Introduction

T cells require multiple signals for activation.1 The first ofthese signals is an antigen receptor-mediated event leading tothe delivery of a transmembrane signal through the CD3complex. Additional activation signals occur once costimula-tory molecules expressed on T cells bind to appropriateligand(s) on professional APC. Of such molecules identifiedto the present, CD28 is the most understood and appears to beimportant during the early phase of antigen-induced activation of naive T cells.2,3 Thus, upon binding of CD28 toB7.1/7.2 (CD80/86) expressed on activated APC,4,5 T cells areprompted to undergo a series of intracellular events leading toclonal expansion and the expression of effector activities.6,7

Fully activated T cells then begin to express CTLA-4 which,due to its greater avidity for CD80/86, displaces CD28binding and initiates a process of self-initiated deactivation.8

Recently, CD43 has been shown to have T-cell costimula-tory properties9 that function independently from CD28.10 Asyet, however, it is unclear what conditions dictate the use ofCD28 versus CD43 in the activation of T cells. Certainly, thismust be determined at least in part by the regulated expres-sion of ligands for CD28 and CD43, although the molecularnature of the CD43 ligand has yet to be established. In addition to direct molecular signalling via T-cell membranemolecules, soluble growth-promoting factors are known to be

important to achieve complete T-cell activation. Theseinclude cytokines that act directly on T cells, as well as thosethat have immunomodulating effects on other cytokines.

T cells in the intestinal epithelium of mice, the intra-epithelial lymphocytes (IEL), display a number of uniquedevelopmental and functional features (reviewed in McGheeand Kiyono11). Unlike resting T cells in the blood, spleen, orlymph nodes, freshly isolated IEL exhibit lytic activity inex vivo redirected cytotoxicity assays.12,13 Moreover, IEL havedistinct activation requirements leading to cell proliferationand clonal expansion. For example, conventional methods ofT-cell activation involving TCR/CD3 stimulation that readilyinduce mitosis in splenic or lymph node T cells have littleeffect on IEL even in the presence of IL-2.14,15 Recently,however, IL-15 has been shown to work in concert with IL-2to augment the in vitro proliferative response of murine IEL,possibly due to the survival-promoting effects of IL-15.16–18

At present, little is known about the extent to which costim-ulatory molecules are used by IEL during activation.However, with the exception of some TCRαβ, CD8αβcells,19 most murine and human small intestine IEL aredevoid of CD28 expression,20–24 thus implying either thatT cells in the gut epithelium do not utilize costimulatorysignals for optimal cell proliferation, or that this is achievedthrough other accessory molecules.

In this report, we demonstrate that CD43 functions as acostimulatory molecule that potentiates the proliferativeresponse of IEL following CD3 triggering, and show that thisis dependent upon IL-2 and IL-15. These findings have directrelevance for understanding a host’s response to gastro-intestinal infection, for elucidating events associated withchronic intestinal inflammation, and for devising strategies tobe used in mucosal vaccine development and delivery.

Immunology and Cell Biology (2001) 79, 303–307

Brief Communication

CD43 potentiates CD3-induced proliferation of murine intestinalintraepithelial lymphocytes

E UMIT BAGRIACIK, 1 MIN TANG, 1 * HEUY-CHING WANG 2 and JOHN R KLEIN 1 , 2

1Department of Biological Science, University of Tulsa, Tulsa, Oklahoma and 2Department of Basic Sciences, Dental Branch, University of Texas Health Science Center, Houston, Texas, USA

Summary The involvement of CD43 in cell proliferation of murine intestinal intraepithelial lymphocytes (IEL)has been studied in in vitro CD3-stimulated cell cultures. In the presence of either IL-2 or IL-15, CD3 stimulationof IEL resulted in low levels of proliferation as measured by thymidine incorporation, whereas no proliferationoccurred upon CD3 stimulation in the absence of cytokines. The combination of both cytokines to IEL cultures syn-ergistically enhanced CD3-induced proliferation by approximately threefold that of cultures supplemented witheither cytokine alone. Most importantly, however, proliferation of IEL was significantly greater when CD3 stimu-lation occurred in conjunction with CD43 triggering, indicating that CD43 functions as a coactivational signal formurine IEL. These findings indicate that a spectrum of potential proliferative responses exist among murine IELdepending on the types and combinations of signals received, and that because under normal conditions murine IEL are largely devoid of CD28 expression, a classical T-cell coactivational molecule, the capacity for high-levelIEL proliferation may reside with CD43.

