of May 20, 2018.This information is current as

Liver Ischemia/Reperfusion InjuryImmune Function in the Pathophysiology of

T Cells Mediate Innate+CD4+CXCR3

Strieter, Ronald W. Busuttil and Jerzy W. Kupiec-WeglinskiQiao, Charles Lassman, John A. Belperio, Robert M. Yuan Zhai, Xiu-da Shen, Wayne W. Hancock, Feng Gao, Bo

http://www.jimmunol.org/content/176/10/6313doi: 10.4049/jimmunol.176.10.6313

2006; 176:6313-6322; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/176/10/6313.full#ref-list-1

, 6 of which you can access for free at: cites 34 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2006 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

CXCR3�CD4� T Cells Mediate Innate Immune Function inthe Pathophysiology of Liver Ischemia/Reperfusion Injury1

Yuan Zhai,* Xiu-da Shen,* Wayne W. Hancock,§ Feng Gao,* Bo Qiao,* Charles Lassman,†

John A. Belperio,‡ Robert M. Strieter,‡ Ronald W. Busuttil,* and Jerzy W. Kupiec-Weglinski2*

Ischemia-reperfusion injury (IRI), an innate immune-dominated inflammatory response, develops in the absence of exogenousAgs. The recently highlighted role of T cells in IRI raises a question as to how T lymphocytes interact with the innate immunesystem and function with no Ag stimulation. This study dissected the mechanism of innate immune-induced T cell recruitment andactivation in rat syngeneic orthotopic liver transplantation (OLT) model. Liver IRI was induced after cold storage (24–36 h) at4°C in University of Wisconsin solution. Gene products contributing to IRI were identified by cDNA microarray at 4-h post-transplant. IRI triggered increased intrahepatic expression of CXCL10, along with CXCL9 and 11. The significance of CXCR3ligand induction was documented by the ability of neutralizing anti-CXCR3 Ab treatment to ameliorate hepatocellular damageand improve 14-day survival of 30-h cold-stored OLTs (95 vs 40% in controls; p < 0.01). Immunohistology analysis confirmedreduced CXCR3� and CD4� T cell infiltration in OLTs after treatment. Interestingly, anti-CXCR3 Ab did not suppress innateimmune activation in the liver, as evidenced by increased levels of IL-1�, IL-6, inducible NO synthase, and multiple neutrophil/monokine-targeted chemokine programs. In conclusion, this study demonstrates a novel mechanism of T cell recruitment andfunction in the absence of exogenous Ag stimulation. By documenting that the execution of innate immune function requiresCXCR3�CD4� T cells, it highlights the critical role of CXCR3 chemokine biology for the continuum of innate to adaptiveimmunity in the pathophysiology of liver IRI. The Journal of Immunology, 2006, 176: 6313–6322.

I schemia-reperfusion (IR)3 injury (IRI) is a common cause ofhepatocellular dysfunction that may develop in various clin-ical settings, including low flow states, surgical procedures,

or during organ procurement for transplantation (1, 2). Indeed, IRI,an Ag-independent component of the harvesting insult represents aserious problem affecting transplantation outcome as it causes upto 10% of early organ failure, and increases the risk of both acuteand chronic rejection (3, 4). Additionally, because of the higher sus-ceptibility of marginal livers to ischemia, IRI contributes to the acuteshortage of livers available for transplantation. Thus, minimizing theadverse effects of IRI might increase the donor pool and the numberof patients that can successfully undergo liver transplantation.

The ischemic insult activates Kupffer cells, and to a lesser de-gree endothelial cells and hepatocytes, leading to the formation ofreactive oxygen species and secretion of proinflammatory cyto-kines/chemokines (5, 6). Oxidant stress directly damages endothe-

lial cells and hepatocytes, while the soluble factors are responsiblefor neutrophil, monocyte, and T cell, recruitment. Putative mech-anisms by which IR activates components of the innate immunesystem remain unclear. We have recently shown that innate im-mune receptor TLR4 signaling is essential for the development ofhepatocellular damage in a murine warm liver IRI model (7). Al-though T lymphocytes have historically been thought marginal tothe process, recent studies have documented their key role in theearly phase of IRI (8). Thus, T cell-deficient and CD4-deficientmice were protected from IRI, despite normal levels of neutrophiland macrophage infiltration (9–12). Moreover, adoptive transfer ofT cells, particularly the CD4� T subset, restored IRI in otherwiseT cell-deficient hosts. The question thus arises as to how T cellsbecome involved in the absence of exogenous Ags, and how Tcells function in the pathogenesis of IRI?

The current study documents that liver IR triggers intrahepaticproduction of multiple chemokines, in particular CXCL9, 10, 11,which target selectively CXCR3�CD4� T cells. Indeed, the ap-plication of anti-CXCR3 Ab ameliorated hepatocellular damagewith resultant increased survival of syngeneic orthotopic livertransplant (OLT) recipients after extended periods of cold preser-vation. Interestingly, anti-CXCR3 Ab did not inhibit IR-inducedinnate immune activation. With the immunohistology demonstra-tion that intrahepatic CD4�CXCR3� lymphocytes were selec-tively depleted in the protected livers, this study reveals a novelmechanism by which CXCR3 T cells mediate the innate immunefunction leading to hepatocellular damage, and highlights the crit-ical role of CXCR3 chemokines in the pathophysiology ofliver IRI.

