- Email: [email protected]
4-1BBL costimulation retrieves CD28 expression in activated T cells
Cellular Immunology 256 (2009) 39–46
Contents lists available at ScienceDirect
Cellular Immunology journal homepage: www.elsevier.com/locate/ycimm
4-1BBL costimulation retrieves CD28 expression in activated T cells Mojtaba Habib-Agahi a,b,*, Mansooreh Jaberipour c, Peter F. Searle b a b c
Immunotherapy Laboratory, Department of Immunology, School of Medicine, Shiraz University of Medical Sciences, Shiraz 71345-3119, Iran Cancer Research UK, Institute for Cancer Studies, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK Shiraz Institute for Cancer Studies, Shiraz University of Medical Sciences, Shiraz, Iran
a r t i c l e
i n f o
Article history: Received 3 September 2008 Accepted 13 January 2009 Available online 12 February 2009 Keywords: T cell CD28 kinetics CD80 CD86 CD137 4-1BBL Costimulation
a b s t r a c t Binding of CD80/86 to CD28 is regarded as the main T cell costimulatory interaction. However, CD28 downregulates soon after T cell activation. To investigate potential cross-interaction between CD137 (4-1BB) and CD28, we stimulated T cells with anti-CD3 in the presence of A549 lung carcinoma cells expressing CD80/CD86 and 4-1BBL molecules, transduced into the cells using recombinant non-replicating adenoviruses. Following initial T cell proliferation, the proportion of CD28+ cells in both CD4+ and CD8+ populations was rapidly reduced by CD80/86 costimulation, whereas cultures costimulated with just 4-1BBL continued to express CD28. CD28 was also downregulated in cultures costimulated with both CD80/86 and 4-1BBL. Interestingly, in cells costimulated with CD80/86 that had downregulated CD28 expression and ceased to proliferate, reactivation of proliferation by 4-1BBL costimulation also restored their CD28 expression. These findings show a positive effect of CD137 signalling on CD28 expression, similar to the effect of CD28 engagement on 4-1BB expression during the initial phases of T cell activation. Moreover, they point to the importance of signals through 4-1BB for the purposes of ex-vivo T cell activation and expansion. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Only a few costimulatory receptors are expressed in a constitutive fashion by un-stimulated T cells. The majority of costimulatory receptors are induced only following cell activation, subsequent to signal transduction through the TCR complex. Among those receptors, there is a central role for CD28 mediated costimulation as T cell activation in the absence of CD28 engagement results in a hyporesponsive state termed anergy [1]. CD28, a homodimeric, integral membrane receptor of the B7 immunoglobulin superfamily, is known as an essential component of the immunological synapse (IS) [2–4]. The majority of human CD4+ and almost half of CD8+ T cells express this receptor [5]. Studies on CD28 costimulation have led to a model by which costimulators enhance TCR engagement with peptide-MHC and prolong intracellular signalling [6]. Signals mediated by CD28 potently accelerate entry into and progression through the cell cycle [7]. Triggering CD28 can initiate other signalling pathways that are distinct from the TCR but lead to the activation of common targets, such as the upregulation of anti-apoptotic members of the Bcl-2 family [8,9]. Evidence shows that reactivation of memory CD4+ T cells may happen independent of the
* Corresponding author. Address: Immunotherapy Laboratory, Department of Immunology, School of Medicine, Shiraz University of Medical Sciences, Shiraz 71345-3119, Iran. Fax: +98 711 2351575. E-mail addresses: [email protected] (M. Habib-Agahi), habibagahim@yahoo. com (M. Habib-Agahi). 0008-8749/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.cellimm.2009.01.003
CD28 ligands, B7.1 (CD80) and B7.2 (CD86) [10]. This suggests that subsequent to initial T cell activation, a switch in the costimulatory requirement occurs. In addition, it has been shown that CD28 expression is significantly downregulated after activation in presence of CD80 costimulation [11], and occurs more rapidly in CD8+ cells [12]. The CD28null phenotype is considered a characteristic sign of replicative senescence of human T cells during repetitive stimulation in vitro, as well as in chronic inflammatory and infectious disease, and in the normal course of aging [13,14]. The exact mechanism of the downregulation of CD28 in T cells is not fully understood but an inoperative transcriptional initiator (INR) at the proximal region of CD28 promoter is involved in the regulation of gene transcription [15]. Microarray analysis comparing CD28null and CD28+ cell populations showed significantly downregulated expression of CD154 (CD40L) in CD28null cells. On the other hand, some other costimulatory receptors such as 4-1BB (CD137), CD244 and SLAMF7 significantly are upregulated in CD28null CD8+ memory T cells [16]. Presence of additional costimulatory interactions could ensure the potential for continued activation of CD28null T cells, as these cells still are functional. Signal transduction by 4-1BB of TNFR superfamily can not only costimulate T cell activation but also have additional effects on T cell survival [17–19]. Sharing many common features with CD28, costimulation via 4-1BB has major effects on CD8 T cells and promotes TH1 differentiation/cytokine production in CD4+ cells. 4-1BB is an inducible receptor and expressed on the surface of T cells after CD28 signalling and initial cell activation [20]. Therefore, we asked
40
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
if 4-1BB in a similar fashion could change the pattern of CD28 expression in activated T cells. Using adenoviral vectors to express CD80, CD86 and 4-1BBL costimulatory ligands on the surface of A549 lung carcinoma cells, we showed that CD28 expression does not change during T cell activation in presence of 4-1BBL and more interestingly, CD28 that was downregulated in response to CD80/ 86 costimulation could be re-expressed in T cells re-activated as a result of 4-1BBL costimulation. These findings may explain possible mechanism for reactivation of CD28null cells and it may suggest employing other costimulatory pathways rather than CD28 for exvivo T cell proliferation.
were used to indicate levels of background staining from different suppliers. 2.4. Statistical analysis The differences between various experimental groups were analyzed by two tailed Student’s t test, using SPSS software. Values of p < 0.05 were considered significant. 3. Results 3.1. CD28 expression profile in cultured lymphocytes
2. Materials and methods 2.1. Expression of costimulatory ligands by adenoviral vectors Production of E1- and E3-deleted, recombinant, non-replicative adenoviruses expressing full length human CD80, CD86, 4-1BBL and enhanced green fluorescent protein (GFP) have been reported previously [21,22]. A549 human lung carcinoma cells were infected with 300–600 virus particles/cell, 48–72 h prior to use to reach to the maximum ligand expression. Surface expression of the ligands was confirmed by flow cytometry analysis using commercial antibodies (data not shown, [22]).
