HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Waldrop, S. L. Articles by Picker, L. J. Search for Related Content PubMed PubMed Citation Articles by Waldrop, S. L. Articles by Picker, L. J.

Normal Human CD4⁺ Memory T Cells Display Broad Heterogeneity in Their Activation Threshold for Cytokine Synthesis¹

Shar L. Waldrop^*, Kenneth A. Davis, Vernon C. Maino and Louis J. Picker^2,*

^* Laboratories of Experimental Pathology and Clinical Immunology, Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75235; and Becton Dickinson Immunocytometry Systems, San Jose, CA 95131

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
CD4⁺ memory T cells coordinate immune responses against viruses and other pathogens via the Ag-induced secretion of potent effector cytokines. The efficacy of these responses depends on both the overall number of pathogen-specific memory T cells and the particular array of cytokines that these cells are programmed to secrete. Here, we provide evidence that heterogeneity in Ag triggering thresholds constitutes an additional critical determinant of memory T cell function. Using a novel assay that allows single-cell detection of Ag-specific T cell cytokine production, we demonstrate that CMV-specific CD4⁺ memory cells from human peripheral blood display pronounced differences in their costimulatory requirements for Ag-induced triggering of IFN- and IL-2 secretion, ranging from cells that trigger with little costimulation (e.g., resting APC alone) to cells requiring potent costimulation through multiple pathways (resting APC plus multiple costimulatory mAbs, or activated APC). These differences in costimulatory requirements are independent of clonal differences in TCR signaling intensity, consistent with an intrinsic activation-threshold heterogeneity that is "downstream" from the TCR. Thus, "effective" frequencies of Ag-specific CD4⁺ memory T cells appear to depend on the activation status of available APC, a dependence that would allow the immune system to rapidly adjust the number of functional Ag-specific memory T cells in a particular effector site according to local conditions.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The memory/effector subset of CD4⁺ T cells comprise a group of Ag-selected, peripherally differentiated cells that are largely, if not completely, responsible for the myriad functions of this lineage in the induction and regulation of immune responsiveness (1, 2). These cells manifest extensive heterogeneity in key functional attributes such as homing behavior and cytokine synthesis capability that provides them with an extraordinary flexibility in appropriately responding to diverse antigenic challenges and in regulating both their own reactivity and the reactivity of other effector cell types (1, 2, 3, 4). Recently, theoretical considerations, as well as studies of TCR transgenic mice and in vitro propagated T cell clones, have suggested the possibility that peripheral T cells display variable, perhaps "tunable," activation thresholds (5, 6, 7, 8, 9). If true, this additional form of functional heterogeneity could play a critical role in governing the initiation, tempo, and outcome of immune effector responses, at once greatly increasing both the complexity of such responses and the capacity for their fine regulation. For example, T cells with low activation thresholds might be expected to be recruited early into effector responses at low Ag densities or low levels of APC activity, whereas T cells with high thresholds might be recruited late, if at all. If such triggering threshold heterogeneity was "superimposed" over effector function heterogeneity, one might imagine a scenario in which certain functional types of effector cells are programmed to be early responders and others late, allowing either graded or sequential responses to a single Ag.

However, to date, triggering threshold heterogeneity has not been definitively demonstrated among physiologic T cells, particularly the Ag-specific memory T cells critical for host defense in the human. Historically, investigation of Ag triggering thresholds among normal human memory T cells has not been experimentally approachable due to the difficulty in assessing Ag-specific T cell triggering on a single-cell basis. We have recently overcome this limitation with the development of a novel flow cytometric assay that visualizes individual Ag- or superantigen-responsive CD4⁺ memory T cells with unprecedented clarity (10). This assay quantitates such T cells by the multiparameter assessment of intracellular cytokine, the activation marker CD69, and T cell subset marker(s) (e.g., CD4) after short-term (6-h) incubation of PBMC with Ag in the presence of the secretion inhibitor Brefeldin A (for the final 5 h). Responses in this assay are measured as frequencies of CD69⁺/cytokine⁺ CD4⁺ T cells and show 1) the expected critical dependence on specific Ag, MHC class II determinants, and APC costimulatory/adhesive interactions, 2) restriction to CD45RA^low/RO^high memory T cells, 3) precise correlation with independent measures of sensitization history (e.g., serologic or skin test reactivity), and 4) coefficients of variation of <10% (10). Moreover, these responses precede activation-induced apoptosis, preventing artifactual alteration of response patterns by this process.

During the early development of this assay, we observed that the provision of supplemental costimulation in the form of CD28 mAbs increased the observed frequencies of cytokine-producing CD4⁺ T cells in response to the viral pathogen CMV by 50–100% but had no stimulatory effect by itself (10). At the time, we hypothesized that a costimulation-mediated synchronization of the response was responsible for this observation. However, recent data indicating that the costimulation through CD28 has the effect of lowering T cell activation thresholds (7, 8, 9, 11) suggested an alternative explanation: that normal CD4⁺ memory T cells may differ in their requirement for costimulation and, thus, in their triggering thresholds. Here, we explore this issue in detail, conclusively demonstrating that a significant proportion (up to 80%) of peripheral blood-derived CD4⁺ memory T cells that are capable of specifically producing IFN- and IL-2 in response to Ag or superantigen cannot be triggered by these stimuli to produce these cytokines in the presence of resting APC alone. The triggering of these cells requires Ag or superantigen and either mAb-mediated ligation of one or more active costimulatory pathways or APC activation with proinflammatory cytokines such as TNF-. Because exogenous costimulation increases the frequency of responding cells at any given level of TCR signaling (as measured by TCR down-regulation), this triggering heterogeneity appears to reflect intrinsic differences in the activation set point that are downstream from the TCR itself.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell preparation and Ag stimulation

PBMC were isolated from heparinized venous blood by density gradient sedimentation using Ficoll-Hypaque (Histopaque, Sigma, St. Louis, MO). Cells were then washed twice in HBSS (Life Technologies, Grand Island, NY) and resuspended in RPMI 1640 media (Life Technologies) supplemented with 10% heat-inactivated FCS (HyClone Sterile Systems, Logan, UT), 1 mM sodium pyruvate (Sigma), and 2 mM L-glutamine (Life Technologies). Ag and superantigen stimulation was performed as described previously (10). Briefly, 1 x 10⁶ PBMC were placed in 12 x 75 mm polystyrene tissue culture tubes containing 2 ml complete media with appropriately titered CMV, herpes simplex virus (HSV),³ varicella-zoster virus (VZV), or matched control Ag preparations (BioWhittaker, Walkersville, MD), or 200 ng/ml of the superantigens staphylococcal enterotoxin (SE) B, SEE, or toxic shock syndrome toxin (TSST) (Toxin Technology, Sarasota, FL). Purified costimulatory or control mAb were added where indicated at a final concentration of 0.5 µg/ml each. The cultures were routinely incubated at a 5° slant at 37°C in a humidified 5% CO₂ atmosphere for 6 h with the final 5 h including 10 µg/ml of Brefeldin A (Sigma). After incubation, cells were harvested by washing once in cold Dulbecco’s PBS (dPBS), resuspending the pellet in 4 ml dPBS containing 0.02% EDTA, incubating at 37°C for 15 min, and again washing in cold dPBS. Harvested cells were then either processed fresh or fixed in 4% paraformaldehyde in dPBS for 5 min at 37°C and frozen, as previously described (10). In some experiments, purified preparations of APC and T cells were produced from PBMC by one to two rounds of neuraminidase-treated SRBC rosetting (12) followed by further purification of the T cell preparations by negative selection using R&D Systems (Minneapolis, MN) T cell purification columns. In these studies, APC preparations (B cells, macrophages, and dendritic cells) contained <1.5% contaminating T cells, and T cell preparations contained <0.5% APC. Purified populations were incubated with or without 15 ng/ml TNF- (R&D Systems) at 37°C in a humidified 5% CO₂ atmosphere for 17 h before reconstitution at an 85:15 T cell/APC ratio and then stimulation with SEB alone vs SEB plus CD28 and CD49d mAbs, as described above.

