Abatacept modulates human dendritic cell-stimulated T-cell proliferation and effector function independent of IDO induction
Introduction
Recognition of antigenic peptides bound to MHC class II molecules expressed on APCs by the TCR is central to T-cell activation in the context of an adaptive immune response. Typically, this TCR-mediated signal alone is not sufficient for full T-cell activation and an independent costimulatory signal provided by contact with the APC plays a critical role in regulating the response to antigen (Ag). The second, antigen-independent or costimulatory signal is required to amplify the Ag-driven response and in its absence, T-cells fail to respond to subsequent Ag stimulation as a direct consequence of reduced IL-2 production [1]. The first identified and best characterized costimulatory receptor on T-cells is CD28 [2], [3]. CD28 is expressed by most CD4+ and, to a lesser extent CD8+ T-cells [4], and is required for both efficient cytokine production and proliferation in response to Ag [2], [5].
Signaling through CD28 induces key biochemical pathways that function in combination with TCR signaling to initiate and maintain T-cell responses [2], [6], [7]. Specifically, ligating CD28 on T-cells with an activating anti-CD28 monoclonal antibody or by using cell lines expressing the CD28 counter-receptors CD80 and CD86, in combination with low levels of anti-CD3 antibody or Ag, increases IL-2 production and cell cycle progression over that seen with anti-CD3 antibody alone [8]. CD28 signaling alone leads to the induction of a subset of genes that are similar to those induced through TCR signaling but are relatively short-lived and non-productive in the absence of concomitant TCR signaling [9]. The majority of these CD28 upregulated genes contain AUUUA sequence motifs in the 3′ untranslated region that are characteristic of unstable transcripts such as those for cytokines or inflammatory mediators. This seemingly unproductive signaling may contribute to the prolonged unresponsiveness or desensitization of T-cells to further CD28 activation [10]. In contrast, T-cell costimulation with anti-CD28 augments and stabilizes the expression of IL-2, TNFα, GM-CSF and IFNγ mRNA following their induction with anti-CD3 [11]. Although the effect of CD28 costimulation on T-cell gene regulation appears to be primarily quantitative by amplifying a set of TCR-induced genes, there are some gene transcripts that appear to be differentially regulated by CD28 [9]. The importance of these differentially regulated gene transcripts for T-cell activation is currently unknown.
CTLA-4, a molecule structurally homologous to CD28, also binds to CD80 and CD86 on the surface of APCs [12]. The binding of CTLA-4 to CD80 was first demonstrated using a soluble fusion protein consisting of the extracellular domain of CTLA-4 and a modified Fc portion of a human IgG1, referred to as CTLA-4Ig [12]. These binding studies demonstrated that CTLA-4Ig has a 10- to 20-fold higher affinity for CD80 than a similarly derived fusion protein of CD28. Later studies demonstrated that CTLA-4Ig binds to both CD80 and CD86 with similar avidities but with CD86 having a faster off rate than CD80 [13]. CTLA-4Ig, by competing for CD80/CD86 binding to CD28, can inhibit in vitro proliferation in the context of an MLR and prolong allograft survival in vivo[14], [15]. Administration of CTLA-4Ig can also prevent disease onset and/or progression in animal models of human autoimmune diseases including systemic lupus erythematosus, collagen-induced arthritis, experimental autoimmune encephalomyelitis and experimental autoimmune glomerulonephritis [16], [17], [18], [19], [20].
In addition to blocking CD28–CD80/86 interactions, it has been proposed that CTLA-4Ig may induce ‘reverse signaling’ through CD80/86 to activate the immunosuppressive pathway of tryptophan catabolism in DCs [21]. The key product of this reverse signaling event is indeolamine 2,3-dioxygenase (IDO), an intracellular enzyme that breaks down tryptophan and prevents cell cycle progression in T-cells. The support for this mechanism comes from studies where CTLA-4Ig treatment of DCs induced the expression and activation of IDO both in vitro and in vivo leading to immunosuppression and long-term transplant survival in mice [21], [22]. However, the role of CTLA-4Ig in reverse signaling through CD80/86 in human DCs remains controversial. Munn et al. [23] published on the generation of a subset of human DCs that constitutively express IDO and inhibit T-cell proliferation in vitro. But Terness et al. [24], using similar experimental conditions, were unable to confirm these findings. They did, however, observe that IDO-positive DCs lose their capacity to sustain stimulation of preactivated T-cells.