Materials and MethodsRats

LEW rats (250–300 g) were obtained from Harlan Sprague Dawley,housed in the University of California (Los Angeles, CA) (UCLA) animal

*The Dumont-University of California Los Angeles (UCLA) Transplant Center, De-partment of Surgery, Division of Liver and Pancreas Transplantation, †Department ofPathology and Laboratory Medicine, ‡Department of Medicine, Division of Pulmo-nary and Critical Care Medicine, David Geffen School of Medicine, UCLA, LosAngeles, CA 90095; and §Biesecker Pediatric Liver Center and Department of Pa-thology and Laboratory Medicine, Children’s Hospital of Philadelphia and Universityof Pennsylvania, Philadelphia, PA 19104

Received for publication September 8, 2005. Accepted for publication February27, 2006.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by National Institutes of Health Grants RO1 DK062357,AI23847, and AI42223 (to J.W.K.-W.), The Roche Transplantation Foundation Grant(to Y.Z.), and The Dumont Research Foundation.2 Address correspondence and reprint requests to Dr. Jerzy W. Kupiec-Weglinski,The Dumont-UCLA Transplant Center, 77-120 CHS, Box 957054, 10833 Le ConteAvenue, Los Angeles, CA 90095. E-mail address: jkupiec@mednet.ucla.edu3 Abbreviations used in this paper: IR, ischemia-reperfusion; IRI, IR injury; OLT,orthotopic liver transplantation; MPO, myeloperoxidase; iNOS, inducible NO syn-thase; sALT, serum alanine aminotransferase.

The Journal of Immunology

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

facility under specific pathogen-free conditions, and received human careaccording to the criteria outlined in the “Guide for the Care and Use ofLaboratory Animals” prepared by the National Academy of Sciences andpublished by the National Institute of Health (National Institutes of Healthpublication 86-23, revised 1985).

Cold ischemia and syngeneic OLT

Livers were harvested from LEW rats, stored for 0, 24, 30, or 36 h at 4°Cin University of Wisconsin (UW) solution, and then transplanted into syn-geneic LEW recipients, with revascularization without hepatic artery re-construction (13, 14). In the treatment group, LEW livers were stored at4°C in UW solution for 30 h before transplantation into syngeneic rats thatwere treated with polyclonal rabbit-anti-murine CXCR3 or normal rabbitserum (24 h before and at the time of OLT; 1 ml i.v.). The production,characterization and in vivo efficacy of this anti-CXCR3 serum in mousemodels have been described (13). However, this antiserum cross-reactswith rat CXCR3. Indeed, the sequence of murine peptide (N-terminal 16-mer of CXCR3: PYDYGENESDFSDSPP) used to generate polyclonal Abis identical with that of the rat CXCR3 peptide. Hence, we assume that theinfused serum binds rat and mouse CXCR3 with similar specificity. OLTrecipients were followed for survival and serum alanine aminotransferase(sALT) levels, an indicator of hepatocellular injury. Separate groups of ratswere sacrificed at 4 or 24 h, and OLT samples were collected for RNA/protein isolation, as well as for histology/immunohistology evaluations.

Immunopathology

For histological examination, portions of each liver were fixed in formalin,paraffin-embedded and stained by H&E. For immunohistology, cryostatsections of snap-frozen samples were stained using mouse anti-rat mAbsdirected against T cell subsets (BD Pharmingen), and polyclonal Abs to ratIFN-�-inducible protein-10 (IP-10) and CXCR3 (Santa Cruz Biotechnol-ogy). Labeling was detected using an Envision immunoperoxidase kit(DakoCytomation) and sections were counterstained with hematoxylin.Control sections were incubated with isotype-matched mAbs, purified rab-bit or goat IgG, and polyclonal Abs preabsorbed overnight using corre-sponding IP-10 and CXCR3 peptides (Santa Cruz Biotechnology). Quan-titation of CD4�, CD8�, and CXCR3� cells was performed blindly, using10 consecutive high power fields of each graft. Sections were cut from twodifferent levels per graft and three grafts per group per time point wereevaluated; data were expressed as mean � SD cells per field.

cDNA microarray and data analysis

Total tissue RNA was prepared using TRIzol reagent (Invitrogen LifeTechnologies). Sample preparation for cDNA microarray and microarrayprocessing were performed by the UCLA Microarray Core, using Rat Ge-nome U34A gene chips from Affymetrix. The raw data was analyzed usingthe Gene Chip Analysis Suite (Affymetrix), and expression data were im-ported into Microsoft Excel 2000 for further analysis and plotting. Mi-croarray data were normalized to the housekeeping gene (GAPDH). Toidentify the genes significantly increased by IR, genes with “absent” call inischemic sample chips were eliminated when comparing two experimentalconditions, average expression levels were calculated and fold increasesbetween groups were used to sort data. The cytokine/chemokine expressiongrid was made with GeneCluster 2 (Whitehead Institute Center for GenomeResearch) using a subset expression database of cytokine/chemokine genesselected from the U34A database.

Quantitative RT-PCR

Five micrograms of RNA was reverse-transcribed into cDNA using ran-dom hexamers and Omniscript reverse transcriptase (Qiagen). QuantitativePCR was performed using the DNA Engine with Chromo 4 Detector (MJResearch). In a final reaction volume of 25 �l, the following were added:1� SuperMix (Platinum SYBR Green qPCR kit; Invitrogen Life Technol-ogies), cDNA and 0.5 mM of each primer. Amplification conditions were:50�C (2 min), 95°C (5 min) followed by 50 cycles of 95°C (15 s), 60°C(30 s). Primers used to amplify a specific fragment of �-actin, CXCL9, 10,11, IL-1�, inducible NO synthase (iNOS), RANTES, CD86, MCP-1,MIP-2 are listed (Table I).

Western blots

Protein was extracted from liver samples with PBS/TDS buffer (50 mMTris, 150 mM NaCl, 0.1% SDS, 1% sodium deoxycholate, and 1% TritonX-100 (pH 7.2)). Proteins (30 �g/sample) in SDS-loading buffer (50 mMTris, (pH 7.6), 10% glycerol, 1% SDS) were subjected to 20% SDS-PAGEand transferred to nitrocellulose membrane (Bio-Rad). The gel was stained

with Coomassie blue to document equal protein loading. Membranes wereblocked with 5% dry milk and 0.1% Tween 20 (USB) in PBS and incu-bated with rabbit anti-rat IP-10 (Cell Science) or �-actin Ab (Santa CruzBiotechnology). Membranes were then washed and incubated with HRP-conjugated donkey anti-rabbit IgG (Amersham Biosciences).

