Cytokine 64 (2013) 71–78

Contents lists available at ScienceDirect

Cytokine

journal homepage: www.journals .e lsev ier .com/cytokine

Liver transplantation and inflammation: Is lipopolysaccharide bindingprotein the link?

1043-4666/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.cyto.2013.07.025

Abbreviations: LPS, lipopolysaccharide; LBP, lipopolysaccharide binding protein;CI, cold ischemia; WI/R, warm ischemia/reperfusion; LTx, liver transplantation; I/R,ischemia/reperfusion; CD14, cluster of differentiation 14; TLR4, toll-like receptor 4;TNF-a, tumor necrosis factor-alpha; ELISA, enzyme-linked immunosorbent assay;PCR, polymerase chain reaction; SDS–PAGE, sodium dodecyl sulfate–polyacryl-amide gel electrophoresis; HPRT, hypoxanthine guanine phosphoribosyltransfer-ase; cDNA, complementary DNA.⇑ Corresponding author. Address: Experimental Transplantation Surgery, Depart-

ment of General, Visceral and Vascular Surgery, Friedrich-Schiller-University Jena,Drackendorfer Straße 1, 07747 Jena, Germany. Tel.: +49 03641 9325350; fax: +4903641 9325352.

E-mail addresses: haoshu.fang@uk-essen.de (H. Fang), anding.liu@uk-essen.de(A. Liu), olaf.dirsch@med.uni-jena.de (O. Dirsch), uta.dahmen@med.uni-jena.de (U.Dahmen).

Haoshu Fang a,b, Anding Liu a,c, Olaf Dirsch d, Uta Dahmen a,⇑a Experimental Transplantation Surgery, Department of General, Visceral and Vascular Surgery, Friedrich-Schiller-University Jena, Jena 07747, Germanyb Department of Pathophysiology, Anhui Medical University, Hefei 236000, Chinac Experimental Medicine Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 243000, Chinad Institute for Pathology, University Hospital of Jena, Jena 07747, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 5 March 2013Received in revised form 19 June 2013Accepted 22 July 2013Available online 16 August 2013

Keywords:LPSLBPWarm ischemia reperfusionCold ischemiaLiver transplantation

Background: Lipopolysaccharide (LPS) binding protein (LBP) is an acute phase protein, which upregulatedin response to surgical interventions. LBP plays an important role in modulating LPS-induced inflamma-tory response. We investigated the expression of LBP and the translocation of LPS in rat models of hepaticischemia reperfusion injury and liver transplantation (LTx). We also elucidated the effect of LBP on theinflammatory response.Methods: In this study, cold ischemia (CI), warm ischemia/reperfusion (WI/R), and LTx models were usedto model relevant physiologic situations during LTx. Serum and effluent protein levels as well as hepatic-mRNA and protein levels of LBP were examined. LBP released into the effluent during CI was used in amacrophage-LPS-stimulation assay to investigate the role of LBP in modulating the LPS-induced inflam-matory response. Blocking experiments using an LBP-inhibitory peptide were performed to confirm therelevance of LPS/LBP for the induction of the inflammatory response. Impairment of the intestinal barrierand translocation of LPS into the liver was visualized by immunohistochemistry. Induction of tumornecrosis factor-alpha (TNF-a) mRNA expression in the liver was taken as indicator of the inflammatoryresponse.Results: Upregulation of LBP in serum and/or liver tissue was observed after WI/R, CI and LTx, respec-tively. The LBP released during CI enhanced the LPS induced inflammatory response in vitro as indicatedby an induction of TNF-a. On the other hand, blocking LBP using LBP inhibitory peptide, suppressed theinduction of TNF-a in vitro markedly. After LTx, elevated serum LBP levels were associated with post-operative LPS translocation and production of inflammatory cytokines.Conclusions: Our findings suggest that translocation of LPS occurs after LTx and that LBP is mediating theLPS-induced inflammatory response after LTx. Blocking LBP using LBP-inhibitory peptide might representa novel strategy to reduce the I/R-induced inflammatory response.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Ischemia/reperfusion (I/R) injury is the main cause of initialgraft dysfunction and primary failure in liver transplantation(LTx) [5,18,22]. Cold ischemia (CI) injury is an inevitable conse-quence of liver explantation and organ preservation prior to trans-plantation. Warm ischemia/reperfusion (WI/R) occurs during theimplantation procedure. In LTx, both CI and WI/R lead to liver dam-age and cause a systemic inflammatory response.

