role of lentivirus-mediated overexpression of programmed death-ligand 1 on corneal allograft...

American Journal of Transplantation 2012; 12: 1313–1322Wiley Periodicals Inc.

C© Copyright 2012 The American Society of Transplantationand the American Society of Transplant Surgeons

doi: 10.1111/j.1600-6143.2011.03948.xBrief Communication

Role of Lentivirus-Mediated Overexpressionof Programmed Death-Ligand 1 on CornealAllograft Survival

M. Nosov†, M. Wilk†, M. Morcos, M. Cregg,

L. O’Flynn, O. Treacy and T. Ritter∗

College of Medicine, Nursing and Health Sciences,Regenerative Medicine Institute, National Centre forBiomedical Engineering Science, National University ofIreland, Galway, Ireland*Corresponding author: Thomas Ritter,[email protected]†Both authors contributed equally to this manuscript.

To investigate the role of lentivirus-mediated overex-pression of programmed death-ligand 1 (PD-L1) on ratcorneal allograft survival. A fully allogeneic rat corneatransplant model was used for in vivo studies. Lentivi-ral (LV) vectors are efficient tools for ex vivo geneticmodification of cultured corneas. LV vector encodingfor PD-L1 (LV.PD-L1) and LV vector encoding for eGFP(LV.eGFP, as control) were constructed and tested. PD-L1 or eGFP expression was increased on corneal cellsupon LV.PD-L1 and LV.eGFP transduction, respectively.Both allogeneic controls and allogeneic LV.eGFP trans-duced corneas were uniformly rejected (MST: 13.8 ± 1.7days and 12.3 ± 1.9 days, respectively). In contrast, al-logeneic LV.PD-L1 transduced corneas showed a highpercentage (83%) of graft survival (MST > 30 days,n = 5, 15 days, n = 1). Graft opacity of PD-L1 transducedcorneas was present but was significantly reducedcompared to control or eGFP expressing corneas.Flow cytometric analysis revealed that percentagesof CD3+CD8+CD161+ and CD3+CD8+CD161– lympho-cytes were decreased in animals receiving LV.PD-L1transduced corneas compared to animals grafted withLV.eGFP transduced corneas. Moreover, reduced ex-pression of proinflammatory cytokines (IFN-c and IL-6)in PD-L1 transduced corneas compared to allogeneiccontrols was also observed. Local PD-L1 gene transferin cultured corneas is a promising approach for the pro-longation of corneal allograft survival and attenuationof graft rejection.

Key words: Cornea, gene therapy, innate immunity,lentivirus, programmed death-ligand 1, transplanta-tion

Abbreviations: APCs, antigen presenting cells; eGFP,enhanced Green Fluorescent Protein; LNs, lymphnodes. LV.PD-L1, lentiviral vector encoding for PD-L1;LV.eGFP, lentiviral vector encoding for eGFP; LV, lentivi-

ral vector; PD-L1, programmed death-ligand 1; PD-L2,programmed death-ligand 2.

Received 20 May 2011, revised 21 November 2011 andaccepted for publication 08 December 2011

Introduction

With more than 100 000 procedures a year, cornea trans-plantation (keratoplasty) is the most frequent transplantprocedure of human tissue. Although keratoplasty is re-garded as a transplant procedure with high acceptancerates, the incidence of graft rejection can vary between5% and 40% (1). Nevertheless, the favorable prognosisof corneal transplants in humans or experimental animalsin the absence of risk factors is attributed to the immuneprivilege of the eye (2). This immune privilege is inducedby the presence of immunomodulatory molecules suchas TGF-b, macrophage migration inhibitory factor, IL-1 re-ceptor antagonist, Fas ligand (CD95L), prostaglandin E2,etc, which are able to suppress both innate and adap-tive immunity (3). Recently, another pathway to maintainimmune privilege has been characterized. Programmeddeath-ligand 1 (PD-1; CD279) is described as an inhibitoryreceptor found on the surface of activated B and T cells,thymocytes and myeloid cells (4). The relevance of PD-1 signaling for lymphocyte inactivation and tolerance in-duction was proven by lymphoproliferative and autoim-mune disease development in PD-1 deficient mice (4,5).Programmed deathligand-1 (PD-L1) and programmed celldeath-ligand 2 (PD-L2) are known ligands for PD-1. Bothare type 1 transmembrane proteins belonging to theB7 family (6). PD-L1 has been detected on lymphoidcells including monocytes, antigen presenting cells (APCs)and B cells, as well as in nonlymphoid tissues such asthe heart, lung, placenta, kidney, liver and cornea (6–9).In contrast, PD-L2 is expressed exclusively on dendriticcells and monocytes (6,9). The immunosuppressive andtolerogenic properties resulting from the interaction be-tween PD-1 with PD-L1 and/or PD-L2 may be a potentialtarget for therapeutic intervention in organ transplan-tation. Systemic application of a chimeric PD-L1.Ig and PD-L2.Ig fusion protein did not induce significant prolongationof mouse cardiac allograft survival whereas, in contrast,application of PD-L1.Ig in combination with immunosup-pressants (cyclosporin A or rapamycin) led to enhanced