Myeloperoxidase (MPO) assay

The presence of MPO, an enzyme specific for neutrophils, was used as anindex of neutrophil accumulation in the liver (14). Briefly, the frozen tissuewas thawed and placed in 4 ml of iced 0.5% hexadecyltrimethyl-ammo-nium bromide and 50 mM potassium phosphate buffer solution with the pHadjusted to 5. Each sample was homogenized for 30 s and centrifuged at15,000 rpm for 15 min at 4°C. Supernatants were mixed with hydrogenperoxide-sodium acetate and tetramethyl-benzidine solutions. The changein absorbance was measured spectrophotometrically at 655 nm. One unit ofMPO activity was defined as the quantity of enzyme degrading 1 �molperoxide/min/g of tissue at 25°C.

Statistical analysis

All values are expressed as mean � SD; data were analyzed with an un-paired two-tailed Student’s t test, and p � 0.05 was considered statisticallysignificant.

ResultsThe development of IRI in syngeneic OLT

Transplanted organs experience “warm” ischemia during organprocurement and surgery, and “cold” ischemia during organ stor-age and transfer. Moreover, transplantation itself triggers a certainstress, which may alter gene expression profiles. Thus, we firstset-up a series of rat OLT groups in which LEW livers were har-vested and stored in UW solution at 4°C for varying periods (0, 24,30, 36 h) before being transplanted into syngeneic LEW recipients(15, 16). One hundred percent of recipients receiving 0 or 24 hcold-preserved OLTs survived long-term (�50 days; n � 6; datanot shown). However, livers stored for 36 h suffered irreversibleand lethal injury, as evidenced by 100% recipient death within 48 h(n � 6). The recipients received 30-h cold-preserved livers had a40% survival rate (n � 12), and was used as a “treatment group.”

Histological analysis showed that the extent and severity of liverparenchyma damage correlated with cold preservation time, andthat the full development of liver IRI was observed after 24-hreperfusion (Fig. 1). At 4 h of reperfusion, 24-h cold-stored OLTs(group no. 24-4) revealed minimal early necrosis with moderatesinusoidal congestion. There were scattered, small hepatocyte

Table I. Quantitative PCR primersa

�-actin A: tgccaacacagtgctgtctgAS: gagccaccaatccacacagag

CXCL10 A: ccaaccttccagaagcaccatAS: accgttcttgcgagagggat

CXCL9 S: gctcataaaggacaaaggtgccAS: acaggagagggtcagcttctt

CXCL11 S: ctagtgataaggaatcagcgatgcAS: tgtttgaggtctttgagggatttg

IL-1� S: gctgacagaccccaaaagatAS: agctggatgctctcatctgg

CD86 S: gctggagcaaggcatgatAS: tccggaacatgtcaacagttaat

iNOS S: ggtctttgaaatccctcctgaAS: agctcctggaaccactcgta

MCP-1 S: gatctctcttcctccaccactatgAS: gaatgagtagcagcaggtgagt

MIP-2 S: actctttggtccagagccatgAS: tggtagggtcgtcaggcatt

RANTES S: ctcaccgtcatcctcgttgAS: gagtggtgtccgagccata

a S, sense primer or 5� primer; AS, antisense primer or 3� primer.

6314 CXCR3 AND INNATE IMMUNITY IN IRI

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

clusters with hypereosinophilic cytoplasm and nuclear hypochro-masia (Fig. 1, a and b). However, the same livers after 24-h reper-fusion (group no. 24-24) developed moderate/severe necrosis withmultiple small foci of necrosis composed of clusters of hepatocyteswith loss of cellular detail, generally midzonal involving fewerthan half of the lobules. There were few large foci of necrosisextending across multiple lobules with periportal sparing (Fig. 1c).Although largely well-preserved at 4-h post-OLT, the 30-h cold-preserved livers after 24 h of reperfusion (group no. 30-24) devel-oped large foci of necrosis extending across multiple lobules (Fig.1d). In contrast, livers after 36 h of cold storage developed paren-chyma perivenular necrosis involving �50% of the lobules asearly as 4 h of reperfusion (group no. 36-4; Fig. 1, e and f). In-terestingly, the areas of necrosis were associated with increasednumbers of inflammatory cells in sinusoids. sALT levels increasedin all three OLT groups, as compared with those from native orliver grafts with no cold ischemia (group no. 0-4), indicating thathepatocellular damage did occur already after 4-h reperfusion andin the absence of histological manifestation of liver parenchymadamage (e.g., sALT � 514 � 186, vs 197 � 42 U/ml in 24-4 and0-4 groups, respectively; p � 0.01, n � 6/group; data not shown).

Cold ischemia induces intrahepatic CXCL10 expression

To identify gene products involved in the initiation of IRI, we usedcDNA microarray to analyze 4-h reperfused OLT samples, whichwere at the early stage of IRI as shown above; native livers and 0-hcold-preserved OLT (group no. 0–4) served as controls. Microar-ray data were uploaded to www.GeneSifter.Net, and pairwiseanalysis of expression data from 24- vs 36-h preserved OLT wasperformed to identify differentially expressed gene products, par-ticularly those related to immune activation. Data from each liversample were normalized to its own global mean expression levels.Four hundred and forty genes were found to have significantchanges (�2-fold) between these two groups. KEGG (Kyoto En-cyclopedia Genes and Genomics) pathway analysis of this gene listrevealed two molecular signaling pathways affected the most byz-score: MAP signaling, and TLR signaling, which is consistentwith our recent finding that TLR4 signaling is critical in triggeringliver IRI cascade (7). There were six chemotaxis-related genesidentified by gene ontology analysis. CXCL10 (IP-10, IFN-induc-ible protein 10) was the most up-regulated one (10-fold increase)associated with T cell chemotaxis. As warm ischemia and surgicalprocedure itself did not trigger its up-regulation, the cold ischemiawas the key factor in CXCL10 induction (group no. 0–4 vs groupno. 24-4, Fig. 2a). The extent of CXCL10 gene induction wascorrelated with the length of cold ischemia time (group no. 24-4 vsgroup no. 30-4 vs group no. 36-4). We confirmed IR-mediatedintrahepatic induction of CXCL10 by using quantitative RT-PCR(see Fig. 4) and immunohistology analyses (Fig. 2b). Among otherchemotaxis or immune-related genes associated with lethal liverIRI after 36 h of cold storage, IL-6, CXCL1 (KC, MIP-2), CCL2(MCP-1), LIX, iNOS and IL-1b were also significantly up-regu-lated (Fig. 3). CXCL1, in particular, was expressed in livers at veryhigh levels, similar to CXCL10. However, unlike CXCL10, itsinduction was initiated by warm and further increased by coldischemia (group no. 0–4 vs group no. 24-4, Fig. 3).

