CTLA-4 engagement and regulatory CD4+CD25+ T cells independently control CD8+-mediated responses under costimulation blockade

Blockade of costimulatory signals is a promising therapeutic target to prevent allograft rejection. In this study, we sought to characterize to what extent CTLA-4 engagement contributes to the development of transplantation tolerance under the cover of CD40/CD40L and CD28/CD86 blockade. In vitro, we found that inhibition of the primary alloresponse and induction of alloantigen hyporesponsiveness by costimulation blockade was abrogated by anti-CTLA-4 mAb. In addition, regulatory CD4(+)CD25(+) T cells (T(REG)) were confirmed to play a critical role in the induction of hyporesponsiveness by anti-CD40L and anti-CD86 mAb. Our data indicated that CTLA-4 engagement is not required for activation or suppressor function of T(REG). Instead, in the absence of either CTLA-4 signaling or T(REG), CD8(+) T cell division was enhanced, whereas the inhibition of CD4(+) T cell division by costimulation blockade remained largely unaffected. In vivo, the administration of additional anti-CTLA-4 mAb abrogated anti-CD40L- and anti-CD86 mAb-induced cardiac allograft survival. Correspondingly, rejection was accompanied by enhanced allograft infiltration of CD8(+) cells. We conclude that CTLA-4 signaling and T(REG) independently cooperate in the inhibition of CD8(+) T cell expansion under costimulation blockade.     

Low CD86 expression in the nonobese diabetic mouse results in the impairment of both T cell activation and CTLA-4 up-regulation

The nonobese diabetic (NOD) mouse spontaneously develops autoimmune insulin-dependent diabetes mellitus and serves as a model for human type I diabetes. NOD spleen cells proliferate to a lesser extent than those from C57BL/6 and BALB/c mice in response to anti-CD3. To investigate the cause of this reduced T cell proliferation, costimulatory molecule expression was investigated. It was found that NOD macrophages, dendritic cells, and T cells, but not B cells, expressed lower basal levels of CD86, but not CD80, CD28, or CD40, compared with C57BL/6 and BALB/c. This low CD86 expression was not dependent on the MHC haplotype or on diabetes development since the NOD-related, diabetes-free mouse strains NON (H-2nb1) and NOR (H-2g7) exhibited similar low levels of CD86 expression and proliferation. Furthermore, following activation, the relative up-regulation of CTLA-4, as compared with CD28, was more pronounced on C57BL/6 and BALB/c T cells as shown by an increased CTLA-4/CD28 ratio. This activation-induced increase in the CTLA-4/CD28 ratio was markedly reduced on NOD T cells compared with the other two strains. The low CD86 expression in NOD mice may account for the reduced increase in both proliferation and the CTLA-4/CD28 ratio, since reducing CD86 expression in C57BL/6 and BALB/c cultures to NOD levels significantly reduces the proliferation and the CTLA-4/CD28 ratio. Therefore, we propose that a low level of CD86 expression in the NOD mouse contributes to a defective regulation of autoreactive T cells by preventing the full activation of T cells and therefore the up-regulation of CTLA-4.     

Blockade of T cell activation using a surface-linked single-chain antibody to CTLA-4 (CD152)

CTLA-4 (CD152) engagement can down-regulate T cell activation and promote the induction of immune tolerance. However, the strategy of attenuating T cell activation by engaging CTLA-4 has been limited by sharing of its natural ligands with the costimulatory protein CD28. In the present study, a CTLA-4-specific single-chain Ab (scFv) was developed and expressed on the cell surface to promote selective engagement of this regulatory molecule. Transfectants expressing anti-CTLA-4 scFv at their surface bound soluble CTLA-4 but not soluble CD28. Coexpression of anti-CTLA-4 scFv with anti-CD3epsilon and anti-CD28 scFvs on artificial APCs reduced the proliferation and IL-2 production by resting and preactivated bulk T cells as well as CD4+ and CD8+ T cell subsets. Importantly, expression of anti-CTLA-4 scFv on the same cell surface as the TCR ligand was essential for the inhibitory effects of CTLA-4-specific ligation. CTLA-4-mediated inhibition of tyrosine phosphorylation of components of the proximal TCR signaling apparatus was similarly dependent on coexpression of TCR and CTLA-4 ligands on the same surface. These findings support a predominant role for CTLA-4 function in the modification of the proximal TCR signal. Using T cells from DO11.10 and 2C TCR transgenic mice, negative regulatory effects of selective CTLA-4 ligation were also demonstrated during the stimulation of Ag-specific CD4+ and CD8+ T cells by MHC/peptide complexes. Together these studies demonstrate that selective ligation of CTLA-4 using a membrane-bound scFv results in attenuated T cell responses only when coengaged with the TCR during T cell/APC interaction and define an approach to harnessing the immunomodulatory potential of CTLA-4-specific ligation.     

