logy 6 (2006) 1889–1894www.elsevier.com/locate/intimp

International Immunopharmaco

Immunosuppressive activity of recombinant ILT3

George Vlad, Zhuoru Liu, Qing-Yin Zhang, Raffaello Cortesini, Nicole Suciu-Foca ⁎

Department of Pathology, Columbia University, 630 West 168th Street, P and S 14-401, New York, NY 10032, USA

Received 18 July 2006; accepted 19 July 2006

Abstract

Tolerogenic antigen presenting cells (APC) are characterized by high expression of the inhibitory receptors ILT3 and ILT4. Wehave engineered ILT3 and ILT4 cytoplasmic deletion mutants (ILT3delta and ILT4delta), which were transfected in the dendritic-likecell line KG1, to investigate ILT3 and ILT4's capacity to signal extracellularly. KG1.ILT3delta, similar to untruncated ILT3, inhibits Tcell responses such as proliferation and cell-mediated cytotoxicity. In contrast, KG1.ILT4delta lost the suppressive activity ofuntruncated ILT4. This indicates that the inhibitory function of ILT4 relies entirely on the cytoplasmic region containing ITIMmotifs.We further demonstrated that recombinant soluble ILT3 inhibits T helper and cytotoxic function while inducing the differentiation ofCD8+ Ts cells. Hence, Ts modulate APC function inducing inhibitory receptors, which in turn elicit the generation of Ts.© 2006 Elsevier B.V. All rights reserved.

Keywords: CD8+ T suppressor cells; Tolerogenic APC; Membrane ILT3; Soluble ILT3

1. Introduction

The interaction between antigen-specific CD8+ CD28−

FOXP3+ T suppressor cells (Ts) with allogeneic dendriticcells (DC) or activated endothelial cells (EC) results intolerization of DC or EC inducing the upregulation of theimmunoglobulin-like transcripts ILT3 and ILT4 [1–14].We have demonstrated that tolerogenic DC or EC anergizealloreactive CD4+CD45RO+CD25+ T cells convertingthem into FOXP3+ regulatory T cells (Treg), whichcontinue the cascade of suppression [5,14]. Interleukin-10(IL-10) and interferon (IFN)-alpha also induce ILT3 andILT4 upregulation in DC and EC [5,11,14]. This implies acommonmechanism of EC andDC-mediated suppressionof T cell alloreactivity. This finding and the observation

⁎ Corresponding author. Tel.: +1 212 305 6941; fax: +1 212 3053429.

E-mail address: ns20@columbia.edu (N. Suciu-Foca).

1567-5769/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.intimp.2006.07.017

that in organ allograft recipients quiescence is associatedwith the presence in the circulation of donor-specific TsandTr emphasize the importance of the cross-talk betweentolerogenic DC and EC with T cells in suppression ofallorecognition [11,13]. This concept was further sub-stantiated in a rat model of heart allotransplantation.Tolerance was induced in ACI recipients by multipletransfusions of UVB-irradiated blood from Lewis heartdonors [12]. CD8+ T cells from tolerant ACI ratsexpressed FOXP3, transferred tolerance to naïve second-ary hosts and induced the upregulation of PIR-B, an ILT4orthologue in Lewis DC and heart EC. When long-termsurviving Lewis heart allografts with PIR-B+ EC wereretransplanted from a primary to a secondary ACIrecipient, they were indefinitely tolerated, demonstratingthat the bi-directional interaction between Ts and APC iscrucial to the induction andmaintenance of tolerance [12].Agents which elicit this interaction, such as humanrecombinant ILT3, may be important for achieving long-

1890 G. Vlad et al. / International Immunopharmacology 6 (2006) 1889–1894

term tolerance. We have explored this possibility bytesting the immunosuppressive activity of membrane andsoluble ILT3 and ILT4 molecules [15].