Anti-CXCR3 Ab ameliorates liver IRI

CXCL10 signals through a G-protein-coupled receptor, CXCR3expressed on activated T cells, and is capable to initiate T cellchemotaxis. There are two other CXCR3 ligands, CXCL9 (mono-kine induced by IFN-�) and CXCL11 (IFN-inducible T cell �chemoattractant), which may also affect CXCR3� T cell recruit-ment. As rat U34 array does not provide their expression data, we

FIGURE 1. Histology of rat IRI in OLT samples harvested at 4 or 24 hposttransplant following 24, 30, or 36 h of cold ischemia. a and b, Twenty-four hour cold-preserved OLTs after 4 h of reperfusion (24-4) withminimal necrosis and moderate sinusoidal congestion, scattered, smallhepatocyte clusters with hypereosinophilic cytoplasm and nuclear hypo-chromasia. c, Twenty-four hour cold-preserved OLTs after 24 h of reper-fusion (24-24) with moderate/severe necrosis and multiple small foci ofnecrosis, clusters of hepatocytes with loss of cellular detail, involvingfewer than half of the lobules. d, Thirty hour cold-preserved livers after24 h of reperfusion (30-24) with large foci of necrosis extending acrossmultiple lobules. e and f, Thirty-six hour cold-preserved OLTs at 4 h ofreperfusion (36-4) with early parenchyma, perivenular necrosis involving�50% of the lobules, and associated increased numbers of inflammatorycells in sinusoids. g and h, Thirty hour cold-preserved OLTs after anti-CXCR3 Ab treatment at 4 or 24 h of reperfusion (30-4T and 30-24T,respectively) with mild necrosis consisting of multiple small foci involvingclusters of 10–20 cells, generally midzonal in approximately one-third ofthe lobules. H&E staining; a, c, d, e, g, and h, Original magnification,�100; b and f, original magnification, �400. Representative of three to fivesamples per group for each time point.

6315The Journal of Immunology

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

undertook quantitative RT-PCR to assess CXCL 9 and 11 levels inour model. As shown in Fig. 4, cold ischemia induced expressionof both genes, similarly to CXCL10. However, the extent ofCXCL11 induction did not correlate well with the cold preserva-tion time. Thus, to inhibit T cell recruitment into livers, we rea-soned that targeting their shared CXCR3 could be more effectivethan neutralizing any single CXCR3 chemokine alone. A rabbitanti-CXCR3 serum (13) was administered to LEW rats of 30-hcold-preserved OLTs. As shown in Fig. 5, 40% of untreated ornormal rabbit serum-treated OLT recipients survived long-term(n � 6–10 rats/group). The CXCR3 blockade improved OLT sur-vival to 95% ( p � 0.01). Histological analysis of liver samplesharvested at 4- and 24-h posttransplant showed that Ab treatmentprevented development of severe hepatic necrosis. There was onlymild necrosis with multiple small foci involving clusters of 10–20cells, generally midzonal in approximately one-third of the lobules(Fig. 1, g and h).

To examine in vivo effects of anti-CXCR3 serum on rat periph-eral lymphocytes, blood samples were collected before, as well as4- and 24-h postserum administration, i.e., the time frame of liverIRI. A commercial rabbit anti-rat CXCR3 polyclonal Ab was usedto analyze cell surface CXCR3 expression on circulating lympho-cytes by FACS. As shown in Fig. 6a, our anti-CXCR3 serum andthe commercial anti-CXCR3 Ab had very similar in vitro stainingpattern of peripheral PBLs. Furthermore, anti-CXCR3 serum usedin this study bound to lymphocytes in vivo at 24 h postinfusion. Todetermine whether CXCR3-targeted therapy depleted peripherallymphocytes, we analyzed CD4 and CD8 T cell frequencies intotal PBLs. Interestingly, although transient decrease of total CD4and CXCR3�CD4 was noted at 4-h postserum injection, both fre-

quencies returned to normal by 24 h (Fig. 6b). The marked de-crease of CXCR3�CD8 T cells was observed at 4 and 24 h, and oftotal CD8 T cells at 24 h. Thus, the binding to CXCR3 Ag ratherthan target cell depletion represents the major mechanism bywhich anti-CXCR3 sera exerts its early acute in vivo effect in ratrecipients.

We performed immunohistology examination of 30-h cold-pre-served syngeneic OLTs that were harvested at various time points.Hepatocyte expression of CXCL10 was observed at 4-h posttrans-plant in the case of both IgG and anti-CXCR3 groups (Fig. 7, a andb). However, whereas isografts from IgG-treated rats showed amoderate and diffusely distributed population of CD4 and smallernumbers of CD8 T cells (Fig. 7, d and j), along with prominentCXCR3 expression (Fig. 7g), corresponding samples from anti-CXCR3 Ab-treated recipients showed �50% reduction in CD4�

( p � 0.01) and CD8� T cells ( p � 0.01), and no expression ofCXCR3 ( p � 0.001, Fig. 7, e, h, and k) (Table II). Analysis ofgrafts from anti-CXCR3 Ab-treated rats 7 days later showed neg-ligible CXCL10 expression and small numbers of CD4 and CD8 Tcells (Fig. 7, c, f, and l), comparable to 4-h samples from anti-CXCR3 Ab group or baseline controls, and no CXCR3 expression(Fig. 7i) (Table II). Hence, anti-CXCR3 Ab prevented local accu-mulation of CXCR3� T cells, critical to facilitate the hepatocel-lular damage in this model.