3.1.1. CD80/CD86 costimulation but not 4-1BBL costimulation downregulate the frequency of CD28+ lymphocytes A549 cells do not show significant expression of costimulatory ligands (0.7–1% for CD80/86 and 3–5% for 4-1BBL) however 48 h after infection with 300 p/cell of recombinant viruses, the majority
2.2. Lymphocyte activation A549 cells growing in Dulbecco’s modified Eagle’s medium with HEPES (DMEM-HEPES, Sigma) supplemented with 10% foetal calf serum (FCS), 2 mM L-glutamine and antibiotics (100 iu/ml penicillin and 100 mg/ml streptomycin, Sigma, UK) were infected with the recombinant adenoviruses to express costimulatory ligands or GFP as control, and cultured in wells of 24-well plates (0.2–0.3 106 cells) as monolayer 48 hrs in advance. Infected cells were co-cultured with 1–1.5 106 cells/well of PBMCs prepared by Ficoll density-cushion centrifugation from blood of healthy donors (obtained with full informed consent) and depleted from plastic adherent cells. Depletion typically reduced CD14+ cells from 10– 15% to 3–5% of mononuclear cell preparations. Cultures were maintained in RPMI 1640 (Life Technologies, UK), supplemented with 7% FCS and 3% human (h)-AB serum (HD supplies, UK) in addition to Lglutamine and antibiotics and stimulated with 100 ng/ml soluble OKT3 anti-CD3 antibody (JANSSEN-CILAG, UK). Lymphocytes were passaged and re-stimulated approximately weekly, by transferring the adjusted number of cells to wells with fresh infected A549 cells expressing the corresponding costimulatory molecules. Lymphocyte proliferation was studied by cell surface marker staining and flow cytometry or viable cell count using hemocytometer and Trypan blue dye exclusion at the appropriate time points. 2.3. Antibody staining and flow cytomery Fresh or cultured PBMCs were incubated with approximately 1 lg/ml of appropriate fluorochrome-conjugated or isotype matched antibodies on ice in dark for 30 min. Un-bound antibodies were removed by two washes with PBS/FCS (2%). Cells were re-suspended in 500 ll buffer for flow cytometry, and analyzed using a four colours Beckman Coulter XL flow cytometer using Coulter System II software for data acquisition and WinMDI 2.8 software for data presentation. Labelled mouse monoclonal antibodies used in this study were sourced as follows: anti-CD4-FITC, anti-CD8-FITC, anti-CD4-PE, anti-CD8- PE, from Beckman Coulter (High Wycombe, UK), antiCD80-FITC, anti-CD80-PE, anti-CD86-PE, anti-CD137-PE, antiCD137L-PE, from BD, Pharmingen (Oxford, UK). Anti human CD28 antibody purchased from Sigma (UK). Appropriate isotype controls
Fig. 1. Effects of different costimulation regimens on CD4+ and CD8+ cell numbers in long-term culture. Peripheral blood mononuclear cells were stimulated by 100 ng/ml anti-CD3 antibody whilst co-cultured with A549 lung carcinoma cells expressing one or two costimulatory transgenes (or GFP as control) as described in Section 2. Cultures were passaged, with adjustment of cell density, and restimulated with anti-CD3 and fresh A549 cells expressing the same co-stimulatory ligands, at weekly intervals (indicated by arrows in B). Viable cell number was determined by haemocytometer and trypan blue dye exclusion, and the proportion of CD4+ (A) and CD8+ cells (B) determined by flow cytometry. Cultures of lymphocytes with either CD80 or CD86 costimulation decline after 7–10 days whilst regimens including 4-1BBL allow extended lymphocyte growth. Error bars indicate standard deviation of the mean. Statistically significant differences (p < 0.05) determined by two-tailed t test.
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
of infected cells (>80–95%) consistently expressed CD80, CD86, 41BBL or GFP transgenes, at levels comparable to those on a lymphoblastoid cell line or dendritic cells as showed before [22]. Using infected A549 cells as the major source of costimulation, previ-
41
ously we reported the different growth kinetics of T cells in response to CD80/86 and 4-1BBL costimulation [22]. In previous report we showed a transient T cell growth by CD80/86 costimulation which stopped eventually by the second week, regardless of
Fig. 2. Flow cytometry density plots of CD28 expression by CD4 and CD8+ cells before and after 4 days anti-CD3 stimulation with different costimulation treatments. Lymphocytes were stimulated as described in Fig. 1 and were stained with either PE conjugated anti-CD8 or anti-CD4 plus FITC conjugated anti-CD28 antibodies and analyzed by flow cytometry. As a reference, cells were also stained at day zero without anti-CD3 stimulation. A significant downregulation of CD28 expression was evident in cultures with CD80 and CD86 costimulation whilst cells with 4-1BBL costimulation kept the receptor presentation.
42
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
re-stimulation with anti-CD3 and CD80/86. In contrast, costimulation with 4-1BBL or its combination with CD80 or CD86 supported both CD4+ and CD8+ cells in continuous growth until the experiment was terminated (up to 5 weeks). Fig. 1A and B depict typical results of similar experiments with relative viable cell counts of CD4+ and CD8+ cells in different culture conditions during 35 days. To see if lymphocytes have different patterns of CD28 expression in response to different costimulations, the expression of CD28 by CD4+ and CD8+ lymphocytes was studied before, and then after 4, 27 and 35 days of stimulation in three independent experiments. Moreover, similar experiments with an EBV-QAK peptide antigen specific CD8+ T cell clone showed comparable results regarding to CD28 expression (data not shown). As a typical flow cytometry result, Fig. 2 shows that almost all CD4+ and 40–50% of CD8+ cells express CD28 before stimulation with OKT3 anti-CD3 antibody. However, by day 4, both CD80 and CD86 costimulations sharply downregulated the frequency of CD28+ expressing cells in CD4+ and CD8+ lymphocytes. As shown in Fig. 3, in CD80 costimulated cultures, CD4+CD28+ cells and CD8+CD28+ cells comprised 25 ± 3% and 18 ± 1% of lymphocytes, respectively, whilst CD86 costimulation caused a greater reduction in frequency of those populations to 14.3 ± 4% and 11.4 ± 1%, respectively (p = 0.01 and p = 0.04, respectively). Dual costimulation regimens of CD80 or CD86 plus 4-1BBL downregulated the frequency of CD28+ cells in both CD4+ and CD8+ cell compartments to the similar extent as observed with CD80 or CD86 alone (10–28% and 6–15%, respectively, p > 0.08 for CD4+ and CD8+ cells in both dual costimulation regimens). Conversely, costimulation with 4-1BBL which resulted in considerable proliferation of lymphocytes did not change the frequency of CD4+ and CD8+ cells with CD28 expression significantly (96.8 ± 0.5% and 47.2 ± 3.1%, respectively, p < 0.001 for both CD4 and CD8 cells). Coculture of lymphocytes with A549/Ad-GFP cells in control cultures also did not change these frequencies either. These effects on CD28 expression were sustained in cultures that were passaged and restimulated weekly with anti-CD3 activation and corresponding costimulation treatments, in experiments lasting up to day 27 or 35. Thus, after 4–5 weeks culture, CD80 or CD86 costimulated cultures had significantly lower proportions of CD28+ lymphocytes compared to cultures with 4-1BBL costimulation or control cultures (p < 0.0007 for CD4 cells, and p < 0.004 for CD8 cells, respectively) (Fig. 3). Significant decrease in CD8+CD28+ also observed compare to GFP control culture (p < 0.007). In parallel to the de-