2. Materials and methods

2.1. Construction and expression of membrane and solubleILT molecules

The cytoplasmic domain of ILT3 and ILT4 which containsthe ITIM motifs was deleted by PCR amplification and cloningin the expression vector pc DNA4/TO/myc-His in frame with ac-myc epitope and polyhistidine (6×His) region. The resultingmutants called ILT3delta (M1-E544) and ILT4delta (M1-E529) encodes proteins which contain the putative leaderpeptide, the extracellular, transmembrane and a stretch of 48and 47 amino acids of the cytoplasmic domain of ILT3 andILT4, respectively, fused with C-terminal myc-His tag (Fig. 1).The ILT3delta and ILT4delta inserts were subcloned into thebicistronic retroviral vector MIG (MSCV-IRES-GFP),sequenced and overexpressed in the dendritic cell line KG1(ATCC, Mannassas, VA). Transfectants were sorted for GFPexpression, cloned and used as stimulators in T cellproliferation and cytotoxicity assays (Fig. 2).

To generate an ILT3-Fc fusion protein, a cDNA fragmentcoding for the entire extracellular domain of human ILT3 wasfused to the Fc portion of human IgG1 heavy chain. Binding toFc receptors was abolished by introducing a mutation in the N-linked glycosylation site, N77 (Asn>Gln) of the Fc domain.Expression vector pcDNA3 containing ILT3-Fc fusion genewas transfected into CHO-S cells. Homogenous cell popula-tions were obtained by limiting dilution and clones with high

Fig. 1. Schematic representation of ILT3 and ILT4 and of c

expression of ILT3 were selected. ILT3-Fc fusion proteinwas purified from the supernatant using a recombinant proteinA-Sepharose fast-flow column and tested in Tcell proliferationand cell-mediated lymphocytotoxicity (CML) assays.

2.2. Proliferation assays

Responding CD3+CD25− T cells were isolated fromperipheral blood mononuclear cells by use of magnetic beads(Miltenyi Biotec, Gladbach, Germany) and plated in 96-well U-bottom cell culture plates (at 5×104 cells/well). The myelomo-nocytic cell line KG1, or the KG1 transfectants (KG1.MIG,KG1.ILT3, KG1.ILT3delta, KG1.ILT4, KG1.ILT4delta) wereirradiated (3000 rad) and used as stimulating cells (at aconcentration of 2.5×104 cells/well). Cultures were harvestedon day 6. [3H]Thymidine was added to the cultures 18 h beforeharvesting and incorporation was determined by scintillationspectrometry using a LKB 1250 Betaplate counter (PerkinElmer). Mean cpm of triplicate cultures and the S.D. from themeanwere calculated. Recombinant human IL-2 (10U/ml, R&DSystems) or anti-ILT3 mAb (5 μg/ml) (clone ZM 3.8, a gift fromDr. M. Colonna, Washington University School of Medicine, St.Louis, MO) were added to the proliferation assays as indicated.The fusion protein ILT3-Fcwas used at various concentrations asindicated. Normal human IgG (Sigma-Aldrich) was used at thesame concentrations in parallel control cultures.

2.3. Induction of CD8+ Ts cells by use of membrane-boundILT3 or ILT3-Fc protein

CD3+CD25− T cells (1×106/ml) were cultured withirradiated allogeneic CD2-depleted PBMCs (0.5×106/ml) inthe presence or absence of ILT3-Fc protein or control human IgG

ytoplamic deletion mutants ILT3delta and ILT4delta.

Fig. 2. Expression of ILT3 and ILT4 in KG1 cells transfected with ILT3, ILT3delta, ILT4 or ILT4delta.

1891G. Vlad et al. / International Immunopharmacology 6 (2006) 1889–1894

(50 μg/ml) or with irradiated KG1, KG1.MIG, KG1.ILT3, orKG1.ILT3 (0.5×106/ml). CD8+ T cells were isolated from thesecultures after 7 days and tested for inhibitory activity. Forsuppression assays, various numbers of alloactivated CD8+ Tcells were added to primary MLC containing autologousCD3+CD25− T cells (5×104 cells/well) and the allogeneicAPC used for Ts priming (2.5×104 cells/well). Proliferation wasmeasured as described above.