1313

Nosov et al.

survival (10). Interestingly, Watson et al. demonstratedthat systemic PD-L1.Ig treatment of animals receiving fullyMHC-mismatched corneal allografts prolonged transplantsurvival (11). However, systemic expression of im-munomodulatory agents may bear the risk of severe sideeffects, which might not be justified in nonlife-threateningdiseases such as cornea transplantation. In contrast, lo-cal application of therapeutic molecules might be moreacceptable. Recently, genetic engineering of ex vivo cul-tured corneas for local gene therapy has shown promise,although long-term graft survival has only rarely been ob-served (12–14). The expression of endogenous PD-L1plays an important role in the regulation of T cells, whichhome to the cornea by modulation of specific chemokinesand chemokine receptors as was shown by using PD-L1deficient mice or PD-L1 blocking antibody in a model ofdry eye syndrome (15). Human corneal endothelial cellscan negatively regulate CD4+ T-cell proliferation by a cellcontact mechanism that is dependent on PD-L1 and PD-1interaction (16). Moreover, a recent study illustrated thatexpression of PD-L1 on multiple myeloma cells attenuatedtherapeutic efficiency with specific cytotoxic natural killer(NK) cells (17). In summary, previous observations led usto investigate the role of PD-L1 overexpression in cornealcells to prolong corneal allograft survival, without the needfor general immunosuppression. Here, we describe the ef-fects of lentivirus-mediated PD-L1 overexpression in donorcorneas on allograft survival and analyze the immunologi-cal characteristics of the cellular infiltration into the trans-planted cornea and in the draining lymph nodes (LNs) andblood. Significant prolongation of transplant survival wasobserved which was associated with a reduction of NK Tcell (NKT) and cytotoxic CD8+ T-cell infiltration, which wasaccompanied by attenuation of proinflammatory cytokineexpression.

Materials and Methods

Animals and corneal transplantation

All procedures performed were conducted under animal license no.B100/3852 and were approved by the Animals Ethics Committee of theNational University of Ireland, Galway. In addition, animal care and manage-ment followed the standard operating procedures of the animal facility atthe National Centre for Biomedical Engineering Science. A well-established,fully allogeneic major histocompatibility complex (MHC) class I/II disparatecornea transplant model was applied for these studies. Male Lewis (LEW,RT-1l) rats served as recipients of male Dark Agouti (DA, RT-1avl) grafts lead-ing to almost 100% rejection in untreated animals. All animals were 8–14weeks old, obtained from Harlan Laboratories UK and housed with foodand water ad lib.

Orthotopic corneal transplantations were performed as reported previously(18). Graft transparency as one indicator of rejection was evaluated everysecond day and graded as follows: 0, completely transparent cornea; 0.5,slight corneal opacity, iris structure easily visible; 1.0, low opacity withvisible iris details; 1.5, modest corneal opacity, iris vessels still visible; 2.0,moderate opacity, only some iris details visible; 2.5, high corneal opacity,only pupil margin visible; 3.0, complete corneal opacity, anterior chambernot visible. Neovascularization was evaluated on the basis of the numberof quaternary segments of donor corneas in which vessels were present.

The opacity score above 2.5 in combination with edema and correlatingchanges of transplant geometry (degree of convex contour, shrinking andsurface roughness of graft) was considered as graft rejection (19). Animalswith surgical complications were excluded.

Ex vivo lentivirus-mediated gene transfer into cultured corneas

Donor corneas were excised and cultured for 4 h at 37◦C and 5% CO2 with20 lL DMEM containing a lentiviral (LV) vector either encoding for PD-L1(LV.PD-L1) or enhanced Green Fluorescent Protein (LV.eGFP) at a concen-tration ranging between 3.4 × 107 and 1 × 108 titration units (TU)/mL.Thereafter, corneas were washed three times with PBS and stored on icein 20 lL of PBS for subsequent transplantation.