CXCR3 blockade does not inhibit early intrahepatic innateimmune activation

To investigate the molecular mechanism of the CXCR3-targetedtherapy in ameliorating liver IRI, we applied microarray analysis

FIGURE 2. IP-10 expression in OLT groups by (a)microarray data: average � SD (n � 2/group) afternormalized to global means; b, immunohistologystaining (anti-IP-10 Ab and control Ig). OLTs wereharvested at 4 h posttransplant after hepatic cold isch-emia ranging from 0 to 36 h.

6316 CXCR3 AND INNATE IMMUNITY IN IRI

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

to identify genes altered by the treatment in the 30-h cold-pre-served OLTs. By pairwise comparison of the treated (group no.30-4T) and control (group no. 30-4) livers, 172 genes were iden-tified to have at least 2-fold changes between the two groups.KEGG pathway analysis of this gene list by z-score revealed thatthree molecular/cellular pathways were affected the most: cyto-kine-cytokine receptor interaction, apoptosis, and TLR-signalingpathway. Bcl2-like 11 was down-regulated by 2-fold, and wild-type p53-induced gene 1 � 10-fold. CCL3 and CCL4, which bindto CCR1, CCR5 and are associated with wide-range lymphocytechemotaxis, were up-regulated by the treatment (Fig. 3). IL-1� andIL-6 were also up-regulated. Interestingly, some of the genes as-sociated with lethal IRI, including CXCL10, CXCL1, CCL2, andLIX, were all up-regulated by the treatment (Fig. 3), indicating afully activated status of innate immunity in the liver despiteCXCR3 Ab treatment. To further investigate the effects ofCXCR3-targeted therapy on innate immune activation, we mea-sured innate immune activation gene products at both 4 h and 24 h

after reperfusion in OLTs by quantitative RT-PCR. As shown inFig. 8, innate immune activation products were induced most sig-nificantly at 4 h, and all but CD86 declined by 24 h. CXCR3 Abtreatment did not down-regulate their induction seen otherwiseduring untreated IRI. Additionally, MPO assay, which reflects neu-trophil infiltration, showed increased enzymatic activity at 4 h inboth control and treated livers ( p � 0.01, vs native livers); therewas a decline in MPO activity at 24 h in the treated OLTs (Fig. 9).Thus, although CXCR3-targeted therapy did not suppress innateimmune activation early during liver IR, it did prevent the devel-opment of hepatocellular damage, indicating that the execution ofinnate immune function in liver IRI required CXCR3�CD4� Tcells.

DiscussionIn this study, we aimed to determine the mechanism responsiblefor the recruitment and activation of T cells in the pathophysiologyof liver IRI in the absence of exogenous Ags. A systemic cDNA

FIGURE 3. Cytokine/chemokine expression in OLT groups by microarray analysis. Gene expression by average � SD (n � 2/group) after normalizedto global means were plotted using the Genesifter.net tool at www.Genesifter.net.

6317The Journal of Immunology

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

microarray was performed in a rat syngeneic OLT model to iden-tify gene products altered early after reperfusion (4 h), in particulargenes encoding T cell-targeted innate immune activation products.Our results demonstrate increased intrahepatic expression ofCXCR3 ligands, including CXCL9, 10, 11, in response to the ex-tended cold ischemia. The functional significance of IRI-inducedchemokines was documented by the blockade of their shared re-ceptor, CXCR3, which not only ameliorated cardinal features ofIRI, but also improved OLT survival from 40 to 95%. Concomitantimmunohistology analysis has confirmed that there was a reduc-tion in the number of OLT infiltrating CD4� T and CXCR3� cells.Interestingly, CXCR3-targeted therapy failed to suppress innateimmune activation early during liver IRI, as manifested by in-creased levels of IL-6, IL-1�, iNOS, and neutrophil/monokine tar-

geted chemokine programs. Together, these results indicate thatliver IR induced-activation of innate immunity results in the pro-duction of T cell-targeting CXCR3 ligands, which are critical forthe selective recruitment of CXCR3�CD4� T cells. These T cellsare required for the execution of innate immune function, leadingto the ultimate hepatocellular damage.

The interlocked roles of proinflammatory cytokines in IRI havebeen extensively studied. Increased TNF-� and IL-1 plasma levelswere observed as early as 5 min of liver reperfusion (17). Theirfunctional role was confirmed by neutralization experiments inwhich IL-1 or TNF-� blockade ameliorated the severity of IRI (18,19). Because IRI develops in the absence of exogenous Ags, mostattention in chemokine studies has been focused on those targetingAg nonspecific leukocytes, in particular macrophages and neutro-phils (20). IL-8 (CXCL8 or CINC in rodents), LIX, and MIP-2, areall involved in neutrophil trafficking, and their neutralization wasprotective in lung, kidney, and myocardial IRI models (21–24).Recently, small molecule-mediated blockade of CXCR1/CXCR2,the neutrophil chemokine receptors that bind IL-8/MIP-2, reducedgranulocyte infiltration and mitigated rat kidney and liver IRI (25,26). Similar effects were recorded in myocardial IRI model byusing CXCR2-deficient mice (27). Chemokines targeted to mono-cytes, such as MCP-1, have also proven to be critical in renal IRI,as evidenced by tubular cytoprotection in CCR2 (binds MCP-1)deficient mice (28). Despite confirmed roles of these proinflam-matory cytokines/chemokines in IRI, the regulation of their induc-tion during IRI and mechanisms of their action other than effectson lymphocyte infiltration have not been fully addressed. Inparticular, the cascade of proinflammatory chemokine/cytokine

FIGURE 5. OLT recipient survival. LEW livers were stored at 4°C inUW solution for 30 h before transplantation into rat syngeneic recipients.Recipients were treated with rabbit anti-rat CXCR3 serum or normal rabbitserum twice: 24 h prior and at the time of liver transplantation (1 ml/injection). Rat survival was monitored daily; n � 6–12 rats/group.