Activated self-MHC-reactive T cells have the cytokine phenotype of Th3/T regulatory cell 1 T cells

In the present study, we show that human self-MHC-reactive (autoreactive) T cell clones are functionally distinct from Ag-specific T cell clones. Self-MHC-reactive T cells exhibited helper function for B cell Ig production when cultured with non-T cells alone, and they exhibit suppressor function when cultured with PWM- or rCD40 ligand (rCD40L)-activated non-T cells, whereas tetanus toxoid (TT)-specific clones exhibited only helper function in the presence of TT with or without PWM or rCD40L. Addition of neutralizing Abs to the cultures showed that the suppression was mediated by TGF-beta but not by IL-10 or IFN-gamma. The self-MHC-reactive clones also inhibited proliferation of primary CD4+ T cells and TT-specific T cell clones, but in this case the inhibition was mediated by both IL-10 and TGF-beta. In further studies, the interactions between self-MHC-reactive T cell clones and non-T cells that led to suppressor cytokine production have been explored. We found that prestimulation of non-T cells for 8 h with PWM or for 48 h for rCD40L results in non-T cells capable of inducing self-MHC-reactive T cell to produce high levels of TGF-beta and IL-10. In addition, these prestimulation times coincided with peak induction of HLA-DR and costimulatory B7 molecule (especially CD86) expression on B cells. Finally, addition of CTLA-4/Fc or blocking F(ab')2 anti-CTLA-4 mAb, plus optimally stimulated non-T cells, to cultures of self-MHC-reactive clones inhibited the induction of TGF-beta but not IL-10 or IFN-gamma production. In summary, these studies show that activated self-MHC-reactive T cells have the cytokine phenotype of Th3 or T regulatory cell 1 and thus may be important regulatory cells that mediate oral and peripheral tolerance and prevent the development of autoimmunity.     

CTLA-4 gene polymorphism at position 49 in exon 1 reduces the inhibitory function of CTLA-4 and contributes to the pathogenesis of Graves' disease

Activation of T cells requires at least two signals transduced by the Ag-specific TCR and a costimulatory ligand such as CD28. CTLA-4, expressed on activated T cells, binds to B7 present on APCs and functions as a negative regulator of T cell activation. Our laboratory previously reported the association of Graves' disease (GD) with a specific CTLA-4 gene polymorphism. In theory, reduced expression or function of CTLA-4 might augment autoimmunity. In the present study, we categorized autoimmune thyroid disease patients and normal controls (NC) by genotyping a CTLA-4 exon 1 polymorphism and investigated the function of CTLA-4 in all subjects. PBMCs and DNA were prepared from GD (n = 45), Hashimoto's thyroiditis (HT) (n = 18), and NC (n = 43). There were more GD patients with the G/G or A/G alleles (82.2% vs 65.1% in NC), and significantly fewer patients with the A/A allele (17.8% vs 34.9% in NC). In the presence of soluble blocking anti-human CTLA-4 mAb, T cell proliferation following incubation with allogeneic EBV-transformed B cells was augmented in a dose-dependent manner. Augmentation induced by CTLA-4 mAb was similar in GD and NC (GD, HT, NC = 156%, 164%, 175%, respectively). We related CTLA-4 polymorphism to mAb augmentation of T cell proliferation in each subgroup (GD, HT, NC). Although PBMC from individuals with the G/G alleles showed 132% augmentation, those with the A/A alleles showed 193% augmentation (p = 0.019). CTLA-4 polymorphism affects the inhibitory function of CTLA-4. The G allele is associated with reduced control of T cell proliferation and thus contributes to the pathogenesis of GD and presumably of other autoimmune diseases.     

