channel in vascular smooth muscle can promote K+ outflow and cell membrane hyperpolarization, which can inactivate the voltage-dependent Ca2+ channel to prevent Ca2+ inflow and reduce the intracellular Ca2+ concentration. In the current study, the KATP channel was involved in hepatic artery ring dilatation.23,24
P38 MAPK is a tyrosine protein kinase composed of 360 amino acids with a molecular weight of 38 kD. As an important intracellular signal transduction system that transduces extracellular stimulation signals to the nucleus, the P38 MAPK pathway plays a key role in the regulation of the inflammatory response. P38 MAPK is first activated by inflammatory factors, and phosphorylated cytoplasmic P38 MAPK then translocates to the nucleus to activate transcription factors that can promote the expression of cytokines and inflammatory mediators.9,10
P38 MAPK is an important kinase upstream of NF-κB; the NF-κB signal transduction pathway is the key link to inflammation and the common intersection point of many signaling pathways.9,10 Activation of P38 MAPK induces NF-κB activation and promotes NO synthesis and release. The phosphorylation of P38 MAPK also enhances the expression of TNFα mRNA. The application of the specific P38 MAPK blocker SB 203580 thus significantly decreases NF-κB activation, reducing the expression of iNOS and TNFα mRNA and the synthesis of NO.9,10
The vascular endothelium is a layer of mononuclear cells between the blood and the vascular wall tissue that can produce NO, PGI2, endothelin-1, and other vasoactive substances to regulate vascular tone.25,26 In the current study, the structure of hepatic artery endothelial cells was damaged by gastrointestinal congestion, and the vascular tension was altered accordingly.
In this study, rings contracted with 60 mM KCl in Krebs solution served as a control, and rings contracted with 60 mM KCl in plasma served as the experimental group. Monoclonal antibodies against TNFα, IL-1, and IL-6 and a P38 MAPK blocker could not entirely restore the ring tension in the latter group relative to that in the control group. We believe that differences in the solutions may have been one reason for this discrepancy; Krebs solution was used in the control group, whereas plasma was used in the experimental group. In the plasma, other cytokines aside from TNFα, IL-1, and IL-6 could play a role in vasorelaxation. Therefore, it remains unclear whether factors other than TNFα, IL-1, IL-6, and P38 MAPK may cause ring relaxation, and in future experiments, we will explore other mechanisms.
We now know that NF-κB, PGI2, NO, the KATP channel, and P38 MAPK are involved in the ring tension decrease induced by the gastrointestinal congestion, but the interactions among these factors remain unclear; thus, further study is needed. We still do not fully understand several of the exact mechanisms by which endothelial cells and vascular smooth muscle maintain ring tension after treatment with plasma from gastrointestinal congestion. In the future, endothelium-denuded rings and endothelial cell cultures may be used to verify these mechanisms.
1. Schumann R, Mandell MS, Mercaldo N, et al.: Anesthesia for liver transplantation
in United States academic centers: Intraoperative practice. J Clin Anesth 2013.25: 542550.
2. Hall TH, Dhir A: Anesthesia for liver transplantation
. Semin Cardiothorac Vasc Anesth 2013.17: 180194.
3. Cao Z, Gao Y, Tao G: Vasoplegic syndrome during liver transplantation
. Anesth Analg 2009.108: 19411943.
4. Cao Z, Tao G: Is it possible to distinguish between vasoplegic syndrome and postreperfusion syndrome during liver graft reperfusion? Anesth Analg 2010.110: 970971.
5. Fukazawa K, Pretto EA: The effect of methylene blue during orthotopic liver transplantation
on post reperfusion syndrome and postoperative graft function. J Hepatobiliary Pancreat Sci 2011.18: 406413.
6. Chung IS, Kim HY, Shin YH, et al.: Incidence and predictors of post-reperfusion syndrome in living donor liver transplantation
. Clin Transplant 2012.26: 539543.
7. Currigan DA, Hughes RJ, Wright CE, Angus JA, Soeding PF:Vasoconstrictor responses to vasopressor agents in human pulmonary and radial arteries: An in vitro study. Anesthesiology 2014.121: 930936.