FIGURE 4. CXCR3 ligand expression in OLT groups by quantitative RT-PCR. Gene expression by ratios of target gene/�-actin (average � SD; n �2/group) were plotted with Microsoft Excel.

6318 CXCR3 AND INNATE IMMUNITY IN IRI

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

FIGURE 7. Immunopathology of rat liver IRI after anti-CXCR3 Ab treatment. Gene expression was analyzed in OLTsamples harvested at 4 h or 7 days posttransplant following30 h of cold ischemia. a–c, Anti-CXCR3 Ab therapy did notabolish IP-10 induction by hepatocytes at 4 h; it had essen-tially fallen to trace levels by 7 days of follow-up. d–f, Levelsof CD4� T cells were higher in IgG-treated rats than in thosereceiving anti-CXCR3 Ab. g–i, Numbers of CXCR3� leuko-cytes were markedly increased in the IgG-treated group. j–l,Numbers of CD8� T cells were comparable between groups.No staining was seen using control IgG or after peptide ab-sorption of polyclonal Abs; data for IP-10 controls are shownas insets in each panel. (Cryostat sections labeled by immu-noperoxidase and counterstained with hematoxylin; originalmagnification �150). Representative of three to four samplesper group per time point.

FIGURE 6. Impact of anti-CXCR3 serum treat-ment on peripheral lymphocytes in vivo. a, PBLswere harvested from normal LEW rats, and stainedwith anti-murine CXCR3 serum (first panel), or acommercial anti-CXCR3 (third panel) rabbit poly-clonal Ab. PBLs were harvested 24 h after i.v. infu-sion of 1 ml of anti-CXCR3 serum, and stained witha FITC-labeled anti-rabbit Ig (second panel). Thefrequency of CD4 expression was screened in paral-lel. b, PBLs were harvested before as well as 4 h, and24 h after i.v. administration of anti-CXCR3 serum.Cells were stained with anti-CD4, anti-CD8, andcommercial anti-CXCR3 polyclonal Abs. Lympho-cytes were gated for the analysis of total CD4,CD4CXCR3�, and total CD8, CD8 CXCR3� ratios.Results are representative of three differentexperiments.

6319The Journal of Immunology

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

programs during IR, and the inflammation status altered by theabsence of any of these chemokines or their receptors has not beenstudied in a comprehensive manner.

Our study is the first that used a systemic approach, i.e., mi-croarray, to identify not only genes induced during liver IRI, butalso those altered by anti-CXCR3 Ab treatment. Results show thatIR-induced hepatic induction of IL-1� did not correlate with theseverity of liver injury. Interestingly, CXCR3-targeted therapy didnot decrease, but actually increased IL-1� levels without causinghepatocellular damage. In parallel, IR-induced neutrophil andmonocyte targeted chemokines, such as CCL-2, CCL-3, CCL-4,CXCL1 or LIX were readily detectable in our model, and anti-CXCR3 Ab treatment did not inhibit their expression. Consistentwith the chemokine expression data, we did find increased neu-

trophil infiltration (assessed by hepatic MPO activity) at 4-h post-transplant despite anti-CXCR3 Ab treatment. These results indi-cate that depletion of CXCR3� cells blocked the development ofinflammation response without affecting the initial innate immuneactivation by liver IRI. This implies that the execution of IR-ac-tivated innate immune function requires CXCR3� T cells, whichmay regulate either the responsiveness of neutrophil/monocytes totheir chemokines, or the functions of subsequently recruited andactivated innate immune cells. Indeed, in the absence of CXCR3�

T cells, IR failed to trigger hepatocellular damage despite activa-tion of the innate immune system.

T cells have been shown to play a key role in the mechanism ofIRI. Indeed, in the absence of T cells, in particular CD4 T cells,both livers and kidneys are largely protected from IRI (9, 10).Interestingly, concomitant neutrophil infiltration was also reducedin the ischemic livers (9). The question arises how Ag-specific Tcells are being recruited into this, by definition Ag-independentinflammatory immune reaction? Our finding that liver IR inducedCXCR3 ligands very early (4 h) after reperfusion, provides a clueto this question. The reduction of CD4 and CD8 infiltration afterCXCR3 blockade further supports the idea that CXCR3 ligands areresponsible for local T cell recruitment in OLTs. Furthermore, ourresults suggest that intrahepatic accumulation of CXCR3� T cellsplays a key role in the pathogenesis of liver IRI. CXCR3 is beingexpressed mostly on preactivated T cells (memory), particularly

FIGURE 8. Gene expression of innate immune activation products in OLT groups by quantitative RT-PCR. Gene expression by ratios of targetgene/�-actin (average � SD; n � 2/group) were plotted with Microsoft Excel.

Table II. Quantitation of CD4�, CD8�, and CXCR3� T cells in 30-hcold-preserved syngeneic OLTa

MarkerIgG treated

(4 h)Anti-CXCR3treated (4 h)

Anti CXCR3treated (7 days)

CD4 17.4 � 5.3 8.3 � 3.6� 6.8 � 2.9�CD8 9.2 � 3.6 4.4 � 2.6� 4.1 � 1.8�CXCR3 7.6 � 2.3 0.2 � 0.1�� 0.1 � 0.1��

a Mean � SD cells/high power field, with quantitation as detailed in Materials andMethods: �, p � 0.01 and ��, p � 0.001 vs IgG-treated group.