Key words: activation, immune regulation, stimulation, T cell.

Correspondence: Dr JR Klein, Department of Basic Sciences,Dental Branch, University of Texas Health Science Center, 6516John Freeman, Houston, TX 77030, USA. Email: jklein@mail.db.uth.tmc.edu

*Present address: M Tang, National Animal Disease Center,Animal Research Services, USDA, Ames, IA 50010, USA.

Received 11 September 2000; accepted 5 February 2001.

Materials and Methods

Mice

Adult female C57BL/6 mice, 8–10 weeks of age, were raised at theUniversity of Tulsa from breeding stocks obtained from the JacksonLaboratories (Bar Harbor, ME, USA). Six–seven-week-old femaleCD43 knockout mice (B6,129-Spntm1Bam) were purchased from theJackson Laboratories; these animals were used at 8–10 weeks of age.

Antibodies and cell isolations

The antibodies used in this study were: rat antimouse CD3(R3/7/159),25 rat antimouse CD43 (R2/60),9,10,26 phycoerythrin (PE)-labelled antirat IgM antibody (G53-238) (PharMingen, SanDiego, CA, USA), PE-labelled anti-CD45 (30-F11) (PharMingen),and rat IgM antibody (R4-22) (PharMingen). Anti-CD3 and anti-CD43 mAb were purified from tissue culture supernatants using agoat antirat IgG sepharose column or a mouse mAb to rat IgMcoupled to a goat antimouse sepharose column, respectively (PierceScientific; Rockford, IL, USA). Concentrations of purified proteinswere estimated spectrophotometrically.

Intraepithelial lymphocytes were isolated from the small intestineusing previously published protocols27 with minor modifications.After recovery of cells from the 40–70% Percoll interface, cells weretreated with mAb G8.828 (American Type Culture Collection,Rockville, MD, USA) for 30 min at 4°C followed by treatment for15 min at 37°C with guinea pig complement (Accurate Chemicals,Westbury, NY, USA) to remove residual epithelial cells. Thisyielded a preparation consisting of >95 CD45+ IEL with 95–98%viability based on trypan blue exclusion.

In vitro proliferation assays

Ninety-six-well tissue culture microtiter plates were coated overnightat 4°C with anti-CD3 mAb diluted in PBS. Wells were washed withPBS, and IEL were added at a concentration of 1.5 × 105 cells/well in 200 µL of RPMI-1640 supplemented with fetal bovine serum (10% v/v), 100 U/mL penicillin-streptomycin, 2 mmol/LL-glutamine, and 5 × 10–5 mol/L 2-mercaptoethanol (Sigma Chemi-cals, St. Louis, MO, USA; all reagents), containing recombinant IL-2 (4 ng/mL) and/or IL-15 (10 ng/mL) (Sigma Chemicals). Forcostimulation experiments, R2/60 or rat IgM antibodies were addedin soluble form to culture wells at varying concentrations as needed.Cells were cultured for 72 h at 37°C in a 5% CO

2environment,

pulsed with 1µCi/well 3H-thymidine (ICN Chemicals, Santa Ana,CA, USA) 6 h prior to harvesting, and collected using an automatedcell harvester (MA Bioproducts, Rockville, MD, USA). Incorporationof cell-associated thymidine was measured in a liquid scintillationcounter (Beckman-Coulter, Fullerton, CA, USA).

