of May 17, 2018.This information is current as

Acquisition of CD80 (B7-1) by T Cells

and Jeffrey SchlomTagaya, Mayumi Naramura, James W. Hodge, John Bernon Helen Sabzevari, Judy Kantor, Adnan Jaigirdar, Yutaka

http://www.jimmunol.org/content/166/4/2505doi: 10.4049/jimmunol.166.4.2505

2001; 166:2505-2513; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/166/4/2505.full#ref-list-1

, 19 of which you can access for free at: cites 38 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

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

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

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2001 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Acquisition of CD80 (B7-1) by T Cells

Helen Sabzevari,* Judy Kantor,* Adnan Jaigirdar,* Yutaka Tagaya,† Mayumi Naramura, ‡

James W. Hodge,* John Bernon,* and Jeffrey Schlom1*

Activation of T cells usually requires two signals. Signal 1 is mediated via a peptide-MHC on the APC; signal 2 is mediated viaa costimulatory molecule on the APC surface. We demonstrate here that naive CD41 T cells actually acquire the costimulatorymolecule CD80 (B7-1) from syngeneic APCs after activation. This phenomenon was demonstrated showing acquisition of CD80by T cells from CD80/CD86 (B7-2) knockout mice, and by treating T cells with cyclohexamide to further rule out endogenousexpression of CD80 by T cells. Moreover, no CD80 mRNA could be detected in T cells that had acquired CD80. The amount ofacquisition of CD80 by T cells was shown to be directly related to both the strength of signal 1 and the amount of CD80 on theAPC. Specificity of this acquisition was also shown by the lack of acquisition by T cells from CD28 knockout mice (implicatingCD28 in this process), the lack of acquisition of CD40 (another molecule on the APC surface) by T cells, and confocal microscopystudies. We demonstrate for the first time that 1) naive T cells, following acquisition of CD80 from APCs, were themselves shownto be capable of acting as APCs; and 2) memory T cells that have acquired CD80 from APCs undergo apoptosis in the presenceof increased levels of signal 1. Thus we demonstrate both immunostimulatory and immunoregulatory functions as a result of CD80acquisition by different T cell populations. The Journal of Immunology,2001, 166: 2505–2513.

T cell activation usually requires at least two different sig-nals: 1) an Ag-specific signal (signal 1); and 2) a second,or costimulatory, signal (signal 2) that is provided by re-

ceptor ligand interaction between T cells and APCs, such as den-dritic cells (DCs).2 Although a variety of molecules reportedlyhave costimulation function, the majority of studies to date havebeen focused on costimulation delivered via CD28 and CD80 in-teractions (1, 2). CD80 is a member of the Ig gene superfamily andis expressed mainly on professional APCs; its receptor, CD28, isconstitutively expressed on thymocytes and on most resting andactivated T cells. CD80 has also been shown to react with CTLA-4on the surface of activated T cells; however, CTLA-4 has beenshown not to be expressed on resting T cells (3, 4). Interactionof CD28 and CD80 prevents the induction of clonal anergy andapoptosis of T cells (5).

Various studies have reported that CD80 can also be detected onactivated human T cell clones, T cells activated repeatedly in vitro,and T cells that have infiltrated skin lesions and are in rheumatoidsynovial fluid (6–11). Studies by Kochli et al. that analyzed theexpression of CD80 and MHC complex class II molecules on cir-culating T cells of HIV-infected individuals have suggested thatthese T cells develop an Ag-presenting phenotype by up-regulatingexpression of HLA and CD80 (12). Moreover, Wolthers et al. (13)

reported increased levels of CD80 expression on both CD41 andCD81 cells in HIV-infected patients and revealed that these T cellswith accessory cell properties can costimulate proliferation in pu-rified responder T cells from healthy control donors. Further stud-ies of lupus patients have demonstrated that these individuals haveincreased amounts of CD80 on PBLs, including T cells (14). How-ever, although all of these studies reported increased levels ofCD80 molecules on T cells, none of the studies have actually dem-onstrated an endogenous up-regulation of CD80 molecules on Tcells (6–14).

In parallel murine studies, the amount of CD80 on T cells wasreported to vary with the state of activation. Resting T cells dis-played little or no CD80; however, upon activation, CD80 levelson the cell surface increased concomitantly, peaking at 72 h (15).In a recent study, Weintraub et al. (16, 17) demonstrated that al-though splenic T cells from normal B6 mice were negative forCD80, freshly isolated T lymphocytes from the spleens of auto-immune B6/gld and B6/lpr mice showed higher levels of CD80.This phenomenon was age dependent and may have contributed tolymphoproliferation and autoimmunity. Although these studiesmentioned “increases in CD80 expression in T cells,” no data wereactually presented in these studies as to whether these increaseswere due to up-regulation of CD80 in T cells or acquisitionof CD80.

Previous studies have demonstrated that T cells and thymocytescan acquire MHC class I and class II determinants from other cellspresent in the local environment (18, 19). Furthermore, it has beendemonstrated that double-negative thymocytes can passively ac-quire a CD41/CD81 determinant from other cells in the thymusthat are actively expressing their CD41/CD81 genes (20). A re-cent study by Huang et al. (19) demonstrated that when peptide-specific T cells interacted with APCs, MHC molecules on APCsformed clusters at the site of T cell contact. Afterward, these clus-ters were acquired by T cells and internalized through TCR endo-cytosis. These studies also showed that the presence of APC-de-rived peptide-MHC molecules on the T cell surface could makethese T cells susceptible to lysis by neighboring T cells. Moreover,it has been reported recently that the activation of T cells with

*Laboratory of Tumor Immunology and Biology, Division of Basic Sciences, Na-tional Cancer Institute;†Metabolism Branch, Division of Clinical Sciences, NationalCancer Institute; and‡Laboratory of Immunology, National Institute of Allergy andInfectious Diseases, National Institutes of Health, Bethesda, MD 20892

Received for publication August 24, 2000. Accepted for publication December4, 2000.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby markedadvertisementin accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 Address correspondence and reprint requests to Dr. Jeffrey Schlom, Laboratory ofTumor Immunology and Biology, National Cancer Institute, National Institutes ofHealth, 10 Center Drive, Building 10, Room 8B09, Bethesda, MD 20892-1402. E-mail address: js141c@nih.gov2 Abbreviations used in this paper: DC, dendritic cell; KO, knockout; B7dKO, CD80/CD86 double-knockout; PCC/TCR-Tg, TCR transgenic for pigeon cytochromec88–104;rF-CD80, recombinant fowlpox virus expressing CD80; FP-WT, wild-type fowlpox;GFP, green fluorescence protein; CHX, cyclohexamide.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Drosophilacells as APCs may result in the rapid transfer of MHCmolecules and other costimulatory molecules, such as CD80, fromAPCs to T cells (21).

Although the concept of molecule acquisition by T cells origi-nated 20 years ago, the actual interactions of receptor ligands andthe factors affecting ligand acquisition still remain unclear. In thisregard, we analyzed CD80 levels on murine T cells upon stimu-lation with various APCs expressing different levels of CD80 ontheir surface. Although CD80 was previously thought to be endo-genously up-regulated, results of studies of CD80/CD86 double-knockout (B7dKO) mice reported here demonstrate that CD80 isphysically acquired by T cells from APCs shortly after T cell ac-tivation. We also demonstrate here that CD80 acquisition by Tcells is mediated by the TCR and its CD28 ligand (confirming therecent data by Hwang et al. (21)) and that the level of CD80 ac-quisition by T cells is related to both the level of CD80 expressionon APCs and the strength of signal 1. Moreover, our data for thefirst time suggest that acquisition of CD80 by T cells might be ofimportance for the Ag-presenting capacity of T cells. Thus, thesefindings might have important implications for a further under-standing of immune regulation and the pathogenesis of immuno-pathological diseases.