The inappropriate activation of T-cells driven by autoantigens and APCs can result in uncontrolled T-cell proliferation, differentiation and cytokine production. This coupled with the body’s inability to modulate these responses can result in chronic autoimmune disorders such as RA, systemic lupus erythematosus, multiple sclerosis and psoriasis. Although T-cells play an important role in animal models of inflammatory arthritis and autoimmunity [25], the contribution and central role for T-cells in human disease was not fully appreciated until recent clinical trials with abatacept demonstrated activity in psoriasis and efficacy in RA [26], [27]. Abatacept is a homodimeric recombinant soluble fusion protein consisting of the extracellular domain of human CTLA-4 linked to a fragment (hinge–CH2–CH3 domains) of the Fc portion of human IgG1. Mutations in the hinge region of the human IgG1 portion of abatacept were engineered to block dimerization via the IgG domain but as a consequence also reduced complement fixation [8], [12], [28], ADCC activity and binding to Fc receptors, CD16 and CD32 [56].
Data from a multidose, phase I clinical trial in patients with psoriasis demonstrated that abatacept reduced the number of intralesional T-cells in a dose-dependent manner. More importantly, the changes in lesional T-cell numbers correlated with a reduction in epidermal proliferation, epidermal thickness, reversion of keratinocyte abnormalities and clinical improvement. In a 1-year RA trial, treatment with abatacept induced improvements in the signs and symptoms of RA and in physical function and slowed the progression of joint damage [27]. In contrast to currently marketed biologic agents such as inflixamab and etanercept that block the activity of a single cytokine, these trials provided the first conclusive evidence that an agent that modulates T-cell CD28-mediated costimulation, an earlier step in the inflammatory cascade, can have an impact on autoimmune diseases.
Despite all the information available from preclinical animal models and recent clinical trials, reports on the effect of abatacept on human peripheral blood T lymphocytes in the context of antigen presentation are extremely limited. In addition, many published studies have utilized versions of CTLA-4Ig that are different from abatacept in that the extracellular domain of CTLA-4 is fused to either IgG2a or IgG4 [29], [30], which results in different Fc effector functions [56]. In the present study, we have demonstrated that abatacept can effectively modulate CD28-mediated costimulation in human lymphocytes resulting in decreased antigen-specific T-cell proliferation and proinflammatory cytokine production in both a primary MLR and a TT recall memory response. Unlike the activity reported with other versions of CTLA-4Ig, the effect of abatacept on T-cell proliferation and cytokine production was independent of IDO generation by the MoDCs.
The medium used was RPMI 1640 (Invitrogen, Grand Island, NY), supplemented with 10% heat-inactivated FCS (Summit Biotechnology, Fort Collins, CO) or 1% autologous serum or 10% heat-inactivated pooled human serum (prepared in-house from normal male donors) as indicated. Abatacept was provided by Bristol–Myers Squibb. Chi L6 is a chimeric fusion protein consisting of the variable region of murine L6 antigen, combined with a human IgG1 constant region, and was used as a control fusion protein (control Ig) in these studies. Tetanus toxoid (TT) was obtained from List Biological Laboratories (Campbell, CA). LPS, PGE2, 1-methyl-d-tryptophan (1-MT), l-tryptophan and l-kynurenine were obtained from Sigma-Aldrich (St. Louis, MO). Recombinant IL-1β, IL-4, IL-6, GM-CSF and TNFα were purchased from Peprotech (Rocky Hill, NJ).
PBMC were obtained by density-gradient separation (Lymphocyte Separation Media; Mediatech Inc., Herndon, VA) of EDTA-treated whole blood from normal healthy donors. T-cells were prepared from E+ fractions of PBMC rosetted with SRBC (Colorado Serum Company; Denver, CO). Mature MoDC were prepared by adherence of monocytes from E− fractions of PBMC from normal donors in 6-well tissue culture plates, followed by extensive gentle washing to remove non-adherent cells. Two methods were used for generation of MoDCs: adherent cells were cultured for 7 days in RPMI containing either 10% FCS (for primary MLRs and IDO assays) or 1% autologous serum (for recall TT responses), together with 100 ng/ml GM-CSF and 50 ng/ml IL-4, with one-half the medium changed every other day and replaced with fresh medium containing the same concentration of cytokines. On day 7, cells that had been cultured with 10% FCS were matured with LPS (1 μg/ml) for 24 h, and cells that had been cultured with 1% autologous serum were matured with a cocktail containing IL-1β (10 ng/ml), TNFα (10 ng/ml), IL-6 (1000 U/ml) and PGE2 (1 μg/ml) for 24 h. MoDCs generated by either method were routinely > 95% CD83+ by day 8 (as described in text). An alternative APC, the EBV-transformed B-cell line from a normal donor that is HLA mismatched to the other blood donors, PM-LCL, was used as a comparator in initial studies.
