Renal tubular epithelial cells modulate T-cell responses via ICOS-L and B7-H1

Kidney International, Vol. 68 (2005), pp. 2091–2102

Renal tubular epithelial cells modulate T-cell responses viaICOS-L and B7-H1

SIMONE DE HAIJ,1 ANDREA M. WOLTMAN, LEENDERT A. TROUW, ASTRID C. BAKKER,SYLVIA W. KAMERLING, SANDRA W. VAN DER KOOIJ, LIEPING CHEN, RICHARD A. KROCZEK,MOHAMED R. DAHA, and CEES VAN KOOTEN

Department of Nephrology, Leiden University Medical Center, Leiden, The Netherlands; Department of Immunology, Mayo Clinic,Rochester, Minnesota; and Department of Molecular Immunology, Robert Koch-Institute, Berlin, Germany

Renal tubular epithelial cells modulate T-cell responses viaICOS-L and B7-H1.

Background. Renal tubular epithelial cells (TECs) play anactive role in renal inflammation. Previous studies have demon-strated the capacity of TECs to modulate T-cell responsesboth positively and negatively. Recently, new costimulatorymolecules [inducible T cell costimulator-L (ICOS-L) and B7-H1] have been described, which appear to be involved inperipheral T-cell activation.

Methods. We characterized expression and regulation of cos-timulatory molecules on primary human TECs and the TEC linehuman kidney-2 (HK-2) with reverse transcription-polymerasechain reaction (RT-PCR) and flow cytometry. Immunohisto-chemistry was performed on human kidney biopsies. The ca-pacity of TECs to modulate T-cell activation was studied inTEC/T-cell cultures.

Results. We demonstrate that TECs express ICOS-L andB7-H1 in vitro and in vivo. Stimulation with interferon-c (IFN-c) resulted in increased expression of B7-H1, whereas ICOS-L expression was marginally increased upon stimulation withCD40L, with no effect of interleukin (IL-1), IL-17, or tumornecrosis factor-a (TNF-a). Furthermore, we show that TECsare able to costimulate T cells that have received signal-1 usingaCD3 antibodies, inducing strong IL-10 production, which waspartially mediated by ICOS-L. In contrast, B7-H1 appeared tobe involved in inhibition of proliferation and cytokine synthe-sis. In addition, TECs were able to alter the cytokine profile offully activated T cells, which were incubated with aCD3 andaCD28 antibodies, resulting in low IFN-c and high IL-10 pro-duction. This activity appeared to be independent of ICOS-Land B7-H1.

Conclusion. Interaction of tubular epithelial cells and kidneyinfiltrating T cells via ICOS-L and B7-H1 may change the bal-

1Current address of Simone de Haij is the Department of Immunother-apy, University Medical Center Utrecht, Utrecht, The Netherlands.

Key words: renal tubular epithelial cells, costimulation, inflammation,human.

Received for publication August 31, 2004and in revised form March 8, 2005, and June 1, 2005Accepted for publication June 13, 2005

C© 2005 by the International Society of Nephrology

ance of positive and negative signals to the T cells, leading toIL-10 production and limitation of local immune responses.

Interstitial infiltration and tubular injury is a hallmarkof renal allograft rejection and many renal inflamma-tory diseases [1, 2]. Renal tubular epithelial cells (TECs)are thought to play a central role in the local inflamma-tory response via cytokine and chemokine production[3]. Furthermore, TECs have been demonstrated to ex-press major histocompatability complex (MHC) class Iand II molecules and also have the capacity to processand present antigens in the context of MHC molecules[4, 5]. Based on these observations, it has been postulatedthat renal epithelial cells may function as nonprofessionalantigen-presenting cells (APCs) and thereby regulate thelocal activation status of T cells [6, 7]. Some studies haveshown that presentation of foreign or alloantigens byTECs may lead to T-cell activation [4, 8, 9], whereas otherstudies have demonstrated that TECs can induce T-cellhyporesponsiveness, suggesting a role for TECs in main-taining peripheral tolerance to self antigens [10–13].

T-cell activation is the net result of stimulatory andinhibitory costimulatory signals. The best characterizedcostimulatory molecules are B7.1 (CD80) and B7.2(CD86), which are predominantly expressed by profes-sional APCs (dendritic cells and monocytes). Whethertubular epithelial cells express B7.1 and B7.2 has been amatter of debate [10, 13–16]. However, TECs were foundto express other accessory molecules, such as CD40, inter-cellular adhesion molecule-1 (ICAM-1), and vascular celladhesion molecule-1 (VCAM-1), which are all involvedin T-cell activation [12, 17]. Importantly, it has been shownthat memory and effector T cells are less dependent oncostimulation than naive cells and can be activated byprofessional and nonprofessional APCs [18].

In addition, new costimulatory molecules have beenidentified, which are more broadly expressed and mayplay a role in the regulation of peripheral T-cell activation

2091

2092 de Haij et al: Renal TECs and T-cell responses via ICOS-L and B7-H1

by tissue cells. One of these molecules is inducible T-cell cotimulator (ICOS) ligand (ICOS-L) (also termedB7h, GL50, B7RP-1, and LICOS), which is expressedin lymphoid (B, T, dendritic cells, monocytes) and non-lymphoid tissue, including kidney [19, 20]. Interaction be-tween ICOS-L and ICOS, which is expressed on activatedT cells, has been implicated in T-cell proliferation andthe regulation of cytokine production by effector T cells[21, 22]. Another member of the B7 family that was re-cently identified is B7-H1 (PD-L1). The receptor for B7-H1 is programmed death-1 (PD-1). Some studies suggestthat B7-H1 is involved in inhibition of T-cell prolifera-tion [23]. However, other studies have shown that B7-H1can increase T-cell proliferation and cytokine production,which was independent of PD-1, indicating the presenceof another, as yet unidentified receptor [24, 25]. Since B7-H1 is also expressed in nonlymphoid tissues, B7-H1 wassuggested to play a role in peripheral tolerance [26, 27].

In the present study we have characterized the expres-sion and regulation of B7.1 and B7.2 as well as the novelB7 molecules ICOS-L and B7-H1 on human proximalTECs. Furthermore, we investigated the role of thesecostimulatory molecules in the capacity of TECs to mod-ulate T-cell activation. We demonstrate that interactionof TECs with T cells favors interleukin (IL)-10 produc-tion and reduces interferon (IFN)-c production, indicat-ing that TECs may alter the effector function of T cellsin renal inflammation.