crease in the frequency of the CD28 positive cells by CD80/86 costimulations, the level of cell-surface CD28 expression in the positive population, as indicated by the mean fluorescent intensities (MFIs), also diminished more rapidly. At time zero of this experiment, the MFI of CD28 expression of the CD4+ and CD8+ cells were 76 ± 7 and 59 ± 3, respectively. After 4 days, anti-CD3 activation with CD80 costimulation, the MFI of CD28 expression in CD4+ and CD8+ cells decreased to 26 ± 5 and 30 ± 3, respectively. At the same day, the intensity of CD28 expression by CD4+ and CD8+ cells in 4-1BBL costimulated cultures were found to be 60 ± 4 and 53 ± 2, respectively (data not shown). 3.1.2. Switching from CD80/CD86 costimulation to 4-1BBL results in CD28 re-expression As demonstrated previously [22], and shown in Fig. 1 for CD4+ and CD8+ cells separately, the proliferation of T cells stimulated with anti-CD3 and CD80/86 is transient, leaving them unresponsive or anergic to continued stimulation, with a gradual decline in viable cells. Provision of 4-1BBL costimulation for those anergic cells, after 3 weeks of culture, resulted in their reactivation and resumption of proliferation. To see if this reactivation associated with 4-1BBL costimulation may cause change in the pattern of CD28 expression we examined the expression of this receptor in CD4+ and CD8+ cells. As shown in Fig. 4, reactivation of lymphocytes was associated with significant increase in the frequency of CD4+ and CD8+ cells with CD28 expression. Switching from CD80 or CD86 costimulation to 4-1BBL increased the frequency of CD4+CD28+ cells in those cultures up to 88 ± 2.3% and 63 ± 3%, respectively. Similarly, there was an increase in the frequency of CD8+CD28+ cells of those cultures up to 48 ± 1.3% and 32 ± 3.6%, respectively (Fig. 5). As the results show, CD80 and CD86 pre-costimulated cultures may have different capacities (p < 0.04 for both CD4 and CD8 cells) in recovering CD28 expression after transferring to the new costimulatory conditions as CD80 treated lymphocytes re-expressed CD28 faster than CD86 pre-stimulated cells in both CD4+ and CD8+ cell compartments. Similarly, transfer of CD80/CD86 pre-costimulated lymphocytes to cultures with double costimulation of CD80/86 plus 4-1BBL also resulted in increase in the frequency of cells with CD28 expression in both CD4+ and CD8+ T cells during first week after transfer. Frequency of CD4+CD28+ cells reached up to 85 ± 2.1% and 61 ± 2.7% for dual costimulation regimens with either CD80 or CD86 plus 4-1BBL,
Fig. 3. Frequency of CD4+ and CD8+ lymphocytes with CD28 expression during 35 days culture with anti-CD3 stimulation and different costimulation treatments. PBMCs were cultured and stained as explained in Figs. 1 and 2 to determine the frequency of CD28 positive cells in CD4+ and CD8+ subpopulations during 35 days culture. The initial downregulation of CD28 in cultures with CD80 or CD86 was persistent until the last day of experiment however, cultures with 4-1BBL costimulation almost maintained the original levels of CD28 expression during this period. Error bars indicate standard deviation of the mean. Statistically significant differences (p < 0.05) determined by twotailed t test.
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
43
Fig. 4. Flow cytometry density plots showing re-expression of CD28 by CD4+ and CD8+ cells after transfer to conditions with 4-1BBL costimulation. PBMCs were prepared and cultured as explained in Fig. 1. After 21 days, half of lymphocytes previously costimulated with CD80 or CD86 were transferred into cultures with 4-1BBL costimulation with or without CD80 and CD86. Expression of CD28 in CD4+ and CD8+ subpopulations was studied 6 days after transfer. Whilst only a minority of lymphocytes in both original CD80 and CD86 costimulated cultures were expressing CD28, transfer of those cells to cultures with 4-1BBL costimulation resulted in significant re-expression of CD28 a week after transfer.
respectively, whilst almost 45 ± 3% and 21 ± 1.2% of CD8+ cells were expressing this receptor in the corresponding culture conditions. Study of these cells on day 35 showed that CD28 is downregulated again in dual costimulated cultures whilst those cells switched to single 4-1BBL costimulation, kept their CD28 expression (p < 0.001 for both CD4 and CD8 cells, Fig. 5). 3.2. Kinetics of 4-1BB expression on lymphocytes Non-activated lymphocytes do not express 4-1BB significantly. However, 4-1BB was moderately upregulated shortly after antiCD3 activation, to allow response to 4-1BBL costimulation. Activation by anti-CD3 in presence of CD86 costimulation showed faster kinetics of 4-1BB expression than CD80 costimulation in similar
condition (2.5 ± 0.5% vs 7.2 ± 1.1% for CD4+ cells, and 12.4 ± 1.8% vs 23.1 ± 2.6% for CD8+ cells, p < 0.01 for both CD4+ and CD8+ cells by day 4) (Fig. 6A and B). As shown, among lymphocytes, CD8+ cells reached a higher frequency of 4-1BB expression. 4-1BBL costimulation was less able to induce its receptor expression during the first week of culture than CD86 costimulation (p = 0.03). However, 41BBL treated and dual costimulated cultures, with continuation of their growth, showed more cells with 4-1BB expression by day 35 (p < 0.003). Moreover, transfer of growth arrested, CD80/86 costimulated cells to conditions with 4-1BBL costimulation or its cooperation with CD80/86, sustained 4-1BB expression in population of lymphocytes as well. Two weeks after transfer, more CD8 cells re-expressed the receptor in cultures with dual costimulation (up to 10-fold increase by 4-1BBL + CD86) (Fig. 6).
44
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
Fig. 5. Frequency of CD28 expressing cells among CD4+ and CD8+ subpopulations, 6 days (day 27) and 14 days after transfer (day 35) to cultures with 4-1BBL costimulation, with or without continued CD80 and CD86 costimulation. PBMCs were prepared and stained as explained in Fig. 4. Transferring lymphocytes previously costimulated with CD80 or CD86 to cultures with 4-1BBL costimulation (with or without continued CD80 and CD86) resulted in re-expression of CD28 in a high proportion of lymphocytes 6 days after transfer. However, that was later downregulated again in dually costimulated cultures. Error bars show SDM. Statistically significant differences (p < 0.05) determined by two-tailed t test.