Immunofluorescent staining and flow cytometric analysis

Frozen cell preparations were rapidly thawed in a 37°C water bath and then washed once with cold dPBS before resuspension in fixation/permeabilization solution (Becton Dickinson Immunocytometry Systems, San Jose, CA) at 2 x 10⁶ cells/ml, and incubated for 10 min at room temperature in the dark. Fixed and permeabilized cells were washed twice with dPBS and then incubated on ice (protected from light) with directly conjugated mAbs for 30 min. In some experiments (those including analysis of CD45RO/RA, HLA-DR, TCR-Vß subsets, or TCR/CD3 down-regulation), cells freshly harvested after Ag or superantigen activation were cell surface stained first (for the markers mentioned above), followed by fixation/permeabilization, washing, and then staining for CD4, intracytoplasmic cytokine, and CD69 (omitting the freezing step). After staining, the cells were washed, resuspended in 1% paraformaldehyde in dPBS, and then kept protected from light at 4°C until analysis on the flow cytometer. Five- or six-parameter flow cytometric analysis was performed on a modified FACSort flow equipped with a second 632 nm line diode laser (Becton Dickinson Immunocytometry Systems) using FITC, phycoerythrin (PE), peridinin chlorophyl protein (PerCP), and allophycocyanin (AP) as the four fluorescent parameters. All analyses included at a minimum assessment of CD4 vs cytokine or isotype-matched control mAb vs CD69. For each analysis, 50,000 events were acquired and gated on CD4 (or CD4 and CD3, or CD4 and TCR-Vß) expression, and a light scatter gate designed to include only viable small lymphocytes. List-mode multiparameter data files (each file with forward scatter, orthogonal scatter, and three to four fluorescent parameters) were analyzed using the PAINT-A-GATE^Plus software program (Becton Dickinson Immunocytometry Systems). The criteria for delineating and quantitating responding (CD69⁺/cytokine⁺) vs non-responding T cells have been previously described in detail (10).

Monoclonal Ab

The following mAbs were used in this study: Leu-3a (CD4; PerCP-, AP-conjugated), Leu-4 (CD3; PerCP, AP); Leu-23 (CD69; PE, PerCP), Leu-28 (CD28; PE), L25.3 (CD49d; PE), clone 16 (CD49e), Leu-1 (CD5; PE), Leu-5b (CD2), Leu-45RO (CD45RO; PE), L130 (CD18), G25.2 (CD11a), LB-2 (CD54), anti-IFN- (FITC, PE, AP), anti-IL-2 (FITC, PE, AP), anti-HLA-DR (L243; PerCP), anti-HLA-DP (B7/21), anti-HLA-DQ (SK10), G1CL (mouse IgG1 control; FITC, PE, PerCP, AP); G2CL (mouse IgG2 control; FITC, PE, PerCP), and streptavidin-AP were obtained from Becton Dickinson Immunocytometry Systems; mAb 2H4 (CD45RA; PE), 4B4 (CD29) and the anti-TCR-Vß2, -Vß3, -Vß8, and -Vß12, and -Vß17 mAbs (PE, biotin) were obtained from Coulter Immunology (Hialeah, FL); mAb TRAP1 (CD40L) and a control rat IgG2a were obtained from PharMingen (San Diego, CA); mAbs BQ16 (CD49f), UMCD2 (CD2), M-T271 (CD27), and 24–31 (CD40L) were obtained from Ancell (Bayport, MN); mAb BRIC 126 (CD47) was obtained from Biosource International (Camarillo, CA); mAb HP2/1 (CD49d) and anti-TCR-V2 (FITC) were obtained from Serotec (Oxford, U.K.); and additional mouse IgG1, G2a, and G2b mAb controls were obtained from Sigma. mAb DREG 56 (CD62L) and Hermes 3 (CD44) were produced in our laboratory from hybridomas available from the American Tissue Culture Collection (Manassas, VA).

The CD28/CD49d tripod Ab (CD28 Fab linked to an intact CD49d mAb) was prepared as follows: 1) pepsin digestion of intact Leu 28 mAb; 2) purification of Leu-28 F(ab')₂ by gel filtration (2-Superdex TM200 HR 10/30 column; nonreducing SDS-PAGE revealed that no intact Ig was detectable after purification); 3) reduction of Leu-28 F(ab')₂ with 20 mM DTT (which was removed by buffer exchange into 50 mM 2-(N-morpholino)ethanesulfonic acid (MES), pH 6.0, on G-50 column); 4) derivatization of CD49d mAb (clone L25.3) with 0.2 mM succinimidyl 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate (SMCC) for 60 min at room temperature followed by buffer exchange into 50 mM sodium phosphate buffer, pH 7.2; 5) coupling of SMCC-derivatized CD49d with reduced CD28 F(ab')₂ at a molar ratio of 1:0.66 for 60 min at room temperature; and 6) purification of the approximately 200-kDa tripod reagent as described in step 2. Consistent with its biochemical structure, binding of (biotinylated) tripod reagent to CD28⁺/CD49d⁺ T cells was completely inhibited by pretreatment with combined CD28 and CD49d mAbs, but only partially (CD49d) or minimally (CD28) inhibited by pretreatment of these cells with each of these mAbs alone. The CD28 Fab was prepared by reduction of Leu-28 F(ab')2 with 20 mM DTT for 45 min followed by a 20-M excess of N-ethyl maleimide for 10 min to block free sulfhydryl groups and then buffer exchange into PBS.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Exogenous "dual-pathway" costimulation reveals maximal frequencies of CMV-specific CD4⁺ effector T cells

To further explore the mechanism by which CD28 mAbs increase the observed frequencies of CMV-triggered, cytokine-producing CD4⁺ T cells (10), we first tested a wide variety of mAbs specific for putative T cell costimulatory molecules (13, 14, 15, 16, 17, 18, 19) for a similar capability. These studies demonstrated that mAbs against CD49d and CD5 enhanced CMV-directed effector frequencies in a manner comparable to that of CD28 mAb (approximately twofold), whereas mAbs specific for CD18, CD49e, CD49f, 4ß7-integrin, CD40L (CD154), CD27, CD44, CD62L, CD2, CD54, and CD47 lacked this activity (Fig. 1, A and B and data not shown). Like CD28 mAbs (10), the CD49d and CD5 mAbs were incapable of eliciting cytokine-producing cells by themselves (i.e., without TCR-directed stimuli; data not shown). The combination of two of these active mAbs (the CD28 plus CD49d combination is shown, but other combinations had analogous effects) further increased effector frequencies, especially for IL-2-producing cells, but the combination of all three active mAbs had no further augmenting effect (Fig. 1A). Thus, the provision of "dual-pathway" exogenous costimulation revealed maximal CMV responder frequencies among CD4⁺ T cells that were unique and reproducible for each individual subject. The enhancement mediated by the active costimulatory mAbs was restricted to responder frequencies: costimulation did not significantly increase the amount of cytokine produced by responding cells (Fig. 2). Importantly, even with maximal exogenous costimulation the CMV-responding cells (both IFN-- and IL-2-producing cells) derived almost exclusively from the resting (HLA-DR^negative), memory (CD45RA^low/RO^high), CD4⁺ T cell subset (Fig. 1B, and data not shown).