METHODS

Cell culture

The immortalized renal proximal TEC line humankidney-2 (HK-2) was kindly provided by M. Ryan, Uni-versity College Dublin, Ireland [28]. Cells were culturedin Dulbecco’s modified Eagle’s medium (DMEM)/HAMF-12 (Bio-Whittaker, Walkersville, MD, USA) supple-mented with insulin (5 lg/mL), transferrin (5 lg/mL),selenium (5 ng/mL), hydrocortisone (36 ng/mL), tri-iodothyronine (40 pg/mL), epidermal growth factor(EGF) (10 ng/mL) (all from Sigma Chemical Co., St.Louis, MO, USA) and 100 U/mL penicillin/100 lg/mLstreptomycin (Invitrogen, Paisley, UK). Primary humanproximal TECs (PTECs) were isolated from kidneys notsuitable for transplantation and from pretransplant biop-sies and cultured as described [17]. The specific outgrowthof PTECs was confirmed by morphologic appearanceand immunofluorescence staining (CD26-positive), as de-scribed [29]. Generation of monocyte-derived immaturedendritic cells has been described previously [30].

Flow cytometry

Cells were harvested by brief trypsinization with 0.02%(wt/vol) ethylenediaminetetraacetic acid (EDTA)/0.05%

(wt/vol) trypsin (Sigma Chemical Co.) to prevent pro-teolysis of surface receptors. Cells were washed influorescence-activated cell sorter (FACS) buffer [1%bovine serum albumin (BSA), 1% decomplemented nor-mal human serum and 0.02% sodium azide in phosphate-buffered saline (PBS)]. FACS analysis was performedusing monoclonal antibody against CD40 (monoclonalantibody 89), B7.1 (monoclonal antibody 104) [31], B7.2(IT2.2) (Pharmingen, San Diego, CA, USA), ICOS-L(HIL-131) [32], and B7-H1 (5H1) or an irrelevant mouseIgG1 control antibody [33]. Staining was visualized withgoat anti-mouse Ig-PE (Dako, Glostrup, Denmark), as-sessed for fluorescence using a FACScalibur and analyzedwith Cell Quest Pro software (Becton Dickinson, Moun-tain View, CA, USA). In stimulation experiments, cellswere incubated for 48 hours with 5 ng/mL IL-1a, 100ng/mL tumor necrosis factor-a (TNF-a), 50 ng/mL IL-17, 100 ng/mL IFN-c (all from R&D Systems, Minneapo-lis, MN, USA) or irradiated (80 Gy) CD40L-transfectedL cells [34] in a 1:1 ratio. To discriminate TECs fromL cells, the cells were incubated with 10 lg/mL digoxi-genin (DIG)-labeled monoclonal antibody 89 (aCD40)followed by aDIG-fluorescein isothiocyanate (FITC)(Boehringer Mannheim, Mannheim, Germany). Acces-sory molecule expression was calculated by the ratio ofmean fluorescence intensity (MFI) (i.e., the quotient ofMFI with and without specific antibody). Stimulation in-dexes (SI) were calculated as follows: SI = MFI stimu-lus/MFI untreated cells. The data were analyzed with apaired Student t test.

Reverse transcription-polymerase chain reaction (RT-PCR) for detection of B7.1, B7.2, ICOS-l, B7-H1, andCD40 transcripts

Total RNA was isolated using RNAzolB (Campro,Veenendaal, The Netherlands) according to manufac-turer’s instructions. Optical density (OD) 260/280 ratioswere measured to determine the quantity and purity ofRNA preparations. Fixed amounts of total cellular RNA(1 lg) were reverse transcribed into cDNA by oligo-dTpriming using Maloney-murine leukemia virus (M-MLV)reverse transcriptase (Gibco/Life Technologies, Breda,The Netherlands). Primer sequences for RT-PCR areshown in Table 1. PCR was performed under standardconditions [50 mmol/L KCl, 10 mmol/L Tris-HCl, pH8.4, 0.06 mg/mL BSA, 1.5 mmol/L MgCl2, 0.25 mmol/Ldeoxynucleosidetriphosphate (dNTP), 25 pmol of eachprimer, 1 U of Taq polymerase] (Perkin Elmer, Norwalk,CT, USA) for 35 cycles. PCR products were analyzed ona 1% agarose gel containing ethidium bromide.

Immunohistochemistry

For detection of B7-H1 and ICOS-L in renal tis-sue, we used samples from three normal kidneys not

de Haij et al: Renal TECs and T-cell responses via ICOS-L and B7-H1 2093

Table 1. Primer sequences

Sequence 5′-3′ bp Reference

B7.1 Forward CCTCTCCATTGTGATCCTGG 565 [55]Reverse GGCGTACACTTTCCCTTCTC

B7.2 Forward GTATTTTGGCAGGACCAGGA 663 [55]Reverse GCCGCTTCTTCTTCTTCCAT

ICOS-L Forward ACCATGCGGCTGGGCAGTC 420Reverse GCACGCTGAAGTTTGCTGCC

B7-H1 Forward GTCACGGTTCCCAAGGACC 450Reverse CAGATGACTTCGGCCTTGGG

CD40 Forward GCCTGGTCTCACCTGGCC 596 [56]Reverse TCTCAGCCGATCCTGGGGACC

b-actin Forward CTACAATGAGCTGCGTGTGG 527 [57]Reverse AAGGAAGGCTGGAAGAGTGC

ICOS-L is inducible T-cell costimulator ligand.

suitable for transplantation and two biopsies taken priorto transplantation of the kidney. Cryostat sections of3 lm were acetone fixed for 10 minutes at room tem-perature. After washing and blocking endogenous per-oxidase with 0.03% H2O2, sections were blocked for 20minutes with 1% normal goat serum in PBS/1% BSA andincubated for 45 minutes with HIL-131 (aICOS-L), 5H1(aB7-H1), or without primary antibody (control). Afterwashing with PBS, sections were subsequently incubatedwith goat-antimouse Ig-horseradish peroxidase (HRP)(Dako), tyramide-FITC in tyramide buffer (NENTM)(Life Science Products, Inc., Boston, MA, USA) andrabbit-anti-FITC-HRP (Dako) for 60 minutes each. An-tibody binding was visualized with Nova RED (Vector,Brunschwig, Amsterdam, The Netherlands) as a sub-strate. Slides were counterstained with hematoxylin andeosin.

Mixed TECs/T-cell culture

T cells were isolated from peripheral blood fromhealthy donors by sheep erythrocyte rosetting and in-cubated with 0.5 lg/mL aCD3 antibody (CLB-T3/4.E)or a combination of aCD3 and 0.1 lg/mL aCD28 anti-body (CLB-CD28/1) to induce T-cell activation [35]. Forcocultures, 1.5 × 105 T cells were mixed with irradiated(60 Gy) primary human TECs or the TEC-derived cellline HK-2 in a ratio of 7.5:1. Cells were cultured in 96-well flat-bottom culture plates (Costar, Badhoevedorp,The Netherlands) in RPMI 1640 containing 10% heat-inactivated fetal calf serum (FCS) (both from Gibco).Cell proliferation was quantified by 6 hr incubation with1 lCi (methyl-[3H]) thymidine (NENTM) (Life ScienceProducts, Inc.). Results are presented as the mean cpm± SD obtained from triplicate cultures. Where indicated,10 ng/mL IL-10 (R&D Systems) or 10 lg/mL neutralizinganti-IL-10R antibody [36] were added to the cultures.