4. Discussion In this study we showed that CD28, which is downregulated by costimulation through its ligands CD80 or CD86, can be re-expressed under the influence of 4-1BBL costimulation, in both CD4+ and CD8+ T cells. Several mechanisms i.e. TCR-mediated activation [23], in vitro replicative senescence [24], and CD80 or CD86 engagement [25] have been shown to be involved in downregulation of CD28 in T cells. Moreover, recent results indicate that the inducible CD80/86 receptor CTLA-4 promotes CD28 internalization and proteolysis, leading to further reduced surface expression [26]. By any mechanism, the net result is the emergence and accumulation of CD28null T cells. Several lines of evidence show that the majority of CD28 effector cells are programmed to die rapidly by apoptosis [27] or have reduced responses to an array of immune challenges, including pathogens [28] and allograft rejection [29]. On the other hand, recent findings show that CD28null cells are not terminally differentiated and could respond to TCR stimulation in a particular condition. Waller et al. showed the proliferative response of CMV specific CD28 CD45RAhi CD8+ T cells to antigen presentation through monocytes or artificial antigen presenting cells expressing CD137L. However, similar antigen presenting cells but with CD80, CD86 or CD275 have not been able to activate CD28 cells [30]. Thus, proliferation of CD28 cells appears to have a specific requirement for 4-1BBL. In our results, we demonstrated decrease in the frequency of CD28+ cells and showed decline in the intensity of the expression of this receptor within a few days after T cells were costimulated with CD80 or CD86 during anti-CD3 stimulation, confirming earlier observations [25]. Of these two B7-family ligands, greater decreases in CD28+ cell population consistently happened when cells were costimulated with CD86. This contrasts with a report by Lewis et al., in which CD86 costimulation of sub-lines of Jurkat T cell line, stimulated with PMA, did not cause significant decrease in CD28+ population [11]. We showed that, provision of 4-1BBL as a sole costimulation in conjunction with anti-CD3 stimulation, did not cause a significant change in CD28 expression even after one month culture. Nevertheless, combination of either CD80 or CD86 with 4-1BBL resulted in downregulation of CD28+ cell frequency, similar to that obtained using just CD80 or CD86 for costimulation. Thus, the greater mag-
nitude and longevity of T cell responses costimulated by the combined use of 4-1BBL with CD80 or CD86 does not appear to result from modulation of CD28 downregulation by 4-1BBL. As shown in Fig. 1 and reported before [22], CD4+ and CD8+ cell counts decline with CD80/86 costimulation after the second week of the culture; however switching of these non-responsive or anergic T cells to 41BBL costimulation condition resulted in reactivation and rapid proliferation within a few days post transfer [22]. We showed that combined costimulation using A549 cells expressing CD80 or CD86 with 4-1BBL allowed rapid and continued lymphocyte proliferation in response to a suboptimal concentration of anti-CD3 antibody, reaching up to 100-fold expansion of the lymphocyte population over 5 weeks without the use of exogenous cytokine or feeder cells. In recent report by Hippen et al. expression of 4-1BBL in conjunction with CD32 on the surface of K562 cells which then display membrane-bound anti-CD3 and anti-CD28 alongside the 4-1BBL, expansion levels of purified umbilical cord blood regulatory T cells exceeded 1250-fold [31]. Similarly, results of Suhoski et al. showed up to 104-fold expansion of CD8 T cells by artificial antigen presenting cells expressing CD80, 4-1BBL and anti-CD3 antibody [32]. These reports and others are in accord with our result regarding to the capability of 4-1BBL costimulation as a potent costimulation for T cell expansion. Lower T cell expansion in our results could be due to the weaker TCR signal from suboptimal, soluble anti-CD3 compared to the cell-attached antibody in other studies. Beside T cell proliferation and then resumption of growth by provision of 4-1BBL costimulation, in this report we showed reexpression of CD28 on the cells after they were transferred from CD80 or CD86 costimulation, and this was then sustained in cultures costimulated with 4-1BBL alone. Anergic T cells that were transferred to the condition of dual costimulation (i.e. 41BBL + CD80/86) also showed substantial subsequent proliferation, however only a transient increase in CD28 expression was observed before the receptor downregulated again, which emphasizes the dominant effect of CD80/CD86 on this process. Clearly, loss of CD28 is not an irreversible mechanism and CD28null cells may re-express this receptor if stimulated appropriately. Warrington et al. also have shown the re-expression of CD28 in CD4+CD28null lymphocytes after the combined action of IL-12 and anti-CD3 stimulation [33]. In their results CD28 expression was
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
45
Fig. 6. Frequency of CD4+ and CD8+ lymphocytes with 4-1BB expression in culture with anti-CD3 stimulation and different costimulation treatments. CD4+ (A) and CD8+ (B) lymphocytes were analyzed for expression of the 4-1BB at the time of culture establishment, and after 4 and 35 days of culture with anti-CD3 and different costimulation regimens. Portions of the cultures initially costimulated with CD80 or CD86 alone were transferred to conditions including 4-1BBL costimulation on day 21. By day 4, 4-1BB is upregulated, particularly with CD86 costimulation and on CD8+ cells. Transferring CD80/86 costimulated cells to regimens including 4-1BBL increased 4-1BB expression on CD8+ cells by day 35, relative to cultures with unchanged costimulation. Error bars show SDM. Statistically significant differences (p < 0.05) determined by two-tailed t test.
temporary, requiring the continued presence of IL-12 in the medium. Different members of TNF receptor superfamily, such as 4-1BB, have been proposed to act subsequent to CD28 downregulation, to compensate and regulate late T cell activation. This signal, at the peak of T cell expansion phase, maintains the survival of activated T cells when antigen becomes limiting [19,34–36]. This appears as one of the several bidirectional interactions among T cell costimulatory pathways. CD28 signal promotes the upregulation of CD40, OX40 and 4-1BB expression in T cells [37]. We also showed that CD80/86 costimulation are able to upregulate 4-1BB receptor, more effectively than 4-1BBL itself. However, growth arrested lymphocytes regained 4-1BB expression when they were costimulated with 4-1BBL alone or in combination with CD80/86. Therefore, as we have shown, there appears to be a regulatory role of 4-1BB for CD28 re-expression in scenario of T cell activation and there is a cross talk between these two costimulatory pathway to regulate their expression. It has been shown that signalling through CD137 and CD134, another member of the TNF receptor family, can upregulate telomerase activity of human CD8+ T cells. This is also an early function of CD28 during T cell activation [38]. Therefore, signalling through
4-1BB after its appearance on the surface of activated T cells, not only can provide costimulatory signal for their survival and further proliferation, but also according to our results, may cause reexpression of CD28 and restoration of the response to CD80/86 costimulation. In the context of aging of the immune system, reexpression of CD28 may be desirable to re-establish the state of immunocompetence. In contrast, induction of CD28 on CD28null T cells in autoimmune diseases may augment the autoimmune response. Mechanisms that restore CD28 expression are, therefore, potential therapeutic targets which need further investigations. Acknowledgment This work was funded by the Ministry of Health and Medical Education of Iran. References [1] Y. Zheng, Y. Zha, T.F. Gajewski, Molecular regulation of T-cell anergy, EMBO Rep. 9 (2008) 50–55. [2] P.G. Andres, K.C. Howland, D. Dresnek, S. Edmondson, A.K. Abbas, M.F. Krummel, CD28 signals in the immature immunological synapse, J. Immunol. 172 (2004) 5880–5886.