View larger version (33K): [in this window] [in a new window]   FIGURE 1. The frequency of human CD4+ memory T cells responding to CMV Ag with production of IFN- and IL-2 increases in the presence of mAb against CD28, CD49d, or CD5, but not other T cell adhesion/costimulatory molecules, and maximum frequencies are observed with dual-pathway (e.g., CD28 plus CD49d) costimulation. PBMC were stimulated with CMV Ag ± mAb(s) for 6 h in the presence of the secretion inhibitor Brefeldin A for the final 5 h and examined for their correlated expression of intracellular cytokine (IFN- and IL-2) vs the activation Ag CD69 vs CD4 (and vs CD45RA, CD45RO, or HLA-DR in some analyses). After gating on CD4+ lymphocytes, responding cells are identified by their parallel up-regulation of (intracellular) CD69 and cytokine expression. Such cells are not present when control Ag (mock purified virus from uninfected cell lines ± costimulatory mAbs) is used to stimulate PBMC (see Fig. 5). A, The ratio (±SEM) of responder frequencies observed with CMV Ag plus 0.5 µg/ml of the designated mAb(s) to responder frequencies observed with CMV Ag alone is shown for three to seven different CMV-seropositive subjects (responder frequencies calculated as CD69+ and cytokine+, CD4+ T cells/total CD4+ T cells; see B). Mean (±SEM) responder frequencies to CMV Ag alone among this overall cohort were 0.88 ± 0.25% and 0.32 ± 0.07% for IFN- and IL-2, respectively. The differences in responder frequencies between T cells costimulated (separately) with CD28, CD49d, and CD5 mAbs vs control mAb were significant at p values between <0.0001 and 0.0022 for IFN- and between 0.003 and 0.024 for IL-2 (two-tailed t test). Costimulation with the combination of CD28 plus CD49d mAbs was significantly different from both control mAb and from CD28, CD49d, and CD5 mAbs alone at similar p value ranges for both cytokines but was not significantly different from costimulation with the combination of CD28 plus CD49d plus CD5 mAbs (p = 0.664 for IFN-; p = 0.184 for IL-2). B, Flow cytometric profiles of a representative CMV-seropositive subject’s CD4+ T cell (IFN-) response to CMV alone, CMV plus either CD28 or CD49d mAbs, CMV plus both mAbs, and CMV plus both mAbs in the presence of MHC class II blockade (anti-HLA-DR, -DP, and -DQ at 5 µg/ml each; class-matched control mAbs for these anti-class II reagents had no effect, not shown). Shown are 30,000 events gated on viable CD4+ lymphocytes, with the events in the IFN-+/CD69+ "response" region of the profiles enlarged and colored black. Note the increasing frequency of responding cells with single- and dual-pathway exogenous costimulation, and that even with maximum costimulation (CD28 plus CD49d) the response is largely restricted to the resting (HLA-DR-), memory (CD45RA-/RO+) subset (percents shown in the figure), and is >90% inhibitable with MHC class II blockade. Similar results were observed for IL-2 (not shown).

View larger version (33K): [in this window] [in a new window]   FIGURE 2. Exogenous costimulation increases the frequency of CMV responding CD4+ T cells but does not increase the amount of cytokine produced by each responding cell. A, The figure demonstrates a representative flow cytometric analysis of PBMC from a CMV-seropositive subject stimulated with CMV ± dual-pathway costimulation and analyzed for the intensity of intracellular IFN- staining within the responding CD4+ T cells. Shown are 30,000 events gated on CD4+/CD3+ lymphocytes in the dot plots on the left, and 1,000 events additionally gated on the responding fraction (i.e., the enlarged black cells in the dot plots) are shown in the histograms on the right (only the first 3 logs of fluorescent intensity are shown in these histograms). Note that whereas the provision of dual-pathway costimulation almost doubles the frequency of CMV-responding cells, the spectrum of responding cell IFN- intensities remains essentially unchanged, as evidenced by the similar mean and peak fluorescent intensities of the responding populations. The mean fluorescent intensity (MFI) of the responding cells are provided in the histograms (in arbitrary fluorescence units); the arrowheads designate the identical positions of the histogram peaks (240 fluorescence units). B, The effect of exogenous costimulation on the cytokine MFI of CMV-responding cells are shown for seven CMV-seropositive subjects. The results are presented as the mean ratio (±SEM) of the cytokine MFIs observed among cells responding to CMV plus the designated mAb(s) to that observed among cells responding to CMV alone (thus, 1.0 = no effect). The corresponding effect of the same mAbs on the frequency of responding cells are shown in Fig. 1A. The slight (16–21%) increase in IL-2 MFI among T cells stimulated with CMV plus CD28 and CMV plus both CD28 and CD49d did not achieve statistical significance (two-tailed t test).

View larger version (24K): [in this window] [in a new window]   FIGURE 3. CD4+ T cells from CMV-naive (seronegative) subjects fail to respond to CMV even in the presence of maximal (dual-pathway) exogenous costimulation. Flow cytometric profiles (as described in Fig. 1B) of the CD4+ T cell response to CMV plus dual-pathway costimulation (CD28 plus CD49d mAbs) in three CMV-naive (seronegative) subjects (representative of seven such subjects). The frequency of CMV-"responding" T cells in these subjects is indistinguishable from what is observed with control Ag or no Ag plus the CD28 and CD49d mAbs (data not shown). Only the IFN- response is shown in the figure, but IL-2 responses were identical. Importantly, in parallel with their negative responses to CMV, all these subjects were demonstrated to respond to one or more positive control stimuli, including the superantigen SEB and/or distinct Ags such as mumps virus, VZV, or HSV (see Fig. 5).

View larger version (14K): [in this window] [in a new window]   FIGURE 4. Responder frequencies to the staphylococcal toxin superantigens SEB, TSST, and SEE are enhanced with dual-pathway costimulation, but remain restricted to T cells bearing the appropriate Vß-defined TCR for each superantigen. PBMC were stimulated with superantigen ± CD28 and CD49d mAbs for 6 h in the presence of the secretion inhibitor Brefeldin A for the last 5 h, and then examined for their correlated expression of (intracellular) IL-2 vs CD69 vs CD4 vs (surface) TCR-Vß subset-specific epitopes, including TCR-Vß-2, -3, -8, -12, and -17. For each of these TCR-Vß subsets, 10–20,000 events (gated on CD4+, TCR-Vß subset+ lymphocytes) were collected and analyzed as shown in Fig. 1. These results are representative of three independent experiments. In each of these experiments, similar results were observed with IFN- (data not shown).