I/R induces an inflammatory response and cytokine releasefrom non-parenchymal cells and passenger leucocytes in the liver[5,18,22]. The inflammatory response following I/R which leadsto organ damage have been studied extensively [25]. In the LTx set-ting, the initial inflammatory is caused by a sterile organ injury in-flicted by the surgical trauma and the I/R injury. However the

72 H. Fang et al. / Cytokine 64 (2013) 71–78

impact of a non-sterile, bacterial damage on the inflammatoryreaction has not been elucidated so far.

Lipopolysaccharid (LPS) and bacterial translocation is observedin case of an impaired intestinal barrier function [6]. LPS is the ma-jor component of the outer membrane of Gram-negative bacteria,and functions as major pathophysiological factor in initiating theinflammatory response. The barrier function of the intestine is im-paired when the portal pressure rises subsequent to clamping thehepatoduodenal ligament as done during the anhepatic phase ofliver transplantation. Portal hypertension as well as hepatic I/R in-jury are associated with bacterial translocation [6,7,27]. Bacterialtranslocation leads to increased levels of circulating LPS, a maincomponent of the cell wall of gram negative bacteria [1].

The inflammatory response to LPS is mediated via binding toLPS binding protein (LBP). LBP recognizes LPS molecules in sys-temic circulation, and then transfers LPS to cluster of differentia-tion 14 (CD14), which in turn initiates the toll-like receptor 4(TLR4) signaling cascade via forming the CD14-TLR4 complex[19,20]. These events result in the activation of the nuclear factorjB (NF-jB) pathways, and cause the production of cytokines,including tumor necrosis factor-alpha (TNF-a) and interleukin(IL)-6 [16].

LBP is an acute phase protein which plays an important role inmodulating the inflammatory response. It is reported that LBP isupregulated in infectious diseases, as well as in surgical stress sit-uations [10]. We previously demonstrated that LBP was upregu-lated after liver resection in rats, and the systemic LBP levels didcorrelate to the remnant functional liver mass [13]. Furthermore,increased LBP levels were associated with an upregulation of thedownstream inflammatory cytokines after LPS challenge.

In this study, we wanted to investigate whether LBP was upreg-ulated and released after hepatic I/R and LTx in rats. We alsowanted to observe whether LBP, released during ischemic damageof liver, was associated with the LPS-induced inflammatory re-sponse after LTx. As LBP is an important host protein for interactionwith LPS, the modulation of LBP may be of potential interest forelucidating mechanisms of post-operative inflammation.

2. Materials and methods

2.1. Experimental design

In this study, three animal models were used to cover differentsituations – WI/R, CI, and LTx – modeling the different pathophys-iological aspects of the surgical procedure during LTx.

Selective warm I/R injury was induced by clamping the vascularblood supply to the median and left lateral lobe of the liver for90 min followed by 0.5 h, 6 h and 24 h reperfusion (n = 6/group).As a CI model, livers were explanted and stored in saline at 4 �C.Effluents were collected every hour during cold ischemia and livertissue was obtained at 0 h, 4 h, 8 h, and 12 h. Six rats were sub-jected to LTx after cold-preservation of the graft for 6 h and sacri-ficed 24 h postoperatively. Another set of 6 rats was included asnormal control group. Rats subjected to 2 mg/kg LPS injection(Escherichia coli serotype O55:B05 type, Sigma Aldrich, St. Louis,CO, US) and an observation time of 1 h, 6 h, 24 h were used as con-trol for LBP elevation (n = 6/group). Serum and effluent protein lev-els as well as hepatic-mRNA and protein levels of LBP wereexamined after warm I/R, CI and LTx. LPS translocation and hepaticmRNA expression of inflammatory cytokines was observed afterLTx.

In in vitro experiment, the effluent was used to co-stimulate ratperitoneal macrophages with LPS. Effluent was collected at definedtime-points of cold liver storage. The LBP in effluent was measuredby western blot. The time point with highest LBP concentration

was chosen for macrophage stimulation. The macrophages wereco-cultured with effluent (50 ll) and 0.33 ng/ml LPS. The same vol-ume 0.9% NaCl and 0.33 ng/ml LPS was added in control group. Inthe LBP blockade experiments, macrophages were stimulated witheffluent and LPS in combination with LBP inhibitory peptide(80 lg/ml). The culture suspensions were taken after 4 h stimula-tion and the TNF-a levels was detected by ELISA.

2.2. Animals

Male inbred Lewis rats, purchased from the Central AnimalFacility of the University Hospital Essen, and weighing 300–350 g, were employed in this study. All animals were housed understandard animal care conditions and had free to access to waterand rat chow ad libitum. All procedures were carried out accordingto the German Animal Welfare Legislation. Animal experimentswere approved by the Bezirksregierung Düsseldorf. All injectionand operative procedures were performed under inhalation anes-thesia with 3 % isoflurane (Sigma Delta, London, UK).