Culture supernatants were collected from MoDC treated for 24 h with control Ig or abatacept in the presence of 200 μM l-kynurenine. Samples were precipitated with TCA, filtered, and 80 μL of each sample was injected onto a C-18 column (Waters Symmetry 300 C-18, 5 μM, 3.9 × 150 mm). The samples were eluted with a linear gradient of 0–80% acetonitrile, 0.1% TFA, at a flow rate of 0.5 ml/min. Absorbance was measured at 254 nm. The peak of kynurenine was determined by comparison to the retention time previously determined with a standard solution of l-kynurenine, extracted with TCA and filtered as above.
RNA was extracted from MoDCs and PM-LCLs treated for 24 h with abatacept or control Ig using RNeasy kit (Qiagen, Valencia, CA), and cDNA was synthesized from RNA using oligo-dT and Superscript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA), according to manufacturer’s instructions. IDO-specific mRNA was quantified relative to GAPDH using the following primers: IDO forward, 5′-GCGCTGTTGGAAATAGCTTC-3′; IDO reverse, 5′-TTTGGGTCTTCCCAGAACC-3′; GAPDH forward, 5′-AATTCCATGGCACCGTCAAG-3′; and GAPDH reverse, 5′-GAAGACGCCAGTGGACTCCA-3′. Quantitative RT-PCR was performed using SYBR GREEN PCR Master Mix (Applied Biosystems, Foster City, CA) according to manufacturer’s instructions. The reactions were run on an ABI PRISM 7000 Sequencing Detection System (Applied Biosystems). The ratio of gene of interest to housekeeping gene was calculated according to the formula: ratio = 2− dCt (dCT = mean Ct gene − mean Ct housekeeping).
For MLR proliferation assays measuring titrations of abatacept, T-cells were cultured at 1 × 105 cells/well in quadruplicate wells together with 2 × 103 of allogeneic MoDC or irradiated PM-LCLs (10,000 rads) as APC in 96-well round-bottom plates in a total volume of 200 μl of 10% FCS-RPMI. Abatacept was added at the indicated concentrations. On day 5 after initiation of the MLR, cultures were pulsed with 1 μCi of 3[H]-thymidine (PerkinElmer, Boston, MA) for 6 h, harvested on a Packard cell harvester (PerkinElmer) and counted by liquid scintillation on a +Packard TopCount NXT (PerkinElmer).
For MLR experiments measuring cytokines at various time points in response to MoDCs as stimulator cells, assays were performed by combining 1 × 105 T-cells/well of a 96-well round-bottom plate with 1 × 104 (for measurement of IL-2 and TNFα), or 2 × 103 (for measurement of IFNγ) allogeneic MoDCs in a total volume of 250 μL of 10% FCS-RPMI. Control Ig or abatacept was added at a final concentration of 30 μg/ml. Assays were performed in triplicate to measure cytokine release at 24, 48 and 72 h after initiation of the MLR.
For TT recall proliferation assays, 1 × 105 T-cells were combined with 2 × 103 autologous MoDCs generated from normal healthy donors recently vaccinated against tetanus and were cultured in quadruplicate in 96-well round-bottom plates in a total volume of 200 μL RPMI containing 10% human serum. Abatacept was added at the indicated concentrations and TT added to a final concentration of 250 ng/ml. On day 5 after initiating the response, cultures were pulsed for 6 h with 3[H]-thymidine, harvested and incorporated label determined by liquid scintillation counting.
For TT assays measuring cytokine production at various time points, 2 × 105 T-cells were combined with 1 × 104 (for IL-2 and TNFα) or 2 × 103 (for IFNγ) autologous MoDCs per well in a total volume of 250 μL of RPMI containing 10% human serum. Control Ig or abatacept was added at a final concentration of 30 μg/ml, and TT was added at a final concentration of 250 ng/ml. Assays were performed in triplicate to measure cytokine release at 24, 48 and 72 h after the addition of TT.
IL-2 and TNFα were detected in supernatants using Duoset ELISA development kits (R&D Systems; Minneapolis, MN). IFNγ was measured using paired antibodies and recombinant cytokine from Pierce Biotechnology (Rockford, IL). All kits and Abs were used according to the manufacturer’s recommendations. For cytokine assays involving various time points, all ELISA plates were set up with standards and controls, and fresh supernatants were added to the ELISA plates as each time point was completed. After samples from all of the time points were added, the assay plates were processed and read on a SpectraMax Plus spectrophotometer (Molecular Devices Corp., Sunnyvale, CA). Data were analyzed using Softmax software by comparison against a standard curve generated using recombinant cytokines at known concentrations.
Measurements of cytokines GM-CSF, IL-1, IL-4, IL-6, IL-8, IL-10, IL-13, IP-10, MCP-1 and MIP-1α in supernatants were performed using a multiplexed immunoassay kit from Linco Research (St. Charles, MO) according to manufacturer’s instructions. Data were generated and analyzed by comparison against a standard curve using recombinant cytokines at known concentrations, using Bioplex software on a Luminex array reader (Bio-Rad, Hercules, CA).