46
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
[3] S.K. Bromley, A. Iaboni, S.J. Davis, A. Whitty, J.M. Green, A.S. Shaw, A. Weiss, M.L. Dustin, The immunological synapse and CD28–CD80 interactions, Nat. Immunol. 2 (2001) 1159–1166. [4] S.A. Wetzel, T.W. McKeithan, D.C. Parker, Live-cell dynamics and the role of costimulation in immunological synapse formation, J. Immunol. 169 (2002) 6092–6101. [5] A. Viola, A. Lanzavecchia, T cell activation determined by T cell receptor number and tunable thresholds, Science 273 (1996) 104–106. [6] M.C. Miceli, M. Moran, C.D. Chung, V.P. Patel, T. Low, W. Zinnanti, Costimulation and counter-stimulation: lipid raft clustering controls TCR signaling and functional outcomes, Semin. Immunol. 13 (2001) 115–128. [7] J.L. Bonnevier, D.L. Mueller, Cutting edge: B7/CD28 interactions regulate cell cycle progression independent of the strength of TCR signaling, J. Immunol. 169 (2002) 6659–6663. [8] L.J. Appleman, A.A. Van Puijenbroek, K.M. Shu, L.M. Nadler, V.A. Boussiotis, CD28 costimulation mediates down-regulation of p27kip1 and cell cycle progression by activation of the PI3K/PKB signaling pathway in primary human T cells, J. Immunol. 168 (2002) 2729–2736. [9] K. Okkenhaug, L. Wu, K.M. Garza, R.J. La, W. Khoo, B. Odermatt, T.W. Mak, P.S. Ohashi, R. Rottapel, A point mutation in CD28 distinguishes proliferative signals from survival signals, Nat. Immunol. 2 (2001) 325–332. [10] A.M. Dahl, C. Klein, P.G. Andres, C.A. London, M.P. Lodge, R.C. Mulligan, A.K. Abbas, Expression of bcl-X(L) restores cell survival, but not proliferation off effector differentiation, in CD28-deficient T lymphocytes, J. Exp. Med. 191 (2000) 2031–2038. [11] D.E. Lewis, M. Merched-Sauvage, J.J. Goronzy, C.M. Weyand, A.N. Vallejo, Tumor necrosis factor-alpha and CD80 modulate CD28 expression through a similar mechanism of T-cell receptor-independent inhibition of transcription, J. Biol. Chem. 279 (2004) 29130–29138. [12] A.N. Vallejo, CD28 extinction in human T cells: altered functions and the program of T-cell senescence, Immunol. Rev. 205 (2005) 158–169. [13] M. Schirmer, C. Goldberger, R. Wurzner, C. Duftner, K.P. Pfeiffer, J. Clausen, G. Neumayr, A. Falkenbach, Circulating cytotoxic CD8(+) CD28( ) T cells in ankylosing spondylitis, Arthritis Res. 4 (2002) 71–76. [14] E. Bryl, A.N. Vallejo, C.M. Weyand, J.J. Goronzy, Down-regulation of CD28 expression by TNF-alpha, J. Immunol. 167 (2001) 3231–3238. [15] A.N. Vallejo, E. Bryl, K. Klarskov, S. Naylor, C.M. Weyand, J.J. Goronzy, Molecular basis for the loss of CD28 expression in senescent T cells, J. Biol. Chem. 277 (2002) 46940–46949. [16] M. Fann, W.K. Chiu, W.H. Wood III, B.L. Levine, K.G. Becker, N.P. Weng, Gene expression characteristics of CD28null memory phenotype CD8+ T cells and its implication in T-cell aging, Immunol. Rev. 205 (2005) 190–206. [17] M.R. Alderson, C.A. Smith, T.W. Tough, T. vis-Smith, R.J. Armitage, B. Falk, E. Roux, E. Baker, G.R. Sutherland, W.S. Din, Molecular and biological characterization of human 4-1BB and its ligand, Eur. J. Immunol. 24 (1994) 2219–2227. [18] C. Takahashi, R.S. Mittler, A.T. Vella, Cutting edge: 4-1BB is a bona fide CD8 T cell survival signal, J. Immunol. 162 (1999) 5037–5040. [19] M. Croft, Co-stimulatory members of the TNFR family: keys to effective T-cell immunity?, Nat Rev. Immunol. 3 (2003) 609–620. [20] K.E. Pollok, Y.J. Kim, Z. Zhou, J. Hurtado, K.K. Kim, R.T. Pickard, B.S. Kwon, Inducible T cell antigen 4-1BB. Analysis of expression and function, J. Immunol. 150 (1993) 771–781. [21] M.G. Gilligan, P. Knox, S. Weedon, R. Barton, D.J. Kerr, P. Searle, L.S. Young, Adenoviral delivery of B7-1 (CD80) increases the immunogenicity of human ovarian and cervical carcinoma cells, Gene Ther. 5 (1998) 965–974.
[22] M. Habib-Agahi, T.T. Phan, P.F. Searle, Co-stimulation with 4-1BB ligand allows extended T-cell proliferation, synergizes with CD80/CD86 and can reactivate anergic T cells, Int. Immunol. 19 (2007) 1383–1394. [23] D.J. Lenschow, T.L. Walunas, J.A. Bluestone, CD28/B7 system of T cell costimulation, Annu. Rev. Immunol. 14 (1996) 233–258. [24] R.B. Effros, N. Boucher, V. Porter, X. Zhu, C. Spaulding, R.L. Walford, M. Kronenberg, D. Cohen, F. Schachter, Decline in CD28+ T cells in centenarians and in long-term T cell cultures: a possible cause for both in vivo and in vitro immunosenescence, Exp. Gerontol. 29 (1994) 601–609. [25] S.C. Eck, D. Chang, A.D. Wells, L.A. Turka, Differential down-regulation of CD28 by B7-1 and B7-2 engagement, Transplantation 64 (1997) 1497– 1499. [26] M. Berg, N. Zavazava, Regulation of CD28 expression on CD8+ T cells by CTLA4, J. Leukoc. Biol. 83 (2008). [27] N.J. Borthwick, M. Lowdell, M. Salmon, A.N. Akbar, Loss of CD28 expression on CD8(+) T cells is induced by IL-2 receptor gamma chain signalling cytokines and type I IFN, and increases susceptibility to activation-induced apoptosis, Int. Immunol. 12 (2000) 1005–1013. [28] H.L. Compton, J.P. Farrell, CD28 costimulation and parasite dose combine to influence the susceptibility of BALB/c mice to infection with Leishmania major, J. Immunol. 168 (2002) 1302–1308. [29] B. Salomon, J.A. Bluestone, Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation, Annu. Rev. Immunol. 19 (2001) 225–252. [30] E.C. Waller, N. McKinney, R. Hicks, A.J. Carmichael, J.G. Sissons, M.R. Wills, Differential costimulation through CD137 (4-1BB) restores proliferation of human virus-specific ‘‘effector memory” (CD28( ) CD45RA(HI)) CD8(+) T cells, Blood 110 (2007) 4360–4366. [31] K.L. Hippen, P. Harker-Murray, S.B. Porter, S.C. Merkel, A. Londer, D.K. Taylor, M. Bina, A. Panoskaltsis-Mortari, P. Rubinstein, R.N. Van, T.N. Golovina, M.M. Suhoski, J.S. Miller, J.E. Wagner, C.H. June, J.L. Riley, B.R. Blazar, Umbilical cord blood regulatory T-cell expansion and functional effects of tumor necrosis factor receptor family members OX40 and 4-1BB expressed on artificial antigen-presenting cells, Blood 112 (2008) 2847–2857. [32] M.M. Suhoski, T.N. Golovina, N.A. Aqui, V.C. Tai, A. Varela-Rohena, M.C. Milone, R.G. Carroll, J.L. Riley, C.H. June, Engineering artificial antigen-presenting cells to express a diverse array of co-stimulatory molecules, Mol. Ther. 15 (2007) 981–988. [33] K.J. Warrington, A.N. Vallejo, C.M. Weyand, J.J. Goronzy, CD28 loss in senescent CD4+ T cells: reversal by interleukin-12 stimulation, Blood 101 (2003) 3543– 3549. [34] P.R. Rogers, J. Song, I. Gramaglia, N. Killeen, M. Croft, OX40 promotes Bcl-xL and Bcl-2 expression and is essential for long-term survival of CD4 T cells, Immunity 15 (2001) 445–455. [35] M. Croft, Costimulation of T cells by OX40, 4-1BB, and CD27, Cytokine Growth Factor Rev. 14 (2003) 265–273. [36] G. Pulle, M. Vidric, T.H. Watts, IL-15-dependent induction of 4-1BB promotes antigen-independent CD8 memory T cell survival, J. Immunol. 176 (2006) 2739–2748. [37] A.H. Sharpe, G.J. Freeman, The B7-CD28 superfamily, Nat. Rev. Immunol. 2 (2002) 116–126. [38] F.J. Plunkett, O. Franzese, H.M. Finney, J.M. Fletcher, L.L. Belaramani, M. Salmon, I. Dokal, D. Webster, A.D. Lawson, A.N. Akbar, The loss of telomerase activity in highly differentiated CD8+CD28 , J. Immunol. 178 (2007) 7710– 7719.