2.4. Cytokine bead array (CBA) and ELISPOT assays

CBA kits for detection of Th1/Th2 cytokines werepurchased from BD Biosciences and used according to themanufacturer's instructions. Briefly, purified T cells wereallo-stimulated by co-incubation with: the KG1 cell line, andits transfectants KG1.ILT3, KG1.ILT3delta, KG1.ILT4 andKG1.ILT4delta. As positive control, solid-phase anti-CD3(clone UCHT1) plus soluble anti-CD28 (clone CD28.2)antibodies were used to stimulate the T cells. T cellresponders or KG1 stimulators alone were used as negativecontrols of cytokine production. Supernatants were removedfrom the cell cultures at days 0, 3 and 7, diluted 1:10 in theassay diluent provided. Cytokine standards, used for thequantification of the results, were prepared as indicated by themanufacturer. Samples and cytokine standards were coin-cubated with capture beads and PE-conjugated detectionreagents for 3 h, protected from light. The beads were then

washed and brought to a suspension volume of 300 μl.Finally, the samples were read and analyzed using theFACSCalibur flow cytometry instrument and the Cell Questsoftware (both from BD Bioscience). ELISPOT assays wereconducted as described in Ref. [15].

3. Results

CD3+ T cells from healthy blood donors were cultured for6 days with irradiated KG1, KG1.ILT3 (transfected with fulllength ILT3) or KG1.ILT3delta (the cytoplasmic deletion mutantof ILT3). KG1 cells induced strong T cell proliferation, whileKG1.ILT3 andKG1.ILT3delta elicited onlyweak reactivity. Tcellresponsiveness to KG1.ILT3 and KG1.ILT3delta was enhancedwhen rIL2was added to the assay indicating that Tcell interactionwith the extracellular domain of ILT3 results in anergy. In contrast,when T cells were stimulated in vitro with KG1, KG1.ILT4 (fulllength) or KG1.ILT4delta, they responded strongly to KG1 andKG1.ILT4delta but not to KG1.ILT4 (Fig. 3). This indicates thatthe inhibitory effect of ILT4 requires the intracellular domain forsignaling. Experiments in which T cells were stimulated inprimary cultures with KG1 and restimulated in secondary cultureswith KG1.ILT3delta or KG1.ILT4delta yielded similar results.Cytokine bead arrays showed that KG1 and KG1.ILT4deltaelicited high production, while the full-length transfectants (KG1.ILT3 and KG1.ILT4) as well as the ILT3 deletion mutant (KG1.ILT3delta) inhibited γ-IFN synthesis (Fig. 4). The data indicated

Fig. 4. IFN-γ production in cultures containing responding T cells andstimulating KG1, KG1.ILT3, KG1.ILT3delta, KG1.ILT4 or KG1.ILT4delta cells.

1892 G. Vlad et al. / International Immunopharmacology 6 (2006) 1889–1894

that while deletion of the intracytoplasmic domain does not alterthe inhibitory activity of ILT3, it abolishes that of ILT4 [15].

Based on these findings, we engineered a soluble ILT3 Fc-fusion protein and studied its effect in vitro. Study of the effectof ILT3-Fc protein and purified human Ig (used as a control) onT cell proliferation in primary and secondary MLC demon-strated that ILT3-Fc inhibits T cell reactivity in a dose-dependent manner. In multiple repeat experiments, 50%inhibition was obtained at 5 μg/ml, while at 50 μg/ml, therewas more than 85% inhibition of T cell responses in primary orsecondary MLC. The protein was not toxic to the cells at anyconcentration, as determined by viability studies. The inhibi-tory effect of ILT3-Fc was abolished by adding to the culturesanti-ILT3monoclonal antibodies or recombinant IL-2, suggest-ing that Tcells were rendered anergic in the presence of soluble(s)ILT3.

To study the effect of sILT3 on the generation of cytotoxicT lymphocytes (CTL), CD3+CD25− T cells were primed toallogeneic APC in the presence of ILT3-Fc or control IgG. Onday 7, CD8+ T cells were isolated from these cultures andtested for CTL activity against priming APC using Annexin V/Propidium iodide for target cell staining. CD8+ T cells primed

Fig. 3. T cell proliferation in response to irradiated KG1, KG1.ILT3,KG1.ILT3delta, KG1.ILT4 and KG1.ILT4delta cells.

in the presence of ILT3-Fc had no cytotoxic activity, while Tcells from control cultures containing human Ig induced dose-dependent target cell lysis [15].