Cutting edge: CTLA-4 (CD152) differentially regulates mitogen-activated protein kinases (extracellular signal-regulated kinase and c-Jun N-terminal kinase) in CD4+ T cells from receptor/ligand-deficient mice

Although CTLA-4 (CD152) has potent inhibitory effects on T cell function, the signaling events affected by this coreceptor remain to be fully defined. Mitogen-activated protein kinases (MAPK) extracellular signal-regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) act as crucial regulators of multiple aspects of cell function. Ab ligation studies have reported an inhibitory effect of CTLA-4 on TCR-induced ERK and JNK activation. In this study, we have re-examined the specificity of CTLA-4 inhibition of MAPKs by using natural ligand with ex vivo-purified CD4(+) T cells deficient in CD80 and CD86 (double knockout), or CTLA-4, CD80, and CD86 (triple knockout). Under these conditions, CTLA-4 ligation was found to up-regulate and sustain JNK activation, while inhibiting ERK activity. At the same time, JNK activation could not account for CTLA-4 induction of TGF-beta production. Our findings demonstrate that CTLA-4 cosignaling is more complex than previously appreciated, with an ability to differentially regulate members of the MAPK family in T cells.     

Induction of polyclonal CD8+ T cell activation and effector function by Pertussis toxin

Pertussis toxin (PTX) has pronounced adjuvant activity and strongly enhances innate and adaptive immune responses, including increased antibody production and Th1/Th2 cytokine production. Adjuvant effects of PTX on Th1 and Th2 cells are primarily mediated via CD80/86 costimulation via enhanced expression of these molecules by APCs. However, it has remained unresolved whether PTX modulates the expression of costimulatory and inhibitory molecules on CD4+ and CD8+ T cells. To address this question, we determined the expression kinetics of CD28, CTLA-4, and CD40L on spleen CD4+ and CD8+ T cells after incubation with PTX. The results show that PTX upregulated the expression of CD28 by CD8+ T cells, but not by CD4+ T cells. In contrast, the expression of CTLA-4 and CD40L was not substantially altered on CD4+ or CD8+ T cells. CD28 upregulation by CD8+ T cells was paralleled by upregulation of CD69 and the induction of IFN-γ, Granzyme B (GrB), and IL-17. CD8+ T cell activation and cytokine production could be substantially blocked with anti-CD80 and CD86 antibodies, consistent with CD28 mediated signaling. Treatment of highly purified CD8+ T cells with PTX resulted in upregulation of CD28 and CD69, and production of IFN-γ. Incubation with CD28 mAb further enhanced this effect, suggesting that PTX has direct effects on CD8+ T cells which are enhanced by CD80/86-mediated costimulation provided by APCs.     

Characterization of human inducible costimulator ligand expression and function

The inducible costimulator (ICOS) is the newest member of the CD28/CD152 receptor family involved in regulating T cell activation. We constructed a soluble-Ig fusion protein of the extracellular domain of human ICOS and used it as a probe to characterize expression patterns of the ICOS ligand (ICOSL). ICOSIg did not bind to CD80- or CD86-transfected Chinese hamster ovary cell lines, demonstrating that ICOSL is distinct from those ligands identified for CD28/CD152. ICOSIg showed selective binding to monocytic and B cell lines, whereas binding was undetectable on unstimulated monocytes and peripheral blood T and B cells. Expression of ICOSL was induced on monocytes after integrin-dependent plastic adhesion. Pretreatment of monocytes with mAb to the beta2-integrin subunit CD18 decreased adhesion and abolished ICOSL up-regulation but had no effect on CD80/86 (CD152 ligand (CD152L)) expression. Both ICOSL and CD152L were up-regulated on monocytes by IFN-gamma but by distinct signaling pathways. Unlike CD152L expression, ICOSL expression did not change when monocytes were differentiated into dendritic cells (DCs) or after DCs were induced to mature by LPS, TNF-alpha, or CD40 ligation. Addition of ICOSIg to allogeneic MLRs between DCs and T cells reduced T cell proliferative responses but did so less efficiently than CTLA4Ig (CD152Ig) did. Similarly, ICOSIg also blocked Ag-specific T cell proliferation to tetanus toxoid. Thus, ICOSL, like CD80/86, is expressed on activated monocytes and dendritic cells but is regulated differently and delivers distinct signals to T cells that can be specifically inhibited by ICOSIg.     