6320 CXCR3 AND INNATE IMMUNITY IN IRI

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Th1, and some NK and B cells (29, 30). Our data that CXCR3�

cells are critical in the mechanism of liver IRI is consistent withpreviously shown cytoprotection against hepatic and renal IRI inStat4- or Th1-type deficient mice (31, 32). Although type-1 proin-flammatory T cells produce IL-2, IFN-�, and TNF-�, and supportmacrophage activation (33), we failed to detect the induction ofthese cytokines in our OLT model (data not shown). This might bedue to the lack of sufficient T cell stimulation during liver IRI, inwhich no specific Ags are present or potent enough to provide thefirst signal for full T cell activation. This raises an intriguing ques-tion as to how T cells function to promote IRI without being fullyactivated? Our preliminary data suggest a role of TRAIL, whichmay be expressed constitutively on activated T cells (data notshown). Additionally, we have shown that CD154 costimulatorypathway is critical for the development of IRI in a mouse liverischemia model (12). Indeed, CD154 expressed on activated CD4T cells, may also be involved in the activation of liver DCs and/ormacrophages (Kupffer cells) via CD40.

One obvious question is which mechanism triggers intrahepaticinduction of CXCR3 ligands in the absence of any detectableIFN-� in livers early during IR. Although this question remainedbeyond the scope of this study, our recent data in a mouse liverischemia model indicate that TLR4 activation might represent theinitiating event (7). We have also shown that MyD88-independentsignaling downstream of TLR4 mediated by IRF3 is critical for thedevelopment of IR-induced hepatocellular damage. IFN-�, one ofthe major players along this pathway (34), can then stimulate mul-tiple liver cell types to elaborate CXCR3 ligands (data not shown).

In conclusion, this study demonstrates a novel mechanism of Tcell recruitment and function in the absence of exogenous Ag stim-ulation. By documenting that the execution of innate immune func-tion requires CXCR3�CD4� T cells, it highlights the critical roleof CXCR3 chemokine biology for the continuum of innate to adap-tive immunity in the pathophysiology of liver IRI.

DisclosuresThe authors have no financial conflict of interest.

References1. Laskowski, I., J. Pratschke, M. J. Wilhelm, M. Gasser, and N. L. Tilney. 2000.

Molecular and cellular events associated with ischemia/reperfusion injury. Ann.Transplant. 5: 29–35.

2. Farmer, D., F. Amersi, J. W. Kupiec-Weglinski, and R. W. Busuttil. 2000. Cur-rent status of ischemia and reperfusion injury in the liver. Transplant. Rev. 14:106–126.

3. Fellstrom, B., L. M. Akuyrek, U. Backman, E. Larsson, J. Melin, and L. Zezina.1998. Postischemic reperfusion injury and allograft arteriosclerosis. Transplant.Proc. 30: 4278–4280.

4. Howard, T. K., G. B. Klintmalm, J. B. Cofer, B. S. Husberg, R. M. Goldstein, andT. A. Gonwa. 1990. The influence of preservation injury on rejection in thehepatic transplant recipient. Transplantation 49: 103–107.

5. Jaeschke, H. 2003. Molecular mechanisms of hepatic ischemia-reperfusion injuryand preconditioning. Am. J. Physiol. 284: G15–G26.

6. Fondevila, C., R. W. Busuttil, and J. W. Kupiec-Weglinski. 2003. Hepatic isch-emia/reperfusion injury—a fresh look. Exp. Mol. Pathol. 74: 86–93.

7. Zhai, Y., X. D. Shen, R. O’Connell, F. Gao, C. Lassman, R. W. Busuttil,G. Cheng, and J. W. Kupiec-Weglinski. 2004. Cutting edge: TLR4 activationmediates liver ischemia/reperfusion inflammatory response via IFN regulatoryfactor 3-dependent MyD88-independent pathway. J. Immunol. 173: 7115–7119.

8. Rabb, H. 2002. The T cell as a bridge between innate and adaptive immunesystems: implications for the kidney. Kidney Int. 61: 1935–1946.

9. Zwacka, R. M., Y. Zhang, J. Halldorson, H. Schlossberg, L. Dudus, andJ. F. Engelhardt. 1997. CD4� T-lymphocytes mediate ischemia/reperfusion-in-duced inflammatory responses in mouse liver. J. Clin. Invest. 100: 279–289.

10. Rabb, H., F. Daniels, M. O’Donnell, M. Haq, S. R. Saba, W. Keane, andW. W. Tang. 2000. Pathophysiological role of T lymphocytes in renal ischemia-reperfusion injury in mice. Am. J. Physiol. 279: F525–F531.

11. Yokota, N., F. Daniels, J. Crosson, and H. Rabb. 2002. Protective effect of T celldepletion in murine renal ischemia-reperfusion injury. Transplantation 74:759–763.

12. Shen, X. D., B. Ke, Y. Zhai, F. Amersi, F. Gao, D. M. Anselmo, R. W. Busuttil,and J. W. Kupiec-Weglinski. 2002. CD154-CD40 T-cell costimulation pathwayis required in the mechanism of hepatic ischemia/reperfusion injury, and itsblockade facilitates and depends on heme oxygenase-1 mediated cytoprotection.Transplantation 74: 315–319.

13. Belperio, J. A., M. P. Keane, M. D. Burdick, J. P. Lynch, 3rd, Y. Y. Xue, K. Li,D. J. Ross, and R. M. Strieter. 2002. Critical role for CXCR3 chemokine biologyin the pathogenesis of bronchiolitis obliterans syndrome. J. Immunol. 169:1037–1049.

14. Mullane, K. M., R. Kraemer, and B. Smith. 1985. Myeloperoxidase activity as aquantitative assessment of neutrophil infiltration into ischemic myocardium.J. Pharmacol. Methods 14: 157–167.

15. Amersi, F., R. Buelow, H. Kato, B. Ke, A. J. Coito, X. D. Shen, D. Zhao, J. Zaky,J. Melinek, C. R. Lassman, et al. 1999. Upregulation of heme oxygenase-1 pro-tects genetically fat Zucker rat livers from ischemia/reperfusion injury. J. Clin.Invest. 104: 1631–1639.