Materials and MethodsAnimals

BALB/c, C57BL/6, B10.A, and CD28KO mice were obtained from TheJackson Laboratory (Bar Harbor, ME) and housed under specific pathogen-free conditions. B7dKO mice were provided by R. Hodes (National Insti-tute on Aging, National Institutes of Health, Bethesda, MD) and wereoriginally obtained from A. Sharpe (Harvard University, Boston, MA).Pigeon cytochromec88–104-transgenic (PCC/TCR-Tg) (5CC7 TCR-Tg)mice, which are TCR transgenic for PCC, were obtained from TaconicFarms (Germantown, NY) and bred in the National Cancer Institute animalfacility under pathogen-free conditions. PCC/TCR-Tg IL-2 green fluores-cence protein (GFP)ki (PCC/I-Ek-specific) mice were provided by M. Nara-mura (National Institute of Allergy and Infectious Diseases, National In-stitutes of Health).

mAbs and flow cytometry

To evaluate CD80 expression on T cells, cell suspensions were stained withdirectly conjugated mAbs (anti-CD4PE; anti-CD80 FITC). The proportionof CD41 cells expressing CD80 was determined by gating on CD41 Tcells, excluding the dead cells. Fluorochrome mAb Abs to CD4, CD80,MHC class II I-Ek, CD28, and purified anti-CD3, anti-CD80, hamster IgG,and anti-CD28 Abs were purchased from PharMingen (San Diego, CA),and samples were analyzed with a FACSCaliber (Becton Dickinson,Mountain View, CA) using CellQuest software for the Macintosh.

APCs

DCs were prepared as described previously (22). Briefly, bone marrowcells from 6- to 8-wk-old mice were depleted of lymphocytes using amixture of magnetic beads specific for CD4, CD8, MHC class I, and MHCclass II (MiniMACS, Miltenyi Biotec, Auburn, CA). Cells were incubatedin six-well plates (53 106 cells/well) with medium supplemented with 10ng/ml GM-CSF and 10 ng/ml IL-4 (R&D Systems, Minneapolis, MN).Cells were replated in fresh cytokine-supplemented medium on days 2 and4. On day 6, cells were harvested for infection and analysis. The followingcell lines were provided by R. Germain (National Institute of Allergy andInfectious Diseases, National Institutes of Health): fibroblast cell lineDCEK, which expressed a low level of CD80; MHC class II I-Ek and 13.9fibroblast cell line, which expressed a high level of CD80; and COS (mon-key-kidney) cells, which expressed MHC class II I-Ek. Murine colonicadenocarcinoma cells (MC38) and the retrovirally transduced MC38/CD80cell line have been described previously (23). A20 cells are murine B celllymphoma cells expressing 37% CD80 on their surface.

Separation of APCs by beads

Fibroblast cells (203 106) were cultured with 250ml of goat anti-mouseDynabeads (Dynal, Lake Success, NY) in the presence of various concen-trations of peptide. After 2 days, the fibroblasts that had absorbed the beadswere separated by a magnet. These cells were then used as APCs (23 105

cells) and cultured with 13 106 effector/memory T cells. The APCs wereseparated from T cells 24 h later by a magnet. This procedure yielded apreparation of T cells devoid of any APCs.

Recombinant fowlpox viruses and infection of DCs

Recombinant fowlpox virus expressing CD80 (rF-CD80) and wild-typefowlpox (FP-WT) have been described previously (24). DCs were har-vested on day 6 of culture and washed with OptiMem (Life Technologies,Gaithersburg, MD). The cells (1.23 106/ml) were infected with FP-WT orrF-CD80 at 50 MOI (multiplicity of infection; PFU/cells) for 2 h at 37°C.Cells were first washed and then incubated in condition medium at 37°Covernight. After 18 h, DCs were harvested for in vitro assays.

Peptide

The I-Ek-restricted PCC peptide was synthesized and HPLC-purified byAmerican Peptide (Sunnyvale, CA).

Cell preparation and culture

CD41 cells were purified using microbeads conjugated to anti-CD4 mAb,according to the manufacturer’s instructions (MiniMACS, Miltenyi Bio-tec), on MACS columns. Isolated CD41 cells (5 3 105 cells/ml) werecultured with APCs (13 105 cells/ml) with or without signal 1, with eitheranti-CD3 Ab or PCC peptide (at concentrations indicated in each figurelegend), in RPMI 1640 supplemented with the following: 10% heat-inac-tivated FBS, 53 1025 M 2-ME, 2 mM L-glutamine, 100 U/ml penicillin,100 mM/ml streptomycin, and 10 mM HEPES (all obtained from LifeTechnologies, Gaithersburg, MD), and 15mg/ml gentamicin (BioWhit-taker, Walkersville, MD).

Effector/memory CD41 T cells

Effector/memory CD41 T cells were generated by activating naive splenicCD41 T cells from PCC/TCR-Tg mice in culture with 10mg/ml PCCpeptide for 3 days. Following centrifugation with Ficoll gradient, the livecells were washed and then rested for 4–6 days in six-well plates with 5IU/ml of murine IL-2 medium. The cells were washed and given IL-2 every3 days following the 3-day PCC activation. The effector/memory CD41

cells were analyzed for markers as follows: FITC-conjugated CD62 ligand,PE-conjugated CD44, FITC-conjugated CD45RB, and FITC-conjugatedCD69. Appropriate isotype control was used for all of the markers.

Apoptosis assay

Effector/memory CD41 cells were cultured for 24 h in the presence ofvarious concentrations of peptide as described inResults. Apoptosis wasassessed using the TUNEL kit (PharMingen) according to manufacturer’sinstructions.

RNA extraction and PCR

RNA was prepared using an RNA STAT-60 kit (Tel-Test, Friendswood,TX), and 2mg of total RNA was used to synthesize cDNA with the Su-perScript Preamplification System (Life Technologies, Gaithersburg, MD),using oligo(dT) according to the manufacturer’s instructions. Oligonucle-otides specific for CD80 andb-actin were prepared by Cruachem (Dulles,VA) and Invitrogen (Carlsbad, CA). Primers for CD80 (517-bp fragment)were 59-ATCCAGGATACACCACTCC-39(sense) and 59-TCCAACCAAGAGAAGCGAGG-39 (antisense). Primers forb-actin (1-kb fragment)were 59-GCTCACCATGGATGATGATATCGC-39(sense) and 59-GGAGGAGCAATGATCTTGATCTTC-39 (antisense). cDNA samples weresubjected to amplification using PCR (model 9600; Perkin-Elmer Cetus,Norwalk, CT). Samples were initially denatured for 2 min at 94°C. CD80andb-actin-specific amplification was conducted for 25 cycles, where eachcycle consisted of denaturation at 94°C for 30 s, annealing at 55°C for 30 s,and polymerization at 72° for 30 s, followed by a finishing step at 72° for10 min. Samples (20ml) were resolved on 1% agarose gels in Tris-borate-EDTA (TBE), electrophoresed at 100 V, and stained with ethidium bro-mide; the cDNA fragments were visualized directly by UV light.