Essential role for both CD80 and CD86 costimulation, but not CD40 interactions, in allergen-induced Th2 cytokine production from asthmatic bronchial tissue: role for alphabeta, but not gammadelta, T cells

CD80 and CD86 interact with CD28 and deliver costimulatory signals required for T cell activation. We demonstrate that ex vivo allergen stimulation of bronchial biopsy tissue from mild atopic asthmatic, but not atopic nonasthmatic, subjects induced production of IL-5, IL-4, and IL-13. Explants from both study groups did not produce IFN-gamma, but secreted the chemokine RANTES without any overt stimulation. In addition to allergen, stimulation of asthmatic explants with mAbs to CD3 and TCR-alphabeta but not TCR-gammadelta induced IL-5 secretion. Allergen-induced IL-5 and IL-13 production by the asthmatic tissue was inhibited by anti-CD80 and, to a lesser extent, by anti-CD86 mAbs. In contrast, the production of these cytokines by PBMCs was not affected by mAbs to CD80, was inhibited by anti-CD86, and was strongly attenuated in the presence of both Abs. FACS analysis revealed that stimulated asthmatic bronchial tissue was comprised of CD4+ T cells that expressed surface CD28 (75. 3%) but little CTLA-4 (4.0%). Neutralizing mAbs to CD40 ligand had no effect on the cytokine levels produced by asthmatic tissue or PBMCs. Collectively, these findings suggest that allergen-specific alphabeta T cells are resident in asthmatic bronchial tissue and demonstrate that costimulation by both CD80 and CD86 is essential for allergen-induced cytokine production. In contrast, CD86 appears to be the principal costimulatory molecule required in PBMC responses. Attenuation of type 2 alphabeta T cell responses in the bronchial mucosa by blocking these costimulatory molecules may be of therapeutic potential in asthma.     

NK cell triggering by the human costimulatory molecules CD80 and CD86.

NK cell-mediated effector functions are regulated by a delicate balance between positive and negative signals. Receptors transmitting negative signals upon engagement with target cell MHC class I molecules have been characterized in detail in recent years. In contrast, less information is available about receptor-ligand interactions involved in the transmission of positive or "triggering" signals to NK cells. Recently, it has been described that murine NK cells are triggered by the costimulatory molecules CD80, CD86, and CD40. Using NK cell lines derived from PBMC as effectors, we demonstrate that the human CD80 and CD86 gene products can function as triggering molecules for NK cell-mediated cytotoxicity. Expression of human CD80 or CD86 molecules in murine B16.F1 melanoma cells rendered these significantly more susceptible to lysis by human NK cell lines. Blocking of the transfected gene products with specific mAb reduced lysis levels to that of nontransfected control cell lines. Triggering of human NK cells by CD80 and CD86 appeared to be independent of CD28 and CTLA-4, at least as determined by the reagents used in the present study, because the expression of these molecules could not be detected on the NK cell lines by either flow cytometry or in redirected lysis assays. Thus, human NK cells may use receptors other than CD28 and CTLA-4 in their interactions with CD80 and CD86 molecules. Alternatively, interactions may involve variants of CD28 (and possibly CTLA-4) that are not recognized by certain anti-CD28 mAb.     

Cutting edge: the relative distribution of T cells responding to chemically dominant or minor epitopes of lysozyme is not affected by CD40-CD40 ligand and B7-CD28-CTLA-4 costimulatory pathways

We examined the frequencies and specificities of the CD4+ T cell responses to the protein hen egg white lysozyme in mice deficient in the CD40-CD40 ligand or B7-CD28 costimulatory pathways. The frequency of T cells was decreased by between 3- and 4-fold in CD40-/- mice, and 12-fold in B7-1/B7-2-/- mice, but surprisingly, the relative distribution of T cells responding to peptides that were presented at levels that differed by >250-fold was similar. We also examined the CD4 response after blocking the regulatory molecule CTLA-4 during immunization. We observed no difference in either the frequency or specificity of the CD4+ T cell response if CTLA-4 was blocking during priming. Thus, the T cell response was generated toward the constellation of chemically dominant and subdominant epitopes as a whole, and did not discriminate among them based on their relative abundance.     