Cytokine production

The production of IFN-c and IL-10 in culture super-natants was measured by enzyme-linked immunosorbent

assay (ELISA). The IFN-c ELISA was performed as pre-viously described [37]. For detection of IL-10, the PeliPairreagent set for human IL-10 (Sanquin, Amsterdam, TheNetherlands) was used according to manufacturer’s in-structions.

Statistical analysis

Results are shown as mean ± SD from triplicate cul-tures from representative experiments or as mean ± SEMfrom composite experiments. Data from different exper-iments were pooled and analyzed with a paired Studentt test. Results were considered significant if P < 0.05.

RESULTS

TECs express ICOS-L and B7-H1

Cell surface expression of costimulatory molecules onprimary TECs and the TEC-derived cell line HK-2 wasdetermined by FACS analysis. Immature dendritic cellswere used as a positive control. Both primary TECs andHK-2 cells were found to express ICOS-L and B7-H1(Fig. 1). Expression of B7.1 and B7.2 was readily detectedon immature dendritic cells, but was not detected on HK-2 cells or primary TECs. To confirm these data, we an-alyzed mRNA expression of costimulatory molecules inseven primary TEC lines and in HK-2 cells (Fig. 2). CD40is known to be expressed by cultured TECs [17] and wasused as a positive control for the mRNA quality. Compat-ible with the FACS results, B7.1 and B7.2 mRNA was notdetected in primary TECs or HK-2. Furthermore, sevenor seven primary TECs were found to express mRNA forICOS-L, whereas five of seven primary TECs expressedB7-H1 mRNA. Both ICOS-L and B7-H1 mRNA weredetected in HK-2. Immature dendritic cells expressedmRNA for all costimulatory molecules investigated.

Regulation of ICOS-L and B7-H1 expression on TECs

Infiltrating monocytes and T cells supply the local en-vironment with a variety of cytokines, including TNF-a, IL-1, IFN-c, and IL-17. Furthermore, CD40-CD40Linteraction has been shown to activate TECs, leadingto increased cytokine and chemokine expression [17].Therefore, we analyzed the effect of these signals onICOS-L and B7-H1 expression by TECs. None of thecytokines tested were able to alter ICOS-L expressionby HK-2 cells (Fig. 3A). Only stimulation with CD40Linduced a marginal but significant (1.6 ± 0.08-fold) in-crease in ICOS-L expression (Fig. 3A). In contrast, B7-H1 expression on HK-2 cells was strongly increased uponstimulation with IFN-c (2.4 ± 0.7-fold), but not with theother cytokines or CD40L (Fig. 3B and C). As shownin Figure 3C, the effect of IFN-c on B7-H1 expressionwas even stronger in primary TECs (4.4 ± 0.6-fold).Stimulation of TECs with cytokines or CD40L did not


010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

010

0

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

immDC HK-2 PTEC

Isotype

CD40

B7.1

B7.2

B7-H1

ICOS-L

Fig. 1. Cell surface expression of costimu-latory molecules by tubular epithelial cells(TECs). Expression of CD40, B7.1, B7.2,inducible T-cell costimulator ligand (ICOS-L) and B7-H1 on the surface of immaturedendritic cells (immDC), the TEC-derivedcell line human kidney-2 (HK-2), and prox-imal human TECs (PTECs) was determinedby fluorescence-activated cell sorter (FACS)analysis. An irrelevant mouse IgG1 antibodywas used as isotype control.

result in the induction of B7.1 or B7.2 expression (datanot shown).

Expression of ICOS-L and B7-H1 in human renal tissue

To investigate whether TECs express ICOS-L andB7-H1 in vivo, immunohistochemistry was performed onnormal human renal tissue obtained from kidneys not

suitable for transplantation due to anatomic reasons orfrom pretransplant biopsies. Expression of ICOS-L wasdetected on the majority of tubules, showing staining inthe cytosol and at the basolateral site of the tubules (Fig.4C and D). This is also the site where infiltrating immunecells will primarily encounter tubular cells after enter-ing the tubulointerstitium. Furthermore, ICOS-L was de-tected on endothelial cells and smooth muscle cells of the


HK-2

PTEC

immDC

B7.

1

B7.

2

ICO

S-L

B7-

H1

CD

40

10-3

10-4

10-5

β-actin

Fig. 2. mRNA expression of costimulatorymolecules by tubular epithelial cells (TECs).Total RNA was isolated from proximal humanTECs (PTECs), human kidney-2 (HK-2) andfrom immature dendritic cells (immDC) as apositive control. Expression of B7.1, B7.2, in-ducible T-cell costimulator ligand (ICOS-L),B7-H1, and CD40 was determined by reversetranscription-polymerase chain reaction (RT-PCR). As a control, serial cDNA dilutions(10−3, 10−4, and 10−5) were used to determinethe expression of b-actin.

large vessels (Fig. 4C). B7-H1 expression was detected onapproximately 30% of the tubules (Fig. 4E and F), whichmight reflect the local activation status of the tubuli. Inaddition, positive staining was observed throughout theentire glomerulus (Fig. 4E), whereas the vessels were neg-ative for B7-H1.

TECs induce IL-10 production by T cells

Next, we examined the role of ICOS-L and B7-H1 onTECs in their capacity to provide costimulatory signalsto T cells, as assessed by T-cell proliferation and cytokineproduction. Therefore, we used a previously describedmodel, in which peripheral blood T cells were triggeredwith an aCD3 antibody (IgE) which does not induce ac-tivation or proliferation unless a costimulatory signal isprovided during culture [35, 38]. In line with this model,incubation of T cells with aCD3 did not induce T-cellproliferation, which was measured by thymidine incor-poration at day 6 of culture (Fig. 5A). As expected, com-bination of aCD3/aCD28 antibodies resulted in strongT-cell proliferation (Fig. 5A) and high levels of IFN-cwere produced (Fig. 5B), ranging from 17.3 ng/mL to36.9 ng/mL in different experiments. Under the condi-tions of aCD3/aCD28 costimulation also production ofIL-10 was induced (Fig. 5C), ranging from 82 pg/mL to1055 pg/mL.