GFP-CD80 fusion protein

A cDNA clone of murine CD80 was cloned and sequenced as describedpreviously (25). Restriction endonuclease digestion of this clone with Kpniand Apai released a 1-Kb fragment that was gel purified and ligated into theKpni/Apai sites of pEGFP-C1, the C-terminal protein-fusion expressionvector (Clontech Laboratories, Palo Alto, CA). DH5a Escherichia colicells were transformed with this construct (Life Technologies, Rockville,MD), and kanemycin-resistant colonies were selected and sequenced for

2506 ACQUISITION OF CD80 BY T CELLS

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

CD80 expression. pEGFP-CD80 DNA was CsCl-banded and subsequentlyused to transfect COS cells for confocal studies.

ResultsCD80 acquisition by T cells in B7dKO mice

To establish the innate expression of CD80 on naive T cells, CD41

cells from four strains of mice were analyzed by two-color FACSanalysis. T cell expression of CD80 is a relatively rare occurrencein these strains, as determined by immunofluorescence and flowcytometry. CD80 expression levels on T cells varied among thedifferent strains of mice, ranging from 2 to 9% in C57BL/6 mice,to very small (1–2%) or undetectable levels in B7dKO, CD28KO,PCC/TCR-Tg mice.

Previous studies have reported that activated T cells expressCD80 molecules on their surface 3–7 days after activation (15).However, none of these studies used B7dKO mice as the source ofT cells. CD41 naive T cells from B7dKO mice were used to in-vestigate whether CD80 expression on T cells shortly after acti-vation is induced endogenously or acquired from APCs. Spleno-cytes from B7dKO mice lack functional genes for both CD80 andCD86; upon activation, these cells are unable to express CD80 orCD86 on their surface (26). CD41 cells from B7dKO mice wereincubated with 10mg/ml of anti-CD3 as signal 1 and A20 cells (aB cell line expressing 37% CD80) as APCs. The T cells weremonitored for 24 h. As shown in Fig. 1A, CD41 cells from B7dKOmice did not express CD80; levels were similar to isotype control.Activation of B7dKO CD41 cells with 10mg of anti-CD3, in theabsence of APCs (A20 cells), did not change CD80 expression(Fig. 1B). As a result of incubating B7dKO CD41 cells with A20cells in the absence of signal 1, 4% of CD41 cells acquired CD80(Fig. 1C). However, when CD41 cells were incubated with A20cells as APCs in the presence of anti-CD3 (10mg/ml) as signal 1,30% of the CD41 cells from the B7dKO mice acquired CD80 (Fig.1D).

To further address the hypothesis that CD80 is acquired by Tcells upon activation, CD41 cells from B7dKO mice were thenincubated for 24 h with either normal DCs or DCs that were in-fected with WT-FP virus or rF-CD80. Normal DCs expressed highlevels of CD80 (70%), and infection with WT virus did not affectCD80 expression. However, infection of DCs with rF-CD80 in-creased expression of CD80 on 90% of cells. In the absence of

signal 1, incubation of T cells from B7dKO mice with both regularDCs and those infected with WT-FP led to the acquisition of rel-atively low levels of CD80 (7–15%) (Table I). However, in thepresence of signal 1 (10mg/ml anti-CD3), 55–57% of CD41 cellsacquired CD80 on their surface (Table I). When B7dKO CD41

cells were incubated with rF-CD80-infected DCs, the CD41 cellsacquired more CD80 in both the absence (53%) and presence(87%) of signal 1 (Table I). These results using T cells fromB7dKO mice further demonstrate that CD80 levels on T cells aftera short period of activation are due to the acquisition of CD80 fromAPCs, and not the constitutive expression and/or up-regulation ofendogenous CD80. Moreover, these results demonstrate that boththe presence of signal 1 and the level of CD80 expression on APCsinfluence this phenomenon.

CD80 is acquired by T cells through CD28 interaction

CD80 acquisition by T cells raises the question of the role of CD28in this process. To examine the role of the CD28 molecule inacquiring CD80, CD41 cells from CD28KO (CD282/2) mice andnormal C57BL/6 mice were incubated for 24 h with MC38/CD80cells; MC38/CD80 are retrovirally transduced murine carcinomacells expressing CD80. These experiments were conducted eitherin the presence or absence of anti-CD3 (10mg/ml) as signal 1. Asshown in Table II, T cells from CD28KO mice did not acquireCD80 upon incubation with MC38/CD80 in the absence or pres-ence of signal 1 (1–2% CD80 acquisition). However, 19% ofCD41 cells from C57BL/6 mice (expressing constitutively highlevels of CD28) did acquire CD80 upon stimulation with MC38/CD80 in the presence of signal 1. These studies demonstrate thatthe acquisition of CD80 by T cells is mediated through the CD28ligand and not by way of a random process. Experiments were alsoconducted to show that the ligand on resting T cells for CD80acquisition is not CTLA-4. T cells from PCC/TCR-Tg mice wereincubated with fibroblasts expressing CD80 in the presence andabsence of anti-CTLA-4 Ab. In both cases, there was no reductionin the acquisition of CD80 by T cells (data not shown).

Physical nature of CD80 transfer to T cells

To investigate the physical nature of the CD80 binding to the sur-face of T cells, COS (monkey-kidney) cells expressing I-Ek MHCclass II were transfected with a construct containing GFP-CD80.As determined by FACS analysis, 30% of COS cells expressedmembrane-bound GFP-CD80 fusion protein (data not shown).Confocal microscopy studies of COS cells demonstrated that whenthese cells were transfected with GFP-CD80 construct, CD80 was

FIGURE 1. Acquisition of CD80 from the A20 cell line by B7dKO Tcells. Two-color FACS analysis was performed on T cells for CD4 andCD80. A, Expression of CD80 on CD41 T cells from B7dKO mice;B,CD80 expression on CD41 cells upon stimulation with 10mg/ml of anti-CD3 Ab for 24 h;C, acquisition of CD80 by B7dKO CD41 cells in thepresence of the A20 cell line as APCs, but in the absence of signal 1; orD,acquisition of CD80 by B7dKO CD41 cells in the presence of A20 cellsand anti-CD3 as signal 1. The data are representative of two independentexperiments. All gates were set based on negative controls.

Table I. CD80 acquisition by CD41 T cells from double knockoutmicea

APCSignal 1

(anti-CD3mg/ml) % CD80

None 0 1.0None 10 1.1DC 0 7.5

10 55DC (FP-WT) 0 15

10 57DC (rF-CD80) 0 53

10 87

a CD41 T cells from B7dKO mice were purified using magnetic beads. CD41

cells (53 105/ml) were cultured with DCs (13 105/ml) or DCs infected with eitherFP-WT or rF-CD80 (as described inMaterials and Methods) in the absence or pres-ence of anti-CD3 as signal 1. Two-color FACS analysis for CD80 and CD41 cellswas performed 24 h later. The data are representative of two independentexperiments.

2507The Journal of Immunology

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

expressed on cell surfaces (Fig. 2A). To further demonstrate cellsurface expression of CD80, these cells were double-stained withPE-anti-CD80. Superimposition of red PE on GFP led to the ap-pearance of a yellow/orange color on cell membranes (Fig. 2,

B–D). When PCC/TCR-Tg CD41 cells were incubated with trans-fected COS cells in the presence of PCC peptide (10mg/ml) assignal 1, T cells were shown to acquire CD80 molecules (Fig. 2,E–H). These results further demonstrate the physical acquisition ofCD80 by T cells.