2.3. Surgical models: selective in vivo liver WI/R, ex vivo liver CI, andLTx model

For the warm ischemia–reperfusion model, the procedure wasperformed as reported before [12]. The left hepatoduodenal liga-ment containing the hepatic artery, portal vein and bile duct ofthe left lateral and median liver lobes was clamped using a microvascular clamp. For the cold ischemia model, livers were explantedand subjected to cold ischemia as described previously in detail[14]. Liver grafts were subjected to 6 h cold ischemia prior to per-forming the transplantation procedure according to the cuff tech-nique described by Kamada [9]. Postoperative analgesia wasachieved by subcutaneous injection of buprenorphine (0.01 mg/kg)(TemgesicTM, Essex Pharma, Munich, Germany).

2.4. Enzyme-linked immunosorbent assay (ELISA)

For analysis of hepatic IL-6 and TNF-a level, commercially avail-able ELISA kits were used (R&D Systems, Minneapolis, US). All pro-cedures were performed according to the instructions of themanufacturers.

For measurement of serum LBP levels, a recently established,novel LPS-LBP ELISA system was used as described before (Manu-script submitted, Fang, et al., 2011 Oct). Standard 96 well ELISAplates were coated with 1 lg LPS (Sigma Aldrich, St. Louis, USA)in PBS. The plates were subsequently washed three times withPBST (0.01% Tween-20, pH 7.4) and blocked with 1% BSA in PBSTfor 2 h. 100 ll of each dilution of the calibrator and samples wereloaded and incubated for 2 h. The plates were incubated withmonoclonal mouse anti-LBP antibody (1:10,000, cell science, Can-ton, MA) for 2 h. After the plates were washed for three times, rab-bit anti-mouse IgG-H&L antibody (1:5000, Abcam, Cambridge, UK)was loaded and incubated for 1 h. And then, 100 ll 1:1 mixed ofH2O2 and Tetramethylbenzidine (BD, Franklin Lakes, US) wasadded to each well and incubated for 12–15 min. The color reac-tion was stopped by adding 50 ll of sulfuric acid (AppliChem,Darmstadt, Germany) and the plates were measured with the ELx808 ELISA plate reader (Bio-Tek Instruments Inc., Winooski, VT,US) at 450 nm. Quantification was based on a non-linear regressionstandard curve using Sigma Plot 10.0 (Systat-Software, Erkrath,Germany).

2.5. Quantitative polymerase chain reaction (PCR)

Total RNA was isolated from liver tissue using the RNeasy kit(Qiagen, Hilden, Germany). cDNA was synthesized by using the

Table 1Characteristics of primers and probes of selected genes.

Gene Forward primer Reverse primer Probea

LBP ATCCGGCTGAACACCAAG TGTCGGGGTACTTTCTGGTT #82IL-6 CCTGGAGTTTGTGAAGAACAACT GGAAGTTGGGGTAGGAAGGA #106TNF-a TGAACTTCGGGGTGATCG GGGCTTGTCACTCGAGTTTT #63IL-1ß GCTGACAGACCCCAAAAGAT AGCTGGATGCTCTCATCTGG #117HPRT GACCGGTTCTGTCATGTCG ACCTGGTTCATCATCACTAATCAC #95

a Universal probe library probes.

Fig. 1. Release of LBP during CI. During CI, hepatic effluents and liver tissues were collected hourly for 24 h cold storage. (A) Release of AST increased slightly, but notsignificantly during 7 h cold ischemic storage (p > 0.05). (B) LBP levels in effluents as quantified by ELISA were highest in samples obtained after 3 h, and then decreaseddrastically. LBP levels were undetectable after 7 h during clod ischemic storage. (C) LBP mRNA expression was not significantly altered during cold storage of liver. Data areshown as mean ± SD. �p < 0.05, ��p < 0.01 vs. 0 h.

H. Fang et al. / Cytokine 64 (2013) 71–78 73

First-Strand complementary DNA (cDNA) synthesis kit (Invitrogen,Carlsbad, USA) and quantified using Agilent bio-analyzer with RNAnano6000 kit (Agilent, Santa Clara, USA). CDNA-stock solution wasdiluted to 1 ng, and equal amounts of cDNA were used for each PCRreaction as described previously [26]. Primers and probes (RocheDiagnostics GmbH, Mannheim, Germany) were mixed with Bril-liant probe-based qPCR Master Mix kit (Agilent, Santa Clara, USA)and then diluted with deionized water up to 20 ll. The sense andantisense primers and the probes from the universal probe libraryare indicated in Table 1. Samples were run on an Mx3000P qPCRSystem (Stratagene, La Jolla, USA). Thermal cycling conditions con-sisted of a 10 min template denaturizing step at 95 �C, followed by50 cycles of 95 �C for 30 s, 50 �C for 30 s and 72 �C for 30 s. Stan-dard curve was generated using a serial dilution of a normal sam-ple. Gene expression was normalized using hypoxanthine guaninephosphoribosyltransferase (HPRT) to compensate for errors whendiluting the cDNA stock solution. The fold change was calculatedusing a normal liver tissue sample as reference sample.