To determine the costimulatory signals of TECs, T cellswere cultured with aCD3 in the presence of irradiatedHK-2 cells at a ratio of 7.5:1. Radiation of HK-2 cellsdid not alter the expression of costimulatory moleculesas determined by FACS analysis (data not shown). NoT-cell proliferation was induced upon cocultures of HK-2 and T cells at day 6 after start of the cultures eitherwith or without aCD3 (Fig. 5A). Strikingly, coculture ofT cells and HK-2 cells in the presence of aCD3 antibodyinduced strong IL-10 production in the absence of IFN-c(Fig. 5B and C). Induction of IL-10 production by HK-2cells was on average 93 (± 26) pg/mL and was significantly

increased compared to T cells stimulated with aCD3 an-tibody only (8 ± 5 pg/mL) (P < 0.005).

IL-10 production is partially dependent on ICOS-L

To determine the kinetics of IL-10 induction, IL-10 wasmeasured at days 1 to 6 after start of the cocultures. IL-10production reached a maximum at day 2 (Fig. 6A), witha mean production of 491 ± 115 ng/mL (Fig. 6C). IFN-c was not detected at any time point (data not shown).Also coculture of T cells with primary human TECs inthe presence of aCD3 resulted in high IL-10 productionat day 2 (Fig. 6B) without an effect on T-cell proliferationand no detectable IFN-c production (data not shown).Induction of IL-10 production by primary TECs fromthree different donors was comparable to the HK-2 cellline (Fig. 6B and C).

To determine the role of B7-H1 and ICOS-L in IL-10 production, we administered blocking antibodies (10lg/mL) to the HK-2/T-cell cocultures and measured [3H]incorporation and IL-10 production at day 2 of culture.Although no proliferation was observed at day 6, we de-tected a low level of [3H] incorporation at day 2 when Tcells were incubated with aCD3, which was significantlyincreased in the presence of HK-2 cells (mean fold in-crease 12.0 ± 3.0) (P < 0.05) (Fig. 7A). Induction ofIL-10 production was strictly dependent on the presenceof HK-2 cells (Fig. 7B). Blocking ICOS-L resulted in asignificant 50.1% (± 4.6) decrease in proliferation (P =0.01) and a concomittant decrease in IL-10 production(29.4% ± 5.3%) (P < 0.05). Blocking B7-H1 resulted inincreased proliferation in three of five experiments, butthis was not statistically significant (P < 0.168). IL-10 pro-duction was increased in five of five experiments, with onaverage 1.8-fold increase (± 0.17) (P < 0.05). Combina-tion of aICOS-L and aB7-H1 antibodies neutralized themodulating effect of these antibodies on proliferation andIL-10 production (data not shown). Taken together, these


4

3

2

1

0

Stim

ulat

ion

inde

x

4

3

2

1

0

Stim

ulat

ion

inde

x

Con

trol

IL-1

TN

F-a

IL-1

7

IFN

-g

CD

40L

Con

trol

IL-1

TN

F-a

IL-1

7

IFN

-g

CD

40L

A

B

IFN-γ

ICOS-L B7-H1

010

00

100

010

00

100

100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104

*

*

C

HK-2

PTEC

Fig. 3. Regulation of inducible T-cell costimulator ligand (ICOS-L)and B7-H1 expression by tubular epithelial cells (TECs). (A and B) Hu-man kidney-2 (HK-2) cells were stimulated for 48 hours with interleukin(IL)-1 (5 ng/mL), tumor necrosis factor-a (TNF-a) (100 ng/mL), IL-17(50 ng/mL), interferon (IFN)-c (100 ng/mL), or CD40L-transfected Lcells in a 1:1 ratio. Nontransfected L cells were used as a negative con-trol for CD40L stimulation. Protein expression of ICOS-L (A) and B7-H1 (B) was measured by comparing relative fluorescence-activated cellsorter (FACS) intensities as described in the Methods section. Data arepresented as mean stimulation index (± SD) of three separate exper-iments. ∗P < 0.01. (C) Representative experiments showing the effectof IFN-c (filled histograms) on ICOS-L and B7-H1 expression by HK-2cells and primary TECs.

data suggest that the outcome of the T-cell response isdependent on the interplay between ICOS-L and B7-H1.

TECs modulate the cytokine profile of activated T cells

Based on the observations described above, we inves-tigated whether TECs may also modulate the responseof fully activated T cells. Therefore, T cells were incu-bated with aCD3 and aCD28 antibodies and coculturedwith irradiated HK-2 cells (Fig. 8). As expected, T cellsactivated via a combination of aCD3 and aCD28 showeda strong T-cell proliferation both at day 2 and at day 6.No difference in T-cell proliferation was observed at day2 in the presence or absence of TECs, whereas at day6, T-cell proliferation was significantly decreased uponcoculture with TECs, with a mean inhibition of 30.2%± 9.2% in seven independent experiments (P < 0.05)(Fig. 8A). IFN-c production initially increased in T cellscocultured with TECs. However, in five of seven experi-ments, production of IFN-c at day 6 was strongly reduced(83.0% ± 1.8%) (P < 0.05) compared to T cells incu-bated with aCD3/aCD28 only (Fig. 8B). Importantly, Tcells incubated in the presence of TECs produced veryhigh amounts of IL-10 compared to T cells incubatedwith aCD3/aCD28 only (Fig. 8C). At day 2, IL-10 pro-duction by aCD3/aCD28-stimulated T cells ranged from98 pg/mL to 2169 pg/mL, which increased to 513 pg/mLto 6804 pg/mL in the presence of HK-2, with a mean foldstimulation of 6.6 ± 0.9 (P < 0.05). Similar results wereobtained with primary human TECs (data not shown).

Administration of blocking antibodies to B7-H1 andICOS-L had no significant effect on proliferation andIFN-c production by T cells activated with aCD3/aCD28antibodies (data not shown). Furthermore, the high IL-10 production by T cells cocultured with HK-2 cells anda combination of aCD3/aCD28 antibodies was not sig-nificantly altered in the presence of blocking antibodiesto ICOS-L, B7-H1 (Fig. 9A), or both (data not shown).

IL-10 is known to inhibit T cell proliferation and IFN-c production [39, 40]. Administration of a neutralizinganti-IL-10R antibody was able to prevent the decrease inT-cell proliferation induced by exogenously added IL-10(10 ng/mL). However, the same antibody showed no ef-fect on T-cell proliferation (Fig. 9B) or IFN-c production(not shown) in HK-2 cocultures. This suggests that the rel-atively late (day 6) block in proliferation in the presenceof HK-2 cells was not mediated by endogenous IL-10 pro-duction. The data depicted in Figure 8 indicate acceler-ated T-cell activation in the presence of HK-2 cells, whichmay result in increased consumption of medium nutri-ents. To investigate this possibility, aCD3/aCD28 stimu-lated T cells were cultured with HK-2 cells at differentratios. As demonstrated in Figure 9C, lowering the T-cellnumber shifted the peak in proliferation from day 2 today 5 (half the amount of T cells) to day 7 (one fourth


Fig. 4. Expression of inducible T-cell costim-ulator ligand (ICOS-L) and B7-H1 by renaltubular epithelial cells (TECs) in vivo. Normalrenal tissue sections were stained for ICOS-L and B7-H1 as described in the Methodssection. (A and B) Negative control (stainingwithout primary antibody). (C and D) ICOS-L. (E and F) B7-H1. Shown are 100-fold (A,C, and E) and 400-fold (B, D, and F) magnifi-cations. Arrows indicate ICOS-L and B7-H1staining at the basal site of renal TECs.

the amount of T cells). Therefore, we cannot exclude thatthe decreased T-cell proliferation at day 6 might at leastin part be due to exhaustion of the culture medium.