Effect of cyclohexamide (CHX) on CD80 acquisition

Recent studies have demonstrated the existence of a new homologof the B7 family of molecules (27). This observation raises thepossibility that there might be a CD80 homolog that can be endo-genously up-regulated rapidly on T cells upon stimulation, and thatthe molecule might interact with the anti-CD80 Ab. To excludethis scenario and further rule out the de novo generation of CD80in short-term activated T cells, CD41 cells from PCC/TCR-Tgmice were treated with the protein synthesis inhibitor CHX (20mg/ml) for 2 h. After CHX treatment, CD41 cells were incubatedwith irradiated fibroblast cells expressing high levels of CD80 andI-Ek MHC class II molecules for 4 h and for 24 h. There were nosignificant differences in CD80 acquisition between the CHX-treated cells and the untreated CD41 cells at either 4 h (22% forCHX-treated and 30% for untreated cells) or 24 h (99% acquisitionof CD80 for both treated and untreated cells). These results furtherindicate that CD80 expression on T cells upon activation is indeeddue to acquisition, rather than up-regulation, of CD80 or any otherhomolog of the B7 family.

Analysis of naive and activated CD41 T cells for CD80 mRNAby PCR

To further demonstrate that CD80 is acquired by T cells uponactivation, and not by the endogenous up-regulation of CD80mRNA, PCR studies were performed (Fig. 3). Purified CD41 cellsof PCC/TCR-Tg mice were incubated with fibroblast cells express-ing high levels of CD80 (Fig. 4C) and 10mg/ml of PCC peptide assignal 1. After 24 h of incubation, CD41 cells were carefully re-moved from supernatant fluid and cultured two additional times(each time recovering cells in the supernatant) to ensure that nofibroblast (an APC that expresses CD80) was contaminating the Tcell population. Purity of CD41 cells, as well as the expression ofCD80, was checked by FACS analysis. The cells were 100%CD41, and 79–80% of these cells had acquired CD80 on theirsurface (Fig. 4D). Simultaneously, as controls, naive PCC/TCR-Tg

Table II. Analysis of CD80 acquisition by T cells from CD281 vsCD28 knockout micea

APCSignal 1

(anti-CD3mg/ml)

Source of CD41 T Cells (%)

C57BL/6 CD28KO

None 0.0 8.0 1.8MC38/CD80 0.0 8.0 1.4

10.0 19.0 0.8

a CD41 T cells were prepared from C57BL/6 mice or CD28KO mice using mag-netic beads. CD41 cells (53 105/ml) were cultured with 13 105/ml MC38 carci-noma cells that had been transduced with CD80 (MC38/CD80) in the absence orpresence of anti-CD3 for 24 h and analyzed with two-color FACS analysis for CD4and CD80 expression. The data are representative of two independent experiments.

FIGURE 2. Physical transfer of CD80 from COS cells (transfected withGFP-CD80 fusion protein) to T cells. Binding and transfer of CD80 fromCOS cells to CD41 cells were visualized by fluorescence and confocalmicroscopy. Transfection of COS cells with GFP-CD80 vector led to cellsurface expression (green) of the CD80 molecule (A) and double-stainingfor CD80 (using PE CD80 Ab) (B–D), demonstrating that CD80 is mem-brane-bound on the COS cell surface. Incubation of GFP-CD80 COS cellswith PCC/TCR-Tg CD41 cells in the presence of PCC peptide (10mg/ml)led to physical acquisition of CD80 from COS cells (E and F).G andH,Higher magnifications ofE andF, respectively.

FIGURE 3. Analysis of CD41 cells for CD80 mRNA. A 517-bp frag-ment of CD80 cDNA was amplified by RT-PCR, as described inMaterialsand Methods.Lane A, DNA-molecular size standard;lane B, fibroblasttransfected with plasmid expressing CD80 (positive control);lanes CandD, two different preparations from naive (unstimulated) PCC/TCR-TgCD41 cells; lane E, PCC/TCR-Tg CD41 cells stimulated for 24 h in thepresence of APCs (fibroblasts expressing CD80) and 10mg/ml of PCCpeptide that had acquired CD80 (see Fig. 4D). These cells were depleted ofAPCs as described inMaterials and Methods.

2508 ACQUISITION OF CD80 BY T CELLS

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

CD41 cells were also purified from PCC/TCR-Tg animals andwere checked for CD80 expression by FACS. These cells had verylittle or no CD80 on their surface. Using RT-PCR, we then ana-lyzed the expression of CD80 mRNA on unstimulated PCC/TCR-Tg CD41 cells or on CD41 cells that were stimulated for24 h. As shown in Fig. 3, a 517-bp CD80 fragment could readilybe amplified from the APC fibroblast cell line that expresses CD80(positive control,lane B). However, CD80 could not be amplifiedfrom two different preparations of unstimulated CD41 cells (Fig.3, lanes CandD) or CD41 cells that were stimulated for 24 h (Fig.3, lane E).b-actin expression demonstrated that equal amounts ofRNA were added in each PCR. To establish the sensitivity of de-tection of CD80 in PCR, a template dilution experiment was per-formed on a CD80-transfected fibroblast cell line. cDNA encodingCD80 was synthesized using 2mg of RNA from transfected fi-broblasts; serial dilutions of cDNA (1:10- to 1:320-fold dilutions)were used as a template for PCR amplification. Analysis of theproduct on an agarose gel revealed that the DNA bands could notbe detected when the template was diluted.1:160-fold, which is#1 ng of RNA. Therefore, we can conclude that the T cells thathave been activated and acquired CD80 express#1 ng of CD80RNA. These results illustrate that upon activation of CD41 cells,there is no up-regulation of detectable levels of mRNA for CD80at the 24-h time point; a 24-h time point was used in the abovestudies to detect the acquisition of CD80.

Acquisition of CD80 by T cells is proportional to CD80expression on APCs

Studies were undertaken to investigate whether increased densityof CD80 on APCs might further stabilize the interaction and, there-fore, lead to more acquisition of CD80 by CD28 on T cells. Toaddress this question, purified CD41 cells of PCC/TCR-Tg micewere incubated for 24 h with fibroblast cells expressing either rel-atively low levels of CD80 (Fig. 4A) or high levels of CD80 (Fig.4C) and PCC peptide (10mg/ml) as signal 1. CD80 acquisition onT cells was proportional to CD80 expression by APCs. AlthoughCD41 cells stimulated with fibroblast cells expressing a low levelof CD80 acquired much less CD80 on their surface (26%, Fig. 4B),79% of CD41 cells that were incubated with fibroblast cells ex-pressing a high level of CD80 acquired CD80 (Fig. 4D).

To further determine whether this level of CD80 on APCs in-fluenced CD80 acquisition by T cells, purified CD41 cells fromPCC/TCR-Tg mice were incubated in the presence of PCC peptide(10 mg/ml) with either uninfected or rF-CD80-infected DCs. Nor-mal DCs express very high levels of CD80 (75%); DCs infectedwith rF-CD80 showed 90% CD80 expression. Incubation of CD41

cells with uninfected DCs led to 35% of CD41 cells acquiringCD80 on their surface. Moreover, CD41 cells incubated with rF-CD80 DCs obtained markedly higher levels of CD80 (81%) ontheir surface.

Studies were then conducted to investigate T cell selectivity inthe acquisition of CD80 molecules; toward this goal, the fate ofanother surface marker, CD40, on APCs was investigated in termsof acquisition by T cells. CD41 cells were incubated with eithernormal DCs or DCs infected with FP-WT or rF-CD80 in the pres-ence or absence of various concentrations of PCC peptide as signal1. FACS analysis was performed on gated T cells, based on sizeand cell surface markers for T cells. An average of 30–40% ofDCs express CD40 on their surface, whereas only 3–4% of T cellsexpress CD40. Incubation of CD41 cells with APCs in the pres-ence of different concentrations of signal 1 PCC peptide led to littleor no increase in the acquisition of CD40 by CD41 cells, whichremained 3–4% positive for CD40. These results further demon-strate that CD80 expression on T cells is not due to contaminatingAPCs, and that CD41 cells show selectivity in the acquisition ofmolecules from APCs.