Contents lists available at ScienceDirect
Cellular Immunology journal homepage: www.elsevier.com/locate/ycimm
4-1BBL costimulation retrieves CD28 expression in activated T cells Mojtaba Habib-Agahi a,b,*, Mansooreh Jaberipour c, Peter F. Searle b a b c
Immunotherapy Laboratory, Department of Immunology, School of Medicine, Shiraz University of Medical Sciences, Shiraz 71345-3119, Iran Cancer Research UK, Institute for Cancer Studies, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK Shiraz Institute for Cancer Studies, Shiraz University of Medical Sciences, Shiraz, Iran
a r t i c l e
i n f o
Article history: Received 3 September 2008 Accepted 13 January 2009 Available online 12 February 2009 Keywords: T cell CD28 kinetics CD80 CD86 CD137 4-1BBL Costimulation
a b s t r a c t Binding of CD80/86 to CD28 is regarded as the main T cell costimulatory interaction. However, CD28 downregulates soon after T cell activation. To investigate potential cross-interaction between CD137 (4-1BB) and CD28, we stimulated T cells with anti-CD3 in the presence of A549 lung carcinoma cells expressing CD80/CD86 and 4-1BBL molecules, transduced into the cells using recombinant non-replicating adenoviruses. Following initial T cell proliferation, the proportion of CD28+ cells in both CD4+ and CD8+ populations was rapidly reduced by CD80/86 costimulation, whereas cultures costimulated with just 4-1BBL continued to express CD28. CD28 was also downregulated in cultures costimulated with both CD80/86 and 4-1BBL. Interestingly, in cells costimulated with CD80/86 that had downregulated CD28 expression and ceased to proliferate, reactivation of proliferation by 4-1BBL costimulation also restored their CD28 expression. These findings show a positive effect of CD137 signalling on CD28 expression, similar to the effect of CD28 engagement on 4-1BB expression during the initial phases of T cell activation. Moreover, they point to the importance of signals through 4-1BB for the purposes of ex-vivo T cell activation and expansion. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Only a few costimulatory receptors are expressed in a constitutive fashion by un-stimulated T cells. The majority of costimulatory receptors are induced only following cell activation, subsequent to signal transduction through the TCR complex. Among those receptors, there is a central role for CD28 mediated costimulation as T cell activation in the absence of CD28 engagement results in a hyporesponsive state termed anergy [1]. CD28, a homodimeric, integral membrane receptor of the B7 immunoglobulin superfamily, is known as an essential component of the immunological synapse (IS) [2–4]. The majority of human CD4+ and almost half of CD8+ T cells express this receptor [5]. Studies on CD28 costimulation have led to a model by which costimulators enhance TCR engagement with peptide-MHC and prolong intracellular signalling [6]. Signals mediated by CD28 potently accelerate entry into and progression through the cell cycle [7]. Triggering CD28 can initiate other signalling pathways that are distinct from the TCR but lead to the activation of common targets, such as the upregulation of anti-apoptotic members of the Bcl-2 family [8,9]. Evidence shows that reactivation of memory CD4+ T cells may happen independent of the
* Corresponding author. Address: Immunotherapy Laboratory, Department of Immunology, School of Medicine, Shiraz University of Medical Sciences, Shiraz 71345-3119, Iran. Fax: +98 711 2351575. E-mail addresses: [email protected] (M. Habib-Agahi), habibagahim@yahoo. com (M. Habib-Agahi). 0008-8749/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.cellimm.2009.01.003
CD28 ligands, B7.1 (CD80) and B7.2 (CD86) [10]. This suggests that subsequent to initial T cell activation, a switch in the costimulatory requirement occurs. In addition, it has been shown that CD28 expression is significantly downregulated after activation in presence of CD80 costimulation [11], and occurs more rapidly in CD8+ cells [12]. The CD28null phenotype is considered a characteristic sign of replicative senescence of human T cells during repetitive stimulation in vitro, as well as in chronic inflammatory and infectious disease, and in the normal course of aging [13,14]. The exact mechanism of the downregulation of CD28 in T cells is not fully understood but an inoperative transcriptional initiator (INR) at the proximal region of CD28 promoter is involved in the regulation of gene transcription [15]. Microarray analysis comparing CD28null and CD28+ cell populations showed significantly downregulated expression of CD154 (CD40L) in CD28null cells. On the other hand, some other costimulatory receptors such as 4-1BB (CD137), CD244 and SLAMF7 significantly are upregulated in CD28null CD8+ memory T cells [16]. Presence of additional costimulatory interactions could ensure the potential for continued activation of CD28null T cells, as these cells still are functional. Signal transduction by 4-1BB of TNFR superfamily can not only costimulate T cell activation but also have additional effects on T cell survival [17–19]. Sharing many common features with CD28, costimulation via 4-1BB has major effects on CD8 T cells and promotes TH1 differentiation/cytokine production in CD4+ cells. 4-1BB is an inducible receptor and expressed on the surface of T cells after CD28 signalling and initial cell activation [20]. Therefore, we asked
40
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
if 4-1BB in a similar fashion could change the pattern of CD28 expression in activated T cells. Using adenoviral vectors to express CD80, CD86 and 4-1BBL costimulatory ligands on the surface of A549 lung carcinoma cells, we showed that CD28 expression does not change during T cell activation in presence of 4-1BBL and more interestingly, CD28 that was downregulated in response to CD80/ 86 costimulation could be re-expressed in T cells re-activated as a result of 4-1BBL costimulation. These findings may explain possible mechanism for reactivation of CD28null cells and it may suggest employing other costimulatory pathways rather than CD28 for exvivo T cell proliferation.
were used to indicate levels of background staining from different suppliers. 2.4. Statistical analysis The differences between various experimental groups were analyzed by two tailed Student’s t test, using SPSS software. Values of p < 0.05 were considered significant. 3. Results 3.1. CD28 expression profile in cultured lymphocytes
2. Materials and methods 2.1. Expression of costimulatory ligands by adenoviral vectors Production of E1- and E3-deleted, recombinant, non-replicative adenoviruses expressing full length human CD80, CD86, 4-1BBL and enhanced green fluorescent protein (GFP) have been reported previously [21,22]. A549 human lung carcinoma cells were infected with 300–600 virus particles/cell, 48–72 h prior to use to reach to the maximum ligand expression. Surface expression of the ligands was confirmed by flow cytometry analysis using commercial antibodies (data not shown, [22]).