Fig. 2. Induction of TNF-a production in peritoneal macrophage assay afterstimulation with LPS. The effluent with the highest LBP concentration, obtainedafter 3 h of cold storage was used. Macrophages were stimulated with LPS (0.33 ng/ml), effluent (50 ll) and LBP inhibitory peptide-LBPK95A 80 lg/ml. The concentra-tion of TNF-a was significantly increased after stimulation with LPS + effluent whencompared with stimulation with LPS only (p < 0.001). The elevation of TNF-a wasinhibited markedly when adding the LBP inhibitory peptide-LBPK95A (p < 0.001).The experiment was performed in triplicates with similar results. Data are shown asmean ± SD.

2.6. Gel electrophoresis and western blotting

Liver tissue was lysed in lysis buffer (Tris 50 mM pH 7.4, NaCl150 mM, 1% Igepal CA-630, 1 mg/ml leupetin, 1 mg/ml pepstatin)on ice. Protein was quantified with the BCA protein Assay Kit(Pierce biotechnology, Rockford, US). An equal amount of protein(15 lg) of the total liver lysate was loaded on 12% sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS–PAGE) gels.

74 H. Fang et al. / Cytokine 64 (2013) 71–78

Samples were separated for 90 min at 120 V and transferred toAmersham Hybond™-P membrane (GE Healthcare, Munich,Germany). The membranes were then probed with a goat poly-clonal antibody to LBP (1:100, Santa Cruz, CA, USA) for LBP expres-sion. Signals were detected with Lumilight western blot substrate(Roche, Basel, Switzerland) and exposed to X-ray film (GEHealthcare, Munich, Germany) for autoradiography. Subsequentlythe membranes were stripped with stripping buffer (Pierce, Rock-ford, IL) for 15 min at room temperature and immunoblotted withrabbit anti-GAPDH (1:20,000, Sigma–Aldrich, St. Louis, MO, US).Digitalization of films was performed using a film scanner (EpsonV750, Nagano, Japan). The signal intensity was quantified withImage J program (NIH, Bethesda, USA), and compared with a cali-bration curve constructed with serially diluted rat serum obtained6 h after LPS injection. The experiments were performed withpooled samples and results confirmed using single samples.

2.7. Isolation and culture of peritoneal macrophages

Isolation and culture of peritoneal macrophages were per-formed as described by Liu et al. [13]. Peritoneal macrophageswere harvested by two times of peritoneal washes with 20 ml ofphosphate buffered saline (PBS) buffer containing 3 units/ml of

Fig. 3. Elevation of LBP expression levels after WI/R. Serum LBP levels were measured byinjection (B) and followed a similar kinetic pattern. After WI/R, Serum LBP levels were siafter reperfusion. The expression of hepatic LBP mRNA (C and D) and protein was investigand protein expression was significantly increased after 6 h of reperfusion. Data are sho

heparin. The cells were washed three times (300 g � 10 min) andthen cultured in RPMI1640 medium supplemented with 10% fetalcalf serum, 2 mM glutamine, 100 IU/ml penicillin and 100 lg/mlstreptomycin. Cells were plated in 24 well plates at a density of3 � 105 cells and cultured at 37 �C under a gas phase of air/CO2

(95:5). The non-adherent cells were discarded after 3 h, and theadherent cells were used for further experiments. Peritoneal mac-rophage purity exceeded 98% as determined by CD68 staining, andviability typically was >96% as determined by trypan blue exclu-sion assay.

2.8. Preparation and treatment of effluent

Effluent was collected every hour during 24 h cold storage insaline. The LBP-concentration in the effluent was determined bywestern blot. The effluent with the highest LBP concentrationwas chosen for the macrophage stimulation assay. The macro-phages were co-cultured with effluent (50 ll) and 0.33 ng/mlLPS. During post-operative LPS translocation, a lower range ofLPS concentrations were usually observed [7] and therefore anLPS concentration of 0.33 ng/ml was chosen [11]. The same volumeof 0.9% NaCl and 0.33 ng/ml LPS, but not effluent, was added incontrol group. In the LBP blockade group, macrophages were

ELISA. Upregulation of serum LBP levels was observed after WI/R (A) and after LPSgnificantly increased as early as 0.5 h, reached a peak at 6 h and remained high 24 hated by quantitative PCR and western blot (E and F), respectively. Hepatic LBP mRNAwn as mean ± SD. �p < 0.05, ��p < 0.01 vs. normal control rats.