Generation of the lentiviral vector encoding for rat PD-L1

Rat PD-L1 mRNA was isolated from LEW rat heart and converted to cDNAusing RT-PCR. Amplified PD-L1 cDNA was cloned into the pLenti6/V5-DEST vector under the control of an ubiquitin promoter (Invitrogen, DunLaoghaire, Ireland) and sequence identity was confirmed by sequencing(Source Bioscience, Dublin, Ireland) using the reference sequence fromthe NCBI database (NM_001191954). Recombinant lentiviruses weregenerated by cotransfection of 293 T cells with plasmids containing eitherPD-L1 or eGFP sequences as well as with gag-pol, rev and VSV-G envelopeplasmids using a standard protocol (Invitrogen, Dun Laoghaire, Ireland).Supernatants were harvested 48 and 72 h posttransfection, filtered through0.45 lm pore size filters and concentrated by ultracentrifugation at 27 000 ×g for 2.5 h at 4◦C. LV was aliquoted and stored at –80◦C. Flow cytometricanalysis was used to determine LV.eGFP virus titer. Two step qRT-PCRusing viral RNA from both LV.eGFP and LV.PD-L1 viruses was performed todetermine LV.PD-L1 titer by comparing Ct values of common shared wood-chuck posttranscriptional regulatory element (WPRE) sequence betweenLV.eGFP with known titer and nontitrated LV.PD-L1. The primers usedfor WPRE amplification were: forward 5’-GGACCTGAAAGCGAAAGGG-3’; reverse 5’-CATCTCTCTCCTTCTAGCCTCCG-3’; probe 5’-Fam-CTCGACGCAGGACTCGGCTTGC-Tamra-3’. All quantitative real time PCRwas performed according to the standard program on the ABI one-stepmachine (Applied Biosystem, Foster City, CA, USA).

Isolation of lymphocytes from transplanted corneas, lymph

nodes and blood

Single cell suspensions from individual transplanted corneas were preparedby digesting corneal buttons with 5% w/v Collagenase D (Sigma–Aldrich,Wicklow, Ireland) in RPMI containing 25 mM HEPES (Lonza, Basel, Switzer-land) plus 1% fetal calf serum (Sigma–Aldrich) at 37◦C for 90 min at 900rpm per minute mixing frequency. Digested corneal tissue was carefullypoured into a 100-lM cell strainer and disrupted by grinding with a syringeplunger. Cell suspensions were transferred into 15 mL falcon tubes, spunat 400 × g for 5 min and washed again with PBS. Ipsi- and contralateralcervical LNs were homogenized with a syringe plunger and passed througha 100-lM cell strainer. Cell suspensions were transferred into 15 mL tubes,spun at 400 × g for 5 min and washed again with PBS. Lymphocytes fromblood were isolated by mixing 5 mL of whole rat blood with 625 lL ofOptiPrepTM (Axis-Shield, Dundee, UK) and adding 500 lL of PBS on top.Blood samples were centrifuged at 1300 × g for 30 min at 20◦C. Periph-eral blood mononuclear cells (PBMCs) were collected from the meniscusdownwards, diluted with two volumes of PBS and spun at 400 × g for 5min. Cell suspensions from individual corneas, LNs and blood were usedfor subsequent multicolor flow cytometric analysis.

Flow cytometry

The following monoclonal antibodies (mAbs) were used for the charac-terization of lymphocytes isolated from transplanted corneas, ipsi- and

1314 American Journal of Transplantation 2012; 12: 1313–1322

PD-L1 Overexpression Prevents Corneal Rejection

contralateral cervical lymph nodes and blood: CD3-FITC, CD8-PE, CD161-AF647, CD4-APC, CD25-FITC, CD134-PE, CD45RA-PE, MHCII-FITC,CD11b/c-APC, CD86-PE (BioLegend, San Diego, CA, USA). For staining,cells were washed with FACS buffer (PBS containing 2% FCS and 0.1%NaN3, all from Sigma–Aldrich). mAbs were diluted in 50 lL FACS buffer,added to the cells and incubated for 30 min at 4◦C. Finally, unbound anti-body was removed by washing twice with FACS buffer and cells were re-suspended in FACS buffer for analysis using a FACS Canto (BD Biosciences,Oxford, UK). Data was analyzed and compensated using FlowJo software(Tree Star, Inc., Ashland, OR, USA). Results are presented as percent of cellpopulation or as absolute cell number in graft. Fluorescent beads (CaliBRITE-PerCP beads; BD bioscience) were added to FACS samples for detectionof absolute number of cells. The absolute number of graft-infiltrated cellswas calculated on the basis of the number of fluorescent beads and cellscounted per sample. A predetermined number of beads had been added toeach sample before analysis on the flow cytometer.

RNA-isolation and RT-PCR

Total RNA from corneas was isolated using TRIzol R© reagent (Invitrogen)according to the manufacturer’s protocol. PCR array was performed us-ing a detection kit according to the manufacturer’s protocol (PARN-011;SABiosciences, Crawley, UK). cDNA was synthesized using RevertAidTM HMinus Reverse Transcriptase (Fermentas, York, UK) with random primers.Two step qRT-PCR based on corneal RNA from control, LV.eGFP and LV.PD-L1 transduced corneas on day 14 after transplantation was performed todetermine IFNc , IL-2, IL-2R and IL-6 mRNA expression by comparing Ct val-ues with the house-keeping gene b-actin. All quantitative real time PCRwas performed according to the standard program on the ABI one-stepmachine.