DISCUSSION

T-cell infiltration is a hallmark of many inflammatoryrenal diseases. Previous studies have demonstrated thecapacity of TECs to modulate T-cell responses both pos-itively and negatively, but the molecular mechanisms arelargely unknown. In the present study we demonstratethat human TECs express the novel B7 family mem-bers ICOS-L and B7-H1. Coculture experiments showedthat TECs-T-cell interaction results in T cells with a low-proliferating, IL-10–producing phenotype, which is par-tially regulated by ICOS-L and B7-H1. Furthermore,TECs were able to promote IL-10 production by fullyactivated T cells, which was accompanied by inhibition

of proliferation and IFN-c production, independent ofICOS-L and B7-H1.

For other nonprofessional APCs, such as endothelialcells, keratinocytes, and fibroblasts, lack of B7.1 and B7.2expression was suggested to be responsible for the lim-ited capacity to induce T-cell activation [32, 41, 42]. Al-though it was suggested that expression of B7.1 and B7.2can be induced in TECs [14, 15], we and others [10, 13,16] did not detect B7.1 and B7.2 expression on eitherresting or activated primary human TECs as well as theTEC-derived cell line HK-2. Recently, novel costimula-tory molecules have been described, such as ICOS-L andB7-H1, which are more broadly expressed and appear tobe involved in peripheral T-cell activation. Expression ofICOS-L has been demonstrated on human umbilical veinendothelial cells (HUVEC), which was further increasedupon stimulation with lipopolysaccharide (LPS), CD40,TNF-a, and IL-1 [32]. Furthermore, ICOS-L expression


200

175

150125

10075

50

25

0

HK-2α CD3

α CD28

T−−−

+−+−

+−++

++−−

−+−−

+++−

+

HK-2α CD3

α CD28

T−−−

+−+−

+−++

++−−

−+−−

+++−

+

HK-2α CD3

α CD28

T−−−

+−+−

+−++

++−−

−+−−

+++−

+

A

B

C

cpm

(x1

000)

ng/m

Lpg

/mL

40

30

20

10

0

1000

750

500

250

0

Fig. 5. Tubular epithelial cells (TECs) induce interleukin (IL)-10 pro-duction without an effect on proliferation and interferon (IFN)-c. Pu-rified T cells were incubated with 0.5 lg/mL aCD3 antibody or witha combination of aCD3 and aCD28 (0.1 lg/mL) as a positive controlfor T-cell activation. For cocultures, T cells were mixed with irradiatedhuman kidney-2 (HK-2) cells at a ratio of 7.5:1. T-cell proliferation wasdetermined by [3H] incorporation at day 6 after start of the cocultures(A). Culture supernatants were harvested for detection of IFN-c (B)and IL-10 (C) by enzyme-linked immunosorbent assay (ELISA). Dataare shown as mean ± SD from triplicate wells. One representative ofseven independent experiments is shown.

was previously demonstrated on murine TECs and HK-2cells [43]. We now extend these findings and show expres-sion of ICOS-L on primary human TECs and HK-2 cells.This expression was not increased after stimulation withIL-1, IL-17, IFN-c, or TNF-a, and was only increased af-ter CD40L stimulation, even though all stimuli stronglyinduced IL-6 production by TECs (data not shown). Inaddition, we observed constitutive B7-H1 expression onHK-2 cells and most of the primary human TEC lines.Expression of B7-H1 was strongly increased upon stimu-lation with IFN-c, as has been described for endothelialcells, keratinocytes, monocytes, and dendritic cells [23, 41,44].

Knowing that TECs express ICOS-L and B7-H1, wedetermined the role of these molecules in the transmis-sion of costimulatory signals by TECs to T cells. Inter-action with TECs resulted in T cells with the capacityto produce IL-10 with a maximum at day 2, which wasaccompanied by a temporary induction of [3H] incor-poration. Blocking studies demonstrated that IL-10 pro-duction and the low level of proliferation were partiallymediated by ICOS-L. Although some studies suggest thatB7-H1 can increase T-cell proliferation associated with anincrease in IL-10 production [24], blockade of B7-H1 re-sulted in increased proliferation and IL-10 production,consistent with an inhibitory role of this molecule [23].Previous studies have demonstrated that IFN-c–treatedTECs can induce T-cell hyporesponsiveness [13]. Our re-sults suggest that induction of B7-H1 expression by IFN-cmight be a mechanism involved.

Using the same T-cell activation model, it has beendemonstrated that melanoma cells induce T-cell prolif-eration in a 3-day coculture, which was dependent onCD40 expression [45]. Furthermore, it has been shownthat CD40 is capable of stimulating T-cell proliferationand favors IL-10 production [46]. As the TECs that weused in our study expressed high levels of CD40, CD40-CD40L interaction may also contribute to the inductionof proliferation and IL-10 production, either directly orvia increased expression of ICOS-L. The absence of pro-liferation at day 6 is most likely due to low IL-2 pro-duction, as ICOS-L and CD40 are weak inducers of IL-2production by T cells [21].