We next explored the possibility that T cell priming in thepresence of ILT3-Fc results in the generation of regulatory cells.For this, CD3+CD25− T cells, which were cultured for 7 dayswith allogeneic APC in medium containing ILT3-Fc or controlIg, were sorted into CD4+ and CD8+ populations. Primed CD4+

or CD8+ Tcells were then added to primaryMLC containing thesame responder–stimulator combination. CD8+ T cells dis-played strong suppressive activity, while CD4+ T cell had noinhibitory effect although proven to be anergic. Diffusionchamber experiments and testing of cytokine production byELISPOT assay for IFN-γ, IL-2 and IL-10 demonstrated thatCD8+ Ts generated in the presence of ILT3-Fc act in a contact,dose-dependent and cytokine-independent manner [15]. Todetermine whether CD8+ Ts interact directly with alloactivatedCD4+ T cells or allostimulatory APC, we challenged the primedTs and Th cells with APC expressing the MHC-class II antigensrecognized by Th, the class I antigens recognized by Ts, or bothtypes of APC. These 3- and 4-cell mixture experimentsdemonstrated unambiguously that ILT3-Fc-induced Ts interactwith bridging APC, rendering them tolerogenic to Th withcognate specificity. These data suggest that sILT3may provide atherapeutic tool for generation and adoptive transfer ofallospecific Ts and/or treatment of rejection.

4. Discussion

Over the last 10 years, the literature has beendominated by studies emphasizing the importance ofnaturally occurring CD4+CD25+ regulatory T cells. Theprevailing dogma has been that this is a unique T cellsubset, which derives from CD4+CD25+ thymocytesand is endowed with the omnipotent capacity of

1893G. Vlad et al. / International Immunopharmacology 6 (2006) 1889–1894

inhibiting auto and allo-reactivity, against self- and non-self-antigens. These natural Tregs were shown to actdirectly on other activated CD4+ and CD8+ Tcells. Highexpression of CD25 was considered a characteristic traitof Tregs since in rodents depletion of CD25+ T cells wasshown to result in autoimmunity, while transfer ofCD25+ T cells inhibited autoimmunity and allograftrejection [17–19]. Paradoxically, in human treatmentwith anti-CD25 mAb prevents transplant rejection, aresult which cannot be reconciled with the deletion ofsuppressors.

More recently, rodent CD4+CD25+ Tregs were shownto express FOXP3, deemed to be another characteristicmarker [17–20]. Much of the dogma that CD4+CD25+

FOXP3+ natural Tregs derive from the thymus andrepresent a distinct, MHC unrestricted population hasbeen shattered by the finding that Treg can be generated inthe periphery fromCD4+ CD25− FOXP3− Tcells, both inrodents and human [21]. Furthermore, it has been shownthat CD4+CD25+ FOXP3+ Treg do not develop in thethymus of MHC class II knockout mice. This indicatesthat their development requires TCR interaction withMHC/peptide complexes [22].

Ignored for a long time, the CD8+ T suppressor cellscame back into the play [23]. We demonstrated for thefirst time that human T suppressor cells are CD8+, MHCclass I restricted, and act directly on APC inducing theup-regulation of inhibitory receptors and down-regula-tion of costimulatory molecules [1–15]. We emphasizedin our studies the fact that the adaptive immune responsecannot be regulated in an antigen non-specific mannerby natural Tregs. The immune response to pathogens orvaccines could not develop if natural Tregs wouldinterfere with it arbitrarily. Similar to helper andcytotoxic T cells, suppressors must act in an antigen-specific manner in order to regulate the immuneresponse [5,14,21].

We further challenged the dogmatic view thatimmature DC are tolerogenic. In our experience,immature DC are excellent stimulators of alloreactivityand become tolerogenic only upon up-regulation ofinhibitory receptors such as ILT3 and ILT4 in human orPIR-B in rodents. We showed that tolerogenic DC arecentral to the generation of CD8+ FOXP3+ Ts as well asof CD4+CD25+ Treg.

