Ctla-4 Activation of Phosphatidylinositol 3-kinase (pi 3- K) and Protein Kinase B (pkb/akt) Sustains T-cell Anergy without Cell Death

The balance of T-cell proliferation, anergy and apoptosis is central to immune function. In this regard, co-receptor CTLA-4 is needed for the induction of anergy and tolerance. One central question concerns the mechanism by which CTLA-4 can induce T-cell non-responsiveness without a concurrent induction of antigen induced cell death (AICD). In this study, we show that CTLA-4 activation of the phosphatidylinositol 3-kinase (PI 3-K) and protein kinase B (PKB/AKT) sustains T-cell anergy without cell death. CTLA-4 ligation induced PI 3K activation as evidenced by the phosphorylation of PKB/AKT that in turn inactivated GSK-3. The level of activation was similar to that observed with CD28. CTLA-4 induced PI 3K and AKT activation also led to phosphorylation of the pro-apoptotic factor BAD as well as the up-regulation of BcL-XL. In keeping with this, CD3/CTLA-4 co-ligation prevented apoptosis under the same conditions where T-cell non-responsiveness was induced. This effect was PI 3K and PKB/AKT dependent since inhibition of these enzymes under conditions of anti-CD3/ CTLA-4 co-ligation resulted in cell death. Our findings therefore define a mechanism by which CTLA-4 can induce anergy (and possibly peripheral tolerance) by preventing the induction of cell death. Copyright: ß 2008 Schneider et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: Wellcome Trust, BBSRC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors have no competing financial interests.


Introduction
CD28 and CTLA-4 have opposing effects on T-cell function by providing positive and negative signals, respectively [1][2][3]. Both bind CD80/86, with CTLA-4 exhibiting a preference for CD80 [4]. CTLA-4 negative regulation was shown by antibody ligation [5], and by the development of autoimmune disease in CTLA-4 deficient mice [6,7]. CD4 positive CTLA-4-/-T-cells are also resistant to anergy induction and tolerance [8]. In this context, CTLA-4 plays a crucial role in autoimmunity and anti-tumor responses [9]. Since anergy often leads to increased apoptosis, one central question concerns how CTLA-4 can anergize and maintain tolerance without inducing T-cell death. This issue could be key in the development of strategies to modulate tumor and transplant tolerance and rejection.
Receptor mediated induction of anergy without cell death is key to the maintenance of immune function and peripheral tolerance. One central question concerns the signaling mechanism used by CTLA-4 to induce long-term anergy and prevent the induction of cell death. In this study, we show that CTLA-4 activation of PI 3-K and PKB/AKT sustains T-cell anergy without cell death. CD3/ CTLA-4 co-ligation rescued cells from apoptosis under the same conditions that induced T-cell anergy, and this occurred in a PI 3K and PKB/AKT dependent manner. Inhibition of the PI 3K-PKB/AKT pathway resulted in cell death without anergy. Overall, our findings provide a novel mechanism to ensure the maintenance of CTLA-4 mediated non-responsiveness and tolerance in the immune system.

Results and Discussion
Given that CTLA-4 can induce non-responsiveness without apoptosis, a key question concerned the underlying mechanism. The maintenance of cell survival during anergy induction is needed for long-term tolerance in transplantation. It was therefore important to investigate whether binding of CTLA-4 to PI 3K leads to activation of PKB/AKT and its downstream targets such as GSK3 a/b and BAD to mediate cell survival. To assess this, pre-activated peripheral CTLA-4 positive T-cells, or a T-cell hybridoma expressing CTLA-4 (DC27.10-CTLA-4) were stimulated with anti-CD3, anti-CD3/CD28 or anti-CD3/CTLA-4 mAbs followed by immunoblotting for phosphorylated PKB/AKT (Thr-308) [14,18] (Fig. 1A, left and right panels). Based on the crystal structure, this site within the activation loop is crucial to the activation of the kinase [18]. In the T-cell hybridoma, anti-CTLA-4 induced phosphorylation of PKB/AKT relative to unstimulated cells (left panel, lane 4 vs. 1; lower band is non-specific; histogram). The level of phosphorylation was comparable to that induced by anti-CD3 and anti-CD28 (lanes 2 and 3, respectively). Co-ligation of CTLA-4 with anti-CD3 revealed phosphorylation at levels similar to anti-CD3/CD28 (lane 6 vs. 5). As a control, immunoblotting with anti-AKT mAb showed equal levels of protein in the cell lysates (Fig. 1A, lower panel). In peripheral Tcells, anti-CD3/CTLA-4 increased phosphorylation of PKB/ AKT when compared to anti-CD3 stimulation (right panel, lane 4 vs 2; histogram). In this case, the phosphorylation was lower than observed for CD3/CD28 (lane 3), most probably due to the fact that CTLA-4 is expressed at lower levels than CD28 and given the fact that there is heterogeneity of CTLA-4 expression in primary T-cells. Importantly, activation of PKB/AKT occurred under conditions where anti-CD3/CTLA-4 inhibited TcR/CD3 mediated IL-2 production and proliferation (Fig. 1B) [5]. These observations indicate that CTLA-4 can activate PKB/AKT as shown by the phosphorylation of Thr-308.