Under inflammatory conditions, TECs are most likelyto interact with recently activated T cells that wereprimed in lymph nodes. Furthermore, T cells might be-come activated locally via infiltrating professional APCs[47]. Therefore, we examined whether TECs are able tomodulate the activation of T cells that were simultane-ously triggered via CD28. Both primary human TECs andthe cell line HK-2 inhibited proliferation of fully activatedT cells. B7-H1 has been shown to inhibit T cell prolifer-ation by inducing cell cycle arrest [23]. However, we didnot observe an increased number of T cells in the G0/G1

phase as measured by propidium iodide staining (data


500

400

300

200

100

0

pg/m

Lpg

/mL

pg/m

L

0 1 2 3 4 5 6Time after stimulation, days

T + aCD3T + aCD3 + HK-2

750

500

250

0T

PTEC1 PTEC2 PTEC3T/aCD3 T T/aCD3 T T/aCD3

750

500

250

0

T/a

CD

3

HK

-2 +

T/a

CD

3

A

B C

***Fig. 6. Primary tubular epithelial cells(TECs) induce interleukin (IL)-10 produc-tion by T cells. (A) T cells were incubated withaCD3 in the presence or absence of humankidney-2 (HK-2) cells. Culture supernatantswere harvested at days 1 to 6 after startof the cultures and IL-10 production wasmeasured by enzyme-linked immunosorbentassay (ELISA). One representative of threeindependent experiments is shown. (B)Similar experiments as described (A) wereperformed with three different proximalTEC lines (PTEC1 to PTEC3). Supernatantswere harvested at day 2. (C) Data fromnine different experiments with HK-2 cells,analyzed for IL-10 production at day 2 afterstart of the cocultures, were pooled andanalyzed with a paired Student t test. ∗∗∗P <

0.005. Results are presented as mean ± SEM.

not shown). Furthermore, administration of a blockingB7-H1 antibody had no effect. However, it should benoted that the antibody we used (5H1) was suggested toblock binding of B7-H1 to an as yet unidentified non-PD-1 receptor [25, 33]. Additional studies, including specificblockade of PD-1-B7-H1 interaction, are required to givemore insight in the role of B7-H1 in TEC/T-cell interac-tion.

In the presence of TECs, fully activated T cells alsodisplayed a different cytokine profile, characterized bylow IFN-c and high IL-10 production. Although differ-ential effects of B7-H1 on IFN-c and IL-10 productionhave been described [41, 44], using the 5H1 antibody, wefound no effect of B7-H1 on IL-10 or IFN-c production.Similarly, while ICOS-L is known to have a profound ef-fect on cytokine production, particularly IL-10 [22, 32],blocking ICOS-L was unable to reverse the cytokine pro-file. Furthermore, endogenous IL-10 production was notinvolved in the inhibition of proliferation and IFN-c pro-duction. Taken together, signals via ICOS-L and B7-H1appear to regulate TEC-induced IL-10 production by T-cell receptor (TCR)–triggered T cells, but are most likelyoverruled by additional signals, including CD28, in theresponse of fully activated T cells. In contrast to our re-

sults, a study with murine TECs has demonstrated thatICOS-L may have a down-modulating role on T-cell ac-tivation [43]. In this study, blocking ICOS-L enhancedIL-2 production by HEL (henn egg lysozy)–specific T-cell hybridomas after antigen presentation by TECs. Inour study, we have investigated the capacity of TECs totransmit a costimulatory signal to T cells bypassing theantigen specificity of the response, which might explainthis contradiction.

The presence of tubulitis is one of the major diagnosticcriteria for scoring rejection in the widely accepted Banffclassification system for allograft rejection [48]. However,by taking protocol biopsies it has been recently demon-strated that renal transplant patients on low-level im-munosuppression did not develop acute rejection despitethe continuous presence of a low-grade T-cell infiltrateand tubulitis [49]. These results suggest a role for TEC/T-cell interaction in the down-modulation of the local T-cellresponse. As IL-10 is a suppressor of T-cell proliferationand cytokine synthesis and functions to limit or termi-nate inflammatory responses [39, 40], our in vitro resultsseem compatible with this hypothesis. Furthermore, wedemonstrate expression of B7-H1 and ICOS-L in tubu-lar epithelial cells in tissue sections of normal human


40

30

20

10

0

1500

1000

500

0

HK

-2

T +

aC

D3

HK

-2 +

T +

aC

D3

alC

OS

-L

aB7-

H1

Con

trol

ab

HK

-2

T +

aC

D3

HK

-2 +

T +

aC

D3

alC

OS

-L

aB7-

H1

Con

trol

ab

A

B

HK-2 + T + aCD3

HK-2 + T + aCD3

cpm

(x1

000)

pg/m

L

Fig. 7. Role for inducible T-cell costimulator ligand (ICOS-L) and B7-H1 in tubular epithelial cell (TEC)/T-cell interaction. T cells were cul-tured with aCD3 and human kidney-2 (HK-2) cells in the presenceof blocking aICOS-L and aB7-H1 antibodies or an irrelevant controlantibody (10 lg/mL). T-cell proliferation (A) and interleukin (IL)-10production (B) were measured at day 2. One representative of five in-dependent experiments is shown.

kidney in vivo. These results support a model in whichparenchymal cells may have an important role in theregulation of the immune response, maintaining toler-ance in healthy conditions and determining the effectortype response during inflammation [50, 51]. In additionto ICOS-L and B7-H1, other B7-like molecules (B7-H3and B7-H4/B7S1) have recently been identified, whichare also expressed in nonlymphoid tissues, including thekidney [52–54]. Although expression of these moleculeson TECs remains to be determined, they might also playa role in modulation of the local immune response.

100

75

50

25

No HK-2HK-2

day 2 day 6

T + aCD3/aCD28

T + aCD3/CD28HK-2 + T + aCD3/CD28

T + aCD3/CD28HK-2 + T + aCD3/CD28

40

30

20

10

00 1 2 3 4 5 6

Time, days

0 1 2 3 4 5 6Time after stimulation, days

5000

4000

3000

2000

1000

0

cpm

(x1

000)

ng/m

Lpg

/mL

A

B

C

Fig. 8. Tubular epithelial cells (TECs) modulate the cytokine profileof activated T cells. T cells were incubated with aCD3 (0.5 lg/mL) andaCD28 (0.1 lg/mL) antibodies and cocultured with irradiated humankidney-2 (HK-2) cells for 6 days. (A) T-cell proliferation was assessedby thymidine incorporation at days 2 and 6 after start of the cocultures.Production of interferon (IFN)-c (B) and interleukin (IL)-10 (C) wasmeasured in culture supernatants by enzyme-linked immunosorbentassay (ELISA). Data are shown as mean ± SD from triplicate wells.One representative of seven independent experiments is shown.