3.1.1. CD80/CD86 costimulation but not 4-1BBL costimulation downregulate the frequency of CD28+ lymphocytes A549 cells do not show significant expression of costimulatory ligands (0.7–1% for CD80/86 and 3–5% for 4-1BBL) however 48 h after infection with 300 p/cell of recombinant viruses, the majority
2.2. Lymphocyte activation A549 cells growing in Dulbecco’s modified Eagle’s medium with HEPES (DMEM-HEPES, Sigma) supplemented with 10% foetal calf serum (FCS), 2 mM L-glutamine and antibiotics (100 iu/ml penicillin and 100 mg/ml streptomycin, Sigma, UK) were infected with the recombinant adenoviruses to express costimulatory ligands or GFP as control, and cultured in wells of 24-well plates (0.2–0.3 106 cells) as monolayer 48 hrs in advance. Infected cells were co-cultured with 1–1.5 106 cells/well of PBMCs prepared by Ficoll density-cushion centrifugation from blood of healthy donors (obtained with full informed consent) and depleted from plastic adherent cells. Depletion typically reduced CD14+ cells from 10– 15% to 3–5% of mononuclear cell preparations. Cultures were maintained in RPMI 1640 (Life Technologies, UK), supplemented with 7% FCS and 3% human (h)-AB serum (HD supplies, UK) in addition to Lglutamine and antibiotics and stimulated with 100 ng/ml soluble OKT3 anti-CD3 antibody (JANSSEN-CILAG, UK). Lymphocytes were passaged and re-stimulated approximately weekly, by transferring the adjusted number of cells to wells with fresh infected A549 cells expressing the corresponding costimulatory molecules. Lymphocyte proliferation was studied by cell surface marker staining and flow cytometry or viable cell count using hemocytometer and Trypan blue dye exclusion at the appropriate time points. 2.3. Antibody staining and flow cytomery Fresh or cultured PBMCs were incubated with approximately 1 lg/ml of appropriate fluorochrome-conjugated or isotype matched antibodies on ice in dark for 30 min. Un-bound antibodies were removed by two washes with PBS/FCS (2%). Cells were re-suspended in 500 ll buffer for flow cytometry, and analyzed using a four colours Beckman Coulter XL flow cytometer using Coulter System II software for data acquisition and WinMDI 2.8 software for data presentation. Labelled mouse monoclonal antibodies used in this study were sourced as follows: anti-CD4-FITC, anti-CD8-FITC, anti-CD4-PE, anti-CD8- PE, from Beckman Coulter (High Wycombe, UK), antiCD80-FITC, anti-CD80-PE, anti-CD86-PE, anti-CD137-PE, antiCD137L-PE, from BD, Pharmingen (Oxford, UK). Anti human CD28 antibody purchased from Sigma (UK). Appropriate isotype controls
Fig. 1. Effects of different costimulation regimens on CD4+ and CD8+ cell numbers in long-term culture. Peripheral blood mononuclear cells were stimulated by 100 ng/ml anti-CD3 antibody whilst co-cultured with A549 lung carcinoma cells expressing one or two costimulatory transgenes (or GFP as control) as described in Section 2. Cultures were passaged, with adjustment of cell density, and restimulated with anti-CD3 and fresh A549 cells expressing the same co-stimulatory ligands, at weekly intervals (indicated by arrows in B). Viable cell number was determined by haemocytometer and trypan blue dye exclusion, and the proportion of CD4+ (A) and CD8+ cells (B) determined by flow cytometry. Cultures of lymphocytes with either CD80 or CD86 costimulation decline after 7–10 days whilst regimens including 4-1BBL allow extended lymphocyte growth. Error bars indicate standard deviation of the mean. Statistically significant differences (p < 0.05) determined by two-tailed t test.
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
of infected cells (>80–95%) consistently expressed CD80, CD86, 41BBL or GFP transgenes, at levels comparable to those on a lymphoblastoid cell line or dendritic cells as showed before [22]. Using infected A549 cells as the major source of costimulation, previ-
41
ously we reported the different growth kinetics of T cells in response to CD80/86 and 4-1BBL costimulation [22]. In previous report we showed a transient T cell growth by CD80/86 costimulation which stopped eventually by the second week, regardless of
Fig. 2. Flow cytometry density plots of CD28 expression by CD4 and CD8+ cells before and after 4 days anti-CD3 stimulation with different costimulation treatments. Lymphocytes were stimulated as described in Fig. 1 and were stained with either PE conjugated anti-CD8 or anti-CD4 plus FITC conjugated anti-CD28 antibodies and analyzed by flow cytometry. As a reference, cells were also stained at day zero without anti-CD3 stimulation. A significant downregulation of CD28 expression was evident in cultures with CD80 and CD86 costimulation whilst cells with 4-1BBL costimulation kept the receptor presentation.
42
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
re-stimulation with anti-CD3 and CD80/86. In contrast, costimulation with 4-1BBL or its combination with CD80 or CD86 supported both CD4+ and CD8+ cells in continuous growth until the experiment was terminated (up to 5 weeks). Fig. 1A and B depict typical results of similar experiments with relative viable cell counts of CD4+ and CD8+ cells in different culture conditions during 35 days. To see if lymphocytes have different patterns of CD28 expression in response to different costimulations, the expression of CD28 by CD4+ and CD8+ lymphocytes was studied before, and then after 4, 27 and 35 days of stimulation in three independent experiments. Moreover, similar experiments with an EBV-QAK peptide antigen specific CD8+ T cell clone showed comparable results regarding to CD28 expression (data not shown). As a typical flow cytometry result, Fig. 2 shows that almost all CD4+ and 40–50% of CD8+ cells express CD28 before stimulation with OKT3 anti-CD3 antibody. However, by day 4, both CD80 and CD86 costimulations sharply downregulated the frequency of CD28+ expressing cells in CD4+ and CD8+ lymphocytes. As shown in Fig. 3, in CD80 costimulated cultures, CD4+CD28+ cells and CD8+CD28+ cells comprised 25 ± 3% and 18 ± 1% of lymphocytes, respectively, whilst CD86 costimulation caused a greater reduction in frequency of those populations to 14.3 ± 4% and 11.4 ± 1%, respectively (p = 0.01 and p = 0.04, respectively). Dual costimulation regimens of CD80 or CD86 plus 4-1BBL downregulated the frequency of CD28+ cells in both CD4+ and CD8+ cell compartments to the similar extent as observed with CD80 or CD86 alone (10–28% and 6–15%, respectively, p > 0.08 for CD4+ and CD8+ cells in both dual costimulation regimens). Conversely, costimulation with 4-1BBL which resulted in considerable proliferation of lymphocytes did not change the frequency of CD4+ and CD8+ cells with CD28 expression significantly (96.8 ± 0.5% and 47.2 ± 3.1%, respectively, p < 0.001 for both CD4 and CD8 cells). Coculture of lymphocytes with A549/Ad-GFP cells in control cultures also did not change these frequencies either. These effects on CD28 expression were sustained in cultures that were passaged and restimulated weekly with anti-CD3 activation and corresponding costimulation treatments, in experiments lasting up to day 27 or 35. Thus, after 4–5 weeks culture, CD80 or CD86 costimulated cultures had significantly lower proportions of CD28+ lymphocytes compared to cultures with 4-1BBL costimulation or control cultures (p < 0.0007 for CD4 cells, and p < 0.004 for CD8 cells, respectively) (Fig. 3). Significant decrease in CD8+CD28+ also observed compare to GFP control culture (p < 0.007). In parallel to the de-