Statistical analyses were conducted using one-way ANOVA with Tukey–Kramer multiple comparison posttest using GraphPad InStat 3.05 software (GraphPad Software, Inc., San Diego, CA). Data are expressed as mean ± SD and p values below 0.05 were considered significant.
The MoDCs used as APCs in our studies had very low expression of CD14 and were > 95% positive for CD83, indicative of mature monocyte-derived dendritic cells. These cells also expressed high levels of HLA-DR, CD80 and CD86 (data not shown). The ability of abatacept to modulate T-cell responses to allogeneic MoDC was evaluated in the context of an MLR. Inhibition with abatacept was dose dependent, with significant inhibition (p < 0.05) of proliferation observed at 0.3 and 3 μg/ml abatacept, with maximum inhibition (> 95%) at 3 μg/ml (Fig. 1A). No additional inhibition (p > 0.05) was observed between 3 and 100 μg/ml of abatacept. Similar results and inhibition profiles were obtained using a B-cell line, PM-LCL, as the APC to drive T-cell proliferation (Fig. 1B). In both instances, the concentration of abatacept required for maximal inhibition of T-cell proliferation in vitro was well within trough plasma levels observed in rheumatoid arthritis patients receiving a clinically effective dose of 10 mg/kg [31].
A potential mechanism for the inhibition of T-cell proliferation seen with abatacept could be attributable to “reverse” signaling through CD80/86 on the APC. Recent reports have demonstrated that ligation of CD80/86 on DCs can trigger, through “reverse” signaling, the functional activation of IDO leading to tryptophan catabolism that is sufficient to inhibit antigen-specific T-cell proliferation [32]. In some instances, CTLA-4 Ig has also been shown to induce the production of IDO in both human and mouse DCs [33], [34]. For this reason, we determined if abatacept, by binding to CD80/86 on APCs, could trigger IDO activation which contributed to the inhibitory effects of abatacept on T-cell proliferation. Minimal, if any, IDO mRNA was detected in PM-LCLs after 24 h treatment with abatacept (Fig 2A). Similarly, treatment of MoDCs with abatacept for 24 h failed to alter the mRNA level of IDO in MoDCs when compared to either no treatment or control Ig (Fig. 2A). In contrast, treatment of MoDCs with a commercially available CTLA-4 Fc (R&D Systems) resulted in a measurable increase in IDO mRNA (data not shown). Additional studies were conducted with abatacept and MoDCs to rule out discrepancies between IDO expression and enzyme activity that could result from posttranslational regulation of the enzyme [35], [36]. The absence of an effect of abatacept on IDO mRNA was confirmed by analysis of IDO enzymatic activity (Fig. 2B), measured as accumulation of kynurenine in the culture supernatant. Kynurenine is a stable downstream metabolite produced by IDO and is a specific qualitative marker for IDO activity [37]. Neither abatacept nor control Ig, as compared to no treatment, altered the amount of kynurenine produced by MoDCs, ranging from 10–50 μM depending on the specific blood donor (Fig. 2B). Consistent with the finding that abatacept did not induce IDO activity in MoDCs, addition of the IDO-inhibitor 1-MT (200 μM) failed to reverse the inhibition of T-cell proliferation observed with abatacept, up to 30 μg/ml (Fig. 2C). Even though T-cell proliferation was unaffected, treatment of cells with 1-MT did substantially reduce kynurenine levels (data not shown). Thus, unlike the activity reported for other versions of CTLA-4 Ig, abatacept does not inhibit T-cell proliferation by inducing IDO activity in the APCs used to drive an MLR in these experiments.
Three important T-cell-derived cytokines that impact TH1-driven autoimmune responses are IL-2, TNFα and IFNγ. Studies using anti-CD28 antibodies and CD28 knockout mice have demonstrated that CD28 signaling can augment the production of these cytokines in response to TCR activation in vitro and play a critical role in IL-2 production and T-cell survival in vivo[2], [38]. The gene transcription of these cytokines can also be generated by CD28 receptor-mediated signaling independent of TCR signaling [9], [39]. Thus, by blocking CD28-mediated signaling by the addition of abatacept, we would expect a significant effect on antigen-driven cytokine production. The extent of cytokine production was dependent upon the APC to T-cell ratio in these assays. Thus, the optimal ratio of APCs to T-cells using either allogeneic MoDC was evaluated to determine the effects of these different ratios on the level of individual cytokine release (data not shown). A ratio of 1:10 MoDC to T-cells was optimal for evaluating the effect of abatacept on IL-2 and TNFα release, whereas a MoDC:T-cell ratio of 1:50 was optimal for evaluating the effect of abatacept on IFNγ production. The concentration of abatacept used in these studies was 30 μg/ml, covering the trough mean plasma levels observed in rheumatoid arthritis patients receiving an effective dose [31].