CONCLUSION

Interaction of TECs and T cells may alter the balanceof positive and negative signals to the T cells. One of theconsequences of this interaction seems to be the devel-opment of IL-10 producing T cells, a process that mightaid in limiting or attenuating local T-cell responses. Our


2000

1750

1500

1250

1000

750

500

250

0

200

150

100

50

0

pg/m

L

cpm

(x1

000)

T +

aC

D3

T +

aC

D3

T +

aC

D3/

CD

28

alL-

10R

lL-1

0

lL-1

0 +

alL

-10R

lL-1

0 +

con

trol

ab

HK

-2

HK

-2 +

alL

-10R

HK

-2 +

con

trol

ab

T +

aC

D3/

CD

28 –

aB7-

H1

alC

OS

-L

alC

OS

-L/a

B7-

H1

Con

trol

ab

T cells 15 x 104

T cells 7.5 x 104

T cells 0.375 x 104

Day 2 Day 5 Day 7

200

150

100

50

0

cpm

(x1

000)

HK-2 + T + aCD3/CD28 T + aCD3/CD28

A

C

B

Fig. 9. Mechanism of tubular epithelial cell (TEC)-induced responses of fully activated T cells. (A) Blocking aB7-H1, a inducible T-cell costimulatorligand (aICOS-L) or control antibodies (10 lg/mL) were added to cocultures of T cells and human kidney-2 (HK-2) cells. Interleukin (IL)-10production was measured at day 2. Data are representative of three separate experiments. (B) T cells were cultured with aCD3/aCD28 antibodiesfor 6 days. IL-10 was added at a concentration of 10 ng/mL with or without a neutralizing anti-IL-10R antibody (aIL-10R) (10 lg/mL) or anirrelevant control antibody. For cocultures, HK-2 cells were added at a ratio of 7.5:1. Data are shown as mean ± SD from triplicate cultures. (C)HK-2 cells were mixed with different amounts of aCD3/aCD28-stimulated T cells. [3H] incorporation was measured at days 2, 5, and 7 after startof the cocultures. One representative of two independent experiments is shown.

results provide new insight in the role of resident cells inthe regulation of peripheral immune responses.

ACKNOWLEDGMENTS

This work was financially supported by a grant from the DutchKidney Foundation (#C98.1749 and #C01-1970) and was performed inthe framework of the EU-grants QLG1-CT-2002-01215 and LSHB-CT-2004-504761.

Reprint requests to Dr. Cees van Kooten, Department of Nephrology,Leiden University Medical Center, Albinusdreef 2, 2333 ZA Leiden, TheNetherlands.E-mail: [email protected]

REFERENCES

1. REMUZZI G, RUGGENENTI P, BENIGNI A: Understanding the natureof renal disease progression. Kidney Int 51:2–15, 1997

2. BECKER GJ, HEWITSON TD: The role of tubulointerstitial injury inchronic renal failure. Curr Opin Nephrol Hypertens 9:133–138, 2000

3. VAN KOOTEN C, DAHA MR, VAN ES LA: Tubular epithelial cells: Acritical cell type in the regulation of renal inflammatory processes.Exp Nephrol 7:429–437, 1999

4. HAGERTY DT, ALLEN PM: Processing and presentation of self andforeign antigens by the renal proximal tubule. J Immunol 148:2324–2330, 1992

5. WILSON JL, PROUD G, FORSYTHE JL, et al: Renal allograft rejection.Tubular epithelial cells present alloantigen in the presence of cos-timulatory CD28 antibody. Transplantation 59:91–97, 1995

6. KELLEY VR, SINGER GG: The antigen presentation function of renaltubular epithelial cells. Exp Nephrol 1:102–111, 1993

7. VAN KOOTEN C, WOLTMAN AM, DAHA MR: Immunological func-tion of tubular epithelial cells: The functional implications of CD40expression. Exp Nephrol 8:203–207, 2000

8. DECKERS JG, BOONSTRA JG, VAN DER KOOIJ SW, et al: Tissue-specificcharacteristics of cytotoxic graft-infiltrating T cells during renal al-lograft rejection. Transplantation 64:178–181, 1997

9. HADLEY GA, ROSTAPSHOVA EA, BARTLETT ST: Dominance of tissue-restricted cytotoxic T lymphocytes in the response to allogeneicrenal epithelial cell lines. Transplantation 62:75–83, 1996

10. SINGER GG, YOKOYAMA H, BLOOM RD, et al: Stimulated renaltubular epithelial cells induce anergy in CD4+ T cells. Kidney Int44:1030–1035, 1993


11. NEILSON EG: Is immunologic tolerance of self modulated throughantigen presentation by parenchymal epithelium? Kidney Int44:927–931, 1993

12. KIRBY JA, RAJASEKAR MR, LIN Y, et al: Interaction between T lym-phocytes and kidney epithelial cells during renal allograft rejection.Kidney Int (Suppl 39):S124–S128, 1993

13. FRASCA L, MARELLI-BERG F, IMAMI N, et al: Interferon-gamma-treated renal tubular epithelial cells induce allospecific tolerance.Kidney Int 53:679–689, 1998

14. NIEMANN-MASANEK U, MUELLER A, YARD BA, et al: B7–1 (CD80)and B7–2 (CD 86) expression in human tubular epithelial cells invivo and in vitro. Nephron 92:542–556, 2002

15. BANU N, MEYERS CM: IFN-gamma and LPS differentially modu-late class II MHC and B7–1 expression on murine renal tubularepithelial cells. Kidney Int 55:2250–2263, 1999

16. YOKOYAMA H, ZHENG X, STROM TB, et al: B7(+)-transfectant tubu-lar epithelial cells induce T cell anergy, ignorance or proliferation.Kidney Int 45:1105–1112, 1994

17. VAN KOOTEN C, GERRITSMA JS, PAAPE ME, et al: Possible role forCD40-CD40L in the regulation of interstitial infiltration in the kid-ney. Kidney Int 51:711–721, 1997

18. CROFT M, BRADLEY LM, SWAIN SL: Naive versus memory CD4 T cellresponse to antigen. Memory cells are less dependent on accessorycell costimulation and can respond to many antigen-presenting celltypes including resting B cells. J Immunol 152:2675–2685, 1994

19. SWALLOW MM, WALLIN JJ, SHA WC: B7h, a novel costimulatory ho-molog of B7.1 and B7.2, is induced by TNFalpha. Immunity 11:423–432, 1999

20. LING V, WU PW, FINNERTY HF, et al: Cutting edge: Identificationof GL50, a novel B7-like protein that functionally binds to ICOSreceptor. J Immunol 164:1653–1657, 2000

21. LIANG L, SHA WC: The right place at the right time: novel B7 familymembers regulate effector T cell responses. Curr Opin Immunol14:384–390, 2002

22. LOHNING M, HUTLOFF A, KALLINICH T, et al: Expression of ICOS invivo defines CD4+ effector T cells with high inflammatory potentialand a strong bias for secretion of interleukin 10. J Exp Med 197:181–193, 2003

23. FREEMAN GJ, LONG AJ, IWAI Y, et al: Engagement of the PD-1 im-munoinhibitory receptor by a novel B7 family member leads tonegative regulation of lymphocyte activation. J Exp Med 192:1027–1034, 2000

24. DONG H, ZHU G, TAMADA K, et al: B7–H1, a third member of theB7 family, co-stimulates T-cell proliferation and interleukin-10 se-cretion. Nat Med 5:1365–1369, 1999

25. WANG S, BAJORATH J, FLIES DB, et al: Molecular modeling and func-tional mapping of B7-H1 and B7-DC uncouple costimulatory func-tion from PD-1 interaction. J Exp Med 197:1083–1091, 2003