Primers for IFNc forward: 5’-AACAGTAAAGCAAAAAAGGATGCATT-3’, re-verse: 5’-TTCATTGACAGCTTTGTGCTGG, probe: 5’-Fam-CGCCAAGTTCGA-GGTGAACAACCC-Tamra-3’; IL-2 forward: 5’-CTCCCCATGATGCTCACGTT-3’, reverse: 5’-TCATTTTCCAGGCACTGAAGATG-3’ probe: 5’-Fam-CAATTCT-GTGGCCTGCTTGGGCAA-Tamra-3’; IL-2R forward: 5’-CACATGCTGTGTACC-AGGAGAACCT-3’, reverse: 5’-CCACGAAGTGGTAGATTCTCTTGG-3’; probe:5’-Fam-CAGGTCACTGCAGGGAGCCCC-Tamra-3’; IL-6 forward: 5’-TCAACT-CCATCTGCCCTTCAG-3’, reverse: 5’- AAGGCAACTGGCTGGAAGTCT-3’,probe: 5’-Fam-AACAGCTATGAAGTTTCTCTCCGCA-Tamra-3’.

Histology and histochemistry

For histological analysis, rat eyes were enucleated on the average day ofrejection (d14) for syngeneic and allogeneic controls, allogeneic LV.eGFPand LV.PD-L1 transduced groups and at the end of the observation periodfor allogeneic LV.PD-L1 transduced grafts (d30). In brief, the eyes embed-ded in paraffin wax were cut into 5-lm-thick sections, dried overnight at56◦C and then deparaffinized twice in xylene for 10 min, followed by hy-dration through graded alcohols. Slides were incubated for 40 s in Harrishematoxylin, washed in tap water for 2 min, then stained in eosin for 7 min,washed again in water for 2 min and dehydrated through graded alcohols.Next, sections were cleared twice for 10 min in xylene and mounted inDPX (Sigma–Aldrich). The thickness of cornea was measured in three differ-ent place of graft of three operated animals on microscope Olympus BX61using CellB software (Olympus, Tokyo, Japan).

Statistics

Statistical analysis was performed with GraphPad Prism software (Graph-Pad, La Jolla, CA, USA) using nonparametric Mann–Whitney or parametricStudent’s t-test. Differences were considered significant if p ≤ 0.05.

Results and Discussion

Local overexpression of PD-L1 leads to prolongation

of corneal allograft survival

A LV-vector encoding for rat PD-L1 (LV.PD-L1) under thecontrol of the ubiquitin promoter was generated frommRNA isolated from rat heart tissue. A LV-vector encodingfor eGFP (LV.eGFP) served as control. First, we addressedthe question of whether the LV-vector would lead to anoverexpression of the therapeutic construct in our targettissue. For this, donor corneas were collected, transducedin vitro with either LV.eGFP or LV.PD-L1 and transplantedinto syngeneic recipients. We found by real time RT-PCRanalysis that high levels of eGFP or PD-L1 mRNA wereexpressed in donor corneas on day 7 after transplantation(Figure 1A). To confirm that elevated levels of PD-L1 ex-pression was because of LV-transduction and not a resultof endogenous upregulation of PD-L1 expression on ocu-lar cells under inflammatory conditions (20), RT-PCR withLV.PD-L1-specific primers was performed. Upregulation ofPD-L1 mRNA was only observed in LV.PD-L1 transducedcorneas and not in LV.eGFP transduced corneas indicatingthe specificity of the approach (Figure 1B). Next, we stud-ied the influence of PD-L1 overexpression on allogeneiccorneal graft survival. Untreated allogeneic controls andLV.eGFP transduced allogeneic corneas were uniformly re-jected (MST: 13.8 ± 1.7 days, n = 11 and 12.3 ± 1.9 days,n = 4, respectively). In contrast, allogeneic LV.PD-L1 trans-duced corneas showed a high percentage (83%) of graftsurvival (MST > 30 days, n = 5, 15 days, n = 1; Figure 1C).In addition, neither LV.eGFP transduced syngeneic corneas(n = 5) nor untreated syngeneic control grafts (n = 8) wererejected (Figure 1C). Graft opacity, as an indicator of cel-lular infiltration and endothelial dysfunction, of allogeneicLV.PD-L1 transduced corneas was present but was re-duced significantly compared to allogeneic control or eGFPexpressing corneas (Figures 1D and S1). LV-transductionand eGFP expression did not change the graft survivaland opacity dynamics compared to untreated controls insyngeneic transplantation. LV.PD-L1 transduced corneaswhen observed at the time of rejection of allogeneic con-trol grafts showed opacity scores in the range of 1–2.5 butthis was not accompanied with changes in the geometryof the transplant, e.g. convex contour, shrinking and sur-face roughness. In contrast, all allogeneic control and eGFPexpressing corneas showed significant signs of rejectionas described above. Evidence of edema developmentwas estimated by Slit Lamp imaging and confirmed bythickness measurement in histological analysis of corneas(Figures S1 and S2). Significantly reduced edema develop-ment and corneal thickness of transplants in the LV.PD-L1treated group was observed at different time points post-transplantation compared to LV.eGFP transduced group(Figures 1C and D and S1 and S2). Levels of neovascular-ization were estimated as described in material and meth-ods and have shown similar severity within allogeneic andsyngeneic groups. The range of neovascularization was

