Blockade of CD86 in BALB/c mice infected withLeishmania major does not prevent the expansion of low avidity T cells

0014-2980/02/1212-3566$17.50+.50/0 © 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Blockade of CD86 in BALB/c mice infected withLeishmania major does not prevent the expansionof low avidity T cells

Monica Moro1, Christophe Filippi1, Alexandra Gallard2, Laurent Malherbe1, GillesFoucras2, Hisaya Akiba3, Hideo Yagita3, Jean-Charles Guery2 and NicolasGlaichenhaus1

1 Centre National de la Recherche Scientifique, Valbonne, France2 Institut National de la Sante et de la Recherche Medicale, Hopital Purpan, Toulouse, France3 Department of Immunology, Juntendo University School of Medicine, Bunkyo-ku, Tokyo, Japan

The interactions between CD28 and its ligand CD86 are critical for the regulation of T cellresponses. However, it is not clear whether CD4+ T cells expressing low and high avidity TCRare equally dependent on CD28 costimulation for their activation and expansion. To addressthis issue, we have used multimers of I-Ad molecules linked to a peptide derived from theLeishmania major homolog for the receptor of activated C kinase (LACK) antigen to comparethe fate of LACK-specific CD4+ T cells in Leishmania-infected BALB/c mice which have beentreated or not with anti-CD86 mAb. Although the administration of anti-CD86 mAb did notcompletely prevent the expansion of LACK-specific T cells, their frequency and numberwere markedly reduced. In mice treated with anti-CD86 mAb as well as in control animals, L.major induced the clonal expansion of LACK-specific T cells which expressed a canonicallow avidity V § 8/V g 4 TCR. Taken together, our results suggest that the molecular interactionsbetween CD28 on T cells and CD86 on APC serve to amplify and modulate T cell responseswithout promoting breadth in the TCR repertoire.

Key words: T lymphocyte / Costimulatory molecule / T cell repertoire / Leishmania major / T celldifferentiation

Received 6/3/02Revised 27/9/02Accepted 17/10/02

[I 22982]

CF and AG contributed equally to this work.

Abbreviations: LACK: Leishmania homolog for the recep-tor of activated C kinase SLA: Soluble Leishmania antigens

1 Introduction

T cell activation and differentiation are dependent on theinteractions between the TCR and its cognate peptide/MHC ligand. However, the outcome of TCR stimulation isregulated by the simultaneous engagement of the co-stimulatory molecule CD28 expressed on T cells with itsligands CD80 and CD86 expressed at the surface of APC[1]. While both CD80 and CD86 bind to CD28 with similarlow affinities, they display different patterns of expres-sion. Thus, CD86 is constitutively expressed at low levelson immature DC [2], resting B cells [3], and resting mono-cytes [3]. In contrast, CD80 is only expressed on acti-vated APC [4]. Furthermore, while both CD80 and CD86are up-regulated upon APC activation [2, 5], CD86 isinduced more rapidly than CD80 [6, 7]. Because of theseresults, it has been proposed that CD86 was more

important than CD80 for the initiation of T cell responses[8, 9].

The role of CD86 in the regulation of T cell responses hasbeen studied both in vitro and in vivo. Thus, studiesusing either cell lines expressing CD86 [10–12], or APCderived from CD86-deficient mice [13, 14], have pro-vided evidence that CD86 promotes T cell proliferationand cytokine production. Likewise, anti-CD86 mAb havebeen shown to prevent the spontaneous development ofdiabetes in non obese diabetic (NOD) mice [15], and toinhibit transplant rejection in mice grafted with allogeneicpancreatic islets [16]. Other studies have shown thatCD86 promotes the development of Th2 cells and down-regulates Th1 responses. Thus, repetitive stimulation ofnaive T cells with CD86, but not with CD80, induced theproduction of IL-4 [17]. Likewise, in experimental allergicencephalomyelitis (EAE), anti-CD86 mAb up-regulatedIFN- + production and exacerbated the disease [18–21].Lastly, blockade of CD86 in BALB/c mice infected with L.major promoted the expansion of Th1 protective cellsand induced the elimination of the parasite and the heal-ing of the lesions [22]. Likewise, CD86-deficient BALB/cmice were resistant to L. major [23].

3566 M. Moro et al. Eur. J. Immunol. 2002. 32: 3566–3575

While CD86 regulates T cell responses in vivo, it is notclear whether all T cells are equally dependent on CD86for their activation and expansion. Thus, it remains to beestablished whether the activation and the expansion oflow avidity T cells is more dependent on CD86/CD28interactions than those of high avidity T cells. To addressthis issue, we sought to analyze the effect of CD86blockade on CD4+ T cells specific for the parasite LACKantigen in mice infected with L. major. To facilitate thedetection of LACK-specific T cells in infected animals,we have used a transgenic mouse strain, 16.2 g , whichcarries the rearranged TCR g gene of a LACK-specific Tcell hybridoma [24]. 16.2 g -transgenic mice were crossedfor 15 generations to BALB/c mice. Although 16.2 gBALB/c-transgenic mice exhibited an increased fre-quency of LACK-specific T cells as compared to theirnegative littermates, these mice mounted a Th2-biasedresponse and developed progressive lesions upon infec-tion with L. major [24]. This experimental model systemwas chosen for three reasons. First, as stated above,CD86 is important for the development of the parasite-specific immune response in BALB/c mice infected withL. major [22, 25]. Secondly, LACK-specific CD4+ T cellscan easily be detected by flow cytometry in the lymphoidorgans of 16.2 g -transgenic mice using multimers of sol-uble peptide/MHC molecules [24]. Lastly, the infection of16.2 g BALB/c-transgenic mice with L. major results inthe expansion of LACK-specific CD4+ T cells whichexpress low avidity TCR [24].

2 Results

2.1 Anti-CD86 mAb protect 16.2 I BALB/c-transgenic mice from L. major

As treatment with anti-CD86 mAb has been shown toconfer resistance to L. major in wild-type (wt) BALB/cmice [22], we first investigated whether this mAb had thesame effect in 16.2 g BALB/c-transgenic mice. To thisaim, wt and 16.2 g -transgenic BALB/c mice wereinfected with L. major and treated with either anti-CD86mAb (GL1) or rat IgG2a isotype control. In agreement withprevious results [22] wt BALB/c mice developed smallerlesions when treated with anti-CD86 mAb than whentreated with IgG2a isotype control (Fig. 1A). Similarly,16.2 g BALB/c-transgenic mice developed progressivefootpad swelling when treated with IgG2a isotype controland small self-healing lesions when treated with anti-CD86 mAb (Fig. 1A). Furthermore, the draining LN of16.2 g -transgenic mice treated with anti-CD86 mAb con-tained 100 to 1,000-fold less parasites than those fromcontrol animals (Fig. 1B).