View larger version (43K): [in this window] [in a new window]   FIGURE 5. Kinetics of human CD4+ T cell responses to CMV plus single- and dual-pathway supplemental costimulation (vs control Ag plus dual-pathway costimulation) in a CMV-seropositive subject, and CMV vs VZV vs HSV (all with dual-pathway costimulation) in a CMV-seronegative subject. PBMC were stimulated with CMV, CMV control, VZV, or HSV Ag ± mAb(s) for 6, 10, 14, 18, and 23 h with Brefeldin A included for the final 5 h of culture and then examined for their correlated expression of intracellular cytokine (IFN- and IL-2) vs CD69 vs CD4 as described in Fig. 1. Note that because of the Brefeldin A incubation "windows", the 6-h time point includes all cells producing cytokine from the beginning of hour 2 through hour 6 (including the earliest cytokine producers), the 10-h time point from the beginning of hour 5 through hour 10, and so forth; thus, essentially all cells producing cytokine up to 23 h are included in this analysis. The profiles shown are representative of results observed in three independent experiments.

As mentioned above, the CMV-responder frequency enhancement resulting from combined CD28 and CD49d (or CD5) costimulation, as compared with single-pathway costimulation, can be explained in two ways. It might reflect a variable threshold effect in which individual cells may require either modest (single-pathway) or marked (dual-pathway) costimulation to achieve triggering (corresponding to intermediate- and high-threshold subsets). Alternatively, there may be a uniform costimulatory threshold for those cells requiring costimulation (potentially achievable through either CD28, CD49d, or CD5) with responsiveness to these signals distributed to different subsets within the overall memory T cell population. In the latter situation, the combined enhancing effect of the CD28 and CD49d mAbs would be the result of an additive effect of these mAbs operating independently on distinct T cell subsets. Three lines of evidence support the "variable threshold" hypothesis and argue against the latter possibility. First, CD49d and CD28 are coexpressed by 80–90% of CD4⁺ memory cells, and CD5 is expressed by nearly 100% (data not shown). Second, we constructed a "tripod" mAb in which a monovalent CD28 (Fab) fragment was chemically "grafted" onto an intact (bivalent) CD49 mAb. As shown in Fig. 6, monovalent CD28 Fab has no enhancing activity in our assay, whereas the tripod reagent is equivalent in activity to the combination of separate CD28- and CD49d-intact mAbs (and reproducibly enhanced over intact CD28 or CD49d mAbs alone). Because CD28 costimulatory activity is obligately linked to that of CD49d with the tripod reagent (i.e., the monovalent CD28 component of the tripod cannot deliver signals without simultaneous ligation of CD49d), these data indicate the existence of cells which are not triggered by CD49d costimulation alone, but that can trigger when the tripod reagent provides CD49d and (linked) CD28 signals together. Thus, the triggering of these cells (the putative "high"-threshold subset) requires the cooperative activity of two distinct costimulatory pathways within a single cell, whereas other cells require only single-pathway costimulation or no exogenous costimulation at all (intermediate- and low-threshold subsets, respectively). The third line of evidence supporting the "variable threshold" hypothesis involves experiments examining the effects of sequentially back-titrating the (intact) CD28 and CD49d mAbs in cultures treated with single mAbs or the combination of both mAbs (Fig. 7). Note that as the concentrations of either CD28 or CD49d mAb are decreased below the saturation point in cultures treated with a single costimulatory mAb, CMV-responder frequencies dropped to baseline; yet at the same mAb concentrations, the combination of both mAbs together still resulted in significant responder-frequency enhancement. In other words, two disparate, subthreshold costimuli can cooperate in individual T cells to facilitate triggering of a subset of CMV-responsive cells that would not be triggered by CMV Ag alone. Taken together, these experiments suggest a continuum of activation thresholds such that some individual cells require two strong (independent) costimulatory signals, others one strong or perhaps two suboptimal costimulatory signal(s), and still others no exogenous costimulation at all.

View larger version (45K): [in this window] [in a new window]   FIGURE 6. Achievement of maximal CMV-response frequencies involves the activity of two costimulatory pathways in individual T cells. A CD49d mAb covalently linked to a CD28 Fab fragment ("Tripod" reagent) provides equivalent (maximal) costimulatory support for CMV effector responses as the combination of intact (individual) CD28 and CD49d mAbs. Monovalent CD28 Fab has no costimulatory effect alone or in combination, indicating that for maximal responder frequencies to be achieved at least a proportion of the responding cells require "cross-linking" of both CD28 and CD49d. As described in Fig. 1, 30,000 events are shown gated on CD4+ lymphocytes with the responding cells enlarged and colored black. The results are representative of four independent experiments.

View larger version (35K): [in this window] [in a new window]   FIGURE 7. Subthreshold signals from distinct costimulatory pathways can act cooperatively in individual T cells to facilitate Ag triggering. PBMC were stimulated with CMV Ag ± progressively decreasing concentrations of CD28 mAb alone, CD49d mAb alone, or the same amount of each mAb combined together for 6 h in the presence of the secretion inhibitor Brefeldin A for the last 5 h, and were examined as described in Fig. 1. The results shown are representative of three independent experiments. mAb saturation was determined by counterstaining an aliquot of the cells treated with titrations of the unconjugated CD28 or CD49d mAbs with PE-conjugates of the same mAb followed by flow cytometric analysis.

As a further refinement of our model, we next asked the question whether the variable activation thresholds noted in these experiments represent 1) heterogeneous signaling through the TCR—perhaps related to heterogeneous avidities of different CMV-derived peptides/class II MHC combinations for different TCRs, or in the case of superantigen, to TCR--chain-dependent variation in superantigen-TCR interaction (21)—or 2) differences in the amount or activity of downstream components of the T cell activation cascade that would together constitute an activation set point. Because signals transmitted through the TCR are proportional to the degree to which cell surface expression of TCR is down-regulated following Ag or superantigen binding (7, 9, 11, 22), we reasoned that these two possibilities could be discriminated by quantitative assessment of TCR signaling (e.g., down-regulation) in the presence or absence of maximal exogenous costimulation. If costimulation simply recruits T cells with lower levels of TCR signaling into the cytokine-producing subset, the expanded cytokine-producing subset observed with Ag or superantigen plus maximal exogenous costimulation would manifest, on average, decreased TCR down-regulation relative to the population of cytokine-producing cells generated with Ag or superantigen alone (due to dilution of the cytokine-producing subset by putative low avidity cells with lower levels of TCR down-regulation in the former group). On the other hand, if the major determinant of the observed activation-threshold heterogeneity is independent from TCR fine specificity (i.e., determined by a downstream activation "set point," such that even cells with identical TCR might potentially have different activation thresholds), exogenous costimulation would likely increase the frequency of cytokine-producing (-responding) cells uniformly across the entire spectrum of (clonally determined) TCR signaling, thus leaving the overall pattern of TCR down-regulation unchanged.