Results and Discussion

Freshly isolated IEL were analysed by flow cytometry forexpression of CD43. As seen in Fig. 1, nearly all (93%) IELexpressed high levels of CD43 as determined by reactivitywith the pan-CD43 mAb, R2/60. This reagent is specific formurine CD43 based on expression cloning,10 and as shownhere by a lack of reactivity for IEL from CD43 knockoutmice (Fig. 1).

To establish optimal in vitro culture conditions for prolif-eration, freshly isolated IEL were seeded into microwellscoated with titrated amounts of anti-CD3 mAb in the pres-ence or absence of IL-2 and/or IL-15. As can be seen inFig. 2a, there was a dose-dependent effect of anti-CD3 stim-ulation of IEL in the presence of IL-2 (4 ng/mL) and IL-15(10 ng/mL), with maximal activity occurring in wells coatedwith anti-CD3 at a concentration of 0.1 µg/mL; optimal con-centrations of cytokines were determined empirically in aseries of titration experiments (data not shown). Little or noproliferative activity was observed for cells cultured withanti-CD3 in the absence of cytokines, with cytokine withoutCD3 stimulation, or when cultured in medium alone(Fig. 2a).

The synergistic effects of IL-2 and IL-15 in the pro-liferative response of IEL were demonstrated in anti-CD3-stimulated cultures supplemented with either or bothcytokines. As can be seen in Fig. 2b, while IL-2 and IL-15were each capable of slightly enhancing the proliferativeactivity of IEL compared to anti-CD3 cultures containing nocytokine supplement, the presence of both IL-2 and IL-15resulted in a threefold increase in proliferation, thus indicat-ing that the combined effects of IL-2 and IL-15 were greaterthan the sum of activity from either cytokine alone.

The effects of CD43 on IEL activation were studied fol-lowing stimulation with 10-fold suboptimal concentrations ofanti-CD3 (0.01 µg/mL per well) in the presence of solubleanti-CD43 or an isotype/species-matched control antibody.

EU Bagriacik et al.304

Figure 1 Reactivity of anti-CD43 mAb R2/60 to intraepithe-lial lymphocytes (IEL) from (a)normal mice, but not (b) CD43knockout mice.

As can be seen in Fig. 2c, costimulation of IEL was signifi-cantly enhanced across a range of anti-CD43 concentrations(0.12–0.03 µg/mL), with optimal activity occurring at0.06 µg/mL. Stimulation of IEL by anti-CD43 mAb in thepresence of cytokines, but in the absence of CD3 stimulation,did not induce IEL proliferation (Fig. 2c), indicating thatCD43 activation is CD3 dependent and that CD43 triggeringby itself does not lead to IEL proliferation. To further rule out a non-specific effect of anti-CD43 mAb on proliferation,IEL from CD43-deficient mice were cultured with anti-CD3mAb in the presence and absence of anti-CD43 mAb.Experiment 1: unstimulated = 3690 c.p.m.; CD3 stimulated =29 120 c.p.m.; CD3 + CD43 stimulated = 31 983 c.p.m.Experiment 2: unstimulated = 156 c.p.m.; CD3 stimulated =47 667 c.p.m.; CD3 + CD43 stimulated = 46 828 c.p.m. Allcells were cultured in the presence of both IL-2 and IL-15 as described. Although CD3-induced proliferation levels differed somewhat between experiments, each fell within therange obtained for normal mice (Fig. 2c). Most importantly,however, within an experiment the addition of CD43 mAbdid not enhance CD3-mediated proliferation in CD43-deficient mice, thus confirming that the effect of anti-CD43costimulation of IEL from normal mice was not an artifact ofaddition of anti-CD43 mAb.

The finding that CD43 has costimulatory properties formurine IEL has relevance for understanding the immuno-biology of intestinal T cells in a number of ways. The datareported here indicate that varying degrees of IEL prolifera-tion can occur depending on the types of signals received byIEL. Although activation through the TCR/CD3 complex isan essential first step, triggering through that pathway alonein the absence of IL-2 and IL-15 is insufficient to achieveproliferation. CD3-stimulation in the presence of either IL-2

or IL-15 results in low-level proliferation, suggesting thatsome proliferation can occur even when activation signals areat suboptimal levels. Considerably higher levels of prolifera-tion occur when both cytokines are present, and the greatestlevel of proliferation occurs when CD3 stimulation is accom-panied by CD43 coactivation in the presence of IL-2 and IL-15.