8. Vidal C, Grassin-Delyle S, Devillier P, et al.: Effect of severe acidosis on vasoactive effects of epinephrine and norepinephrine in human distal mammary artery. J Thorac Cardiovasc Surg 2014.147: 16981705.
9. Gaundar SS, Bendall LJ: The potential and limitations of p38MAPK as a drug target for the treatment of hematological malignancies. Curr Drug Targets 2010.11: 823833.
10. Gaundar SS, Bradstock KF, Bendall LJ: p38MAPK inhibitors attenuate cytokine production by bone marrow stromal cells and reduce stroma-mediated proliferation of acute lymphoblastic leukemia cells. Cell Cycle 2009.8: 29752983.
11. Biancofiore G, Bindi L, Miccoli M, et al.: Balance of pro- and anti-inflammatory cytokines in cirrhotic patients undergoing liver transplantation
. Transpl Immunol 2013.28: 193197.
12. Kita Y, Iwaki Y, Noguchi K, et al.: Daily serum inflammatory cytokine
(tumor necrosis factor-alpha, interleukin-6) monitoring in liver transplantation
focusing on allograft rejection: a five-case report. Transplant Proc 1996.28: 12371240.
13. Ohkawa F, Ikeda U, Kanbe T, Kawasaki K, Shimada K: Effects of inflammatory cytokines on vascular tone. Cardiovasc Res 1995.30: 711715.
14. Marceau F, deBlois D, Petitclerc E, et al.: Vascular smooth muscle contractility assays for inflammatory and immunological mediators. Int Immunopharmacol 2010.10: 13441353.
15. Levin MA, Lin HM, Castillo JG, Adams DH, Reich DL, Fischer GW: Early on-cardiopulmonary bypass hypotension and other factors associated with vasoplegic syndrome. Circulation 2009.120: 16641671.
16. Shanmugam G: Vasoplegic syndrome—The role of methylene blue. Eur J Cardiothorac Surg 2005.28: 705710.
17. Ríha H, Ríhová L, Pind’ák M, Brezina A, Pirk J: [Methylene blue in the therapy of vasoplegic syndrome after cardiac surgery procedure]. Cas Lek Cesk 2006.145: 322324.
18. Evora PR, Ribeiro PJ, Vicente WV, et al.: Methylene blue for vasoplegic syndrome treatment in heart surgery: Fifteen years of questions, answers, doubts and certainties. Rev Bras Cir Cardiovasc 2009.24: 279288.
19. Valera FC, Umezawa K, Brassesco MS, et al.: Suppression of inflammatory cytokine
secretion by an NF-ĸB inhibitor DHMEQ in nasal polyps fibroblasts. Cell Physiol Biochem 2012.30: 1322.
20. Nikolai LM, Mark PB, Andrei VK, David W: MAP3K-related kinase involved in NF-KB induction by TNF, CD95 and IL-1. Nature 1997.385: 540544.
21. Lim EJ, Lee SH, Lee JG, et al.: Toll-like receptor 9 dependent activation of MAPK and NF-kB is required for the CpG ODN-induced matrix metalloproteinase-9 expression. Exp Mol Med 2007.39: 239245.
22. Ivanenkov YA, Balakin KV, Lavrovsky Y: Small molecule inhibitors of NF-kB and JAK/STAT signal transduction pathways as promising anti-inflammatory therapeutics. Mini Rev Med Chem 2011.11: 5578.
23. Quayle JM, Standen NB: KATP channels in vascular smooth muscle. Cardiovasc Res 1994.28: 797804.
24. Ko EA, Han J, Jung ID, Park WS: Physiological roles of K+ channels in vascular smooth muscle cells. J Smooth Muscle Res 2008.44: 6581.
25. Ando J, Yamamoto K: Vascular mechanobiology: Endothelial cell responses to fluid shear stress. Circ J 2009.73: 19831992.
26. Tian J, Wong WT, Tian XY, Zhang P, Huang Y, Wang N: Rosiglitazone attenuates endothelin-1—Induced vasoconstriction by upregulating endothelial expression of endothelin B receptor. Hypertension 2010.56: 129135.