Indeed, as illustrated in Fig. 8 for superantigen-stimulated, IFN--producing CD4⁺ memory cells, the pattern of TCR down-regulation exhibited by the population triggered to produce cytokine with SEB alone is essentially identical to that observed on the nearly twofold larger population of T cells triggered to produce cytokine with SEB plus dual-pathway costimulation. In other words, all along the broad spectrum of TCR down-regulation produced by SEB stimulation (see histograms), the addition of exogenous costimulation homogeneously increases the frequency of responding cells, indicating that at any given level of TCR signaling there are cells with differing costimulatory requirements and, thus, differing activation thresholds. This finding was highly reproducible: in five independent experiments, the mean (±SEM) ratio of (IFN--defined) responder frequencies between Vß-17⁺/CD4⁺ (memory) T cells stimulated with SEB alone vs with SEB plus CD28 and CD49d mAbs was 1.74 ± .09, whereas the percent TCR down-regulation exhibited by these two groups was essentially identical (mean ratio ± SEM of Vß-17 down-regulation between SEB alone vs SEB plus CD28 and CD49d mAbs = 1.01 ± .03). Similar results were observed for the other superantigen and Vß subsets shown in Fig. 4 (data not shown). Moreover, the same independence of costimulatory requirements for triggering and TCR signaling strengths could be demonstrated for the Vß-17⁺/CD4⁺ memory T cell response to CMV Ag (Fig. 9A) as well as the overall CD4⁺ memory T cell responses to CMV (Fig. 9B, using CD3 down-regulation as a measure of TCR signaling; see figure legend). In the latter figure, note particularly that the mean CD3 down-regulation remained constant despite the progressive increase in CMV-specific responder frequencies promoted by the addition of single- and dual-pathway costimulation.

View larger version (53K): [in this window] [in a new window]   FIGURE 8. The activation-threshold heterogeneity observed in superantigen responses is not associated with differences in TCR signaling strength. The figure demonstrates a representative flow cytometric analysis (12,000 events) of PBMC stimulated with SEB ± dual-pathway costimulation and gated on CD4+/Vß-17+ T cells (Vß-17 intensities remain high enough after maximal down-regulation to allow accurate gating) with the dot plots on the left delineating the CD69 and IFN--defined activation subsets (and adjacent to each subset-defining box, the percent of total cells within that subset) and the histograms on the right demonstrating the corresponding spectrum of TCR-Vß-17 intensities for each population. The percent down-regulation reflects the decrease in mean fluorescent intensity of TCR-Vß-17 expression from the CD69-/IFN-- subset (which is essentially identical to the TCR-Vß-17 intensity profile of unstimulated T cells; data not shown) to the CD69+/IFN-+ subset. The lower histograms showing the Vß-17 profiles of the CD69+/IFN-+ subsets are reproduced at x5.0 and x3.1 magnification so as to facilitate comparison of the distribution of Vß-17 intensities. Note that not only is the mean percent down-regulation identical between the CD4+ memory T cells stimulated with SEB and SEB plus CD28/CD49d mAbs to produce IFN-, but (with the scale of each histogram modified to equalize the number of events shown) the distribution of TCR intensities on the IFN- producing cells are almost superimposable.

View larger version (24K): [in this window] [in a new window]   FIGURE 9. The activation-threshold heterogeneity observed in Ag responses is also not associated with clonal heterogeneity in TCR signaling strength. A, In a subject with preferential representation of CMV reactivity in the Vß-17 subset, TCR down-regulation is essentially equivalent among T cells responding to CMV alone and the >twofold larger population responding to CMV plus maximal costimulation. PBMC were stimulated with CMV alone vs CMV plus CD28 and CD49d mAbs and assessed as described in Fig. 8. Shown are 3,000 events gated on CD4+/Vß-17+ T cells with responding cells colored black, and the percent responding and percent mean TCR down-regulation designated in the figure. A second independent experiment with this subject’s cells revealed similar results with the (IFN--defined) responder frequencies for Vß-17+/CD4+ T cells stimulated with CMV alone vs CMV plus CD28 and CD49d mAbs being 6.5% and 12.3%, respectively and the corresponding Vß-17 down-regulation being 71.2% and 72.4%, respectively. B, In a different subject, the degree of CD3 down-regulation (CD3 is down-regulated concomitantly with TCR during T cell activation; Ref. 7) on T cells responding to CMV alone with either IFN- or IL-2 production is essentially identical to that observed on the progressively larger population of responding cells induced by CMV plus single- or dual-pathway co-stimulation. In this experiment, PBMC stimulated with CMV alone, CMV plus CD28 or CD49d mAbs, and CMV plus both mAbs were stained on the surface with CD3 mAb, fixed and permeabilized, and then stained intracellularly for CD4, CD69, and cytokine. Responder frequencies and CD3 down-regulation on the responding fraction were determined on CD4+ T cells analogous to the procedure described in Fig. 8 for Vß-17. Identical results were obtained in two different CMV-seropositive subjects.

Inflammatory activation of APC with TNF- allows activation of high-threshold T cells without exogenous (mAb-mediated) costimulation

TNF- has been reported to up-regulate costimulatory molecules and MHC class II determinants on APC and thus facilitate Ag presentation (23, 24, 25). To provide a physiologic context for the inherent memory T cell-threshold heterogeneity demonstrated above with costimulatory mAbs, we asked whether inflammatory activation of APC by TNF- could mimic the effects of these mAbs. Because the effects of TNF- on nominal Ag uptake and processing are complex (23, 24), we restricted our focus to T cell stimulation with superantigen, for which such processing is not required. PBMC were thus examined fresh for CD4⁺ T cell cytokine-synthesis responses to SEB vs SEB plus CD28 and CD49d mAbs, or after a 17-h preincubation in the presence or absence of TNF-. As demonstrated in Fig. 10A, preincubation in media alone had little effect on either the frequencies of IFN- and IL-2-producing T cells in response to SEB, or the increases in these frequencies associated with the addition of dual-pathway exogenous costimulation. In contrast, preincubation with TNF- dramatically increased the frequency of cytokine-producing T cells observed in response to SEB alone, nearly to the level observed among control PBMC with maximal costimulation. Moreover, the provision of CD28 and CD49d mAbs to TNF--treated PBMC had little additional effect, suggesting that TNF- pretreatment alone effectively substitutes for mAb-mediated costimulation, and allows near maximal responses to superantigen. Finally, to determine whether the effect of TNF- is directed toward T cells or APC, these populations were separated, preincubated independently with or without TNF-, and then remixed in various combinations before assessment of CD4⁺ T cell cytokine responses to SEB. As shown in Fig. 10B, these studies confirm that the APC are the primary target for the TNF--mediated augmentation of CD4⁺ T cell SEB response frequencies.