Consistent with the observed inhibition of T-cell proliferation, IL-2 release was significantly reduced (p < 0.01 and p < 0.05, respectively) at 24, 48 and 72 h, with maximal inhibition ranging from 90% to 100% (Fig. 3A). TNFα release was similarly reduced (p < 0.05) at all 3 time points, with maximal inhibition of 89% observed at 72 h (Fig. 3B). The impact of abatacept on IFNγ was significant but somewhat less dramatic than IL-2 or TNFα, with a maximal inhibition of 92% observed at 48 h but reduced to 74% inhibition by 72 h as IFNγ production increased (Fig. 3C). These findings are consistent with previous reports where CTLA-4Ig blocked the induction of IL-2 and IFNγ mRNA transcripts in human peripheral blood T-cells by 4 h following activation with a specific alloantigen [14]. Similarly, CTLA-4Ig administration diminished a DTH response in KLH-challenged mice by blocking IL-2 and IFNγ production by activated T-cells [40]. Surprisingly, there have been no reports on the effects of CTLA-4Ig on TNFα.
A more extensive analysis of MLR-dependent cytokine production revealed that abatacept significantly reduced the levels of GM-CSF, MIP-1α, IP-10, IL-6, IL-10 and IL-13, while having little impact on IL-1 and IL-4 (Fig. 4, upper and lower panels). More importantly, abatacept had no effect on the MoDC-derived cytokines, MCP-1 and IL-8, suggesting that abatacept was not directly modulating APC cytokine production in these assays. These data strongly suggest that abatacept can alter T-cell responses without affecting signaling into the APCs.
To determine the importance of CD28–CD80/86 interactions in memory T-cell responses, we compared the effect of abatacept on proliferation and cytokine production in the context of a memory TT recall response, using isolated T-cells plus autologous MoDC. MoDCs for these experiments were differentiated and matured in autologous serum as described in Materials and methods in order to avoid T-cell responses to xenogeneic serum antigens. As was observed in the primary MLR, increasing concentrations of abatacept resulted in a dose-dependent inhibition of T-cell proliferation in response to TT using isolated T-cells plus autologous MoDC (Fig. 5). Significant inhibition of proliferation (p < 0.01) was observed with 1.0 μg/ml abatacept, with maximal inhibition of proliferation at 10 μg/ml. No additional inhibition (p > 0.05) was observed between 10 and 100 μg/ml of abatacept. The concentration of abatacept required for maximal inhibition of a recall response appeared to be slightly greater than that required for the maximal inhibition of a primary MLR (3 μg vs. 10 μg/ml). In addition, similar to what was observed in a primary MLR, the maximal inhibition of T-cell proliferation using MoDC was almost complete (90%).
The kinetics of cytokine release in a TT recall response was similar to that of the primary MLR (Fig. 6). However, the levels of IL-2 produced in a TT recall response were about 2-fold greater, and the levels of TNFα and IFNγ about 2- and 5-fold lower, respectively, than that observed in a primary MLR. Abatacept treatment significantly reduced the levels of IL-2 and IFNγ (p < 0.05) but, in contrast to what was seen in a primary MLR, had no effect on TNFα production. Thus, TNFα release in a TT recall response, unlike IL-2 and IFNγ production, appears to be independent of CD28 signaling.
A substantial body of experimental data, using either neutralizing antibodies, CTLA-4 fusion proteins and CD80/CD86/CD28 mouse knockouts, has demonstrated that T-cell activation can be effectively reduced by blocking the interaction of CD80/CD86 with CD28. However, many of these studies, and specifically studies with CTLA-4-Ig, have been conducted in rodents, with limited studies on human peripheral blood T-cells. In the present report, the effect of abatacept on human T-cell proliferation and cytokine production in both a primary MLR and a TT recall response was examined. This is important as abatacept differs from the other versions of CTLA-4Ig through the modifications engineered into the Fc portion of the molecule that reduce Fc effector function [41].
The maximal inhibition of proliferation observed with abatacept in an MLR using MoDCs as APCs was ≥ 95%, reached at a concentration of 3 μg/ml. This concentration is well below the mean trough plasma levels (15–30 μg/ml) of patients successfully treated with abatacept [27], [31]. This is somewhat different from what has been reported using allogeneic monocytes, PBMCs or EBV-transformed B cells as APCs in a primary MLR where ∼60–80% inhibition is achieved with CTLA-4Ig [14], [31], [42]. One explanation could be the result of differences in the expression of costimulatory molecules on monocytes or B cells versus MoDC, with the incomplete inhibition related to compensatory, T-cell costimulatory pathways that provide functional redundancy in the absence of CD28 signaling. For example, the ICOS /ICOS–ligand interaction is another costimulatory pathway that upregulates T-cell activation. ICOS is expressed on the T-cell surface after initial activation and its expression is partially dependent upon CD28 signaling [43], [44]. Although we observed similar levels of CD152 (CTLA-4), CD273 (PD-L2), CD274 (PD-L1), CD275 (ICOS-L), CD278 (ICOS) and CD279 (PD-1) on MoDcs or B cells and T-cells at various times during an MLR, both in the presence and absence of abatacept (data not shown), studies have shown that CTLA4-Ig can act in an additive fashion with ICOS-Ig to suppress T-cell proliferation [45]. This suggests that signaling through other costimulatory molecules may occur in the presence of CD28 inhibition.