It was previously shown that the blockade of CD86 in wtBALB/c mice abrogated Th2 differentiation and pro-moted the development of a Th1 immune response [22].To determine whether a similar phenomenon was occur-ring in 16.2 g -transgenic BALB/c mice, these animalswere treated with either anti-CD86 mAb or IgG2a isotypecontrol, infected with L. major, and LN CD4+ cells werepurified 4 weeks later. In a first set of experiments, cellswere stimulated with PMA and ionomycin and analyzedby flow cytometry for intracellular staining of IL-4 andIFN- + . In both wt and 16.2 g -transgenic mice, the fre-quency of IL-4-secreting CD4+ T cells was lower in micetreated with anti-CD86 mAb than in control mice(Fig. 1C). In 16.2 g -transgenic BALB/c mice, the fre-quency of IL-4-secreting CD4+ T cells was 0.3% in micetreated with anti-CD86 mAb as compared to 1.0% inthose treated with IgG2a isotype control (Fig. 1C). Toindependently confirm this result, LN CD4+ T cells wereincubated with mitomycin C-treated BALB/c spleno-cytes with or without an optimal concentration of solubleLeishmania antigen (SLA). Cellular supernatants wereharvested 72 h later and IFN- + , IL-4 and IL-5 contentswere measured by ELISA. Cells from mice treated withanti-CD86 mAb secreted similar amounts of IFN- + inresponse to LACK as compared to cells from micetreated with IgG2a isotypic control (Fig. 1D). In contrast,cells from anti-CD86 mAb-treated mice secreted four- tofivefold less IL-4 and eight- to tenfold less IL-5 thanthose from the animals that have been treated with IgG2a

isotypic control (Fig. 1D). Thus, as previously reportedfor wt BALB/c mice, blockade of CD86 in 16.2 g -transgenic BALB/c mice resulted in the down-regulationof the counter-protective Th2 response directed to L.major and in the development of a healing phenotype.

2.2 Anti-CD86 mAb inhibit the expansion of I-Ad/LACK+ CD4+ T cells

To further analyze the effects of CD86 blockade on theactivation and expansion of LACK-specific CD4+ T cells,16.2 g -transgenic mice were treated with either anti-CD86 mAb or IgG2a isotype control, and infected with L.major. At different times after infection, LN CD4+ cellswere purified and stained with I-Ad/LACK multimers, andmAb directed to the activation markers CD44 or CD69. Inagreement with previous results [24], L. major induced arapid increase in the frequency and the number of I-Ad/LACK+ cells in mice treated with IgG2a isotype control.The frequency of I-Ad/LACK+ CD4+ T cells started toincrease at day 1, peaked to 5% at day 3, and droppedat later time points to reach background levels 4 weeksafter infection (Fig. 2A). While the administration of anti-CD86 mAb did not completely prevent the expansion of

Eur. J. Immunol. 2002. 32: 3566–3575 CD86 blockade does not prevent the expansion of low avidity T cells 3567

P Fig. 1. Effect of anti-CD86 mAb on the course of L. majorinfection in 16.2 g -transgenic mice. Wt and 16.2 g -transgenicBALB/c mice were injected with 2×106 stationary phase pro-mastigotes in the left hind footpad. At the time of infectionand subsequently three times a week, mice were injectedi.p. with either anti-CD86 mAb or IgG2a isotype control(150 ? g/injection). (A) Footpad swelling was measured at theindicated times after infection using a metric caliper by sub-tracting the thickness of the uninfected footpad from that ofthe infected one. Data show the arithmetic mean ± SD of fiveindividual mice. (B) Mice treated with either IgG2a isotypecontrol (empty bars) or anti-CD86 mAb (filled bars) werekilled 3 weeks after infection and parasite numbers weremeasured in their draining LN. Data show the arithmeticmean ± SD of three individual mice. (C) Wt (left panels) and16.2 g -transgenic (right panels) mice were killed 4 weeksafter infection. LN CD4+ cells were purified, stimulated withPMA and ionomycin, permeabilized and stained with anti-CD4, anti-IL-4 and anti-IFN- + mAb. Typical flow cytometryprofiles are shown after gating on CD4+ T cells. The percent-ages of IL-4- and IFN- + -secreting cells are indicated formice treated with IgG2a isotype control (upper panels) oranti-CD86 mAb (lower panels). Data are representative ofone out of three experiments. (D) 16.2 g -transgenic mice(three mice in each group) were treated with anti-CD86 mAb(solid bars) or IgG2a isotype control (empty bars). LN werepooled and cells were incubated in triplicates with SLA. Cel-lular supernatants were harvested 72 h later and IFN- + , IL-4and IL-5 contents were measured by ELISA. Data show thearithmetic mean ± SD for each triplicate. Data are represen-tative of one out of two experiments.

I-Ad/LACK+ CD4+ T cells, their frequency at the peak ofthe response was reduced by almost twofold (3.2 versus5%). Furthermore, from day 3 after infection to the end ofthe experiment, the LN of mice treated with anti-CD86mAb contained two- to threefold less I-Ad/LACK+ CD4+ Tcells than those of control animals (Fig. 2B). The fre-quency of I-Ad/LACK+ cells progressing through the cellcycle was reduced in mice treated with anti-CD86 mAb.Thus, while the frequency of I-Ad/LACK+ cells exhibiting alarge size increased from 3 to 5% at day 0 to 75% atday 3 in control mice, only 56% of I-Ad/LACK+ cellsexhibited a large size 3 days after infection in the animalsinjected with anti-CD86 mAb (Fig. 2C).