Immune modulation through 4-1BB enhances SIV vaccine protection in non-human primates against SIVmac251 challenge

Costimulatory molecules play a central role in the development of cellular immunity. Understanding how costimulatory pathways can be directed to positively influence the immune response may be critical for the generation of an effective HIV vaccine. Here, we evaluated the ability of intravenous administration of a blocking monoclonal antibody (mAb) directed against the negative costimulatory molecule CTLA-4, and an agonist mAb directed against the positive costimulatory molecule 4-1BB, either alone or in combination, to augment intramuscular SIV DNA immunizations. We then tested the ability these of these responses to impact a high-dose SIVmac251 challenge. Following immunization, the groups infused with the anti-4-1BB mAb exhibited enhanced IFN-γ responses compared to the DNA vaccine only group. Interestingly, although CTLA-4 blockade alone did not enhance IFN-γ responses it did increase the proliferative capacity of the CD4⁺ and CD8⁺ T cells. The combination of both mAbs enhanced the magnitude of the polyfunctional CD8⁺ T cell response. Following challenge, the group that received both mAbs exhibited a significant, ∼2.0 log, decrease in plasma viral load compared to the naïve group the included complete suppression of viral load in some animals. Furthermore, the use of the CTLA-4 blocking antibody resulted in significantly higher viral loads during chronic infection compared to animals that received the 4-1BB mAb, likely due to the higher CD4⁺ T cell proliferative responses which were driven by this adjuvant following immunization. These novel studies show that these adjuvants induce differential modulation of immune responses, which have dramatically different consequences for control of SIV replication, suggesting important implications for HIV vaccine development.     

Cutting edge: monovalency of CD28 maintains the antigen dependence of T cell costimulatory responses

CD28 and CTLA-4 are the major costimulatory receptors on naive T cells. But it is not clear why CD28 is monovalent whereas CTLA-4 is bivalent for their shared ligands CD80/86. We generated bivalent CD28 constructs by fusing the extracellular domains of CTLA-4 or CD80 with the intracellular domains of CD28. Bivalent or monovalent CD28 constructs were ligated with recombinant ligands with or without TCR coligation. Monovalent CD28 ligation did not induce responses unless the TCR was coligated. By contrast, bivalent CD28 ligation induced responses in the absence of TCR engagement. To extend these findings to primary cells, we used novel superagonistic and conventional CD28 Abs. Superagonistic Ab D665, but not conventional Ab E18, predominantly ligates CD28 bivalently at low CD28/Ab ratios and induces Ag-independent T cell proliferation. Monovalency of CD28 for its natural ligands is thus essential to provide costimulation without inducing responses in the absence of TCR engagement.     

Lymphoproliferative disorder in CTLA-4 knockout mice is characterized by CD28-regulated activation of Th2 responses

Mice lacking CTLA-4 die at an age of 2-3 wk due to massive lymphoproliferation, leading to lymphocytic infiltration and destruction of major organs. The onset of the lymphoproliferative disease can be delayed by treatment with murine CTLA4Ig (mCTLA4Ig), starting day 12 after birth. In this study, we have characterized the T cells present in CTLA-4-deficient mice before and after mCTLA4Ig treatment. The T cells present in CTLA-4-deficient mice express the activation markers, CD69 and IL-2R; down-regulate the lymphoid homing receptor, CD62L; proliferate spontaneously in vitro and cannot be costimulated with anti-CD28 mAb consistent with a hyperactivated state. The T cells from CTLA-4-deficient mice survive longer in culture correlating with higher expression of the survival factor, Bcl-xL, in these cells. Most significantly, the CD4+ T cell subset present in CTLA-4-deficient mice secretes high levels of IL-4 and IL-5 upon TCR activation. Treatment of CTLA-4-deficient mice treated with mCTLA4Ig reverses the activation and hyperproliferative phenotype of the CTLA-4-deficient T cells and restores the costimulatory activity of anti-CD28 mAb. Furthermore, T cells from mCTLA4Ig-treated mice are not skewed toward a Th2 cytokine phenotype. Thus, CTLA-4 regulates CD28-dependent peripheral activation of CD4+ T cells. This process results in apoptosis-resistant, CD4+ T cells with a predominantly Th2 phenotype that may be involved in the lethal phenotype in these animals.     