View larger version (25K): [in this window] [in a new window]   FIGURE 10. APC activation by TNF- increases CD4+ T cell superantigen responder frequencies to similar levels as exogenous, mAb-mediated dual-pathway costimulation. A, PBMC were examined fresh or after a 17-h preincubation (at 37°C) with media alone or media plus 15 ng/ml TNF- for their CD4 T cell response to SEB alone or SEB plus CD28 and CD49d mAbs. Note that preincubation with TNF- increases the frequency of CD4+, IFN--, and IL-2-producing SEB responders to levels approximating those observed with SEB plus CD28 and CD49d mAbs in fresh or control incubated PBMC, and that the addition of CD28 and CD49d mAbs to TNF- preincubated PBMC provides little further enhancement. The experiment shown is representative of three identical experiments using different subjects. B, T cells and APC were prepurified by E-rosetting and negative selection columns (see Materials and Methods) and then separately incubated with media alone or media plus 15 ng/ml TNF- for 17 h at 37°C. Control and TNF--treated T cells were then reconstituted with either control or TNF--treated APC at an 85:15 ratio and then examined for their CD4+ T cell response to SEB or, for comparative purposes, to SEB plus CD28 and CD49d mAbs (the latter in assays with control APC). Note that TNF- pretreatment of T cells has little effect on the response, whereas TNF- pretreatment of APC increases SEB responder frequencies to the level of control APC plus dual-pathway costimulation. These results are representative of two independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Multiple lines of evidence indicate that the additional cytokine-producing cells observed in costimulation supplemented assays are indeed specific for CMV determinants and represent memory T cells that are primarily reactive with CMV (i.e., memory cells that differentiated in response to CMV) and not memory cells cross-reactive with CMV determinants that arose in response to distinct Ag(s). First, the expanded population of CMV responders generated in the presence of supplemental costimulation is still inhibitable by MHC class II blockade, indicating that the costimulation-dependent response remains contingent on classical Ag presentation. Second, increasing the activity of distinct costimulatory pathways progressively increases the responder frequencies to a maximum point (achieved with dual-pathway costimulation and which is characteristic of each cytokine and individual subject) and no further, a finding more consistent with a discrete Ag-specific population than a nonspecific response. Third, the observation that superantigen responses with maximal costimulation remain exquisitely restricted to the appropriate TCR-Vß-defined subsets for each superantigen rules out any significant bystander activation and, when combined with observation that costimulatory mAbs have no cytokine-inducing activity by themselves, leaves no obvious mechanism for a completely nonspecific induction of cytokine-producing cells. Finally, there is the issue of cross-reactivity. If a non-CMV-associated Ag or group of Ags (i.e., Ags to which an individual’s exposure is independent of CMV) were common enough to account for the ubiquitous augmentation of responder frequencies brought about by increasing costimulation in CMV-exposed subjects, the responses to these putative cross-reactive Ags should also be observable in CMV-nonexposed (seronegative) individuals when their PBMC are incubated with CMV plus maximal costimulation. However, all seven CMV-seronegative subjects examined in this manner failed to manifest such a response, effectively ruling out such cross-reactivity.

Taken together, these data strongly support the conclusion that resting, CMV-specific memory CD4⁺ T cells taken directly from the peripheral blood of CMV-exposed subjects vary in their costimulatory requirements. Given the emerging concept that costimulation promotes T cell activation by lowering the number of TCR-derived signals required for any given activation response (7, 8, 9, 11, 22), this would translate into a heterogeneity among normal CMV-specific memory T cells in their activation threshold for Ag-induced cytokine synthesis. Indeed, detailed analysis of the costimulatory requirements of this Ag-specific population, particularly the requirement for the cooperative activity of distinct costimulators for the triggering of some T cells (Figs. 1, 6, and 7), suggest a spectrum of activation thresholds. While this spectrum is likely continuous, our data would suggest that the CMV-specific CD4⁺ memory T cells of the cohort of subjects studied here can be divided into three broad subsets: 1) cells with low costimulatory requirements ("low-threshold" cells; triggerable without exogenous costimulation), accounting for 20% and 36% of total CMV-specific cells capable of producing IL-2 and IFN-, respectively; 2) cells with moderate costimulatory requirements ("intermediate-threshold" cells; requiring optimal single-pathway or suboptimal dual-pathway costimulation for triggering), accounting for 29% and 35% of total CMV-specific cells capable of producing IL-2 and IFN-, respectively; and finally, 3) cells with high costimulatory requirements ("high-threshold" cells; requiring optimal dual-pathway costimulation for triggering), accounting for 51% and 29% of total CMV-specific cells capable of producing IL-2 and IFN-, respectively.

Although the memory T cell-triggering thresholds for both IFN- and IL-2 were analogously heterogeneous in these studies, there were consistent differences in the relative triggering thresholds for these two cytokines. These differences can best be appreciated by observing that maximum dual-pathway costimulation increased the frequency of IFN--producing CMV responders by 2.75-fold, whereas concurrently measured IL-2-producing responders increased by 4.9-fold (Fig. 1A). Multiparameter analyses simultaneously examining IFN- and IL-2 on these cells confirm that the vast majority of CMV-triggered cells produce IFN-, with the relative fraction of these cells that also produce IL-2 increasing with increasing levels of exogenous costimulation—from a mean 36% of cells triggered by CMV alone to 67% of cells triggered by CMV plus CD28 and CD49d mAbs (data not shown). Thus, in our hands, IL-2 synthesis is relatively more dependent on costimulation as compared with IFN-. These data mirror the observations of Itoh et al. (9), who demonstrated that cloned murine CD4⁺ T cells specific for pigeon cytochrome c show hierarchical set points for different cytokines. In their clones, IFN- synthesis was also triggered by lower levels of TCR signaling or CD28-mediated costimulation than IL-2, such that, for example, the secretory profiles of the clones changed from IFN- alone to IFN- plus IL-2 as costimulation intensity was increased. Thus, the relative Ag density and availability of costimulation affects not only the number of triggerable T cells, but also the cytokine profile of the triggered cells.

The observation that CD4⁺ T cell stimulation by superantigens shows similar patterns of activation-threshold heterogeneity as CMV-specific responses strongly argues that such threshold heterogeneity is a common feature of CD4⁺ memory T cells and not unique to CMV. Indeed, preliminary data reveals analogous (although not identical) variability in the costimulatory requirements of CD4⁺ memory T cell responses to other Ags, both viral and nonviral (L.J.P. and V.C.M., unpublished observations). In addition, the demonstration of threshold heterogeneity with TCR-Vß-mediated superantigen stimulation suggests that differences in the fine specificity of the CMV-specific T cells does not account for the threshold heterogeneity observed among these cells. However, the strength of superantigen signaling can be differentially influenced by the TCR--chains (21) and perhaps other factors. We initially explored this issue by examining the differences in (cytokine-producing) responding cell fractions to SEB with or without dual-pathway costimulation on T cells expressing TCR-Vß-17 in combination with a single TCR-V-chain (V-2; 0.2% of CD4⁺ T cells express this TCR heterodimer, which were analyzed by acquisition gating after staining PBMC stimulated with SEB ± costimulation for CD4 vs cytokine vs Vß-17 vs V-2); our results indicated a similar responder cell-frequency enhancement with exogenous costimulation on this Vß-17/V-2 subset as demonstrated for the overall Vß-17 expressing population (data not shown).

These data argue against the possibility that heterogeneity in superantigen signaling through different V-chains accounts for the activation-threshold heterogeneity among superantigen-stimulated cells, but they did not completely address the broader issue of whether the observed threshold heterogeneity for superantigen and even Ag-activated cells primarily reflects differences in TCR signaling (due to variability in TCR avidities in the polyclonal populations examined) or, alternatively, differences in a downstream activation "set point." We thus investigated this issue more thoroughly by comparing patterns of TCR down-regulation (which is proportionate to and reflective of TCR signaling strength, see Results) among T cells triggered by superantigen (or Ag) alone vs superantigen (or Ag) plus maximal costimulation. These results clearly demonstrate that in the context of our system (i.e., using optimal Ag or superantigen concentrations), costimulation is not acting to recruit cells with lower levels of TCR signaling into the cytokine-producing subset but rather increases the cytokine-producing subset homogeneously across the spectrum of TCR signal strengths (Figs. 8 and 9). These down-regulation data strongly argue that the predominant mechanism for the activation-threshold heterogeneity described herein is independent of TCR signaling strength, and thus, of clonally determined differences in TCR fine specificity.