We have previously shown that I-Ad/LACK+ cells undergorapid surface phenotypic changes upon infection with

L. major [24]. Likewise, in control mice, the frequency ofCD69+ I-Ad/LACK+ cells increased from 10–15% at day 0to 60% at day 1 (Fig. 3A). In the same mice, the propor-tion of CD44high I-Ad/LACK+ cells increased from 15–20%at day 0 to 70% at day 3 (Fig. 3B). Surprisingly, anti-CD86 mAb had only a small effect on the proportion ofI-Ad/LACK+ exhibiting an activated phenotype (Fig. 3Aand B). For example, the frequency of CD69+ I-Ad/LACK+

cells at day 1 was 54% in mice treated with anti-CD86mAb as compared to 60% in control mice. Similarly, theproportion of CD44high I-Ad/LACK+ cells at day 3 was57% in mice treated with anti-CD86 mAb as comparedto 70% in control animals. Thus, while treatment withanti-CD86 mAb had only a marginal effect on the surfacephenotype of I-Ad/LACK+ CD4+ T cells, it significantlyreduced their number in the draining LN and their fre-quency among CD4+ T cells.

2.3 Anti-CD86 mAb do not promote breadth inthe T cell repertoire

We have recently shown that LACK-specific T cellsexpressing high affinity TCR are rapidly selected in16.2 g -transgenic B10.D2 mice following infection with


Fig. 2. Frequency, number and size of I-Ad/LACK+ CD4+ cells in mice treated with anti-CD86 mAb. 16.2 g -transgenic BALB/c micewere treated with either anti-CD86 mAb (filled bars) or IgG2a isotype control (empty bars) and infected with L. major as describedin Fig. 1. At the indicated times after infection, LN CD4+ cells were purified and stained with I-Ad/LACK multimers and anti-CD4mAb. Cells were analyzed by flow cytometry after gating on CD4+ live cells. Data show (A) frequencies of I-Ad/LACK+ amongCD4+ cells, (B) numbers of I-Ad/LACK+ CD4+ cells per LN, (C) frequencies of I-Ad/LACK+ cells exhibiting a large size as determinedby FSC/SSC analysis.

L. major [24]. Interestingly, this selection process isimpaired in BALB/c animals that mainly exhibit low affin-ity T cells even several weeks after infection [24].Because of this latter result, and because T cells arebelieved to integrate signals mediated through both theTCR and co-stimulatory molecules, it was possible thatthe blockade of CD86 in 16.2 g -transgenic BALB/c micewould prevent the expansion of T cells expressing lowavidity TCR. To test this hypothesis, we have comparedthe TCR § chains expressed by LACK-specific T cells inmice treated with anti-CD86 mAb and in control animals.Because LACK-specific T cells preferentially express aV § 8 TCR chain in normal BALB/c mice [26, 27], wedecided to focus on T cells that had rearranged the AV8gene segment. 16.2 g -transgenic BALB/c mice weretreated with either anti-CD86 mAb or IgG2a isotype con-trol, infected with L. major, and LN CD4+ T cells werepurified 3 days later. Cells were stained with I-Ad/LACKmultimers, I-Ad/LACK+ and I-Ad/LACK– cells were sortedby flow cytometry, and the amounts of TCR AV8 tran-scripts in the sorted populations were measured usingreal time quantitative PCR and normalized to the amountof TCR § chain constant region (AC) transcripts. In micetreated with anti-CD86 mAb as well as in control mice,the ratios of AV8 chain copies/AC copies were six- totenfold higher in I-Ad/LACK+ as compared to I-Ad/LACK-

cells (Fig. 4A). Most importantly, the ratios of AV8 chaincopies/AC copies in I-Ad/LACK+ cells were similar inmice treated with anti-CD86 mAb and in control animals,further suggesting that CD86 blockade did not preventthe selective expansion of cells which had rearrangedthe AV8 gene segment (Fig. 4A). To further investigatethe effect of CD86 blockade on LACK-specific T cells,we have analyzed the CDR3 size distribution of rear-

ranged TCR AV8 chain in I-Ad/LACK+ and in I-Ad/LACK–

sorted cells. As expected, the analysis of I-Ad/LACK–

cells in mice treated with anti-CD86 mAb and in controlanimals revealed a Gaussian-like distribution of the AV8-AC rearranged TCR. In striking contrast, in both groupsof mice, the Immunoscope analysis of AV3-AC run-offproducts in I-Ad/LACK+ cells showed a single peak corre-sponding to a CDR3 region of eight amino acids(Fig. 4B). Bacterial cloning and sequencing of the AV8-AC PCR products revealed the preferential usage of theAV8-J8 gene segments with a preferred eight-aminoacid-long CDR3 sequence (SEDMGYKL). Interestingly,we had previously described a LACK-specific T cellhybridoma, LMR4.11, which expressed the same TCR §chain [24]. This hybridoma was generated from the LN of16.2 g -transgenic BALB/c mice and was eventually dem-onstrated to express a low avidity TCR [24]. Thus, takentogether, our results show that the infection of 16.2 g -transgenic BALB/c mice with L. major results in theclonal expansion of low avidity LACK-specific T cellswhich have rearranged the AV8 gene segment andexpress the canonical SEDMGYKL CDR3 sequence.Furthermore, the expansion of these cells was not selec-tively prevented in mice that had been treated with anti-CD86 mAb.

We have then compared the apparent KD values formultimer binding to I-Ad/LACK+ cells in mice treated withanti-CD86 mAb and in control animals. To this aim,16.2 g -transgenic BALB/c mice were treated with eitheranti-CD86 mAb or IgG2a isotype control and infected withL. major. Control 16.2 g -transgenic B10.D2 mice were notinjected with antibodies and were infected at the sametime. LN CD4+ T cells were purified 3 and 21 days after


Fig. 3. Surface phenotype of I-Ad/LACK+ CD4+ cells in micetreated with anti-CD86 mAb. 16.2 g -transgenic BALB/c micewere treated with either anti-CD86 mAb or IgG2a isotype con-trol, and infected with L. major as described in Fig. 1. (A, B)LN CD4+ cells were purified at the indicated times after infec-tion, stained with I-Ad/LACK multimers, anti-CD4 mAb andeither (A) anti-CD69 or (B) anti-CD44 mAb. Typical flow cyto-metry profiles are shown after gating on live CD4+ T cells.

infection and stained with different concentrations of I-Ad/LACK multimers under equilibrium staining conditionsto measure the apparent KD values for multimer bindingto I-Ad/LACK+ cells. In agreement with our previousresults [24], I-Ad/LACK+ cells from untreated B10.D2mice exhibited a lower apparent KD value than thosefrom BALB/c mice treated with IgG2a isotype control(6.5±0.4 versus 17±5 nM, Fig. 5). In contrast, the appar-ent KD values for multimer binding to I-Ad/LACK+ cellswere similar in BALB/c mice treated with anti-CD86 mAband in those treated with IgG2a isotype control (19±2 ver-sus 17±5 nM). Similar results were obtained when micewere analyzed 3 days after infection. Thus, at day 3 postinfection, the apparent KD values for multimer binding toI-Ad/LACK+ cells were 18±2 nM in control BALB/c mice,16±2.4 nM in BALB/c mice treated with anti-CD86 mAb,and 10±0.2 nM in untreated B10.D2 animals (Fig. 5).Thus, while anti-CD86 mAb partially prevented thedevelopment of a Th2 response in 16.2 g transgenicmice, CD86 blockade did not prevent the clonal expan-sion of low avidity LACK-specific T cells which normallyoccurs in 16.2 g -transgenic BALB/c mice infected by L.major.