American Journal of Transplantation 2012; 12: 1313–1322 1315

Nosov et al.

Figure 1: Lentivirus mediated overexpression of PD-L1 in cultured corneas leads to prolongation of corneal allograft survival.

(A) mRNA expression levels of eGFP and PD-L1 in corneas transduced by LV.eGFP or LV.PD-L1 7 days posttransplantation as measured byreal time RT-PCR. (B) Conventional PCR for the detection of lentivirally transferred PD-L1 in transduced corneas with specific primers forLV-mediated PD-L1 expression. (C) Graft survival curves of syngeneic control (n = 8) or LV.eGFP transduced corneas (n = 5), allogeneiccontrol (n = 11), LV.eGFP (n = 4) and LV.PD-L1 (n = 5) transduced corneas. (D) Opacity score of rats transplanted with syngeneic control,LV.eGFP transduced corneas and allogeneic control, LV.eGFP and LV.PD-L1 transduced corneas. (E) Neovascularization score in syngeneiccontrol or LV.eGFP transduced corneas, allogeneic control, LV.eGFP or LV.PD-L1 transduced corneas.

significantly higher in allogeneically transplanted groupscompared to syngeneic groups (Figure 1E). Expression ofeGFP or lentiviral transduction itself has been shown pre-viously to initiate minimal/moderate immune responses inmouse models (21,22). However, our study indicated thatneither eGFP expression nor lentiviral transduction causedchanges in observed parameters during graft rejection or

its survival. Similar results were described in the appli-cation of eGFP expressing stem cells or LV based genetherapy in other rat models (22,23).

In summary, our data suggest that the local overexpres-sion of PD-L1 does not protect the allogeneic cornea fromdevelopment of opacity but rather attenuates rejection



(3,16). Furthermore, gene engineered corneal grafts ex-pressing PD-L1 have been shown to significantly prolongsurvival upon transplantation in allogeneic recipients, a re-sult which has only rarely been observed after local genetherapy (12–14).

Characterization of immune cell types infiltrating

corneal allografts

To study the influence of PD-L1 overexpression on in-flammatory cell populations within the allogeneic cornea,allogeneic transplants were collected from the threeexperimental groups at a time-point corresponding to theaverage day of rejection of the control group (betweenday 13 and 14). Inflammatory cells were isolated fromexplanted corneas by collagenase digestion and analyzedby flow cytometry as described previously (18). Moreover,cells from ipsi- and contralateral submandibular andsuperficial cervical LNs and PBMCs were collected fromthe same animals and subjected to flow cytometry. First,a significantly higher frequency of CD3+CD8+CD161+,CD3–CD8+CD161+ and CD3–CD8+CD161++ cells whichcorrespond to NKT, NK and activated NK cells, respec-tively, were observed in rejecting control corneas whencompared to their presence in the LNs and PBMCs(Figures 2–4). This increase in cell numbers may indicatea preferential infiltration or clonal expansion within thetarget organ and confirm the important role of innate im-munity in the corneal graft rejection process as describedpreviously (18,24,25). The percentage of CD4+CD134+

cells which refers to an activated CD4+ T-cell populationis also increased in rejecting control corneas comparedto LNs and blood (39.5 ± 2.5% vs. 16.9 ± 2.1% and8.8 ± 2.3%; Figures 2–4). However, the most strikingdifferences in infiltrating cell populations were found in theCD3+CD8+CD161+ (NKT cells) compartment, which mayplay a “dual role” in eye immunology (26). Further investi-gation will be required in upcoming research to thoroughlycharacterize the NKT cell population because of their poten-tial importance in corneal allograft rejection. The frequencyof CD3+CD8+CD161+ cells was significantly reducedfrom 18.7 ± 5.8% in control or 15.5 ± 2.0% in LV.eGFPtransduced allografts to 10.4 ± 1.7% in LV.PD-L1 trans-duced corneas (Figure 2). This reflects an overall reductionof 44% in graft infiltration by this particular cell population.A significant reduction of infiltrating cells in LV.PD-L1 trans-duced corneas was also found in the CD3+CD8+CD161–