Most importantly, we found that Ts induce the up-regulation of inhibitory receptors and down-regulationof adhesion and costimulatory molecules not only onDC but also on non-professional APC such as EC. Thefinding that the interaction of antigen-specific Ts withAPC and, indeed, EC is the true source of immuneregulation explains what otherwise would be a para-

doxical and counterproductive aspect of immuneregulation. If immune regulation depended on non-Ag-specific T cells alone, it would be broadly suppressiveand thus possibly more detrimental to the well-beingthan the autoimmune reactions it presumably controls.In contrast, the focus on APC and EC allows immuneregulation to be tissue specific and site specific, and inthis way, it provides much more efficient controlmechanism.

DC and EC which have been “licensed to heal” arelikely to inhibit the direct allorecognition pathway. Inthe absence of inflammation, resulting from the directattack of the graft by host T cells, indirect allo-recognition may be of no consequence. This view issupported by our findings that in rats, tolerated heartallografts whose EC express high levels of inhibitoryreceptors such as PIR-B are resistant to rejection whenre-transplanted in secondary, naïve recipients. There-fore, these allografts do not elicit either direct or indirectT cell alloreactivity [12].

Finally, our data suggest that tolerance to allogeneictransplants can be induced by treatment with solubleILT3. While sILT3 displayed strong inhibitory activity,sILT4 does not according to Beinhauer et al. [16]. Thisconfirms the conclusion that ILT3 signals both intra- andextracellularly, while ILT4 exerts its inhibitory effectsignaling only intracellularly and blocking the differ-entiation of mature immunogenic APC. This hypothesisis supported by our finding that recombinant sILT3inhibits T cell alloreactivity in vitro and converts allo-activated CD8+ T cells into antigen-specific FOXP3+ Tscells. sILT3 is unlikely to induce nonspecific immuno-suppression because it acts only on activated T cells.

Another field that may benefit from the use oftolerogenic DC or sILT3 is autoimmunity. It is believedthat under inflammatory conditions, DC which presenttissue- or organ-specific peptides derived from cellsundergoing necrosis can prime autoreactive T cells. Theprogression of the autoimmune response may beprevented by treatment of patients with autologousILT3high ILT4high DC that have processed ex vivo(apoptotic cells containing) the putative autoantigen.Alternatively, administration of sILT3 may convertproliferating autoreactive T cells into Ts cells, whichinitiate and maintain the cascade of suppressive events.Similarly, cell-therapy using antigen specific CD8+ Tsprimed in vitro in the presence of sILT3 may inhibitautoimmunity.

It is our belief that licensing APC to heal, by inducingthem to express inhibitory receptors ILT3 and ILT4,opens new avenues to induction of immunologicaltolerance in transplantation and autoimmune disease.

1894 G. Vlad et al. / International Immunopharmacology 6 (2006) 1889–1894

Acknowledgements

This work was supported by grants from the NIH(AI25210-19, AI55234-03) and the InteruniversityOrgan Transplantation Consortium, Rome, Italy.

References

[1] Liu Z, Tugulea S, Cortesini R, Suciu-Foca N. Specificsuppression of T helper alloreactivity by allo-MHC class I-restricted CD8+CD28− T cells. Int Immunol 1998;10:775–83.

[2] Cortesini R, LeMaoult J, Ciubotariu R, Cortesini NS.CD8+CD28− T suppressor cells and the induction of antigen-specific, antigen-presenting cell-mediated suppression of Threactivity. Immunol Rev 2001;182:201–6.

[3] Suciu-Foca N, Manavalan JS, Cortesini R. Generation andfunction of antigen-specific suppressor and regulatory T cells.Transpl Immunol 2003;11:235–44.

[4] Chang CC, Ciubotariu R, Manavalan JS, Yuan J, Colovai AI,et al. Tolerization of dendritic cells by T(S) cells: the crucialrole of inhibitory receptors ILT3 and ILT4. Nat Immunol2002;3:237–43.

[5] Manavalan JS, Rossi PC, Vlad G, Piazza F, Yarilina A, et al. Highexpression of ILT3 and ILT4 is a general feature of tolerogenicdendritic cells. Transpl Immunol 2003;11:245–58.