The mechanism(s) by which costimulatory signals accomplish activation-threshold reduction are controversial. It has been proposed that costimulators induce signal transduction pathways that ultimately act to provide transcription factors for cytokine gene promoters (or promoters for genes involved in other aspects of activation), thereby reducing the number of such factors required to be generated by TCR-derived signals (9, 19, 26). Another proposal suggests costimulators act at the cell surface to alter topography of TCR signaling units and thus enhance the signal-transduction capability of these units (27). In either case, it is possible that differences in the constitutive level or activity of downstream signaling components or transcription factors determines the relative requirement for costimulation after TCR ligation and thus "presets" activation thresholds. We propose that these putative downstream thresholds are subject to regulation during T cell differentiation and that this regulation accounts for the memory T cell-threshold heterogeneity demonstrated here.

Whatever the precise molecular mechanism of this memory T cell-threshold heterogeneity, its existence has profound implications for the operation and regulation of immune effector responses in vivo. Most importantly, it follows that the number of pathogen-specific T cells triggered at any given tissue site of pathogen invasion and replication, and consequently the degree of immune activity initiated by these T cells, will depend not only on the absolute frequency of pathogen-specific T cells migrating into that site, but also the Ag-presenting/costimulatory capabilities of local APC. The number and mix of local APC types (predominantly dendritic cells, macrophages, and B cells for CD4⁺ memory T cells, but perhaps in certain circumstances MHC class II-expressing nonhematolymphoid cells as well (28)) varies among tissues, and their functionality can be regulated by activation- or inflammatory cytokine-induced changes in Ag-presenting capabilities (23, 24, 25). Sites with few professional APC or in which pathogen invasion and replication has initiated only a low level of inflammation-induced up-regulation of APC costimulatory activity will be able to maximally trigger only relatively few pathogen-specific T cells (the low-threshold subset), analogous to the in vitro stimulation of PBMC containing resting APC with Ag alone. In Th1-type responses, these T cells would largely produce the effector cytokine IFN- (as well as TNF-, which closely follows IFN- in its synthesis patterns; Ref. 10 and data not shown) and relatively little IL-2, which would limit clonal expansion and favor early apoptosis of the triggered effector cells. If this low-intensity, self-limiting immune response controls the invasion, the microbial threat would be thwarted with little danger of a potentially damaging, over-zealous immune response or the initiation of autoimmunity. In contrast, if the initial pathogen invasion is overwhelming, or if the pathogen is not controlled by this low-intensity response, the inflammatory up-regulation of APC activity induced by the pathogen and/or pathogen-mediated damage to host tissues will potentially activate all available pathogen-specific T cells, analogous to our demonstration of maximal effector frequencies after APC pretreatment with TNF- (Fig. 10). As discussed above, relatively more of these effector cells would produce IL-2, favoring the continuation and expansion of the response. In this situation, control of pathogen replication would require bringing the maximum force of that individual’s pathogen-specific immune response into play, despite the risk of collateral damage to host tissues, or the potential for inducing autoimmunity. Thus, in this model, activation-threshold heterogeneity and hierarchical activation thresholds for different cytokines allow the immune system to instantly adjust "functional" Ag-specific memory CD4⁺ T cell frequencies to local conditions upon the tissue extravasation of these cells and thereby provide a degree of immune responsiveness that best conforms to local needs.

Given this model, it becomes particularly important to understand the origin of memory T cell-threshold heterogeneity and its regulation. The spectrum of memory T cell activation thresholds might, like cytokine synthesis or homing heterogeneity, be "tunable" on an Ag- or pathogen-specific basis, such that effector responses to a particular pathogen may be preprogrammed to "lean" toward maximum immune responsiveness, or alternatively, toward prevention of pathologic immunity. The naive to memory T cell transition in secondary lymphoid tissue is a prime candidate for such regulation as the microenvironmental milieu supporting this process differentially regulates the development of new cytokine synthesis and homing functions among differentiating memory cells (1, 2, 3, 4), and this process is thought to include the broad change in activation requirements that differentiate naive from memory cells (1, 2, 29, 30). It is possible that, like changes in cytokine synthesis and homing function, a set of specific regulatory factor(s)—cytokines, adhesive interactions, Ag densities, etc.—act to differentially lower the activation thresholds of "transitioning" T cells. Alternatively, or additionally, memory-cell activation thresholds might be regulated after the naive to memory transition by the periodic exposure of recirculating memory T cells to activating and/or subthreshold TCR stimuli derived from interactions with specific Ag peptides, cross-reactive peptides, or perhaps only self peptides in the context of autologous MHC class II molecules (5). With this proposal, the type, duration, frequency, and microenvironment of such stimuli would determine whether a cell has a high or low activation threshold. Finally, it cannot necessarily be presumed that naive T cell activation thresholds are homogeneous, and it remains formally possible that at least part of the observed memory T cell activation-threshold heterogeneity originated during thymic differentiation, and persisted through the naive to memory cell differentiation process (31).

Also, the observation that three distinct costimulatory pathways—CD28, CD49d, and CD5—can apparently function interchangeably and cooperatively in recruiting high-threshold T cells into an effector response has interesting physiologic implications. First, the fact that there are at least three such pathways provides a possible explanation for conflicting claims regarding the dependence (or lack thereof) of memory/effector T cell responses on costimulation via CD28 and its major APC ligands CD80 and CD86 (32). Clearly, depending on availability of the other active costimulatory ligands in local microenvironments, a particular response may not require CD28. In addition, this potential redundancy brings into question the clinical value of therapeutic blockade of a single costimulatory pathway (e.g., CD28), at least in regard to memory/effector T cell responses. Because of the other costimulatory pathways, significant and perhaps even maximal effector responses may still be achievable in the face of such blockade. Indeed, our data would suggest that therapeutic interference with memory/effector T cell costimulation might require blockade of multiple pathways for optimal clinical effectiveness.

Finally, the existence of memory T cell activation threshold heterogeneity also has the effect of introducing considerable additional complexity to the measurement and interpretation of Ag-specific T cell function in clinical situations. Our results strongly argue that if costimulation intensity is not taken into specific account, measurements made by such T cell assays (both the frequency of Ag-specific T cells and, given the hierarchical activation thresholds of different cytokines, their function) will be influenced by APC-activation status and function. One might envision, for example, two subjects with identical T cell frequencies and functional characteristics with respect to a given Ag manifesting quantitatively and qualitatively different responses if one subject had experienced activation of circulating APC with consequent up-regulation of costimulatory ligands. One solution to this problem would be to provide supplemental maximal costimulation in all assays so that results reflect the overall cohort of Ag-specific memory T cells. However, in any given clinical situation it remains to be determined whether the overall memory T cell cohort or some "lower"-threshold subset of the overall cohort is the most immunologically relevant. Thus, detailed understanding of immunologic status may ultimately require utilization of assays that standardize APC function and then define the spectrum of activation-threshold heterogeneity for each cohort of Ag-specific memory T cells.