As a read-out for PKB/AKT activity, we next assessed the phosphorylation of the PKB/AKT target GSK-3 a and b on inhibitory serine 21 and 9, respectively [14]. Anti-CTLA-4 readily induced GSK-3 a phosphorylation in DC27.10-CTLA-4 and activated peripheral T-cells ( Fig. 2A, left and right panel, lane 4. histograms). The level of phosphorylation was comparable to that induced by anti-CD3 and anti-CD28 (lanes 2 and 3, respectively). Co-ligation of CTLA-4 with CD3 or CD28 often led to increased  Both PKB/AKT and GSK-3 generate pro-survival signals that rescue cells from AICD or apoptosis [19,20,21]. Therefore, we first investigated whether CTLA-4 could induce pro-survival signals and secondly, whether these effects would be dependent on PKB/AKT activity. To assess this, peripheral T-cells were preactivated with anti-CD3 to express surface CTLA-4 ( Fig. 3), rested for two days and then re-stimulated with anti-CD3 and anti-CD3/ CTLA-4. After 48 hours cells were analysed for cell death. As reported by many labs [22], anti-CD3 induced significant levels of AICD (Fig. 3). About 55 percent of cells underwent AICD in response to CD3 ligation as assessed by AnnexinV/PI staining. However, co-ligation of CTLA-4 increased cell survival to 74 percent versus 44 percent cell survival with anti-CD3 alone (Fig. 3A, upper right panel vs. upper left panel). This finding was also observed at various post-ligation time points (data not shown). Of importance, this rescue from cell death occurred using the same conditions by which CTLA-4 co-ligation inhibited anti-CD3 induced IL-2 production and proliferation (Fig. 1B).
To next assess whether this CTLA-4 induced cell survival was dependent on PKB/AKT activity, cells were cultured in the presence of the AKT inhibitor II. The increase in cell survival that was induced by CTLA-4 was reversed by the addition of the AKT inhibitor (Fig. 3A, lower right panel). In fact, the inhibitor reduced CTLA-4 mediated cell survival to levels comparable to those induced by anti-CD3 (71 percent vs. 77 percent cell death). The inhibition of PKB/AKT activity therefore blocked the induction of the survival signals induced by CTLA-4 co-ligation such that the level of cell survival was similar to anti-CD3. The remaining cells were capable of proliferation as measured by the incorporation of [ 3 H]thymidine (data not shown). This observation showed that the ability of CTLA-4 to rescue from cells death occurred in a PKB/ AKT dependent manner.
Given the activation of PKB/AKT by CTLA-4, we assessed whether CTLA-4 might rescue anti-CD3 induced AICD by modulating the BAD/BcL-XL pathway. To assess whether this pathway was engaged by CTLA-4, phosphorylation of BAD at Ser-136 was investigated upon CTLA-4 co-ligation. BAD Ser-136 is the major site for BAD phosphorylation by PKB/AKT [23]. As shown in Figure 4A 6 vs. 4). The total level of BAD expression was monitored by blotting with an anti-BAD antibody (lower panel). These observations indicated that CTLA-4 can activate PKB/AKT in a pathway that phosphorylates and inactivates BAD in its inhibition of BcL-XL and BcL-2 function.
A decrease in BAD function might result in an increase in the expression of BcL-XL. Given this, we next assessed whether anti-CTLA-4 could increase BcL-XL expression as monitored by anti-BcL-XL immunoblotting (Fig. 4B). Up-regulation of BcL-2 by IL-2 has been reported to depend on the PI 3K-PKB/AKT pathway [24]. Further, CD3/CD28 co-ligation leads to an increase in BcL-XL expression [25]. Indeed, anti-CD3/CTLA-4 increased BcL-XL expression relative to anti-CD3 as assessed 24 hours following ligation in both DC27.  Figure 4C shows the up-regulation of BcL-2 in primary T-cells after 24 and 48 hour stimulation with anti-CD3, anti-CD28 and anti-CTLA-4. These findings indicate that CTLA-4 can increase BcL-XL and BcL-2 expression in a manner similar to CD28, but under conditions that accompany a blockade of anti-CD3 mediated IL-2 production and proliferation (Fig. 1B).