To further investigate the selectivity in CD80 acquisition, naiveCD41 cells from PCC/TCR-Tg mice were incubated with a fibro-blast cell line expressing high levels of CD80 and PCC peptide (10mg/ml) as signal 1 for 24 h (as described previously) in the pres-ence of CD81 cells from non-Tg mice (B10.A mice). As shown inFig. 4E, non-Tg CD81 cells acquired relatively little CD80 (18%)above their 6–7% normal level of expression, whereas PCC/TCR-Tg CD41 cells acquired high levels of CD80 (77%, Fig. 4F).These results further demonstrate that CD80 acquisition by T cellsis specific and not due to random acquisition or incorporation.

CD80 acquisition by T cells is directly related to the strength ofsignal 1

Naive CD41 cells from PCC/TCR-Tg mice were incubated withfibroblast cells expressing low and high levels of CD80 in thepresence of various concentrations of PCC peptide to investigatethe effect of the concentration of signal 1 on T cell acquisition ofCD80. Acquisition of CD80 by CD41 cells upon stimulation withfibroblast cells expressing a low level of CD80 was shown at thehigh level of signal 1 (Fig. 5A). When PCC/TCR-Tg CD41 cellswere stimulated with fibroblast cells expressing high levels ofCD80, a direct relationship between the strength of signal 1 and theacquisition of CD80 by T cells could be more clearly defined (Fig.

FIGURE 4. CD80 acquisition by T cells is related to the level of CD80expression on APCs. When PCC/TCR-Tg CD41 cells (53 105 cells/ml)were incubated with a fibroblast cell line with lower expression levels ofCD80 (13 105/ml) in the presence of 10mg/ml of PCC peptide for 24 h(A), T cells acquired relatively lower levels of CD80 (B). Incubation ofthese cells with a fibroblast cell line expressing higher levels of CD80 (C)in the presence of 10mg/ml PCC peptide led to higher acquisition of CD80by T cells (D). CD41 cells were double-stained with PE anti-CD4 Ab andFITC anti-CD80 Ab. These experiments were repeated three times. Gateswere set based on negative controls.E and F, Selective acquisition ofCD80 by T cells. Naive PCC/TCR-Tg CD41 (5 3 105) cells and non-transgenic, non-PCC-restricted B10.A CD81 (MHC I-Ek-restricted) cells(5 3 105/ml) were incubated with an irradiated 13.9 fibroblast cell lineexpressing high levels of CD80 (13 105 cells/ml) and 10mg of PCCpeptide in 48-well plates for 24 h.E, CD80 acquisition by B10.A CD81

cells after 24 h.F, CD80 acquisition by PCC/TCR-Tg CD41 cells after24 h. Acquisition of CD80 by T cells was monitored by three-color FACSanalysis. The data are representative of two different experiments. All gateswere set based on negative controls.

2509The Journal of Immunology

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

5B). These results further demonstrate that the acquisition of CD80by T cells can be influenced by the level of signal 1.

Differential acquisition of CD80 by naive and effector/memoryCD41 T cells

To investigate the biological consequence of CD80 acquisition byeffector/memory T cells, effector/memory CD41 cells were gen-erated from naive CD41 cells, as described above, and then cul-tured with fibroblasts expressing high levels of CD80 as APCs inthe presence of various concentrations of PCC peptide for 24 h.Before incubation with APCs, 8–9% of effector T cells were pos-itive for CD80 expression, as compared with 1–2% of naive T cells(data not shown). When effector/memory CD41 cells were cul-tured with fibroblasts expressing high levels of CD80 as APCs,;40% of the cells acquired CD80. This is in contrast to the rela-tively low percent CD80 acquisition by naive CD41 cells when

cultured with the same APCs expressing high levels of CD80 (Fig.6). Based on our observations, the probable reason for this differ-ence between effector/memory and naive T cell acquisition is thateffector/memory T cells express higher levels of CD28, the CD80ligand, than naive T cells. Indeed, the effector/memory CD41 cellsfrom PCC/TCR-Tg mice used here expressed much higher levelsof CD28 than naive T cells from these mice (70% for memory vs32% for naive). When both naive and effector/memory CD41 cellswere cultured with APCs (expressing high amounts of CD80) andvarious concentrations of peptide, the effector/memory cells al-ways acquired more CD80 at a given peptide concentration (Fig.6). At a peptide concentration of 0.1mg/ml, ;90% of naive T cellsacquired CD80; no cell death was observed in these T cells.

Acquisition of CD80 by memory cells results in apoptosis

To determine whether the cell death in effector/memory T cellpopulations (as shown in Fig. 6) was due to apoptosis, effector/memory CD41 T cells were generated by activating naive cellswith PCC peptide for 3 days and then resting them with IL-2 for4 days. These effector/memory cells were then divided into twogroups. Cells in the first group were cultured alone with variousconcentrations of PCC peptide for 24 h. As shown in Fig. 7A–C,these cells demonstrated minimal apoptosis when cultured with orwithout peptide. The second group consisted of effector/memory Tcells that were cultured with APCs (fibroblasts expressing highamounts of CD80) in the presence of 0.001mg/ml of PCC peptidefor 24 h. Approximately 80% of these T cells acquired CD80 (asshown in Fig. 6), and the T cells were then separated from thefibroblast APCs by magnetic beads and cultured further in thepresence of various concentrations of PCC peptide (i.e., 0, 0.001,or 0.1 mg/ml) for 24 h. Upon acquisition of CD80, 34% of theseeffector/memory T cells underwent spontaneous apoptosis (Fig.7D) in the absence of exogenous peptide. Although the addition of

FIGURE 5. Influence of signal 1 on acquisition of CD80 by T cells.Purified PCC/TCR-Tg CD41 cells (53 105/ml) were incubated with 13105/ml irradiated fibroblast APCs expressing low CD80 (A) or high levelsof CD80 in the presence of different amounts of PCC peptide (B) for 1 h(open column), 4 h (hatched column), and 24 h (filled column). The dataare representative of three independent experiments.

FIGURE 6. Differential acquisition of CD80 by naive and effector/memory CD41 T cells. Either naive (f) or effector/memory CD41 T cells(L) (5 3 105) were cultured with 13 105 fibroblasts expressing highlevels of CD80 as APCs with varying amounts of PCC peptide (0–0.1mg/ml) as signal 1 in 48-well plates for 24 h. CD41 cells were carefullyremoved without disturbing the adherent fibroblast layer, and were ana-lyzed for the presence of surface CD4 and CD80 by two-color FACS anal-ysis.p, Denotes that a high number of effector/memory T cells were deadwhen analyzed by FACS analysis and trypan blue assay (as shown in Fig.7). No cell death was seen in naive CD41 T cells. The data represent fourdifferent experiments.

2510 ACQUISITION OF CD80 BY T CELLS

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

0.001mg/ml PCC peptide had minimal effect (Fig. 7E), the addi-tion of 0.1 mg/ml of PCC clearly augmented the occurrence ofapoptosis of CD41 T cells (Fig. 7F). To further ascertain that theacquisition of CD80 caused the effector/memory cells apoptoticdeath, CD80 Ab was included in the assay, with an even higherconcentration (1mg/ml) of peptide as signal 1. As seen in Fig. 8Aand B, isotype control Ab showed no inhibitory effects on theapoptotic death of these CD41 T cells, whereas effector/memory

CD41 T cells treated with anti-CD80 Ab demonstrated a mean-ingful decrease in their apoptosis (46%, Fig. 8C). Based on theseresults, we propose that CD80 acquisition can in part lead toapoptosis of effector/memory cells and the level of apoptosis isenhanced with increased levels of signal 1.