2.4 Blocking costimulatory signals throughCD27, OX40 and 4–1BB limits the expansionof parasite-specific T cells in vivo

The residual expansion of I-Ad/LACK+ cells observed inmice injected with anti-CD86 mAb suggested that othercostimulatory molecules could compensate for CD86 inthese animals. To test this hypothesis, 16.2 g -transgenicBALB/c mice were infected with L. major and injected atthe time of infection with either anti-CD86 mAb or block-ing antibodies against the ligands of CD27, CD30,4–1BB and OX40. Draining LN were harvested 3 dayslater and CD4+ T cells were stained with I-Ad/LACK multi-mers, anti-CD4 mAb and anti-CD69 or anti-CD44 mAb.In contrast to anti-CD86 mAb, no other antibody causeda decrease in the frequency of I-Ad/LACK+ cells in thedraining LN (Table 1). However, mice treated with anti-CD86, anti-OX40L, anti-CD27L and anti-4–1BBL mAbexhibited lower numbers of I-Ad/LACK+ cells/LN thanthose injected with either anti-CD30L or control mAb.Thus, in a representative experiment, the number of I-Ad/LACK+ cells/LN was reduced by twofold in mice injectedwith either anti-CD27L or anti-4-IBBL mAb(p X 0.001) and by 50% in animals injected with anti-OX40L mAb (p X 0.009). Despite these latter results, thefrequency of CD69+ among I-Ad/LACK+ cells was similarin all groups of mice. Furthermore, the proportion ofCD44high I-Ad/LACK+ cells was only slightly reduced inmice injected with anti-OX40L, anti-CD27L and anti-4–1BBL mAb as compared to those injected with eitheranti-CD30L or control mAb. Lastly, CD44high I-Ad/LACK+

cells from all mice exhibited comparable mean fluores-cent intensity (mfi) upon multimer staining suggestingthat they expressed TCR of similar avidity. Thus, asobserved with anti-CD86 mAb, the injection of anti-bodies against OX40L, CD27L and 4–1BBL caused areduction in the total number of I-Ad/LACK+ cells withoutpreventing the expansion of low avidity T cells.

3 Discussion

Although the injection of anti-CD86 mAb intoLeishmania-infected mice caused a reduction in the fre-quency and number of I-Ad/LACK+ CD4+ T cells, neitherthe activation nor the expansion of these cells wascompletely prevented. Furthermore, anti-CD86 mAb didnot prevent I-Ad/LACK+ cells to up-regulate the earlyactivation marker CD69 and to express high surface lev-els of CD44. Thus, the proportion of I-Ad/LACK+ cellsexpressing high levels of CD44 was similar in micetreated with anti-CD86 mAb and in control animals. It ispossible that the residual response which was observedin mice treated with anti-CD86 mAb was simply theresult of incomplete blockade of CD86. Although we


Fig. 4. Repertoire analysis of I-Ad/LACK+ in mice treated with anti-CD86 mAb. 16.2 g -transgenic BALB/c mice were treated witheither anti-CD86 mAb or IgG2a isotype control, and infected with L. major as described in Fig. 1. CD4+ LN cells were purified3 days after infection. Cells were stained with anti-CD4 mAb and I-Ad/LACK multimers, and I-Ad/LACK+ and I-Ad/LACK- CD4+

cells were sorted by flow cytometry. (A) The relative amounts of TCR AV8 transcripts in I-Ad/LACK+ cells and I-Ad/LACK– cellswere measured by real time quantitative RT-PCR and normalized to total TCR § chain (AC) transcripts. Data are expressed asmean ± SD of the ratio of AV8 transcripts over AC transcripts from three independent experiments. (B) The CDR3 size distributionof rearranged AV8-AC gene segments was analyzed in I-Ad/LACK- (left panels) and in I-Ad/LACK+ (right panel) CD4+ T cells. cDNAwere synthesized and PCR was performed using AV8- and AC-specific primers, followed by a run-off reaction with a nested fluo-rescent AC-specific primer. The run-off products were size fractionated in an automated DNA sequencer, and the CDR3 size dis-tribution was analyzed using the Immunoscope software. (C) The AV8-AC PCR products corresponding to the single peaks indi-cated by the arrows were cloned into the pGEM-T vector. Clones (15: 8 for mice treated with IgG2a isotype control, and 7 formice treated with anti-CD86 mAb) were picked randomly and the nucleotide sequence of their insert was determined. Resultsshow the deduced amino acid sequences of the CDR3 region of these clones, and the number of clones exhibiting the samesequence.

Fig. 5. Scatchard analysis of I-Ad/LACK multimer binding to I-Ad/LACK+ T cells. 16.2 g -transgenic BALB/c mice were treated witheither anti-CD86 mAb (empty diamonds) or IgG2a isotype control (empty circles) as described in Fig. 1. 16.2 g -transgenic B10.D2mice were left untreated (filled squares). Mice were infected with L. major, and analyzed 3 days (left panel) and 21 days (rightpanel) later. LN CD4+ T cells were purified and stained with anti-CD4 mAb and I-Ad/LACK multimers over a range of concentra-tions. Apparent KD values were derived from Scatchard plots of bound multimer/free multimer versus bound multimer. Data showthe arithmetic mean ± SD for three individual mice in each group.