The capacity of CD43 to augment IEL proliferation raisesa number of new questions, including the nature of the ligandmolecules with which CD43 interacts, the cells of the intes-tine that express those molecules, and the conditions thatserve to control or regulate their expression. However, thestrong negative charge of CD43 has been cited for its poten-tial to inhibit rather than facilitate the cell contact and adhe-sion that would be expected for activation.29 Consequently,activation via CD43 may occur as part of a multistagedprocess involving the redistribution of CD43 on the surfaceof T cells, resulting in the exposure of other activation deter-minants. Some evidence in support of this idea comes fromstudies using CD43-deficient mice in which homing toperipheral lymph nodes was enhanced due to an increase inthe exposure of surface L-selectin,30 and from studies demon-strating that on resting T cells CD43 inhibits interactionsbetween the TCR and APC.31

The overall absence of CD28 expression on murine IEL,as seen by the lack of cell staining,20–22 by the inability ofCD28 stimulation to enhance IEL proliferation or cytokinesynthesis,20,21 and by the failure of CTLA-4-Ig to immuno-modulate TCR-mediated proliferation of IEL,20 indicates thatthe process of CD28 signalling is largely omitted from theactivation scheme of most IEL. Although the reason for thisremains unclear, it is possible that combined T-cell signallingvia CD28 and CD43 leads to hyperactivation or other effects

CD43 coactivation of IEL 305

Figure 2 (a) Dose-dependent activity of anti-CD3-induced intraepithelial lymphocytes (IEL) proliferation in the presence of 4 ng/mLIL-2 and 10 ng/mL IL-15 compared to proliferation of CD3-stimulated cultures without cytokines, with cytokines in the absence of CD3 stimulation, or in the absence of cytokines and CD3 stimulation. (b) IL-2 (4 ng/mL) and IL-15 (10 ng/mL) synergistically augmentCD3-mediated IEL proliferation compared to CD3 stimulation alone; microwells were coated with 0.1 µg/mL anti-CD3 mAb. (c) In the presence of 4 ng/mL IL-2 and 10 ng/mL IL-15, CD43 costimulation resulted in significantly greater (*P < 0.01) CD3-induced pro-liferation compared to cultures containing isotype/species-matched control antibody; wells were coated with 0.01 µg/mL anti-CD3. Notethat anti-CD43 stimulation in the presence of cytokines without CD3 stimulation did not induce significant levels of IEL proliferation.Data are mean values ± SEM of two independent experiments.

that are undesirable for IEL, given the propensity for excessiveinflammatory responses within the intestine.32 Interestingly,CD28 has been shown to be required for the development ofautoimmune encephalomyelitis in mice, further implying a rolefor CD28 in inflammation.33 Alternatively, the lack of CD28expression on IEL may have to do with the fact that antigenpresentation to gut T cells can occur from non-professionalAPC, such as MHC class-II+ epithelial cells,34 that expresslittle or no CD80/86,35 thereby rendering CD28 costimulationuseless along that pathway. Interestingly, it was recentlydemonstrated that high levels of CD43 expression on T cellscorrelated with resistance to antigen-induced cell death andwere associated with a memory CD4+ T-cell phenotype,36

a finding that may help to explain how CD43-mediated co-stimulation is downregulated after activation has occurred.

In summary, this study provides additional informationinto the mechanisms by which T cells of the intestinal epithe-lium are regulated. Experiments can now be done to explorethe extent to which CD43 signalling contributes to intestinalinflammation under natural and experimental conditions.Conversely, knowledge of CD43 coactivational properties hasdirect relevance to situations in which IEL activation isdesired, for example, vaccine delivery to the gastrointestinaltract.