CD80 acquisition can lead to Ag presentation by T cells

Studies were then conducted to determine the biologic conse-quence that results when naive T cells acquire CD80. We askedspecifically whether T cells that have acquired CD80 can nowfunction as APCs to naive CD41 T cells. To this end, IL-2 pro-duction was measured as an indicator of T cell activation. How-ever, one cannot simply measure IL-2 levels in the supernatant,because IL-2 can be produced not only from T cells functioning asAPCs but also from T cells receiving the costimulatory signal, ifany. We took advantage of an IL-2-GFPki mouse (28) in which theIL-2 exon from one of the IL-2 alleles was replaced by GFP. Inthese mice, IL-2 production can be monitored by cell-associated(of the producer cell) fluorescence from GFP expression driven bythe endogenous IL-2 promoter.

PCC/TCR-Tg CD41 cells that have acquired CD80 were pre-pared in the manner described above. Fibroblasts were used asAPCs of choice to minimize the possibility of contamination ofCD41 cells that had acquired CD80 from professional APCs be-cause they can easily be separated from T cells by magnetic beadseparation. Isolated CD41 cells (2–3% expressing CD80 sponta-neously) were irradiated and used as control APCs. In the absenceof PCC peptide, these control APCs were unable to activate theresponding T cells (Fig. 9A); in the presence of PCC peptide (10mg/ml), they could activate 16% of PCC/TCR-Tg/IL-2-GFPki

FIGURE 7. Analysis of apoptosis in effector/memory CD41 T cellsbefore and after acquisition of CD80. Naive CD41 cells from PCC/TCR-Tg mice were cultured with 1mg/ml of PCC peptide for 3 days andthen rested with IL-2 (5mg/ml) for 4 days to generate the effector/memoryT cells. These cells were divided into two groups. Cells in the first groupwere cultured overnight with various concentrations of peptide (i.e., 0,0.001, or 0.1mg/ml) (A–C, respectively). Cells in the second group werefirst cultured with fibroblasts (which express high levels of CD80) for 24 hto acquire CD80 and were then separated from APCs as described inMa-terials and Methods. The T cells were then cultured alone in the presenceof various concentrations of peptide for 24 h (0, 0.001, or 0.1mg/ml) (D–F,respectively). All of the above cells (in groups I and II) were analyzed forapoptosis using the TUNEL assay. Percent apoptosis is given in each panel.The data represent three different experiments.

FIGURE 8. CD80 acquisition can lead to apoptosis of effector/memoryCD41 T cells. Cells were cultured with fibroblasts for 24 h to acquireCD80 and were then separated from APCs as described inMaterials andMethods. The CD41 cells were then treated with either no Ab (A), isotypecontrol Ab (hamster IgG) (B), or anti-CD80 Ab (10mg/ml) (C) for 30 minand replated with 1mg/ml of PCC peptide for 24 h. Cells were analyzed forapoptosis using the TUNEL assay. Percent apoptosis is given in each panel.

FIGURE 9. T cells that have acquired CD80 can themselves becomeAPCs. Naive CD41 cells from PCC/TCR-Tg mice were purified, irradiated(as APCs), and cultured with CD41 cells from PCC/TCR-Tg/IL-2-GFPki

mice as effector cells for 24 h in the absence of PCC peptide (A) or pres-ence of 10mg/ml of PCC peptide (B). Furthermore, PCC/TCR-Tg CD41

cells that had acquired CD80 from a fibroblast cell line expressing highlevels of CD80 were purified, irradiated, and used as APCs in culture withPCC/TCR-Tg/IL-2-GFPki CD41 cells as effectors.C andD, Activation ofthe effector cells (PCC/TCR-Tg/IL-2-GFPki) in the absence of peptide (C)and in the presence of 10mg/ml of PCC peptide (D). Moreover, PCC/TCR-Tg CD41 cells that acquired CD80 were treated with either 10mg/mlof isotype-control Ab (E) or purified anti-CD80 Ab (F) for 30 min on ice,irradiated, and then incubated with PCC/TCR-Tg/IL-2-GFPki CD41 cellsin the presence of peptide (10mg/ml) for 24 h. Two-color FACS analysiswas performed to study the presence of activated PCC/TCR-Tg/IL-2-GFPki

CD41 cells. The data represent two different experiments.

2511The Journal of Immunology

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

CD41 cells, possibly due to the spontaneous expression of CD80(Fig. 9B). In contrast, CD41 cells that had acquired CD80 fromfibroblasts were much more potent as stimulators of PCC/TCR-Tg/IL-2-GFPki cells to produce IL-2 at 24 h in the absence (Fig.9C) or presence (Fig. 9D) of peptide. The possible reason for thiswill be discussed below. To further demonstrate that CD80 acqui-sition allows naive CD41 cells to become APCs, anti-CD80 Abwas included in the assay. As seen in Fig. 9F, APCs treated withanti-CD80 Ab led to a marked decrease of GFP-IL-2 CD41 cellsbecoming activated (Fig. 9F), when compared with APCs treatedwith control Ab (Fig. 9E).

DiscussionIn this report, we demonstrate for the first time the acquisition ofthe CD80 costimulatory molecule by naive CD41 T cells usingsyngeneic APCs. Several studies were conducted to rule out theendogenous expression of CD80 by T cells. When T cells fromB7dKO mice were incubated with either a B cell line or DCs in thepresence of signal 1, they acquired CD80 on their cell surface (Fig.1; Table I). Because the CD80 and CD86 genes are deleted inB7dKO mice, neither one of these molecules can be expressed(26). In these studies, we also used normal DCs and a B cell lineas APCs to eliminate the possibility that CD80 acquisition mightbe due to some abnormal phenomenon on vector-transfected APCmembranes. The PCR studies (Fig. 3) and the studies with CHX-treated T cells further demonstrated that the expression of CD80on T cells is not de novo. In view of the recent discovery of B7h(a new homolog of CD80 and CD86), these results also ruled outthe possibility of interaction between anti-CD80 Ab and this hom-olog. The confocal microscopy studies using GFP/CD80 fusionprotein further strengthened the notion that CD80 molecules areactually acquired upon T cell/APC interaction (Fig. 2). Moreover,we demonstrated here that the level of acquisition by T cells isrelated to the strength of signal 1 (Fig. 5).

It has been demonstrated that shortly after T cell/APC interac-tion, immunological synapses or supramolecular activation clus-ters form at the contact site (29). Formation of these immunolog-ical synapses is associated not only with TCRs and MHCmolecules, but also with various costimulatory molecules such asLFA/ICAM, CD2/CD48, and CD28/B7 receptor ligands (30–32).Wade et al. (33) have demonstrated that truncation of the intracy-toplasmic tail of MHC molecules enhances lateral movement ofthese complexes but decreases immunogenicity. Presentation ofMHC/peptide complexes in a cross-linked form can lead to overtT cell activation (33). It has been suggested that the accumulationand interaction of costimulatory molecules with their receptorsand, perhaps, cytoskeletal proteins may promote the stabilizationof TCR/MHC interactions. In addition, costimulatory molecules atthe early stages of T cell activation may function by enhancingTCR cross-linking, therefore intensifying signal 1. In light of theabove-mentioned studies, it is possible that increased CD80 ex-pression on APCs, along with increased signal 1 in our models,may stabilize TCR/MHC interactions, contribute to the overallavidity of T cell/APC interactions, and thus lead to the increasedtransfer of CD80 to T cells. The data reported here on CD80 ac-quisition by CD282/2 cells support the hypothesis by Hwang et al.(21) that a high enough affinity for adhesion between CD28 andCD80 may actually result in the transfer of CD80 molecules fromAPCs to T cells.