16. Anselmo, D. M., F. F. Amersi, X. D. Shen, F. Gao, M. Katori, C. Lassman, B. Ke,A. J. Coito, J. Ma, V. Brinkmann, et al. 2002. FTY720 pretreatment reduceswarm hepatic ischemia reperfusion injury through inhibition of T-lymphocyteinfiltration. Am. J. Transplant. 2: 843–849.

17. Suzuki, S., and L. H. Toledo-Pereyra. 1994. Interleukin 1 and tumor necrosisfactor production as the initial stimulants of liver ischemia and reperfusion injury.J. Surg. Res. 57: 253–258.

18. Shirasugi, N., G. Wakabayashi, M. Shimazu, A. Oshima, M. Shito, S. Kawachi,T. Karahashi, Y. Kumamoto, M. Yoshida, and M. Kitajima. 1997. Up-regulationof oxygen-derived free radicals by interleukin-1 in hepatic ischemia/reperfusioninjury. Transplantation 64: 1398–1403.

19. Shito, M., G. Wakabayashi, M. Ueda, M. Shimazu, N. Shirasugi, M. Endo,M. Mukai, and M. Kitajima. 1997. Interleukin 1 receptor blockade reduces tumornecrosis factor production, tissue injury, and mortality after hepatic ischemia-reperfusion in the rat. Transplantation 63: 143–148.

20. Koo, D. D. H., and S. V. Fuggle. 2002. Chemokines in ischemia/reperfusioninjury. Curr. Opin. Organ Transplant. 7: 100–106.

21. Sekido, N., N. Mukaida, A. Harada, I. Nakanishi, Y. Watanabe, andK. Matsushima. 1993. Prevention of lung reperfusion injury in rabbits by a mono-clonal antibody against interleukin-8. Nature 365: 654–657.

22. Boyle, E. M., Jr., J. C. Kovacich, C. A. Hebert, T. G. Canty, Jr., E. Chi,E. N. Morgan, T. H. Pohlman, and E. D. Verrier. 1998. Inhibition of interleukin-8blocks myocardial ischemia-reperfusion injury. J. Thorac. Cardiovasc. Surg.116: 114–121.

FIGURE 9. Intrahepatic neutrophil accumulation inischemic lobes at 4, or 24 h of reperfusion, as analyzedby myeloperoxidase (MPO) enzymatic activity (interna-tional units per gram). There were two to four animalsper group. Means and SD are shown.

6321The Journal of Immunology

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

23. Chandrasekar, B., J. B. Smith, and G. L. Freeman. 2001. Ischemia-reperfusion ofrat myocardium activates nuclear factor-�B and induces neutrophil infiltration vialipopolysaccharide-induced CXC chemokine. Circulation 103: 2296–2302.

24. Miura, M., X. Fu, Q. W. Zhang, D. G. Remick, and R. L. Fairchild. 2001. Neu-tralization of Gro � and macrophage inflammatory protein-2 attenuates renalischemia/reperfusion injury. Am. J. Pathol. 159: 2137–2145.

25. Bertini, R., M. Allegretti, C. Bizzarri, A. Moriconi, M. Locati, G. Zampella,M. N. Cervellera, V. Di Cioccio, M. C. Cesta, E. Galliera, et al. 2004. Noncom-petitive allosteric inhibitors of the inflammatory chemokine receptors CXCR1and CXCR2: prevention of reperfusion injury. Proc. Natl. Acad. Sci. USA 101:11791–11796.

26. Cugini, D., N. Azzollini, E. Gagliardini, P. Cassis, R. Bertini, F. Colotta,M. Noris, G. Remuzzi, and A. Benigni. 2005. Inhibition of the chemokine re-ceptor CXCR2 prevents kidney graft function deterioration due to ischemia/reperfusion. Kidney Int. 67: 1753–1761.

27. Tarzami, S. T., W. Miao, K. Mani, L. Lopez, S. M. Factor, J. W. Berman, andR. N. Kitsis. 2003. Opposing effects mediated by the chemokine receptor CXCR2on myocardial ischemia-reperfusion injury: recruitment of potentially damagingneutrophils and direct myocardial protection. Circulation 108: 2387–2392.

28. Furuichi, K., T. Wada, Y. Iwata, K. Kitagawa, K. Kobayashi, H. Hashimoto,Y. Ishiwata, M. Asano, H. Wang, K. Matsushima, et al. 2003. CCR2 signaling

contributes to ischemia-reperfusion injury in kidney. J. Am. Soc. Nephrol. 14:2503–2515.

29. Homey, B., A. Muller, and A. Zlotnik. 2002. Chemokines: agents for the immu-notherapy of cancer? Nat. Rev. Immunol. 2: 175–184.

30. Sallusto, F., A. Lanzavecchia, and C. R. Mackay. 1998. Chemokines and che-mokine receptors in T-cell priming and Th1/Th2-mediated responses. Immunol.Today 19: 568–574.

31. Shen, X. D., B. Ke, Y. Zhai, F. Gao, D. Anselmo, C. R. Lassman, R. W. Busuttil,and J. W. Kupiec-Weglinski. 2003. Stat4 and Stat6 signaling in hepatic ischemia/reperfusion injury in mice: HO-1 dependence of Stat4 disruption-mediated cy-toprotection. Hepatology 37: 296–303.

32. Yokota, N., M. Burne-Taney, L. Racusen, and H. Rabb. 2003. Contrasting rolesfor STAT4 and STAT6 signal transduction pathways in murine renal ischemia-reperfusion injury. Am. J. Physiol. 285: F319–F325.

33. Liblau, R. S., S. M. Singer, and H. O. McDevitt. 1995. Th1 and Th2 CD4� T cellsin the pathogenesis of organ-specific autoimmune diseases. Immunol. Today 16:34–38.

34. Takeda, K., T. Kaisho, and S. Akira. 2003. Toll-like receptors. Annu. Rev. Im-munol. 21: 335–376.

6322 CXCR3 AND INNATE IMMUNITY IN IRI

by guest on May 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from