	Acknowledgments

 
We acknowledge the invaluable assistance of the following individuals: C. Pitcher, Dr. R. Scheuermann, Dr. M. Siegelman, and Dr. J. Uhr for helpful discussion and invaluable advice, M. Suni and J. Ruitenberg for the preparation and testing of fluorochrome-conjugated anti-cytokine mAbs, and B. Abrams for synthesis of the Tripod Ab reagent.

	Footnotes

 
¹ This work was supported by National Institutes of Health Grant AI31545 and by a grant from the Texas Higher Education Coordinating Board.

² Address correspondence and reprint requests to Dr. Louis J. Picker, Department of Pathology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, Texas 75235-9072. E-mail address:

³ Abbreviations used in this paper: HSV, herpes simplex virus; AP, allophycocyanin; dPBS, Dulbecco’s PBS; PE, phycoerythrin; PerCP, Peridinin chlorophyll protein; SE, staphylococcal enterotoxin; SMCC, succinimidyl 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate; TSST, toxic shock syndrome toxin; VZV, varicella zoster virus.

Received for publication March 27, 1998. Accepted for publication July 13, 1998.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Picker, L. J., M. H. Siegelman. 1993. Lymphoid tissues and organs. W. E. Paul, ed. Fundamental Immunology 3rd Ed.145. Raven Press, New York.
2. Ahmed, R., D. Gray. 1996. Immunologic memory and protective immunity: understanding their relation. Science 272:54.[Abstract]
3. Paul, W. E., R. A. Seder. 1994. Lymphocyte responses and cytokines. Cell 76:241.[Medline]
4. Butcher, E. C., L. J. Picker. 1996. Lymphocyte homing and homeostasis. Science 272:60.[Abstract]
5. Grossman, Z., W. E. Paul. 1992. Adaptive cellular interactions in the immune system: the tunable activation threshold and the significance of subthreshold responses. Proc. Natl. Acad. Sci. USA 89:10365.[Abstract/Free Full Text]
6. Goodnow, C. C.. 1996. Balancing immunity and tolerance: deleting and tuning lymphocyte repertoires. Proc. Natl. Acad. Sci. USA 93:2264.[Abstract/Free Full Text]
7. Viola, A., A. Lanzavecchia. 1996. T cell activation determined by T cell receptor number and tunable thresholds. Science 273:104.[Abstract]
8. Bachmann, M. F., E. Sebzda, T. M. Kundig, A. Shahinian, D. E. Speiser, T. K. Mak, P. S. Ohashi. 1996. T cell responses are governed by avidity and costimulatory thresholds. Eur. J. Immunol. 26:2017.[Medline]
9. Itoh, Y., R. N. Germain. 1997. Single cell analysis reveals regulated hierarchical T cell antigen receptor signaling thresholds and intraclonal heterogeneity for individual cytokine responses of CD4⁺ T cells. J. Exp. Med. 186:757.[Abstract/Free Full Text]
10. Waldrop, S. L., C. J. Pitcher, D. Peterson, V. C. Maino, L. J. Picker. 1997. Determination of antigen-specific memory/effector CD4⁺ T cell frequencies by flow cytometry: evidence for a novel, antigen-specific homeostatic mechanism in HIV-associated immunodeficiency, J. Clin. Invest. 99:1739.[Medline]
11. Wells, A. D., H. Gudmundsdottir, L. A. Turka. 1997. Following the fate of individual T cells throughout activation and clonal expansion: signals from T cell receptor and CD28 differentially regulate the induction and duration of a proliferative response. J. Clin. Invest. 100:3173.[Medline]
12. Galili, U., M. Schlesinger. 1974. The formation of stable E rosettes after neurominidase treatment of either human peripheral blood lymphocytes or sheep red blood cells. J. Immunol. 112:1628.[Abstract/Free Full Text]
13. Shevach, E. M.. 1993. Accessory molecules. W. E. Paul, ed. Fundamental Immunology 3rd Ed.531. Raven Press, New York.
14. Shimizu, Y., G. A. Van Seventer, K. J. Horgan, S. Shaw. 1990. Costimulation of proliferative responses of resting CD4⁺ T cells by the interaction of VLA-4 and VLA-5 with fibronectin or VLA-6 with laminin. J. Immunol. 145:59.[Abstract]
15. Cayabyab, M., J. H. Phillips, L. L. Lanier. 1994. CD40 preferentially costimulates activation of CD4⁺ T lymphocytes. J. Immunol. 152:1523.[Abstract]
16. Hintzen, R. Q., S. M. A. Lens, K. Lammers, H. Kuiper., M. P. Beckmann, R. A. W. van Lier. 1995. Engagement of CD27 with its ligand CD70 provides a second signal for T cell activation. J. Immunol. 154:2612.[Abstract]
17. Reinhold, M. I., F. P. Lindberg, G. J. Kersh, P. M. Allen, E. J. Brown. 1997. Costimulation of T cell activation by integrin-associated protein (CD47) is an adhesion-dependent, CD28-independent pathway. J. Exp. Med. 185:1.[Abstract/Free Full Text]
18. Murakawa, Y., Y. Minami, W. Strober, S. P. James. 1992. Association of human lymph node homing receptor (Leu 8) with the TCR/CD3 complex. J. Immunol. 148:1771.[Abstract]
19. Udagawa, T., D. G. Woodside, B. W. McIntyre. 1996. 4ß1 (CD49d/CD29) integrin co-stimulation of human T cells enhances transcription factor and cytokine induction in the absence of altered sensitivity to anti-CD3 stimulation. J. Immunol. 157:1965.[Abstract]
20. Marrack, P., J. Kappler. 1990. The staphylococcal enterotoxins and their relatives. Science 248:705.[Abstract/Free Full Text]
21. Smith, H. P., P. Le, D. L. Woodland, and M. A. Blackman. T cell receptor -chain influences reactivity to Mls-1 in Vß-8.1 transgenic mice. J. Immunol. 149:887.
22. Valitutti, S., S. Muller, M. Cella, E. Padovan, A. Lanzavecchia. 1995. Serial triggering of many T-cell receptors by a few peptide-MHC complexes. Nature 375:148.[Medline]
23. Sallusto, F., M. Cella, C. Danieli, A. Lanzavecchia. 1995. Dendritic cells use macropinocytosis and the mannose receptor to concentrate macromolecules in the major histocompatibility complex class II compartment: downregulation by cytokine and bacterial products. J. Exp. Med. 182:389.[Abstract/Free Full Text]
24. Cella, M., A. Engering, V. Pinet, J. Pieters, A. Lanzavecchia. 1997. Inflammatory stimuli induce accumulation of MHC class II complexes on dendritic cells. Nature 388:782.[Medline]
25. Reiser, C., G. Bock, H. Klocker, G. Bartsch, M. Thurnher. 1997. Prostaglandin E₂ and tumor necrosis factor cooperate to activate human dendritic cells: synergistic activation of interleukin 12 production. J. Exp. Med. 186:1603.[Abstract/Free Full Text]
26. Fraser, J. D., D. Straus, A. Weiss. 1993. Signal transduction events leading to T-cell lymphokine gene expression. Immunol. Today 14:357.[Medline] <