Acknowledgements

This work was supported by grant DK35566 from theNational Institute of Health (NIH) and by a grant from the Oklahoma Center for the Advancement of Science andTechnology (OCAST).

References

1 Langman RE, Cohn M. Two signal models of lymphocyte activation. Immunol. Today 1993; 14: 235–7.

2 Chambers CA, Allison JP. Costimulatory regulation of T cellfunction. Curr. Opin. Cell Biol. 1999; 11: 203–10.

3 Rudd CE. Upstream-downstream: CD28 cosignaling pathwaysand T cell function. Immunity 1996; 4: 527–34.

4 Liu Y, Janeway CA. Cells that present both specific ligand andco-stimulatory activity are the most efficient inducers of clonalexpansion of normal CD4 T cells. Proc. Natl Acad. Sci. USA1992; 89: 3845–9.

5 Lenchow DJ, Walunas TL, Bluestone JA. CD28/B7 system of T cell costimulation. Ann. Rev. Immunol. 1996; 14: 233–58.

6 Fraser JD, Irving BA, Crabtree GR, Weiss A. Regulation ofinterleukin-2 gene enhancer activity by the T-cell accessory molecule CD28. Science 1991; 251: 313–16.

7 Ward GS. CD28: a signaling perspective. Biochem. J. 1996; 318:361–77.

8 Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA,Sharpe AH. Loss of CTLA-4 leads to massive lymphoprolifera-tion and fetal multiorgan tissue destruction, revealing a criticalnegative regulatory role of CTLA-4. Immunity 1995; 3: 541–7.

9 Hamad M, Mosley RL, Wang J, Klein JR. Stimulation via theCD43 coreceptor augments T cell proliferation during the early phase of antigen-induced activation. Dev. Comp. Immunol.1996; 201: 77–82.

10 Sperling AI, Green JM, Mosley RL et al. CD43 is a murine T cell costimulatory receptor that functions independently ofCD28. J. Exp. Med. 1995; 182: 139–46.

11 McGhee JR, Kiyono H. The mucosal immune system. In: PaulWE (ed.). Fundamental Immunology. New York: Raven Press,1999; 909–45.

12 Mosley RL, Styre D, Klein JR. Differentiation and functionalmaturation of bone marrow-derived intestinal epithelial T cellsexpressing membrane T cell receptor in athymic radiationchimeras. J. Immunol. 1990; 145: 1369–75.

13 Viney JL, Kilshaw PJ, MacDonald TT. Cytotoxic α/β+ and γ/δ+

T cells in murine intestinal epithelium. Eur. J. Immunol. 1990;20: 1623–6.

14 Mosley RL, Klein JR. Proliferative properties of murine intesti-nal intraepithelial lymphocytes (IEL): IEL expressing TCRαβor TCRγδ are largely unresponsive to proliferative signals mediated via conventional stimulation of the CD3-TCRcomplex. Int. Immunol. 1991; 3: 563–9.

15 Sydora BC, Mixter PF, Holcombe HR et al. Intestinal intra-epithelial lymphocytes are activated and cytolytic but do notproliferate as well as other T cells in response to mitogenicsignals. J. Immunol. 1993; 150: 2179–91.

16 Inagaki-Ohara K, Nishimura H, Mitani A, Yoshikai Y. Inter-leukin-15 preferentially promotes the growth of intestinalintraepithelial lymphocytes bearing γδ T cell receptor in mice.Eur. J. Immunol. 1997; 27: 2885–91.

17 Chu CL, Chen SS, Wu TS, Kuo SC, Liao NS. Differential effects of IL-2 and IL-15 on the death and survival of activated TCRγδ+ intestinal intraepithelial lymphocytes. J. Immunol. 1999;162: 1896–903.

18 Lai YG, Gelfanov V, Gelfanov V et al. IL-15 promotes survivalbut not effector function differentiation of CD8+ TCRαβ+

intestinal intraepithelial lymphocytes. J. Immunol. 1999; 163:5843–50.