Table 1. Frequency, number and phenotype of I-Ad/LACK+ cells in mice treated with blocking antibodies to costimulatoryligandsa)

Treatment Infection % I-Ad/LACK+ I-Ad/LACK+

(nb./LN × 10–4)% CD69+ % CD44high % Large cells mfi

None No 0.6 ± 0.3 0.6 ± 0.1 18 ± 2 14 ± 2 5 ± 1 48 ± 2

IgG2a control Yes 2.8 ± 0.4 8.0 ± 0.2 35 ± 7 49 ± 1 58 ± 2 50 ± 4

Anti-CD86 Yes 1.6 ± 0.2* 2.6 ± 0.5** 30 ± 4 40 ± 2 40 ± 4 52 ± 2

Anti-OX40L Yes 2.9 ± 0.3 6.0 ± 0.2*** 36 ± 1 33 ± 4 45 ± 4 50 ± 2

Anti-CD27L Yes 2.1 ± 1.4 3.0 ± 0.2** 36 ± 4 33 ± 4 48 ± 2 49 ± 6

Anti-CD30L Yes 2.3 ± 0.5 9.0 ± 0.3 32 ± 8 43 ± 4 60 ± 3 53 ± 8

Anti-4-IBBL Yes 2.6 ± 0.5 4.0 ± 0.6** 33 ± 4 35 ± 2 56 ± 1 51 ±1a) 16.2 g -transgenic BALB/c mice were infected or not with 2×106 stationary phase promastigotes and injected i.p. at the same

time and 24 h later with 150 ? g of the indicated antibodies. Mice were killed 3 days after infection and CD4+ LN cells werestained with I-Ad/LACK multimers, anti-CD4 and anti-CD69 or anti-CD44 mAb. Cells were analyzed by flow cytometry aftergating on live lymphocytes. Data show the arithmetic mean ± SD of three individual mice. *p X 0.001, compared with thefrequency of I-Ad/LACK+ cells in mice treated with IgG2a control. **p X 0.001, compared with the number of I-Ad/LACK+ cells inmice treated with IgG2a control. ***p X 0.009, compared with the number of I-Ad/LACK+ cells in mice treated with IgG2a control.

cannot rule out this hypothesis, the residual responsethat we have observed in this study is in agreement withprevious reports which have shown that CD86 may bedispensable for T cell activation [13]. Our results alsosuggest that other costimulatory molecules may com-pensate for the blockade of CD86 in animals treated withanti-CD86 mAb. In agreement with this hypothesis, theincrease in the number of I-Ad/LACK+ cells whichoccurred in the draining LN of 16.2 g -transgenic mice3 days after infection was partially inhibited by anti-OX40L, anti-CD27L or anti-4–1BBL mAb, but not by anti-CD30L mAb. The fact that anti-OX40L antibodies par-tially prevented the expansion of I-Ad/LACK+ cells inBALB/c-derived 16.2 g -transgenic mice is in agreementwith previous studies which have shown that the interac-tions between OX40 and its ligand play a critical role inthe development of a counterprotective Th2 response ingenetically susceptible BALB/c mice [28]. Furthermore,our results suggest that the ligands of two other co-stimulatory molecules, CD27L and 4–1BBL play a role inthe expansion of parasite-specific CD4+ T cells. How-ever, blockade of these molecules did not induce thedevelopment of a protective T cell response as observedin the case of CD86 or CD40L blockade [28]. Thus, sev-eral costimulatory molecules seem to cooperate to pro-mote the expansion of parasite-specific CD4+ T cells inmice infected with L. major.

According to a widely accepted model, the strength ofsignal through the TCR and co-stimulatory molecules iscritical in determining the outcome of TCR engagement[29–31]. This was demonstrated in experiments in which

naive T cells from TCR-transgenic mice were stimulatedin vitro with different doses of antigen and APC express-ing or not costimulatory molecules. Likewise, severalstudies have shown that blockade of costimulatory mol-ecules in vivo could bias T cell responses towards Th1 orTh2. For example, in another study as well as this one,the administration of anti-CD86 mAb to L. major-infecteddown-regulated Th2 responses and promoted the devel-opment of Th1 cells [22]. Likewise, anti-CD86 mAbblocked the early production of IL-4 that was induced byL. major in the LN of wt, but not CD28-deficient, BALB/cmice [25]. In agreement with previous studies, we haveshown here that blockade of CD86 in 16.2 g -transgenicBALB/c mice resulted in the development of a healingphenotype which was associated with reduced numbersand frequencies of IL-4-secreting cells. However, treat-ment with anti-CD86 mAb did not selectively prevent theactivation and the expansion of I-Ad/LACK+ T cellsexpressing low avidity TCR. Thus, our results do notsupport a model in which costimulatory molecules wouldpromote the development of Th2 responses by recruitingadditional T cell subsets.

According to the two-signal theory of T lymphocyte acti-vation [32] both TCR engagement and other costimula-tory signals from APC are essential for T cell activationand proliferation. Although this theory was supported bya large number of in vitro and in vivo experiments[33–36], a recent study has proposed a new model of Tcell activation in which T cells convert independentlyreceived signals into linear additive effects on divisiontimes which, in turn, amplify T cell numbers exponen-


tially [37]. According to this new model, molecules suchas CD28 would act independently of signals deliveredthrough the TCR by reducing the average time requiredby T cells to traverse the first division. In contrast, co-stimulatory molecules would have little effect on the sub-sequent division rate. Interestingly, blockade of CD86 in16.2 g -transgenic mice did not alter the proportion ofI-Ad/LACK+ cells which up-regulated CD69 or CD44upon infection with L. major. However, injection of anti-CD86 mAb resulted in a significant reduction in the num-ber of I-Ad/LACK+ cells and in the proportion of thesecells which were progressing through the cell cycle.Blockade of other costimulatory molecules such asOX40L, CD27L and 4–1BBL had qualitatively similar, butquantitatively less dramatic, effects. Thus, our results arein agreement with a stochastic model in which the prob-ability for a given cell to be activated, and to eventuallyenter the cell cycle, would be solely determined by thestrength of signal through the TCR but not by costimula-tory molecules. Although the engagement of costimula-tory molecules do not seem to qualitatively influence theT cell repertoire recruited during the course of theimmune response, these signals may serve to increaseimmune responses by amplifying T cell numbers.

4 Materials and methods

4.1 Mice and parasites

BALB/c mice were purchased from Harlan, GB. 16.2 g BALB/c and B10.D2-transgenic mice have been described [24].Animals were sex-matched and used between 7 and10 weeks of age. L. major promastigotes (World HealthOrganization strain WHOM/IR/-/173) were grown in M199medium containing 20% FCS and used as described [27].Disease progression was assessed by weekly measuring ofthe swelling of the infected left footpad compared with theuninfected right footpad. Parasite numbers were measuredas described [38]. Briefly, cell suspensions were preparedfrom draining LN of infected mice and serial dilutions wereplated in 96-well plates on blood containing completeM199-agar. Cultures were scored 5 days later for the pres-ence of parasites and the absolute number of parasites perLN was calculated.