In summary, a central question in the induction of T-cell nonresponsiveness concerns the mechanism by which non-responsiveness can be induced without an increase in cell death or apoptosis. Without an ability to protect against cell death, anergy would not be possible due to cell deletion such as in the case of peripheral deletion. In this report, we have uncovered a mechanism by which CTLA-4 can prevent apoptosis without incurring anergy/nonresponsiveness (Fig. 4D). We found that the PI 3K-PKB/AKT anti-apoptotic pathway was engaged under the same conditions that led to the classic induction of T-cell non-responsiveness by anti-CD3/CTLA-4 antibody co-ligation. CTLA-4 binds directly to the p85 subunit of PI 3K [26]. While the mechanism by which CTLA-4 induces anergy is still under investigation [3,[11][12][13], it required the PI 3K-PKB/AKT pathway to maintain cell survival under conditions of anergy induction. This was shown by the fact that anti-CD3/CTLA-4 ligation led to extensive apoptosis in the presence of PI 3K and AKT/PKB inhibitors, factors whose phosphorylation and activation was increased by CTLA-4 coligation. In another pathway, PP2A, a phosphatase reported to associate with CTLA-4 [13], dephosphorylates BcL-2 and protects this anti-apoptotic protein from proteasome-dependent degradation [28]. The ability of CTLA-4-PI 3K-PKB/AKT to maintain cell survival under conditions of anergy induction would provide a mechanism to ensure long-term tolerance in immunity, as has been observed using CTLA-4 and anti-CD45RB reagents in transplant rejection [29,30].
Anergy induction is often associated with increased susceptibility to apoptosis [31]. Our findings clearly demonstrated that CTLA-4 ligation can increase survival of cells responding to anti-CD3 (Fig. 3). Reversal of the pro-survival signals by inhibition of PKB/AKT indicates that the PKB/AKT pathway is responsible of GSK-3 by immunoblotting with anti-phospho-GSK-3 a/b antibody (lanes 1-6). Lower panel: Equal amounts of cell lysates were immunoblotted for total GSK-3 a/b (lanes 1-6) for this event in a manner similar to CD28 [32]. Interestingly, CTLA-4 induced PKB/AKT activation also led to two events, increased BAD phosphorylation and increased levels of BcL-XL/ Bcl-2 expression (Fig. 4). BAD phosphorylation leads to binding and degradation by 14-3-3 proteins and to reduced BAD binding and sequestration of Bcl-XL [33]. CTLA-4 therefore targets BcL-XL by BAD inactivation leading to the increased presence of proapoptotic BcL-XL and less degradation. Alternatively, the effects of increasing Bcl-XL expression may be mediated via a different pathway. In either case, CTLA-4 mediates its effects by a  combination of providing more active BcL-XL (i.e. less BAD) and greater levels of BcL-XL expression. These findings with primary cells complement previous work by ourselves and others on CTLA-4 effects on Fas-FasL signaling [34,35]. One prediction from this work is that the modulation of the CTLA-4-PKB/AKT pathway may be exploited to reverse CTLA-4 induced anergy in various physiological conditions such as in the case of tumour rejection. The outcome will depend on the sensitivity of the TcR/ CD3 versus CTLA-4 pathways on PKB/AKT signaling. Unlike with CTLA-4, the TcR employs multiple pathways in the activation of PKB/AKT [36]. Future studies will be needed to modulate this CTLA-4 survival pathway for modulating tumor and transplant tolerance and rejection.
To assess proliferation, peripheral T-cells were either left unstimulated or stimulated with anti-CD3 (2 mg/ml), anti-CTLA-4 (20 mg/ml) or anti-CD3/CTLA-4 (2 mg/ml/20 mg/ml) for 48 hours. 1 mCi of [ 3 H] thymidine was added for the last 12 hours of the incubation time. All cell culture groups were done in triplicates.

Measurement of apoptosis
PBLs were pre-activated with anti-CD3 (5 mg/ml), rested for 48 hrs and re-stimulated for 48 hrs with anti-CD3 (2 mg/ml) or anti-CD3/CTLA-4 (2 mg/ml/20 mg/ml) in the absence or presence of 15 mM AKT inhibitor II. To measure apoptosis, cells were stained after 48 hrs with Annexin V-Cy5 and propidium iodide (PI) according to the manufacturer's protocol. PI was used to identify late apoptotic and dead cells.