(cytotoxic T cells) compartment. The frequency of cellswas reduced from 5.9 ± 2.1% in control and 6.5 ± 1.2%in LV.eGFP transduced allografts to 2.9 ± 0.9% in LV.PD-L1transduced corneas (Figure 2). Finally, a profound, albeitnot significant, reduction in CD3–CD8+CD161+ (NK cells)cells was observed in LV.PD-L1 transduced corneascompared to untreated and LV.eGFP transduced corneas.No significant differences in cellular distribution wereobserved in our analysis of LNs and PBMCs (Figures 3and 4), which was expected because of the nature of thelocal gene therapy application. Furthermore, no significant

differences in cellular distribution could be found in ouranalysis of contralateral LNs (Figure S3). In addition,calculation of absolute numbers of cornea-infiltratingcells was performed to estimate levels of immune cellinfiltration. A significant reduction of absolute cell numberswas detected in all analyzed cell populations in LV.PD-L1transduced corneas compared to control group (Figure S4).This observation also correlates well with results from his-tological analysis after H&E staining (Figures 5B–D and Fand S4).

In summary, our data indicate that local overexpressionof PD-L1 reduces cornea infiltration as confirmed by de-tection of absolute numbers of graft-infiltrating cells byboth FACS and histology. Moreover, modulation of graft-infiltrating cell populations of both innate (NKT) and adap-tive (cytotoxic T cells) immunity was detected by analyz-ing the percentage of graft-infiltrating cells. Results shownhere also support previous reports on the role of PD-L1 intumor immunology (27).

To further investigate the mechanisms of PD-L1 medi-ated reduction of inflammatory cell infiltration, a PCR-arrayanalysis of proinflammatory markers was performed. Forthis control, LV.eGFP and LV.PD-L1 transduced allograftswere collected at the average time point of rejection (be-tween day 13 and 14) and subjected to RNA-isolation fol-lowed by PCR-array analysis. We found that the reduc-tion of the CD3+CD8+CD161+ cell population in LV.PD-L1transduced corneas was accompanied by a significant at-tenuation of proinflammatory gene expression (data notshown). To confirm the data generated by PCR-array, realtime RT-PCR analysis was carried out for selected genes.A significant reduction of IFNc and IL-6 mRNA was foundin LV.PD-L1 transduced corneas compared to both con-trols (Figure 5A). Interestingly, no significant differencesin IL-2 or IL-2R expression were detected. The reductionof IFNc mRNA expression levels may be a result of thedepletion of the CD3+CD8+CD161+ population, which isknown to express IFNc (26). IL-6 is a cytokine with pro-and antiinflammatory activity, which is expressed by Tcells, macrophages, fibroblasts and endothelial cells andhas been shown to be activated immediately after cornealtransplantation and during rejection. IL-6 blockade in theexperimental autoimmune uveoretinitis model resulted inan inhibition of Th1 and Th17 responses and in an ame-lioration of disease development (28). Decreased levels ofIL-6 production in LV.PD-L1 transduced corneas could or-chestrate in a similar way to modulate proinflammatoryresponses.

As mentioned above, we found that prolongation of cornealgraft survival by LV.PD-L1 transduction of allografts is asso-ciated with a decreased frequency of CD3+CD8+CD161+

cells and levels of IFNc and IL-6 mRNA expression inthe graft tissue which occurs without prevention of opac-ity development. Histological analysis showed massiveinfiltration in control and LV.eGFP transduced corneas,


Nosov et al.

Figure 2: Analysis of graft-infiltrating cells by flow cytometry. Upper line: Gating strategy of subpopulations of graft-infiltrating cellsfrom allogeneic control (n = 4), LV.eGFP (n = 3) and LV.PD-L1 (n = 8) groups on day 14 after transplantation. Graphics: Frequency ofdifferent subpopulations of cells in transplanted corneas from control, LV.eGFP and LV.PD-L1 groups as shown as the percentage fromCD3+, CD3– or lymphocyte fraction as mentioned on the graph. ∗p ≤ 0.05, ∗∗p ≤ 0.01.

however, cell infiltration in LV.PD-L1 transduced corneaswas markedly reduced which correlates with the severityof opacity score (Figures 1 and 5B–E). Local overexpres-sion of PD-L1 in corneal allografts may in addition to theendogenous expression of PD-L1 on corneal cells, aid thefurther development of an immunomodulatory milieu aftertrauma or cornea transplantation (3). Taken together ourresults demonstrate that lentivirus-mediated local overex-pression of the therapeutic gene PD-L1 was efficient inprolonging corneal graft survival through the modulationof CD3+CD8+CD161– and CD3+CD8+CD161+ cells which

revealed the importance of these cells in the process ofcorneal allograft rejection.