26. NISHIMURA H, HONJO T: PD-1: An inhibitory immunoreceptor in-volved in peripheral tolerance. Trends Immunol 22:265–268, 2001

27. SHARPE AH, FREEMAN GJ: The B7-CD28 superfamily. Nat Rev Im-munol 2:116–126, 2002

28. RYAN MJ, JOHNSON G, KIRK J, et al: HK-2: an immortalized proximaltubule epithelial cell line from normal adult human kidney. KidneyInt 45:48–57, 1994

29. VAN KOOTEN C, LAM S, DAHA MR: Isolation, culture, characteri-zation and use of human renal tubular epithelial cells. J Nephrol14:204–210, 2001

30. WOLTMAN AM, DE FIJTER JW, KAMERLING SW, et al: The effect ofcalcineurin inhibitors and corticosteroids on the differentiation ofhuman dendritic cells. Eur J Immunol 30:1807–1812, 2000

31. VALLE A, GARRONE P, YSSEL H, et al: mAb 104, a new monoclonalantibody, recognizes the B7 antigen that is expressed on activatedB cells and HTLV-1-transformed T cells. Immunology 69:531–535,1990

32. KHAYYAMIAN S, HUTLOFF A, BUCHNER K, et al: ICOS-ligand, ex-pressed on human endothelial cells, costimulates Th1 and Th2 cy-tokine secretion by memory CD4+ T cells. Proc Natl Acad Sci USA99:6198–6203, 2002

33. DONG H, STROME SE, SALOMAO DR, et al: Tumor-associated B7-H1 promotes T-cell apoptosis: A potential mechanism of immuneevasion. Nat Med 8:793–800, 2002

34. GARRONE P, NEIDHARDT EM, GARCIA E, et al: Fas ligation induces

apoptosis of CD40-activated human B lymphocytes. J Exp Med182:1265–1273, 1995

35. VAN LIER RA, BOOT JH, VERHOEVEN AJ, et al: Functional studieswith anti-CD3 heavy chain isotype switch-variant monoclonal an-tibodies. Accessory cell-independent induction of interleukin 2 re-sponsiveness in T cells by epsilon-anti-CD3. J Immunol 139:2873–2879, 1987

36. LIU Y, DE WAAL MR, BRIERE F, et al: The EBV IL-10 homologueis a selective agonist with impaired binding to the IL-10 receptor. JImmunol 158:604–613, 1997

37. DE FIJTER JW, DAHA MR, SCHROEIJERS WE, et al: Increased IL-10 production by stimulated whole blood cultures in primary IgAnephropathy. Clin Exp Immunol 111:429–434, 1998

38. POUW-KRAAN T, VAN KOOTEN C, RENSINK I, et al: Interleukin (IL)-4production by human T cells: Differential regulation of IL-4 vs. IL-2production. Eur J Immunol 22:1237–1241, 1992

39. AKDIS CA, BLASER K: Mechanisms of interleukin-10-mediated im-mune suppression. Immunology 103:131–136, 2001

40. MOORE KW, DE WAAL MR, COFFMAN RL, et al: Interleukin-10 andthe interleukin-10 receptor. Annu Rev Immunol 19:683–765, 2001

41. CAO Y, ZHOU H, TAO J, et al: Keratinocytes induce local toleranceto skin graft by activating interleukin-10-secreting T cells in thecontext of costimulation molecule B7-H1. Transplantation 75:1390–1396, 2003

42. LANING JC, DELUCA JE, ISAACS AND CM, et al: In vitro analysis ofCD40-CD154 and CD28-CD80/86 interactions in the primary T-cellresponse to allogeneic “nonprofessional” antigen presenting cells.Transplantation 71:1467–1474, 2001

43. WAHL P, SCHOOP R, BILIC G, et al: Renal tubular epithelial expres-sion of the costimulatory molecule B7RP-1 (inducible costimulatorligand). J Am Soc Nephrol 13:1517–1526, 2002

44. MAZANET MM, HUGHES CC: B7-H1 is expressed by human en-dothelial cells and suppresses T cell cytokine synthesis. J Immunol169:3581–3588, 2002

45. PIROZZI G, LOMBARI V, ZANZI D, et al: CD40 expressed on hu-man melanoma cells mediates T cell co-stimulation and tumor cellgrowth. Int Immunol 12:787–795, 2000

46. ROGERS NJ, JACKSON IM, JORDAN WJ, et al: CD40 can costimulatehuman memory T cells and favors IL-10 secretion. Eur J Immunol33:1094–1104, 2003

47. PENFIELD JG, WANG Y, LI S, et al: Transplant surgery injury recruitsrecipient MHC class II-positive leukocytes into the kidney. KidneyInt 56:1759–1769, 1999

48. RACUSEN LC, SOLEZ K, COLVIN RB, et al: The Banff 97 workingclassification of renal allograft pathology. Kidney Int 55:713–723,1999

49. STARZL TE, MURASE N, ABU-ELMAGD K, et al: Tolerogenic immuno-suppression for organ transplantation. Lancet 361:1502–1510, 2003

50. MATZINGER P: The danger model: a renewed sense of self. Science296:301–305, 2002

51. MARELLI-BERG FM, LECHLER RI: Antigen presentation byparenchymal cells: A route to peripheral tolerance? Immunol Rev172:297–314, 1999

52. CHAPOVAL AI, NI J, LAU JS, et al: B7–H3: a costimulatory moleculefor T cell activation and IFN-gamma production. Nat Immunol2:269–274, 2001

53. PRASAD DV, RICHARDS S, MAI XM, et al: B7S1, a novel B7 fam-ily member that negatively regulates T cell activation. Immunity18:863–873, 2003

54. SICA GL, CHOI IH, ZHU G, et al: B7-h4, a molecule of the b7family, negatively regulates T cell immunity. Immunity 18:849–861,2003

55. MELICHAR B, NASH MA, LENZI R, et al: Expression of costimula-tory molecules CD80 and CD86 and their receptors CD28, CTLA-4on malignant ascites CD3+ tumour-infiltrating lymphocytes (TIL)from patients with ovarian and other types of peritoneal carcino-matosis. Clin Exp Immunol 119:19–27, 2000

56. STAMENKOVIC I, CLARK EA, SEED B: A B-lymphocyte activationmolecule related to the nerve growth factor receptor and inducedby cytokines in carcinomas. EMBO J 8:1403–1410, 1989

57. VAN KOOTEN C, VAN DER LINDE X, WOLTMAN AM, et al: Synergisticeffect of interleukin-1 and CD40L on the activation of human renaltubular epithelial cells. Kidney Int 56:41–51, 1999

Documents

Renal tubular epithelial cells modulate T-cell responses via ICOS-L and B7-H1