Fig. 4. Elevation of LBP expression levels after LTx. LBP levels were observed 24 h after rat LTx. Serum LBP (A), hepatic LBP mRNA (B), and protein (C) were measured usingELISA, quantitative PCR and western blot, respectively. Both serum and hepatic LBP levels were significantly increased after LTx. Data are shown as mean ± SD. �p < 0.01 vs.normal control rats.

H. Fang et al. / Cytokine 64 (2013) 71–78 75

stimulated with effluent, 0.33 ng/ml LPS and 80 lg/ml LBP inhibi-tory peptide-LBPK95A (RVQGRWKVRASFFK) [2]. The peptide wassynthesized in-house using an Fmoc standard procedure on anABI 433A-Peptide-Synthesizer. Before adding the stimulators(effluent, LPS, and/or peptide) to the medium, all reagents weremixed and incubated together briefly.

2.9. LPS immunohistochemistry

Liver tissue was fixed in 4.5% buffered formalin for at least 24 h.Paraffin embedding was performed using standard techniques andsections (4 lm) were cut. After de-paraffinization and rehydration,antigen retrieval was performed in a water bath using citrate buf-fer (10 mM Citric Acid, pH 6.0) for 20 min at 100 �C. Slides werewashed 3 times with Tris Buffered Saline with Tween 20 (TBST).Nonspecific protein binding was blocked using 100 ll serum freeblocking buffer (Dako, Glostrup, Denmark). Sections were incu-bated with diluted (1/100) polyclonal mouse anti-LPS antibody(Abcam, Cambridge, UK) for 15 min at room temperature. Slideswere washed 3 times with TBST. Signals were amplified followingthe instructions of CSA II biotin-free tyramide signal amplificationsystem (Dako, Glostrup, Denmark). Sections were counterstainedwith Hematoxylin for 5 min. Slides were visualized using a Virtualslide scanner (Hamamatsu Electronic Press Co., Ltd., Lwata, Japan).The hepatocytes with cytoplasmic staining were identified as LPSpositive cells.

2.10. Statistical analysis

All values were expressed as mean ± SD. All statistical calcula-tions were performed by using Sigma Stat (ver. 3.5.54; Systat Soft-ware GmbH, Erkarth, Germany). Groups of animals were comparedemploying Student’s t-test in case of normal distribution of thedata. If data were not normally distributed, the Mann–Whitneyrank sum test was employed to compare sets of data in different

experimental groups. A p-value below 0.05 was considered statis-tically significant.

3. Results

3.1. CI caused release of LBP into the effluent

Effluent AST levels were slightly increased with time 7 h duringclod ischemic storage of livers (Fig. 1A, p > 0.05). Cold ischemicstorage caused hepatic injury, which became apparent after 8 has indicated by a release of AST into the effluent (data not shown).To determine whether LBP was also released during cold ischemicstorage of liver, the effluent was subjected to the LBP-ELISA. Asshown in Fig. 1B, the release of LBP was detected as early as 1 h,and then peaked at 3 h. After 7 h, LBP was undetectable in theeffluent during cold storage (data not shown).

We also investigated whether cold ischemia upregulated LBP-expression in the cold preserved organ. As shown in Fig. 1C, hepaticLBP mRNA expression was not induced during cold ischemicstorage.

3.2. LBP in effluent enhanced LPS induced pro-inflammatory cytokinesynthesis

LBP has been implicated in LPS induced inflammation. Wewanted to determine whether LPS released from ischemic livercould modulate LPS induced inflammation. Effluent with the high-est concentration of LBP, obtained after 3 h of cold storage, was se-lected for the macrophage stimulation assay. Effluent with orwithout addition of LBP inhibitory peptide-LBPK95A was addedto cultured peritoneal macrophage prior to LPS stimulation. After6 h of stimulation with LPS, the cell culture supernatant was col-lected and the release of TNF-a was measured by ELISA. As shownin Fig. 2, TNF-a levels were significantly increased after stimulationwith 3 h-effluent and LPS, when compared with LPS only

Fig. 5. Elevation of LPS translocation and hepatic inflammatory cytokines after LTx. (A) Immunohistochemical staining of LPS demonstrated that LPS was translocated intohepatocytes after LTx (original magnification �200). Representative images from six rats/group were selected. Inflammatory cytokines TNF-a, IL-6 and IL-1ß mRNAexpression levels were measured by quantitative PCR in normal liver tissue, and tissue obtained after LTx. mRNA levels for TNF-a (B), IL-6 (C) and IL-1ß (D) were significantlyelevated after LTx. Data are shown as mean ± SD. �p < 0.05 vs. normal control rats.