Acknowledgments

The authors would like to thank Mr. Gerry Fahy, Consultant Ophthal-mologist, University College Hospital Galway for helpful advice and dis-cussions. The authors would also like to thank Dr. Fabio Quondamat-teo and Mr. Marc Canney, College of Medicine, Nursing and Health Sci-ences, Discipline of Anatomy, National University of Ireland, Galway and



Figure 3: Analysis of cell distribution in draining lymph nodes. Upper line: Gating strategy of cell subpopulations from ipsilateralLNs of animals transplanted with allogeneic control (n = 4), LV.eGFP (n = 3) and LV.PD-L1-transduced (n = 8) corneas on day 14 aftertransplantation. Graphics: Frequency of different subpopulations of cells in animals transplanted with control, LV.eGFP and LV.PD-L1-transduced corneas as shown as the percentage from CD3+, CD3– or lymphocyte fraction as mentioned in the graph.


Nosov et al.

Figure 4: Analysis of cell distribution in PBMCs. Upper line: Gating strategy of PBMCs in animals transplanted with allogeneic control(n = 4), LV.eGFP (n = 3) and LV.PD-L1-transduced (n = 8) corneas. Graphics: Frequency of different subpopulations of cells in animalstransplanted with control, LV.eGFP and LV.PD-L1 transduced corneas as shown as the percentage from CD3+, CD3– or lymphocytefraction as mentioned on the graph.



Figure 5: Overexpression of PD-L1 in transplanted corneas leads to attenuation of cytokine expression and severity of infiltration.

(A) mRNA expression of IFNc , IL-6, IL-2 and IL-2R in transplanted corneas from control, LV.eGFP and LV.PD-L1 groups on day 14 aftertransplantation. H&E staining of infiltrating cells into corneas of allogeneic control, (B) LV.eGFP, (C) or LV.PD-L1, (D) transduced corneas orsyngeneic control, (E) on day 14 and allogeneic LV.PD-L1 transduced corneas, (F) on day 30 after transplantation. Representative imagesare shown. ∗p ≤ 0.05.

Ms Aoife Duffy, College of Medicine, Nursing and Health Sciences, Re-generative Medicine Institute, for their help with histology. This work issupported by Science Foundation Ireland (SFI) Principal Investigator GrantNo 07/IN.1/B925.

Disclosure

The authors of this manuscript have no conflicts of inter-est to disclose as described by the American Journal ofTransplantation.


Nosov et al.

Conflict of Interest

None declared.

References

1. Ritter T, Pleyer U. Novel gene therapeutic strategies for the in-duction of tolerance in cornea transplantation. Expert Rev ClinImmunol 2009; 5: 749–764.

2. Niederkorn JY, Larkin DF. Immune privilege of corneal allografts.Ocul Immunol Inflamm 2010; 18: 162–171.

3. Hori J, Wang M, Miyashita M, et al. B7-H1-induced apoptosis as amechanism of immune privilege of corneal allografts. J Immunol2006; 177: 5928–5935.

4. Nishimura H, Honjo T. PD-1: An inhibitory immunoreceptor in-volved in peripheral tolerance. Trends Immunol 2001; 22: 265–268.

5. Nishimura H, Nose M, Hiai H, Minato N, Honjo T. Developmentof lupus-like autoimmune diseases by disruption of the PD-1 geneencoding an ITIM motif-carrying immunoreceptor. Immunity 1999;11: 141–151.

6. Dong H, Zhu G, Tamada K, Chen L. B7-H1, a third member ofthe B7 family, co-stimulates T-cell proliferation and interleukin-10secretion. Nat Med 1999; 5: 1365–1369.

7. Latchman Y, Wood CR, Chernova T, et al. PD-L2 is a second ligandfor PD-1 and inhibits T cell activation. Nat Immunol 2001; 2: 261–268.

8. Yamazaki T, Akiba H, Iwai H, et al. Expression of programmeddeath 1 ligands by murine T cells and APC. J Immunol 2002; 169:5538–5545.

9. Ishida M, Iwai Y, Tanaka Y, et al. Differential expression of PD-L1and PD-L2, ligands for an inhibitory receptor PD-1, in the cells oflymphohematopoietic tissues. Immunol Lett 2002; 84: 57–62.

10. Ozkaynak E, Wang L, Goodearl A, et al. Programmed death-1 tar-geting can promote allograft survival. J Immunol 2002; 169: 6546–6553.

11. Watson MP, George AJ, Larkin DF. Differential effects of costim-ulatory pathway modulation on corneal allograft survival. InvestOphthalmol Vis Sci 2006; 47: 3417–3422.

12. Gong N, Pleyer U, Vogt K, et al. Local overexpression of nervegrowth factor in rat corneal transplants improves allograft survival.Invest Ophthalmol Vis Sci 2007; 48: 1043–1052.