4.2 Reagents and antibodies

The following mAb were purchased from Becton DickinsonSA (Le Pont de Claix, France): GK1.5, anti-CD4; 53–6.7,anti-CD8; KT4.1, anti-V g 4; H1.2F3, anti-CD69; M1/70, anti-CD11b; RA3–6B2, anti-B220; 2G9, anti-I-Ad/I-Ed; anti-CD28(37.51); GL1, CD86; IM7, anti-CD44; XMG1.2, anti-IFN- + ;11B11, anti-IL-4. Other mAb were prepared as described[28]: RM134L, anti-OX40L; FR70, anti-CD27L; RM153, anti-

CD30L; TKS-1, anti-4–1BBL. Soluble I-Ad/LACK dimerswere produced and used as described [24]. IgG2a isotypecontrol, PMA, ionomycin, brefeldin A were purchased fromSigma SARL (St. Quentin Fallavier, France). SLA were pro-duced as described [27].

4.3 Purification of CD4+ cells

CD4+ cells were purified by negative depletion of CD8+,B220+, CD11b+, and I-Ad+ cells using Dynabeads coatedwith sheep anti-rat total IgG (Dynal S.A., Compiegne,France) as described [39]. The resulting populations werepure to more than 95% as assessed by flow cytometry.

4.4 Cyokine assay

LN cells were purified and 8×105 cells were incubated with50 ? g / ml of SLA in the wells of round-bottom 96-well platesin triplicates. Cellular supernatants were harvested after72 h and IL-4, IL-5 and IFN- + contents were measured byELISA as described [39].

4.5 Flow cytometry analysis

Staining with I-Ad/LACK multimers was performed asdescribed [24]. Briefly, 1.4 ? l of either Alexa 488-coupled orbiotin-coupled protein A (Molecular Probes Inc., Eugene) atthe concentration of 0.5 mg/ml in PBS was incubated for30 min at 20°C with 8 ? g of I-Ad/LACK dimers. Purified CD4+

cells (106/sample) were incubated with protein A-boundI-Ad/LACK at the final concentration of 60 ? g/ml for 1 h onice in PBS supplemented with 0.5% bovine serum albumin(PBS/BSA). Cells were washed twice and incubated withanti-CD4 and anti-CD44 or anti-CD69 mAb. Lymphocyteswere gated by forward and side-scatter analysis and ana-lyzed on a FACScan flow cytometer (Becton Dickinson SA,Le Pont de Claix, France). Scatchard analysis for KD deter-mination was performed as described [40] with a differencethat cells were stained for 2 h on ice.

For intracellular cytokine staining, 106 CD4+ cells wereseeded in round-bottom 96-well plates and incubated for2 h at 37°C with PMA (50 ng/ml) and ionomycin (1 ? g/ml).Brefeldin A was added at the final concentration of 10 ? g/mland incubated at 37°C for additional 2 h. Cells were washedtwice with PBS/BSA supplemented with 10 ? g/ml of brefel-din A, and stained overnight at 4°C with 60 ? g/ml of I-Ad/LACK multimers in brefeldin A-containing PBS/BSA. Cellswere fixed in 2% paraformaldehyde for 20 min at room tem-perature, permeabilized and stained with anti-IL-4 and anti-IFN- + mAb as described [41].


4.6 T cell repertoire analysis

RNA was prepared from sorted cell populations using theTrizol reagent (Gibco-BRL, Gaithersburg, MD), and cDNAwas synthesized using M-MLV reverse transcriptase (Gibco-BRL). To measure the ratio between AV8 and AC transcripts,real time quantitative PCR was performed on cDNA usingthe ABI Prism 5700 sequence detector (Perkin Elmer Bio-systems). PCR were performed using AV8 and AC primersand TaqMan Universal PCR master mix in a final volume of25 ? l according to the manufacturer’s conditions. AV8-specific primers have been described [42]. The sequence forthe AC primer is 5’-CTG TCC TGA GAC CGA GGA TCT TT-3’. The AC specific probe 5’-ATC CAG AAC CCA GAA CCTGCT GTG TAC CAG T-3’ was labeled with the reporter dyeFAM (6-carboxyfluoresceine, covalently linked to the 5’-endof the probe) and a quencher TAMRA (6-carboxyltetramethyl-rhodamine, attached at the 3’-end via alinker arm). The cDNA encoding the TCR § chain of theLACK-specific LMR 4.1 hybridomas was cloned into pGEMTand used as an external standard. To measure the amountsof transcripts containing the TCR C § region, PCR was per-formed on cDNA using AC5 (5’-CTG CCT GTT CAC CGACTT TGA-3’) and AC3 (5’-TCC ATG GTT TTC GGC ACA TT-3’) primers. The reaction was performed using the SybGreenuniversal PCR master mix in a final volume of 25 ? l. Data areexpressed as the ratio between the AV-AC rearrangementover TCR AC mRNA copies. To analyze the size of the CDR3regions, PCR was performed on cDNA using AV8-specificprimers and analyzed on a 2% agarose gel stained with ethi-dium bromide as described [43]. AV8-ACa PCR product(5 ? l) was incubated for 40 min at 37°C with 0.5 U of ShrimpAlkaline Phosphatase and 10 U of Exonuclease I (AmershamPharmacia Biotech Europe GbmH) in 10 ? l. Sequencingreactions were performed using ACb internal primer and theABI PRISM Dye Terminator Cycle Sequencing Ready Reac-tion Kit (Perkin Elmer Biosystems, Foster City, CA).Sequences were run on a 373A DNA sequencer (PerkinElmer Biosystems) and the Immunoscope analysis was per-formed as described [43].

4.7 Statistics

Statistical significance was determined using a two-tailedStudent’s t-test.

Acknowledgements: This study was funded by the Minis-tere de l’Education Nationale, de la Recherche et de l’En-seignement Superieur and the EU (contract QLGI-CT-1999–00050). M.M. and L.M. were supported by fellowshipsfrom the Association Italiana for Cancer Research (AIRC)and the Ligue Nationale Contre le Cancer (LNCC), respec-tively.

References

1 Bluestone, J., Khattri, R. and van Seventer, G., Accessory mol-ecules. In Paul, W. E. (Ed.) Fundamental immunology, 4th edn,Lippincott-Raven, Philadelphia 1999, pp 449–478.