Previous studies have shown that varying the strength of co-stimulation dramatically reduces the Ag dose required for T cellactivation (34). Our observations indicate that increasing the den-sity of CD80 on APCs may facilitate a strong interaction withCD28, which would lead to stabilization of the T cell/APC inter-

action and, therefore, acquisition of higher levels of costimulatorymolecules in the presence of less signal 1.

Several groups have previously reported the expression of CD80on T cell clones as well as on T cells from patients with autoim-mune diseases or HIV infection (6–13). Interestingly, there hasbeen no study to address the up-regulation of CD80 message in Tcells. T cells in these studies have always been in the presence ofAPCs either in vitro or in vivo; therefore, it is reasonable to inferthat reported CD80 expression on T cells in these studies mighthave actually been due to acquisition, rather than up-regulation ofexpression, of CD80. The studies reported here demonstrate that atthe time of acquisition of CD80 by CD41 cells (24 h), no CD80mRNA could be detected by PCR (Fig. 3).

Although two different biological consequences of CD80 acqui-sition by T cells are demonstrated here, this phenomenon will mostlikely be the subject of numerous investigations in the future. Dur-ing an initial encounter with Ag, naive Ag-specific lymphocytesproliferate and differentiate to become activated effector/memoryT cells. The current dogma states that most of these activated ef-fector T cells die after a brief life span (35–37). Mechanisms un-derlying the regulated expansion of effector/memory T cells arenot well understood. However, two of the findings reported heremay indicate negative regulatory consequences upon acquisition ofCD80 by memory cells: 1) effector/memory T cells can acquiresignificantly higher levels of CD80 in the presence of low levels ofsignal 1; and 2) these cells undergo apoptosis upon acquisition ofCD80 and increased levels of signal 1. The apoptosis observed inthe results of Figs. 7 and 8 is most likely not due to a consequenceof the way cells were handled because the apoptosis was depen-dent on the amount of peptide present (including no apoptosis withno peptide), and all cells were handled in the same way. Also, itshould be mentioned that our data do not exclude the possibilitythat effector/memory CD41 cells could also acquire the MHC/peptide complex; therefore, these MHC complexes might play arole in the apoptosis of these cells. It has been demonstrated thatactivated T cells can process and present Ags to other T cells invitro (38). The studies reported here using CD41 cells from PCC/TCR-Tg mice that have acquired CD80 show not only that thesecells expressed MHC class II molecules, but also that they wereable to present soluble PCC peptide to T cells from PCC/TCR-Tg/IL-2-GFPki mice more efficiently than CD41 cells from PCC/TCR-Tg mice that had not acquired CD80; this led to activationand production of IL-2. Interestingly, the CD41 cells that had ac-quired CD80 were able to activate the PCC/TCR-Tg/IL-2-GFPki

CD41 cells without adding additional peptide. One possible ex-planation of this is that, because these CD41 cells had been incu-bated with fibroblast cells in the presence of peptide and had up-regulated their MHC class II molecules (data not shown), these Tcells actually also acquired peptide-MHC and were able to activatePCC/TCR-Tg/IL-2-GFPki CD41 cells in the absence of the addi-tion of exogenous PCC peptide. These results suggest that acti-vated T cells that have acquired CD80 molecules have all the char-acteristics of professional APCs. However, these functions are notexpressed constitutively; instead, they occur only after T cell ac-tivation, when class II synthesis is up-regulated and T cells haveacquired the CD80 ligand. Studies are now in progress investigat-ing whether the APC function of CD41 cells that had acquiredCD80 as reported here is dependent on other acquired moleculessuch as MHC/peptide complexes or other costimulatory molecules.However, the data presented here provide preliminary evidencethat upon acquiring CD80, T cells can more efficiently activatebystander T cells.

There are at least two possible reasons why T cells becomeAPCs in vivo. One potential advantage is that activated T cells

2512 ACQUISITION OF CD80 BY T CELLS

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

would deliver both signals 1 and 2 to other T cells, thus amplifyingan immune response; the other is that T cells could deliver onlysignal 2 to T cells that have received signal 1 from nonprofessionalAPCs (38). Consequently, although the initiation of the T cell re-sponse is dependent on professional APCs, it is possible that Tcells may costimulate each other by acquiring CD80 under certainconditions (e.g., crowding of many T cells around the same DC).This would allow the APCs to interact further with other T cells.This mechanism of “T cell to T cell” costimulation could facilitatethe regulation of the immune response.

In conclusion, the studies reported here demonstrate for the firsttime that naive CD41 cells acquire CD80 upon activation by syn-geneic APCs. This acquisition was shown to be directly related toboth the strength of signal 1 and the level of signal 2 on APCs.Moreover, we demonstrate that CD80 acquisition is CD28 medi-ated. The specificity of the process of CD80 acquisition was dem-onstrated using both CD80 and CD28KO mice. Moreover, the ac-quisition of high levels of CD80 by effector/memory CD41 cellsand the resulting increased apoptosis of these cells together indi-cate a negative regulatory function. In contrast, when naive T cellsacquire CD80, these cells themselves can act as more efficientAPCs. Therefore, CD80 acquisition by T cells might play an im-portant role in activating and/or regulating the immune response.The specific consequences of CD80 acquisition by T cells in theregulation of the immune response to pathogens, in anti-tumorresponses, and in diseases of the immune system will require fur-ther investigation.

AcknowledgmentsWe thank Earl Smith for technical assistance, and Debra Jacobs and NicoleRyder for editorial assistance.

References1. June, C. H., J. A. Bluestone, L. M. Nadler, and C. B. Thompson. 1994. The B7

and CD28 receptor families.Immunol. Today 15:321.2. Linsley, P. S., E. A. Clark, and J. A. Ledbetter. 1990. T-cell antigen CD28

mediates adhesion with B cells by interacting with activation antigen B7/BB-1.Proc. Natl. Acad. Sci. USA 87:5031.

3. Freeman, G. J., J. G. Gribben, V. A. Boussiotis, J. W. Ng, V. A. Restivo, Jr., L.A. Lombard, G. S. Gray, and L. M. Nadler. 1993. Cloning of B7-2: a CTLA-4counter-receptor that costimulates human T cell proliferation.Science 262:909.

4. Hathcock, K. S., G. Laszlo, H. B. Dickler, J. Bradshaw, P. Linsley, and R. J.Hodes. 1993. Identification of an alternative CTLA-4 ligand costimulatory for Tcell activation.Science 262:905.

5. Gimmi, C. D., G. L. Freeman, J. G. Gribben, G. Gray, and L. M. Nadler. 1993.Human T cell clonal anergy is induced by antigen presentation in the absence ofB7 costimulation.Proc. Natl. Acad. Sci. USA 90:6586.

6. Sansom, D. M., and N. D. Hall. 1993. B7/BB.1, the ligand for CD28, is expressedon repeatedly activated human T cells in vitro.Eur. J. Immunol. 23:295.

7. Azuma, M., H. Yssel, J. H. Phillips, H. Spits, and L. L. Lanier. 1993. Functionalexpression of B7/BB.1 on activated T lymphocytes.J. Exp. Med. 177:845.