CD40 ligand and CTLA-4 are reciprocally regulated in the Th1 cell proliferative response sustained by CD8(+) dendritic cells

Subsets of murine dendritic cells (DCs) from the spleen differ in their ability to induce proliferative responses in both primary and secondary CD4(+) T cells. Recent evidence indicates that lymphoid-related CD8(+) DCs fail to provide appropriate signals to freshly isolated secondary CD4(+) T cells to sustain their proliferation in vitro. In the present study, we examined peptide-pulsed CD8(-) and CD8(+) DCs for ability to stimulate Th1 and Th2 cell clones with the same Ag specificity. Defective ability to induce proliferation was selectively shown by CD8(+) DCs presenting Ag to the Th1 clone. The deficiency in CD8(+) DCs was overcome by CD40 triggering before peptide pulsing. When exposed to CD8(+) DCs in the absence of CD40 activation, the Th1 clone expressed low levels of CD40 ligand and high levels of surface CTLA-4. Neutralization of CTLA-4 during the DC/T cell coculture resulted in increased CD40 ligand expression and proliferation of T cells. Remarkably, the activation of CD40 on DCs under conditions that would increase Th1 cell proliferation, also resulted in down-regulation of surface CTLA-4. These results confirm differential effects of CD8(+) and CD8(-) DCs in the stimulation of Ag-primed Th cells. In addition, they suggest that reciprocal regulation of CD40 ligand and CTLA-4 expression occurs in Th1 cells exposed to CD8(+) DCs.     

Cutting edge: Cbl-b: one of the key molecules tuning CD28- and CTLA-4-mediated T cell costimulation

Cbl-b negatively regulates CD28-dependent T cell activation. In this report, we tested the hypothesis that CD28 and CTLA-4 have opposite roles in tuning T cell activation threshold by controlling the levels of Cbl-b protein expression. We demonstrate that CD28 costimulation potentiates TCR-induced Cbl-b degradation, whereas CTLA-4-B7 interaction is required for Cbl-b re-expression. In support of this finding, Cbl-b expression in CTLA-4 knockout (KO) T cells is significantly reduced, and treating CTLA-4KO mice with human CTLA-4Ig to block CD28-B7 interaction restores Cbl-b expression on T cells. Furthermore, CD28 and CTLA-4 costimulatory effects are compromised in Cbl-bKO T cells. These observations indicate that CD28 and CTLA-4 tightly regulate Cbl-b expression which is critical for establishing the threshold for T cell activation.     

Intercellular adhesion molecule-2, a second counter-receptor for CD11a/CD18 (leukocyte function-associated antigen-1), provides a costimulatory signal for T-cell receptor-initiated activation of human T cells.

Activation of T cells often requires both activation signals delivered by ligation of the TCR and those resulting from costimulatory interactions between certain T cell surface accessory molecules and their respective counter-receptors on APC. CD11a/CD18 complex on T cells modulate the activation of T cells by interacting with its counter-receptors intracellular adhesion molecule (ICAM-1) (CD54) and/or ICAM-2 on the surface of APC. The costimulatory ability of ICAM-1 has been demonstrated. Using a soluble ICAM-2 Ig fusion protein (receptor globulin, Rg) we demonstrate the costimulatory effect of ICAM-2 during the activation of CD4+ T cells. When coimmobilized with anti-TCR-1 mAb ICAM-2 Rg induced vigorous proliferative response of CD4+ T cells. This costimulatory effect of ICAM-2 was dependent on its coimmobilization with mAb directed at the CD3/TCR complex but not those directed at CD2 or CD28. Both resting as well as Ag-primed CD4+ T cells responded to the costimulatory effects of ICAM-2. The addition of mAb directed at the CD11a or CD18 molecules almost completely inhibited the responses to ICAM-2 Rg. These results are consistent with the role of CD11a/CD18 complex as a receptor for ICAM-2 mediating its costimulatory effects. Stimulation of T cells with coimmobilized anti-TCR-1 and ICAM-2 resulted in the induction of IL-2R (CD25), and anti-Tac (CD25) mAb inhibited this response suggesting the contribution of endogenously synthesized IL-2 during this stimulation. These results demonstrate that like its homologue ICAM-1, ICAM-2 also exerts a strong costimulatory effect during the TCR-initiated activation of T cells. The costimulatory effects generated by the CD11a/CD18:ICAM-2 interaction may be critical during the initiation of T cell activation by ICAM-1low APC.