2 Inaba, K., Witmer-Pack, M., Inaba, M. et al., The tissue distribu-tion of the B7–2 costimulator in mice: abundant expression ondendritic cells in situ and during maturation in vitro. J. Exp. Med.1994. 180: 1849–1860.

3 Azuma, M., Ito, D., Yagita, H., Okumura, K., Phillips, J. H.,Lanier, L. L. and Somoza, C., B70 antigen is a second ligand forCTLA-4 and CD28. Nature 1993. 366: 76–79.

4 Freedman, A. S., Freeman, G. J., Rhynhart, K. and Nadler, L.M., Selective induction of B7/BB-1 on IFN- + stimulated mono-cytes: a potential mechanism for amplification of T cell activationthrough the CD28 pathway. Cell Immunol. 1991. 137: 429–437.

5 Larsen, C. P., Ritchie, S. C., Hendrix, R., Linsley, P. S., Hath-cock, K. S., Hodes, R. J., Lowry, R. P. and Pearson, T. C., Reg-ulation of immunostimulatory function and costimulatory mole-cule (B7–1 and B7–2) expression on murine dendritic cells. J.Immunol, 1994. 152: 5208–5219.

6 Lenschow, D. J., Sperling, A. I., Cooke, M. P., Freeman, G.,Rhee, L., Decker, D. C., Gray, G., Nadler, L. M., Goodnow, C.C. and Bluestone, J. A., Differential up-regulation of the B7–1and B7–2 costimulatory molecules after Ig receptor engagementby antigen. J. Immunol. 1994. 153: 1990–1997.

7 Boussiotis, V. A., Freeman, G. J., Gribben, J. G., Daley, J.,Gray, G. and Nadler, L. M., Activated human B lymphocytesexpress three CTLA-4 counterreceptors that costimulate T cellactivation. Proc. Natl. Acad. Sci. USA 1993. 90: 11059–11063.

8 Freeman, G. J., Borriello, F., Hodes, R. J., Reiser, H., Hath-cock, K. S., Laszlo, G., McKnight, A. J., Kim, J., Du, L., Lom-bard, D. B. and et al., Uncovering of functional alternative CTLA-4 counter-receptor in B7-deficient mice. Science 1993. 262:907–911.

9 Freeman, G. J., Gribben, J. G., Boussiotis, V. A., Ng, J. W.,Restivo, V. A., Jr., Lombard, L. A., Gray, G. S. and Nadler, L. M.,Cloning of B7–2: a CTLA-4 counter-receptor that costimulateshuman T cell proliferation. Science 1993. 262: 909–911.

10 Lanier, L. L., O’Fallon, S., Somoza, C., Phillips, J. H., Linsley, P.S., Okumura, K., Ito, D. and Azuma, M., CD80 (B7) and CD86(B70) provide similar costimulatory signals for T cell proliferation,cytokine production, and generation of CTL. J. Immunol. 1995.154: 97–105.

11 Levine, B. L., Ueda, Y., Craighead, N., Huang, M. L. and June,C. H., CD28 ligands CD80 (B7–1) and CD86 (B7–2) induce long-term autocrine growth of CD4+ T cells and induce similar pat-terns of cytokine secretion in vitro. Int. Immunol. 1995. 7:891–904.

12 Natesan, M., Razi-Wolf, Z. and Reiser, H., Costimulation of IL-4production by murine B7–1 and B7–2 molecules. J. Immunol.1996. 156: 2783–2791.

13 Schweitzer, A. N., Borriello, F., Wong, R. C., Abbas, A. K. andSharpe, A. H., Role of costimulators in T cell differentiation: stud-ies using antigen-presenting cells lacking expression of CD80 orCD86. J. Immunol. 1997. 158: 2713–2722.

14 Schweitzer, A. N. and Sharpe, A. H., Studies using antigen-presenting cells lacking expression of both B7–1 (CD80) andB7–2 (CD86) show distinct requirements for B7 molecules duringpriming versus restimulation of Th2 but not Th1 cytokine produc-tion. J. Immunol. 1998. 161: 2762–2771.


15 Lenschow, D. J., Ho, S. C., Sattar, H., Rhee, L., Gray, G.,Nabavi, N., Herold, K. C. and Bluestone, J. A., Differentialeffects of anti-B7–1 and anti-B7–2 monoclonal antibody treat-ment on the development of diabetes in the nonobese diabeticmouse. J. Exp. Med. 1995. 181: 1145–1155.

16 Lenschow, D. J., Zeng, Y., Hathcock, K. S., Zuckerman, L. A.,Freeman, G., Thistlethwaite, J. R., Gray, G. S., Hodes, R. J.and Bluestone, J. A., Inhibition of transplant rejection followingtreatment with anti-B7–2 and anti-B7–1 antibodies. Transplanta-tion 1995. 60: 1171–1178.

17 Freeman, G. J., Boussiotis, V. A., Anumanthan, A., Bernstein,G. M., Ke, X. Y., Rennert, P. D., Gray, G. S., Gribben, J. G. andNadler, L. M., B7–1 and B7–2 do not deliver identical costimula-tory signals, since B7–2 but not B7–1 preferentially costimulatesthe initial production of IL-4. Immunity 1995. 2: 523–532.

18 Kuchroo, V. K., Das, M. P., Brown, J. A., Ranger, A. M., Zamvil,S. S., Sobel, R. A., Weiner, H. L., Nabavi, N. and Glimcher, L.H., B7–1 and B7–2 costimulatory molecules activate differentiallythe Th1/Th2 developmental pathways: application to autoim-mune disease therapy. Cell 1995. 80: 707–718.

19 Racke, M. K., Scott, D. E., Quigley, L., Gray, G. S., Abe, R.,June, C. H. and Perrin, P. J., Distinct roles for B7–1 (CD-80) andB7–2 (CD-86) in the initiation of experimental allergic encephalo-myelitis. J. Clin. Invest. 1995. 96: 2195–2203.

20 Perrin, P. J., Scott, D., Quigley, L., Albert, P. S., Feder, O., Gray,G. S., Abe, R., June, C. H. and Racke, M. K., Role of B7:CD28/CTLA-4 in the induction of chronic relapsing experimental allergicencephalomyelitis. J. Immunol. 1995. 154: 1481–1490.