8. Wyss-Coray, T., D. Mauri-Hellweg, K. Baumann, F. Bettens, R. Grunow, and W.J. Pichler. 1993. The B7 adhesion molecule is expressed on activated human Tcells: functional involvement in T-T cell interactions.Eur. J. Immunol. 23:2175.

9. Nickoloff, B. J., R. S. Mitra, K. Lee, L. Turka, J. Green, C. Thompson, and Y.Shimizu. 1993. Discordant expression of CD28 ligands, BB.1 and B7 on kera-tinocytes in vitro, and psoriatic cells in vivo.Am. J. Pathol. 142:1029.

10. Pichler, W. G., and T. Wyss-Coray. 1994. T cells as antigen-presenting cells.Immunol. Today 15:312.

11. Verwilghen, J., R. Lovis, M. De Boer, P. S. Linsley, G. K. Haines, A. E. Koch,and R. M. Pope. 1994. Expression of functional B7 and CTLA4 on rheumatoidsynovial T cells.J. Immunol. 153:1378.

12. Kochli, C., T. Wendland, K. Frutig, R. Grunow, S. Merlin, and W. J. Pichler.1999. CD80 and CD86 costimulatory molecules on circulating T cells of HIV-infected individuals.Immunol. Lett. 65:197.

13. Wolthers, K. C., S. A. Otto, S. M. A. Lens, D. N. Kohlbach, R. A. W. van Lier,F. Miedema, and L. Meyaard. 1996. Increased expression of CD80, CD86 andCD70 on T cells from HIV-infected individuals upon activation in vitro: regu-lation by CD41 T cells.Eur. J. Immunol. 26:1700.

14. Takasaki, Y., K. Abe, Y. Tokano, and H. Hashimoto. 1999. The expression ofLFA-1, ICAM-1, CD80 and CD86 molecules in lupus patients: implications forimmunotherapy.Intern. Med. 38:175.

15. Prabhu Das, M. R., S. S. Zamvil, F. Borriello, H. L. Weiner, A. H. Sharpe, andV. K. Kuchroo. 1995. Reciprocal expression of costimulatory molecules B7.1 andB7.2 on murine T cells following activation.Eur. J. Immunol. 25:207.

16. Weintraub, J. P., R. A. Eisenberg, and P. L. Cohen. 1997. Upregulation of Fasand costimulatory molecules B7-1 and B7-2 on peripheral lymphocytes in auto-immune B6/GLD mice.J. Immunol. 159:4117.

17. Weintraub, J. P., and P. L. Cohen. 1999. Ectopic expression of B7-1 (CD80) onT lymphocytes in autoimmune lpr and gld mice.Clin. Immunol. 91:302.

18. Lorber, M. I., M. R. Loken, A. M. Stall, and F. W. Fitch. 1982. I-A antigens oncloned alloreactive murine T lymphocytes are acquired passively.J. Immunol.128:2798.

19. Huang, J., Y. Yang, H. Sepulveda, W. Shi, I. Hwang, P. A. Peterson, M. R.Jackson, J. Sprent, and Z. Cai. 1999. TCR-mediated internalization of peptide-MHC complexes acquired by T cells.Science 286:952.

20. Shores, E. W., S. O. Sharrow, and A. Singer. 1991. Presence of CD4 and CD8determinants on CD42 CD82 murine thymocytes: passive acquisition of CD8accessory molecules.Eur. J. Immunol. 21:973.

21. Hwang, I., J.-F. Huang, H. Kishimoto, A. Brunmark, P. A. Peterson, M. R. Jack-son, C. D. Surh, Z. Cai, and J. Sprent. 2000. T cells can use either T-cell receptoror CD28 receptors to absorb and internalize cell surface molecules derived fromantigen-presenting cells.J. Exp. Med. 191:1137.

22. Inaba, K. 1992. Generation of large numbers of dendritic cells from mouse bonemarrow cultures supplemented with granulocyte/macrophage colony-stimulatingfactor.J. Exp. Med. 176:1693.

23. Hodge, J. W., and J. Schlom. 1999. Comparative studies of a retrovirus versus apoxvirus vector in whole tumor cell vaccines.Cancer Res. 59:5106.

24. Hodge, J. W., H. Sabzevari, A. G. Yafal, L. Gritz, M. G. Lorenz, and J. Schlom.1999. A triad of costimulatory molecules synergize to amplify T-cell activation.Cancer Res. 59:5800.

25. Hodge, J. W., S. I. Abrams, J. Schlom, and J. Kantor. 1994. Induction of anti-tumor immunity by recombinant vaccinia viruses expressing B7.1 or B7.2 co-stimulatory molecules.Cancer Res. 54:5552.

26. Borriello, F., M. P. Sethna, S. D. Boyd, A. N. Schweitzer, E. A. Tivol, D. Jacoby,T. B. Strom, E. M. Simpson, G. J. Freeman, and A. H. Sharpe. 1997. B7.1 andB7.2 have overlapping, critical roles in immunoglobulin class switching and ger-minal center formation.Immunity 16:303.

27. Swallow, M. M., J. J. Wallin, and W. C. Sha. 1999. B7h, a novel costimulatoryhomolog of B7.1 and B7.2, is induced by TNF-a. Immunity 11:423.

28. Naramura, M., R. J. Hu, and H. Gu. 1998. Mice with a fluorescent marker forinterleukin-2 gene activation.Immunity 9:209.

29. Monks, C. R., B. A. Freiberg, H. Kupfer, N. Sciaky, and A. Kupfer. 1998. Three-dimensional segregation of supramolecular activation clusters in T cells.Nature395:82.

30. Wulfing, C., and M. M. Davis. 1998. A receptor/cytoskeletal movement triggeredby costimulation during T-cell activation.Science 282:2266.

31. Viola, A. S., Y. Schroeder, and A. Lanzavecchia. 1999. T lymphocyte costimu-lation mediated by reorganization of membrane microdomains.Science 283:680.

32. Grakoui, A., S. K. Bromley, C. Sumen, M. M. Davis, A. S. Shaw, P. M. Allen,and M. L. Dustin. 1999. The immunological synapse: a molecular machine con-trolling T cell activation.Science 285:221.

33. Wade, W. F., E. D. Ward, E. F. Rosloniec, B. G. Barisas, and J. H. Freed. 1994.Truncation of the Aa chain of MHC class II molecules results in inefficientantigen presentation to antigen-specific T cells.Int. Immunol. 6:1457.

34. Murtaza, A., V. K. Kuchroo, and G. J. Freeman. 1999. Changes in the strengthof costimulation through the B7/CD28 pathway alter functional T cell responsesto altered peptide ligands.Int. Immunol. 11:407.

35. Levy, M., P. Vieira, A. Coutinho, and A. Freitas. 1987. The majority of “natural”immunoglobulin secreting cells are short-lived and the progeny of cycling lym-phocytes.Eur. J. Immunol. 17:849.

36. Russell, J. H., C. L. White, D. Y. Loh, and P. Meleedy-Rey. 1991. Receptor-stimulated death pathway is opened in antigen in mature T cells.Proc. Natl.Acad. Sci. USA 88:2151.

37. Zhang, X., L. Giangreco, H. E. Broome, C. M. Dargan, and S. L. Swain. 1995.Control of CD4 effector fate: transforming growth factorb1 and interleukin 2synergize to prevent apoptosis and promote effector expansion.J. Exp. Med.182:699.

38. Barnaba, V., C. Watts, M. de Boer, P. Lane, and A. Lanzavecchia. 1994. Pro-fessional presentation of antigen by activated human T cells.Eur. J. Immunol.24:71.

2513The Journal of Immunology

by guest on May 17, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from