13. Klebe S, Sykes PJ, Coster DJ, Krishnan R, Williams KA. Prolon-gation of sheep corneal allograft survival by ex vivo transfer ofthe gene encoding interleukin-10. Transplantation 2001; 71: 1214–1220.

14. Beutelspacher SC, Pillai R, Watson MP, et al. Function of in-doleamine 2,3-dioxygenase in corneal allograft rejection and pro-longation of allograft survival by over-expression. Eur J Immunol2006; 36: 690–700.

15. El Annan J, Goyal S, Zhang Q, Freeman GJ, Sharpe AH, DanaR. Regulation of T-cell chemotaxis by programmed death-ligand 1(PD-L1) in dry eye-associated corneal inflammation. Invest Oph-thalmol Vis Sci; 51: 3418–3423.

16. Sugita S, Usui Y, Horie S, et al. Human corneal endothelial cellsexpressing programmed death-ligand 1 (PD-L1) suppress PD-1+T helper 1 cells by a contact-dependent mechanism. Invest Oph-thalmol Vis Sci 2009; 50: 263–272.

17. Benson DM, Jr., Bakan CE, Mishra A, et al. The PD-1/PD-L1 axismodulates the natural killer cell versus multiple myeloma effect: atherapeutic target for CT-011, a novel monoclonal anti-PD-1 anti-body. Blood; 116: 2286–2294.

18. Maenz M, Morcos M, Ritter T. A comprehensive flow-cytometricanalysis of graft infiltrating lymphocytes, draining lymph nodes and

serum during the rejection phase in a fully allogeneic rat corneatransplant model. Molecular Vision 2011; 17: 420–429.

19. Matoba AY, Peeler JS, Niederkorn JY. T cell subsets in the im-mune rejection of murine heterotopic corneal allografts. InvestOphthalmol Vis Sci 1986; 27: 1244–1254.

20. Yang W, Li H, Chen PW, et al. PD-L1 expression on human ocularcells and its possible role in regulating immune-mediated ocularinflammation. Invest Ophthalmol Vis Sci 2009; 50: 273–280.

21. Rosenzweig M, Connole M, Glickman R, et al. Induction of cyto-toxic T lymphocyte and antibody responses to enhanced green flu-orescent protein following transplantation of transduced CD34(+)hematopoietic cells. Blood 2001; 97: 1951–1959.

22. Baekelandt V, Eggermont K, Michiels M, Nuttin B, Debyser Z.Optimized lentiviral vector production and purification procedureprevents immune response after transduction of mouse brain.Gene Ther 2003; 10: 1933–1940.

23. Moloney TC, Dockery P, Windebank AJ, Barry FP, Howard L, DowdE. Survival and immunogenicity of mesenchymal stem cells fromthe green fluorescent protein transgenic rat in the adult rat brain.Neurorehabil Neural Repair; 24: 645–656.

24. Flynn TH, Mitchison NA, Ono SJ, Larkin DF. Aqueous humor al-loreactive cell phenotypes, cytokines and chemokines in cornealallograft rejection. Am J Transplant 2008; 8: 1537–1543.

25. Claerhout I, Kestelyn P, Debacker V, Beele H, Leclercq G. Role ofnatural killer cells in the rejection process of corneal allografts inrats. Transplantation 2004; 77: 676–682.

26. Niederkorn JY. Role of NKT cells in anterior chamber-associatedimmune deviation. Expert Rev Clin Immunol 2009; 5: 137–144.

27. Chang WS, Kim JY, Kim YJ, et al. Cutting edge: Programmeddeath-1/programmed death ligand 1 interaction regulates the in-duction and maintenance of invariant NKT cell anergy. J Immunol2008; 181: 6707–6710.

28. Haruta H, Ohguro N, Fujimoto M, et al. Blockade of interleukin-6signaling suppresses not only Th17 but also interphotoreceptorretinoid binding protein-specific Th1 by promoting regulatory TCELLS in experimental autoimmune uveoretinitis. Invest Ophthal-mol Vis Sci 2011; 52: 3264–3271.

Supporting Information

Additional Supporting Information may be found in the on-line version of this article:

Figure S1: (A) Clinical evaluation of rat corneal allo-

graft by slit lamp biomicroscopy at postoperative days

(6,10,14,18,22,26,30).

Figure S2: Histological analysis of transplanted

corneas.

Figure S3: Analysis of cell distribution in contralateral

draining lymph nodes.

Figure S4: Analysis of absolute number of graft-

infiltrating cells by flow cytometry.

Please note: Wiley-Blackwell is not responsible for the con-tent or functionality of any supporting materials suppliedby the authors. Any queries (other than missing material)should be directed to the corresponding author for thearticle.


role of lentivirus-mediated overexpression of programmed death-ligand 1 on corneal allograft...

Documents