[6] Ciubotariu R, Colovai AI, Pennesi G, Liu Z, Smith D, et al.Specific suppression of human CD4+ Th cell responses to pigMHC antigens by CD8+CD28− regulatory T cells. J Immunol1998;161:5193–202.

[7] Jiang S, Tugulea S, Pennesi G, Liu Z, Mulder A, et al. Inductionof MHC-class I restricted human suppressor T cells by peptidepriming in vitro. Hum Immunol 1998;59:690–9.

[8] Scotto L, Naiyer AJ, Galluzzo S, Rossi P, Manavalan JS, et al.Overlap between molecular markers expressed by naturallyoccurring CD4+CD25+ regulatory T cells and antigen specificCD4+CD25+ and CD8+CD28− T suppressor cells. Hum Immunol2004;65:1297–306.

[9] Liu Z, Tugulea S, Cortesini R, Lederman S, Suciu-Foca N.Inhibition of CD40 signaling pathway in antigen presenting cellsby T suppressor cells. Hum Immunol 1999;60:568–74.

[10] Li J, Liu Z, Jiang S, Cortesini R, Lederman S, Suciu-Foca N.T suppressor lymphocytes inhibit NF-kappa B-mediated tran-scription of CD86 gene in APC. J Immunol 1999;163:6386–92.

[11] Manavalan JS, Kim-Schulze S, Scotto L, Naiyer AJ, Vlad G,et al. Alloantigen specific CD8+CD28− FOXP3+ T suppressorcells induce ILT3+ ILT4+ tolerogenic endothelial cells,inhibiting alloreactivity. Int Immunol 2004;16:1055–68.

[12] Liu J, Liu Z, Witkowski P, Vlad G, Manavalan JS, et al. RatCD8+ FOXP3+ T suppressor cells mediate tolerance to allogeneicheart transplants, inducing PIR-B in APC and rendering the graftinvulnerable to rejection. Transpl Immunol 2004;13:239–47.

[13] Ciubotariu R, Vasilescu R, Ho E, Cinti P, Cancedda C, et al.Detection of T suppressor cells in patients with organ allografts.Hum Immunol 2001;62:15–20.

[14] Vlad G, Cortesini R, Suciu-Foca N. License to heal: bidirectionalinteraction of antigen-specific regulatory T cells and tolerogenicAPC. J Immunol 2005;174:5907–14.

[15] Kim-Schulze S, Scotto L, Vlad G, Piazza F, Lin H, et al.Recombinant Ig-like transcript 3-Fc modulates T cell responsesvia induction of Th anergy and differentiation of CD8+ Tsuppressor cells. J Immunol 2006;176:2790–8.

[16] Beinhauer BG, McBride JM, Graf P, Pursch E, Bongers M, et al.Interleukin 10 regulates cell surface and soluble LIR-2 (CD85d)expression on dendritic cells resulting in T cell hyporesponsive-ness in vitro. Eur J Immunol 2004;34:74–80.

[17] Sakaguchi S. Naturally arising CD4+ regulatory T cells forimmunologic self-tolerance and negative control of immuneresponses. Annu Rev Immunol 2004;22:531–62.

[18] Shevach EM. CD4+CD25+ suppressor T cells: more questionsthan answers. Nat Rev Immunol 2002;2:389–400.

[19] Wood KJ, Sakaguchi S. Regulatory T cells in transplantationtolerance. Nat Rev Immunol 2003;3:199–210.

[20] Shortman K, Heath WR. Immunity or tolerance? That is thequestion for dendritic cells. Nat Immunol 2001;2:988–9.

[21] Bluestone JA, Abbas AK. Natural versus adaptive regulatoryT cells. Nat Rev Immunol 2003;3:253–7.

[22] Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, FarrAG, Rudensky AY. Regulatory T cell lineage specification bythe forkhead transcription factor FOXP3. Immunity 2005;22:329–341.

[23] Zimring JC, Kapp JA. Identification and characterization ofCD8+ suppressor T cells. Immunol Res 2004;29:303–12.