Alternatively, it has been suggested that CTLA-4Ig and CTLA4-expressing T regulatory cells can induce IDO in APCs, specifically MoDCs, by signaling through CD80/86 [22], [46]. The majority of the studies supporting a role for IDO in CTLA-4Ig-induced immunosuppression have come from in vitro and in vivo mouse studies. Grohmann et al. [21] demonstrated that CTLA-4Ig indirectly upregulated IDO in murine DCs by induction of IFNγ. Consistent with this finding, the authors showed that antagonism of IDO with 1-methyltryptophan in vivo could block the protective effect of CTLA-4Ig on allograft survival. About the same time, Munn et al. [23] reported that a subset of MoDCs constitutively express IDO and can suppress an allogeneic T-cell response. In a subsequent set of studies [32], these authors demonstrated that CTLA-4Ig could induce IDO in this subpopulation of MoDCs resulting in inhibition of T-cell proliferation, and that this inhibition could be reversed by 1MT. However, in a more recent study, Terness et al. [24] were unable to show that this subset of MoDCs produced IDO or could suppress a T-cell response. Similarly, our studies failed to demonstrate induction of IDO with abatacept treatment or the ability of 1-MT to reverse the abatacept-mediated inhibition of T-cell proliferation. Overall, with limited studies in human cells and the varying structures of CTLA-4Ig used in mouse studies (different IgG tails with different capacities for binding to Fc receptors), the biologic relevance of the upregulation of IDO in APCs in CTLA-4Ig-mediated immunosuppression remains debatable.
IL-2, TNFα and IFNγ are cytokines that have been associated with autoimmune diseases and chronic transplant rejection. Inhibition of IL-2 release by abatacept in the primary MLR correlated well with the lack of T-cell proliferation under the same experimental conditions. This is not unexpected as T-cells require IL-2 for proliferation. This lack of IL-2 production by T-cells incubated with CTLA-4Ig may also result in the induction of reversible anergy in memory T-cells [47]. Our findings correlate well with earlier studies by Tan et al. [14], demonstrating that CTLA-4Ig blocks antigen-mediated activation of IL-2 and IFNγ mRNA expression in human T-cells, but they differ from those reported by Zhang et al. [42] where IL-2 and IFNγ release, in response to an EBV-transformed B cell line (ARC), were not inhibited by CTLA-4Ig unless there was concomitant treatment with CsA. The explanation provided by Zhang et al. [42] was that the strength of the response provided by using a transformed B cell line as the APC was too great to be inhibited by CTLA-4Ig alone. Since other APC to T-cell ratios were not explored in the Zhang et al. [42] study, it is difficult to determine whether the differences observed between these two studies reflect differences in the strength of response delivered by the APCs or a difference in the properties of the CTLA-4Ig fusion proteins (such as affinities for CD80/CD86) used in these studies.
TNFα release was also measured in the context of a primary MLR driven by MoDCs. It has been presumed that most of the TNFα production in autoimmune diseases such as RA is derived from monocytes/macrophages at the site, or sites, of inflammation. However, T-cell-derived TNFα can be produced in response to antigen and has been shown to make a significant contribution to the morbidity and mortality observed in mouse models of arthritis, graft-versus-host-disease and autoimmune hepatitis [48], [49], [50]. In those studies, significant numbers of both activated memory CD4+ and CD8+ T-cells expressed intracellular TNFα. Our studies are the first observation that abatacept can effectively inhibit the production of TNFα from T-cells in the context of an MLR. This was confirmed by evaluating the effect of abatacept on intracellular TNFα production in T-cells during an MLR (data not shown). More importantly, abatacept had no direct effect on TNFα production in monocytes activated with either LPS or immune complexes (data not shown).