19 Ohteki T, MacDonald HR. Expression of the CD28 costimula-tory molecule on subsets of murine intestinal intraepitheliallymphocytes correlates with lineage and responsiveness. Eur. J.Immunol. 1993; 23: 1251–5.

20 Gelfanov V, Lai YG, Gelfanov V, Dong JY, Su JP, Liao NS. Differential requirement of CD28 costimulation for activation of murine CD8+ intestinal intraepithelial lymphocyte subsetsand lymph node cells. J. Immunol. 1995; 155: 76–82.

21 Gramzinski RA, Adams E, Gross JA, Goodman TG, Allison JP,Lefrancois L. T cell receptor-triggered activation of intra-epithelial lymphocytes in vitro. Int. Immunol. 1993; 5: 145–53.

22 Boll G, Rudolphi A, Spiess S, Reimann J. Regional specializa-tion of intraepithelial T cells in the murine small and large intestine. Scand. J. Immunol. 1995; 41: 103–13.

23 Anumanthan A, Bensussan A, Boumsell L et al. Cloning ofBY55, a novel Ig superfamily member expressed on NK cells,CTL, and intestinal intraepithelial lymphocytes. J. Immunol.1998; 161: 2780–90.

24 Sillett HK, Southgate J, Howdle PD, Trejdosiewicz LK. Expres-sion of activation and costimulatory elements by human intestinal intraepithelial lymphocytes. Scand. J. Immunol. 1999;50: 52–60.

25 Bagriacik EU, Miller KS. Cell surface sialic acid and the regu-lation of immune cell interactions: the neuraminidase effectreconsidered. Glycobiology 1999; 9: 267–75.

26 Mosley RL, Hamad M, Whetsell M, Klein JR. A novel markerof murine bone marrow hematopoietic stem cells that isexpressed on peripheral T cells and is associated with a func-tionally important molecule on activated cytotoxic T lympho-cytes. Hybridoma 1994; 13: 353–8.

27 Mosley RL, Klein JR. A rapid method for isolating murine intestine intraepithelial lymphocytes with high yield and purity.J. Immunol. Methods 1992; 156: 19–26.

EU Bagriacik et al.306

28 Farr A, Nelson A, Truex J, Hosier S. Epithelial heterogeneity in the murine thymus: a cell surface glycoprotein expressed by subcapsular and medullary epithelium. J. Histochem. Cyto-chem. 1991; 39: 645–53.

29 Ostberg JR, Barth RK, Frelinger JG. The Roman god Janus: aparadigm for the function of CD43. Immunol. Today 1998; 19:546–50.

30 Stockton BM, Cheng G, Manjunth N, Ardman B, von AndrianUH. Negative regulation of T cell homing by CD43. Immunity1998; 8: 373–81.

31 Sperling AI, Sedy JR, Manjunath N, Kupfer A, Ardman B,Burkhardt JK. TCR signaling induces selective exclusion ofCD43 from the T cell-antigen-presenting cell contact site. J. Immunol. 1998; 161: 6459–62.

32 MacDonald TT, Monteleone G, Pender SL. Recent developments

in the immunology of inflammatory bowel disease. Scand. J.Immunol. 2000; 51: 2–9.

33 Oliveira-dos-Santos AJ, Ho A, Tada Y et al. CD28 costimulationis crucial for the development of spontaneous autoimmuneencephalomyelitis. J. Immunol. 1999; 162: 4490–5.

34 Kaiserlian D. Intestinal epithelial cells are nonconventionalantigen presenting cells. In: Kaiserlian D (ed.). Antigen Presen-tation by Intestinal Epithelial Cells. New York: Springer, 1996;39–52.

35 Sanderson IR, Oulette AJ, Carter EA, Walker WA, Harmatz PR.Differential regulation of B7 mRNA in enterocytes and lym-phoid cells. Immunology 1993; 79: 434–8.

36 He YW, Bevan MJ. High level expression of CD43 inhibits T cell receptor/CD3-mediated apoptosis. J. Exp. Med. 1999;190: 1903–11.

CD43 coactivation of IEL 307