21 Perrin, P. J., Scott, D., Davis, T. A., Gray, G. S., Doggett, M. J.,Abe, R., June, C. H. and Racke, M. K., Opposing effects ofCTLA4-Ig and anti-CD80 (B7–1) plus anti-CD86 (B7–2) on experi-mental allergic encephalomyelitis. J. Neuroimmunol. 1996. 65:31–39.

22 Brown, J. A., Titus, R. G., Nabavi, N. and Glimcher, L. H.,Blockade of CD86 ameliorates Leishmania major infection bydown-regulating the Th2 response. J. Infect. Dis. 1996. 174:1303–1308.

23 Brown, J. A., Greenwald, R. J., Scott, S., Schweitzer, A. N.,Satoskar, A. R., Chung, C., Schopf, L. R., Woude Dv, D.,Sypek, J. P. and Sharpe, A. H., T helper differentiation in resis-tant and susceptible B7-deficient mice infected with Leishmaniamajor. Eur. J. Immunol. 2002. 32: 1764–1772.

24 Malherbe, L., Filippi, C., Julia, V., Foucras, G., Moro, M.,Appel, H., Wucherpfennig, K., Guery, J. C. and Glaichenhaus,N., Selective activation and expansion of high-affinity CD4+ Tcells in resistant mice upon infection with Leishmania major.Immunity 2000. 13: 771–782.

25 Elloso, M. M. and Scott, P., Expression and contribution of B7–1(CD80) and B7–2 (CD86) in the early immune response to Leish-mania major infection. J. Immunol. 1999. 162: 6708–6715.

26 Reiner, S. L., Wang, Z. E., Hatam, F., Scott, P. and Locksley, R.M., TH1 and TH2 cell antigen receptors in experimental leish-maniasis. Science 1993. 259: 1457–1460.

27 Mougneau, E., Altare, F., Wakil, A. E., Zheng, S., Copolla, T.,Wang, Z.-E., Waldmann, R., Locksley, R. and Glaichenhaus,N., Expression cloning of a Leishmania major protective T cellantigen. Science 1995. 268: 563–566.

28 Akiba, H., Miyahira, Y., Atsuta, M., Takeda, K., Nohara, C.,Futagawa, T., Matsuda, H., Aoki, T., Yagita, H. and Okumura,K., Critical contribution of OX40 ligand to T helper cell type 2 dif-ferentiation in experimental leishmaniasis. J. Exp. Med. 2000.191: 375–380.

29 Bluestone, J. A., New perspectives of CD28-B7-mediated T cellcostimulation. Immunity 1995. 2: 555–559.

30 Tao, X., Constant, S., Jorritsma, P. and Bottomly, K., Strengthof TCR signal determines the costimulatory requirements for Th1and Th2 CD4+ T cell differentiation. J. Immunol. 1997. 159:5956–5963.

31 Manickasingham, S. P., Anderton, S. M., Burkhart, C. andWraith, D. C., Qualitative and quantitative effects of CD28/B7-mediated costimulation on naive T cells in vitro. J. Immunol.1998. 161: 3827–3835.

32 Lafferty, K. J. and Woolnough, J., The origin and mechanism ofthe allograft reaction. Immunol. Rev. 1977. 35: 231–266.

33 Chambers, C. A. and Allison, J. P., Co-stimulation in T cellresponses. Curr. Opin. Immunol. 1997. 9: 396–404.

34 Sperling, A. I., Auger, J. A., Ehst, B. D., Rulifson, I. C., Thomp-son, C. B. and Bluestone, J. A., CD28/B7 interactions deliver aunique signal to naive T cells that regulates cell survival but notearly proliferation. J. Immunol. 1996. 157: 3909–3917.

35 Shahinian, A., Pfeffer, K., Lee, K. P., Kundig, T. M., Kishihara,K., Wakeham, A., Kawai, K., Ohashi, P. S., Thompson, C. B.and Mak, T. W., Differential T cell costimulatory requirements inCD28-deficient mice. Science 1993. 261: 609–612.

36 Green, J. M. and Thompson, C. B., Modulation of T cell prolifer-ative response by accessory cell interactions. Immunol. Res.1994. 13: 234–243.

37 Gett, A. V. and Hodgkin, P. D., A cellular calculus for signal inte-gration by T cells. Nat. Immunol. 2000. 1: 239–244.

38 Titus, R. G., Ceredig, R., Cerrottini, J. C. and Louis, J. A., Ther-apeutic effect of anti-L3T4 monoclonal GK1.5 on cutaneousleishmaniasis in genetically susceptible BALB/C mice. J. Immu-nol. 1985. 135: 2108–2114.

39 Julia, V., Rassoulzadegan, M. and Glaichenhaus, N., Resis-tance to Leishmania major induced by tolerance to a single anti-gen. Science 1996. 274: 421–423.

40 Savage, P. A., Boniface, J. J. and Davis, M. M., A kinetic basisfor T cell receptor repertoire selection during an immuneresponse. Immunity 1999. 10: 485–492.

41 Martin-Fontecha, A., Moro, M., Crosti, M. C., Veglia, F., Caso-rati, G. and Dellabona, P., Vaccination with mouse mammaryadenocarcinoma cells coexpressing B7–1 (CD80) and B7–2(CD86) discloses the dominant effect of B7–1 in the induction ofantitumor immunity. J. Immunol. 2000. 164: 698–704.

42 Casanova, J. L., Romero, P., Widmann, C., Kourilsky, P. andMaryanski, J. L., T cell receptor genes in a series of class I majorhistocompatibility complex-restricted cytotoxic T lymphocyteclones specific for a Plasmodium berghei nonapeptide: implica-tions for T cell allelic exclusion and antigen-specific repertoire. J.Exp. Med. 1991. 174: 1371–1383.

43 Pannetier, C., Cochet, M., Darche, S., Casrouge, A., Zoller, M.and Kourilsky, P., The sizes of the CDR3 hypervariable regionsof the murine T cell receptor g chains vary as a function of therecombined germ-line segments. Proc. Natl. Acad. Sci. USA1993. 90: 4319–4323.

Correspondence: Nicolas Glaichenhaus, Institut de Phar-macologie Moleculaire et Cellulaire, 660 Route des Lucioles,F-06560 Valbonne, FranceFax: +33-4-9395-7708e-mail: glaichenhaus — ipmc.cnrs.fr


Blockade of CD86 in BALB/c mice infected withLeishmania major does not prevent the expansion of low avidity T cells

Documents