Cytokine profiling of the supernatants from the MLRs revealed the presence of a number of T-cell cytokines and chemokines in addition to IL-2, TNFα and IFNγ. These included both the TH1 cytokines IL-6 and GM-CSF, and the TH2 cytokines IL-4, IL-10 and IL-13. In addition, there were significant levels of MoDC-derived IL-1, MCP-1 and IL-8. We did not observe an increase in IL-12p70 with MoDC as APC as others have reported [51], suggesting that the extent of MoDC activation during the MLR was somewhat limited. We also failed to detect IL-1α, IL-5, IL-7, IL-15, IL-17 and G-CSF. The levels of GM-CSF, IL-6, IL-10, IL-13, MIP-1α and IP-10 were significantly reduced following treatment with abatacept. In contrast, abatacept had no effect on IL-1, IL-4, IL-8 or MCP-1 production. The inability of abatacept to impact the levels of the MoDC-derived cytokines/chemokines IL-1, IL-8 and MCP-1 suggests that abatacept, by binding to CD80/86, does not initiate “reverse signaling” into the APC. This, taken together with the inability of abatacept to induce IDO or interfere with LPS-induced TNFα production in monocytes, suggests that it is unlikely that abatacept will have a substantial impact on innate immunity.
It has been reported that antigen-primed, memory T-cells are less dependent upon CD28-mediated costimulation [52]. However, some memory T-cells remain dependent on CD28 for their effector function. For example, the antigen-specific proliferation of peripheral circulating memory T-cells from patients with chronic berylliosis was inhibited when treated ex vivo with CTLA-4Ig [53]. However, the proliferation of memory T-cells from the lungs of these patients was unaffected by CTLA-4Ig. This indicates that in these patients, circulating memory T-cells remain dependent upon CD28 costimulation for effector function. Additionally, murine studies with CTLA-4Ig have demonstrated that inhibiting CD28 results in inhibition of memory responses and decreases in effector memory populations [54]. The requirement for a slightly higher concentration of abatacept (maximal inhibition of 10 μg/ml) compared to a primary MLR is consistent with the notion that memory cells are less dependent upon CD28 costimulation. In contrast to published findings where inhibition of proliferation by CTLA-4Ig in a tuberculin-recall response using PBMC as APC was about 60–70% [42], [47], we saw almost complete inhibition of a memory response driven by MoDCs. This suggests that the impact of abatacept on a memory response is likely to be dependent upon the context and strength of the T-cell:APC interaction. Taking all the study data into consideration, it is still reasonable to conclude that abatacept is likely to have a greater effect on naïve rather than memory or effector T-cells but that the effector function of circulating memory T-cells will still be significantly impacted.
Not only was the inhibition of proliferation with abatacept somewhat different between a primary and a TT recall response, but the impact on cytokine release was also different between these two responses. There was less overall inhibition of IL-2 and IFNγ production, with no impact on TNFα. The level of inhibition of IL-2 and IFNγ obtained with abatacept was not substantially different (65–75%) from the CTLA-4Ig-mediated inhibition of a tuberculin recall response where the maximal inhibition of these cytokines observed with CTLA-4Ig was 55–60% [47]. The inability of abatacept to block TNFα production in a TT recall response suggests a differential affect of abatacept on a naive versus memory T-cell response. As with the primary MLR assays, the TNFα produced in the TT recall response is most likely derived from T-cells, as others have reported that monocyte-derived TNFα is detectable by 2 h with a peak response within 3–4 h following challenge of PBMC with TT [55].
In summary, we have demonstrated that abatacept, a selective CD28-mediated costimulation modulator, can effectively modulate T-cell activation, as measured by cytokine production, and proliferation of human T-cells in the context of both a primary and a memory immune response. An important aspect of this study was our ability to demonstrate that abatacept does not mediate this effect by increasing IDO in the APC. Thus, these data confirm that abatacept inhibits peripheral T-cell proliferation and effector function via the modulation of CD28 signaling rather than through “reverse signaling” into the APC. This is also the first demonstration that abatacept reduces APC-driven TNFα production in a primary MLR. More importantly, the concentrations of abatacept that resulted in maximal inhibition of proliferation and cytokine production in all our in vitro studies were well within the abatacept trough plasma levels observed in rheumatoid arthritis patients receiving a clinically effective dose. Taken together, these data increase our understanding of the immunological mechanism of action of abatacept in humans.
References (56)
- et al.
Role of the CD28 receptor in T-cell activation
Immunol. Today
(1990) - et al.
Human B7-1 (CD80) and B7-2 (CD86) bind with similar avidities but distinct kinetics to CD28 and CTLA-4 receptors
Immunity
(1994) - et al.
Active suppression of allogeneic proliferative responses by dendritic cells after induction of long-term allograft survival by CTLA4Ig
Blood
(2003) - et al.
Regulation of human auto- and alloreactive T cells by indoleamine 2,3-dioxygenase (IDO)-producing dendritic cells: too much ado about IDO?
Blood
(2005) - et al.
CTLA4Ig: a novel immunoglobulin chimera with immunosuppressive properties
Methods Companion Methods Enzymol.
(1995) - et al.