76 H. Fang et al. / Cytokine 64 (2013) 71–78

(3 h-effluent + LPS group: 11.09 ± 0.58 ng/ml, LPS group:9.07 ± 0.22 ng/ml, p < 0.001). However, the synthesis of TNF-awas significantly attenuated when pretreating the effluent withLBP inhibitory peptide-LBPK95A (3 h-effluent + LPS group: 11.09 ±0.58 ng/ml, 3 h-effluent + LBPK95A + LPS group: 5.14 ± 0.91 ng/ml,p < 0.001). This finding indicated that LBP released during CI couldindeed modulate the LPS induced inflammatory response.

3.3. WI/R caused release of LBP into serum and upregulation ofintrahepatic mRNA and protein levels

Rats subjected to 2 mg/kg LPS injection and an observation timeof 1 h, 6 h, 24 h were used as control for LBP elevation. Both serumand hepatic LBP levels were significantly increased at 6 h and 24 hafter LPS injection (Fig. 3).

To assess the extracellular release of LBP after WI/R injury, ser-um LBP concentrations were quantified using the newly developedELISA-Assay. LBP was released into the serum as early as 0.5 h, andreached a peak 6 h after reperfusion (0.5 h: 13.83 ± 1.39 lg/ml, 6 h:45.00 ± 9.20 lg/ml, 24 h: 20.30 ± 7.16 lg/ml, vs. normal controls:2.05 ± 1.00 lg/ml, p < 0.05). We also detected the hepatic LBP-mRNA and protein expression after WI/R by quantitative PCR andwestern blot, respectively. A significant elevation of hepatic mRNA(fold increase to normal: 7.16 ± 1.93, p < 0.01) and protein (fold in-crease to normal: 3.27 ± 2.00, p < 0.05) was observed after 6 h, and

then increased with reperfusion time up to 24 h after reperfusion(fold increase to normal: mRNA 12.77 ± 3.04; protein: 10.30 ±7.69, p < 0.01) (Fig. 3).

3.4. LTx caused release of LBP into serum and upregulation ofintrahepatic mRNA and protein levels

To determine whether LBP expression was upregulated afterLTx, rats were subjected to LTx, and serum and hepatic LBP weredetected. As shown in Fig. 4, serum LBP levels were elevatedapproximately ten fold at 24 h after LTx using ELISA-assay (24 h:24.90 ± 3.47 lg/ml, vs. normal controls: 2.05 ± 1.00 lg/ml,p < 0.01). In parallel, the hepatic expression levels of LBP, bothmRNA and protein were also significantly upregulated, more than5-fold increase at 24 h after LTx (fold increase to normal: mRNA9.08 ± 3.01, p < 0.01; protein: 6.59 ± 5.18, p < 0.01).

3.5. LTx caused the translocation of LPS into hepatocytes

To determine whether LPS was translocated after LTx, LPSimmunohistochemical staining was employed in normal liver tis-sues and in livers subjected to 6 h CI, LTx and 24 h observationtime. As shown in Fig. 5, no positive staining signals were detectedin normal liver tissue. However, in liver sections obtained after LTx,cytoplasmic staining in almost all hepatocytes, but not in other cell

H. Fang et al. / Cytokine 64 (2013) 71–78 77

types became visible. This finding suggested that the LPS did trans-locate to the liver and was internalized into hepatocytes.

The expression of mRNA of TNF-a, IL-6 and IL-1ß in liver graftswas determined by quantitative PCR. As expected, an increasedexpression of these cytokines was observed after LTx in compari-son to the control group (p < 0.05). Taken together, these data sug-gest that the increased hepatic inflammatory cytokines expressionmay be associated with LPS translocation into hepatocytes.

4. Discussion

Since the first description of LBP by Ulevitch and colleagues[23], many studies were performed to investigate the role of LBPin the LPS-induced inflammatory response, mainly in the contextof infectious diseases. Increased serum levels of LBP are not limitedto infections, but do also occur after surgical procedures as re-ported by Kudlova et al. They observed that LBP-levels increasedsubsequent to cardiac surgery [10]. We previously demonstratedthat upregulation of LBP also occurred after partial hepatectomy,and that serum LBP levels were associated with the remnant livermass [3]. These data together with similar findings in the past ledto the perception that LBP was an acute phase protein andprompted investigation of LBP as biomarker.

However, recent reports suggest that LBP is more than a purebiomarker after major surgery [8,17]. Hiki et al. demonstrated thata significant increase of serum LBP levels was observed in patientsafter major abdominal surgery, and correlated this finding with thesystemic levels of inflammatory cytokines. The observation impliesthat LBP may play a role in regulating the biologic activity of circu-lating LPS [8].

Here, we provide additional evidence for the importance of LBPby investigating the role of LBP after LTx and I/R injury. We dem-onstrated that LBP was upregulated during CI storage in vitroand was also increased after WI/R and LTx in vivo. Of note, wefound that the release of LBP peaked at 3 h after cold storage,which was similar as the CI time in living donor liver transplanta-tion of the patients [15].

The capacity of LBP released by the ischemic liver to augmentan inflammatory response in vitro was determined in the macro-phage stimulation assay and further clarified by the attenuationof the LPS induced inflammatory response via LBP inhibitory pep-tide. Our finding is inline with the observation of Lamping et al.,who reported that LBP augmented the LPS-induced inflammatoryresponse in vitro, indicated by an increase in TNF-a level ofcultured macrophages [11]. These observations were furthersupported by Arana et al. [2]. They reported that the release ofTNF-a was inhibited when using the LBP inhibitory peptide. Theseresults were confirmed by others in vivo. Su et al. found that block-ing LBP-LPS interactions using LBP inhibitory peptide protectedfrom liver injury induced by acetaminophen [21]. Minter et al. re-ported that high LBP-levels were deleterious in a mouse bile ductligation model, where LPS-levels are presumingly high [17]. In aseparate study, we observed that the inflammatory response toLPS was massively enhanced in rats with elevated LBP-levels [4].Blockade of LBP using inhibitory peptide attenuated inflammatoryinjury and improved the outcome.

LPS and bacterial translocation are observed after liver trans-plantation, as reported by Yokohama et al. [27] in patients andby Goto [7] and Tsoulfas [24] in an experimental study. Tsoulfasobserved that, the upregulation of serum LPS levels after reperfu-sion was accompanied with the activation of LBP-TLR-NF-kB signalaxis [24]. In the present study, we demonstrated that LBP wasupregulated during CI storage, and the released LBP augment LPSinduced inflammatory response. We detected intracytoplasmic sig-nals for LPS in hepatocytes by immuno-histochemical staining,

indicating that LPS was translocated to the cytoplasm of hepato-cytes after LTx. We could demonstrate that LPS translocation wasassociated with an increased hepatic expression of LBP. Increasedhepatic expression of LBP leads to increased LPB protein concentra-tions. Increased LBP concentrations are associated with an en-hanced inflammatory response induced by LPS.

5. Conclusions

In conclusion, as LTx causes LPS-translocation as well as upreg-ulation of LBP, this proinflammatory pathway could contribute tothe ischemia–reperfusion associated inflammatory response. Inter-fering with this pathway by blocking LBP could represent a novelanti-inflammatory strategy in this LTx setting.

Acknowledgements

We gratefully thank Dr. Roland Kaufmann for providing theLBPK95A peptide. This study was supported by German FederalMinistery for Education and Research (BMBF) Virtual LiverNetwork

References

[1] Alexander C, Rietschel ET. Bacterial lipopolysaccharides and innate immunity. JEndotoxin Res 2001;7:167–202.

[2] Arana MJ, Vallespi MG, Chinea G, Vallespi GV, Rodriguez-Alonso I, Garay HE,et al. Inhibition of LPS-responses by synthetic peptides derived from LBPassociates with the ability of the peptides to block LBP–LPS interaction. JEndotoxin Res 2003;9:281–91.

[3] Fang H, Liu A, Dirsch O, Sun J, Jin H, Lu M, et al. Serum LBP levels reflect theimpaired synthetic capacity of the remnant liver after partial hepatectomy inrats. J Immunol Methods 2012.

[4] Fang H, Liu A, Sun J, Kitz A, Dirsch O, Dahmen U. Granulocyte colonystimulating factor induces lipopolysaccharide (LPS) sensitization viaupregulation of LPS binding protein in Rat PLoSOne 8:e56654; 2013.

[5] Fondevila C, Busuttil RW, Kupiec-Weglinski JW. Hepatic ischemia/reperfusioninjury – a fresh look. Exp Mol Pathol 2003;74:86–93.

[6] Garcia-Tsao G, Albillos A, Barden GE, West AB. Bacterial translocation in acuteand chronic portal hypertension. Hepatology 1993;17:1081–5.

[7] Goto M, Takei Y, Kawano S, Tsuji S, Fukui H, Fushimi H, et al. Tumor necrosisfactor and endotoxin in the pathogenesis of liver and pulmonary injuries afterorthotopic liver transplantation in the rat. Hepatology 1992;16:487–93.