Regulation of indoleamine 2,3-dioxygenase and tryptophanyl-tRNA-synthetase by CTLA-4-Fc in human CD4+ T cells
Blood
(2005) - et al.
Characteristics of interferon induced tryptophan metabolism in human cells in vitro
Biochim. Biophys. Acta
(1989) - et al.
CTLA4Ig: a novel immunoglobulin chimera with immunosuppressive properties
Methods Companion Methods Enzymol.
(1995) - et al.
The CD28-related molecule ICOS is required for effective T cell-dependent immune responses
Immunity
(2000) - et al.
Distinct and nonredundant in vivo functions of TNF produced by t cells and macrophages/neutrophils: protective and deleterious effects
Immunity
(2005)
Donor T cell-derived TNF is required for graft-versus-host disease and graft-versus-tumor activity after bone marrow transplantation
Blood
Clonal anergy is induced in vitro by T cell receptor occupancy in the absence of proliferation
J. Immunol.
CD28 activation pathway regulates the production of multiple T-cell-derived lymphokines/cytokines
Proc. Natl. Acad. Sci. U. S. A.
T-cell antigen CD28 mediates adhesion with B cells by interacting with activation antigen B7/BB-1
Proc. Natl. Acad. Sci. U. S. A.
Monoclonal antibody 9.3 and anti-CD11 antibodies define reciprocal subsets of lymphocytes
Eur. J. Immunol.
CD28/B7 interactions deliver a unique signal to naive T cells that regulates cell survival but not early proliferation
J. Immunol.
CD28 delivers a costimulatory signal involved in antigen-specific IL-2 production by human T cells
J. Immunol.
Binding of the B cell activation antigen B7 to CD28 costimulates T cell proliferation and interleukin 2 mRNA accumulation
J. Exp. Med.
Modulation of TCR-induced transcriptional profiles by ligation of CD28, ICOS, and CTLA-4 receptors
Proc. Natl. Acad. Sci. U. S. A.
CD28 engagement by B7/BB-1 induces transient down-regulation of CD28 synthesis and prolonged unresponsiveness to CD28 signaling
J. Immunol.
Regulation of lymphokine messenger RNA stability by a surface-mediated T cell activation pathway
Science
CTLA-4 is a second receptor for the B cell activation antigen B7
J. Exp. Med.
Induction of alloantigen-specific hyporesponsiveness in human T lymphocytes by blocking interaction of CD28 with its natural ligand B7/BB1
J. Exp. Med.
Treatment of murine lupus with CTLA4Ig
Science
Collagen-induced arthritis in the BB rat. Prevention of disease by treatment with CTLA-4-Ig
J. Clin. Invest.
Prevention and amelioration of collagen-induced arthritis by blockade of the CD28 co-stimulatory pathway: requirement for both B7-1 and B7-2
Eur. J. Immunol.
Inhibition by CTLA4Ig of experimental allergic encephalomyelitis
J. Immunol.
CD28-B7 blockade prevents the development of experimental autoimmune glomerulonephritis
J. Clin. Invest.
Cited by (61)
New structural formats of therapeutic antibodies for rheumatology
2018, Joint Bone SpineNew structural formats of therapeutic antibodies for rheumatology
2017, Revue du Rhumatisme (Edition Francaise)EFIS Lecture: Understanding the CTLA-4 checkpoint in the maintenance of immune homeostasis
2017, Immunology LettersCitation Excerpt :However this hypothesis was challenged by transcriptional profiling analysis that failed to identify gene expression changes in CD80/CD86 expressing cells after treatment with soluble CTLA-4 fusion proteins (including abatacept and belatacept) [56]. Furthermore, it was shown that abatacept inhibits T cell proliferation without increasing expression of IDO [57], consistent with a role for CTLA-4 in limiting CD28 engagement via binding to their shared ligands, rather than triggering the IDO pathway. In addition, CTLA-4 blockade was even more effective in IDO deficient mice than wildtype mice [58], an observation that is hard to reconcile with the notion that CTLA-4 operates via IDO induction.
A transendocytosis perspective on the CD28/CTLA-4 pathway
2014, Advances in ImmunologyCitation Excerpt :Signaling has also been reported to trigger IDO activity (Grohmann et al., 2002; Munn, Sharma, & Mellor, 2004) where it has been suggested that engagement of ligands via CTLA-4-Ig or by Treg can trigger the induction of the tryptophan-degrading enzyme IDO, with resultant immune suppression (Fallarino et al., 2003). Whether CTLA-4-Ig consistently performs this function is unclear (Mayer et al., 2013; Pree et al., 2007; Sucher et al., 2012) and it is possible that the Fc region of the reagents used can have impacts on APC (Davis, Nadler, Stetsko, & Suchard, 2008). One study was unable to identify changes in gene expression subsequent to CTLA-4-Ig binding (Carman et al., 2009).