[8] Hiki N, Berger D, Mimura Y, Frick J, Dentener MA, Buurman WA, et al. Releaseof endotoxin-binding proteins during major elective surgery: role of solubleCD14 in phagocytic activation. World J Surg 2000;24:499–506.

[9] Kamada N, Calne RY. Orthotopic liver transplantation in the rat. Techniqueusing cuff for portal vein anastomosis and biliary drainage. Transplantation1979;28:47–50.

[10] Kudlova M, Kunes P, Kolackova M, Lonsky V, Mandak J, Andrys C, et al.Lipopolysaccharide binding protein and sCD14 are not produced as acutephase proteins in cardiac surgery. Mediators Inflamm 2007;2007:72356.

[11] Lamping N, Dettmer R, Schroder NW, Pfeil D, Hallatschek W, Burger R, et al.LPS-binding protein protects mice from septic shock caused by LPS or gram-negative bacteria. J Clin Invest 1998;101:2065–71.

[12] Liu A, Dirsch O, Fang H, Sun J, Jin H, Dong W, et al. HMGB1 in ischemic andnon-ischemic liver after selective warm ischemia/reperfusion in rat.Histochem Cell Biol 2011;135:443–52.

[13] Liu A, Fang H, Dirsch O, Jin H, Dahmen U. Early release of macrophagemigration inhibitory factor after liver ischemia and reperfusion injury in rats.Cytokine 2012;57:150–7.

[14] Liu A, Jin H, Dirsch O, Deng M, Huang H, Brocker-Preuss M, et al. Release ofdanger signals during ischemic storage of the liver: a potential marker of organdamage. Mediators Inflamm 2010;2010:436145.

[15] Malago M, Testa G, Frilling A, Nadalin S, Valentin-Gamazo C, Paul A, et al. Rightliving donor liver transplantation: an option for adult patients: singleinstitution experience with 74 patients. Ann Surg 2003;238:853–62.

[16] Martin TR, Mathison JC, Tobias PS, Leturcq DJ, Moriarty AM, Maunder RJ, et al.Lipopolysaccharide binding protein enhances the responsiveness of alveolarmacrophages to bacterial lipopolysaccharide. Implications for cytokineproduction in normal and injured lungs. J Clin Invest 1992;90:2209–19.

[17] Minter RM, Bi X, Ben Josef G, Wang T, Hu B, Arbabi S, et al. LPS-binding proteinmediates LPS-induced liver injury and mortality in the setting of biliaryobstruction. Am J Physiol Gastrointest Liver Physiol 2009;296:G45–54.

[18] Montalvo-Jave EE, Escalante-Tattersfield T, Ortega-Salgado JA, Pina E, GellerDA. Factors in the pathophysiology of the liver ischemia-reperfusion injury. JSurg Res 2008;147:153–9.

78 H. Fang et al. / Cytokine 64 (2013) 71–78

[19] Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, Du X, et al. Defective LPSsignaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science1998;282:2085–8.

[20] Poltorak A, Ricciardi-Castagnoli P, Citterio S, Beutler B. Physical contactbetween lipopolysaccharide and toll-like receptor 4 revealed by geneticcomplementation. Proc Natl Acad Sci USA 2000;97:2163–7.

[21] Su GL, Hoesel LM, Bayliss J, Hemmila MR, Wang SC. Lipopolysaccharidebinding protein inhibitory peptide protects against acetaminophen-inducedhepatotoxicity. Am J Physiol Gastrointest Liver Physiol 2010;299:G1319–25.

[22] Teoh NC, Farrell GC. Hepatic ischemia reperfusion injury: pathogenicmechanisms and basis for hepatoprotection. J Gastroenterol Hepatol2003;18:891–902.

[23] Tobias PS, Soldau K, Ulevitch RJ. Isolation of a lipopolysaccharide-bindingacute phase reactant from rabbit serum. J Exp Med 1986;164:777–93.

[24] Tsoulfas G, Takahashi Y, Ganster RW, Yagnik G, Guo Z, Fung JJ, et al. Activationof the lipopolysaccharide signaling pathway in hepatic transplantationpreservation injury. Transplantation 2002;74:7–13.

[25] Vardanian AJ, Busuttil RW, Kupiec-Weglinski JW. Molecular mediators of liverischemia and reperfusion injury: a brief review. Mol Med 2008;14:337–45.

[26] Xing W, Deng M, Zhang J, Huang H, Dirsch O, Dahmen U. Quantitativeevaluation and selection of reference genes in a rat model of extended liverresection. J Biomol Technol 2009;20:109–15.

[27] Yokoyama I, Todo S, Miyata T, Selby R, Tzakis AG, Starzl TE. Endotoxemia andhuman liver transplantation. Transplant Proc 1989;21:3833–41.