crease in the frequency of the CD28 positive cells by CD80/86 costimulations, the level of cell-surface CD28 expression in the positive population, as indicated by the mean fluorescent intensities (MFIs), also diminished more rapidly. At time zero of this experiment, the MFI of CD28 expression of the CD4+ and CD8+ cells were 76 ± 7 and 59 ± 3, respectively. After 4 days, anti-CD3 activation with CD80 costimulation, the MFI of CD28 expression in CD4+ and CD8+ cells decreased to 26 ± 5 and 30 ± 3, respectively. At the same day, the intensity of CD28 expression by CD4+ and CD8+ cells in 4-1BBL costimulated cultures were found to be 60 ± 4 and 53 ± 2, respectively (data not shown). 3.1.2. Switching from CD80/CD86 costimulation to 4-1BBL results in CD28 re-expression As demonstrated previously [22], and shown in Fig. 1 for CD4+ and CD8+ cells separately, the proliferation of T cells stimulated with anti-CD3 and CD80/86 is transient, leaving them unresponsive or anergic to continued stimulation, with a gradual decline in viable cells. Provision of 4-1BBL costimulation for those anergic cells, after 3 weeks of culture, resulted in their reactivation and resumption of proliferation. To see if this reactivation associated with 4-1BBL costimulation may cause change in the pattern of CD28 expression we examined the expression of this receptor in CD4+ and CD8+ cells. As shown in Fig. 4, reactivation of lymphocytes was associated with significant increase in the frequency of CD4+ and CD8+ cells with CD28 expression. Switching from CD80 or CD86 costimulation to 4-1BBL increased the frequency of CD4+CD28+ cells in those cultures up to 88 ± 2.3% and 63 ± 3%, respectively. Similarly, there was an increase in the frequency of CD8+CD28+ cells of those cultures up to 48 ± 1.3% and 32 ± 3.6%, respectively (Fig. 5). As the results show, CD80 and CD86 pre-costimulated cultures may have different capacities (p < 0.04 for both CD4 and CD8 cells) in recovering CD28 expression after transferring to the new costimulatory conditions as CD80 treated lymphocytes re-expressed CD28 faster than CD86 pre-stimulated cells in both CD4+ and CD8+ cell compartments. Similarly, transfer of CD80/CD86 pre-costimulated lymphocytes to cultures with double costimulation of CD80/86 plus 4-1BBL also resulted in increase in the frequency of cells with CD28 expression in both CD4+ and CD8+ T cells during first week after transfer. Frequency of CD4+CD28+ cells reached up to 85 ± 2.1% and 61 ± 2.7% for dual costimulation regimens with either CD80 or CD86 plus 4-1BBL,
Fig. 3. Frequency of CD4+ and CD8+ lymphocytes with CD28 expression during 35 days culture with anti-CD3 stimulation and different costimulation treatments. PBMCs were cultured and stained as explained in Figs. 1 and 2 to determine the frequency of CD28 positive cells in CD4+ and CD8+ subpopulations during 35 days culture. The initial downregulation of CD28 in cultures with CD80 or CD86 was persistent until the last day of experiment however, cultures with 4-1BBL costimulation almost maintained the original levels of CD28 expression during this period. Error bars indicate standard deviation of the mean. Statistically significant differences (p < 0.05) determined by twotailed t test.
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
43
Fig. 4. Flow cytometry density plots showing re-expression of CD28 by CD4+ and CD8+ cells after transfer to conditions with 4-1BBL costimulation. PBMCs were prepared and cultured as explained in Fig. 1. After 21 days, half of lymphocytes previously costimulated with CD80 or CD86 were transferred into cultures with 4-1BBL costimulation with or without CD80 and CD86. Expression of CD28 in CD4+ and CD8+ subpopulations was studied 6 days after transfer. Whilst only a minority of lymphocytes in both original CD80 and CD86 costimulated cultures were expressing CD28, transfer of those cells to cultures with 4-1BBL costimulation resulted in significant re-expression of CD28 a week after transfer.
respectively, whilst almost 45 ± 3% and 21 ± 1.2% of CD8+ cells were expressing this receptor in the corresponding culture conditions. Study of these cells on day 35 showed that CD28 is downregulated again in dual costimulated cultures whilst those cells switched to single 4-1BBL costimulation, kept their CD28 expression (p < 0.001 for both CD4 and CD8 cells, Fig. 5). 3.2. Kinetics of 4-1BB expression on lymphocytes Non-activated lymphocytes do not express 4-1BB significantly. However, 4-1BB was moderately upregulated shortly after antiCD3 activation, to allow response to 4-1BBL costimulation. Activation by anti-CD3 in presence of CD86 costimulation showed faster kinetics of 4-1BB expression than CD80 costimulation in similar
condition (2.5 ± 0.5% vs 7.2 ± 1.1% for CD4+ cells, and 12.4 ± 1.8% vs 23.1 ± 2.6% for CD8+ cells, p < 0.01 for both CD4+ and CD8+ cells by day 4) (Fig. 6A and B). As shown, among lymphocytes, CD8+ cells reached a higher frequency of 4-1BB expression. 4-1BBL costimulation was less able to induce its receptor expression during the first week of culture than CD86 costimulation (p = 0.03). However, 41BBL treated and dual costimulated cultures, with continuation of their growth, showed more cells with 4-1BB expression by day 35 (p < 0.003). Moreover, transfer of growth arrested, CD80/86 costimulated cells to conditions with 4-1BBL costimulation or its cooperation with CD80/86, sustained 4-1BB expression in population of lymphocytes as well. Two weeks after transfer, more CD8 cells re-expressed the receptor in cultures with dual costimulation (up to 10-fold increase by 4-1BBL + CD86) (Fig. 6).
44
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
Fig. 5. Frequency of CD28 expressing cells among CD4+ and CD8+ subpopulations, 6 days (day 27) and 14 days after transfer (day 35) to cultures with 4-1BBL costimulation, with or without continued CD80 and CD86 costimulation. PBMCs were prepared and stained as explained in Fig. 4. Transferring lymphocytes previously costimulated with CD80 or CD86 to cultures with 4-1BBL costimulation (with or without continued CD80 and CD86) resulted in re-expression of CD28 in a high proportion of lymphocytes 6 days after transfer. However, that was later downregulated again in dually costimulated cultures. Error bars show SDM. Statistically significant differences (p < 0.05) determined by two-tailed t test.
4. Discussion In this study we showed that CD28, which is downregulated by costimulation through its ligands CD80 or CD86, can be re-expressed under the influence of 4-1BBL costimulation, in both CD4+ and CD8+ T cells. Several mechanisms i.e. TCR-mediated activation [23], in vitro replicative senescence [24], and CD80 or CD86 engagement [25] have been shown to be involved in downregulation of CD28 in T cells. Moreover, recent results indicate that the inducible CD80/86 receptor CTLA-4 promotes CD28 internalization and proteolysis, leading to further reduced surface expression [26]. By any mechanism, the net result is the emergence and accumulation of CD28null T cells. Several lines of evidence show that the majority of CD28 effector cells are programmed to die rapidly by apoptosis [27] or have reduced responses to an array of immune challenges, including pathogens [28] and allograft rejection [29]. On the other hand, recent findings show that CD28null cells are not terminally differentiated and could respond to TCR stimulation in a particular condition. Waller et al. showed the proliferative response of CMV specific CD28 CD45RAhi CD8+ T cells to antigen presentation through monocytes or artificial antigen presenting cells expressing CD137L. However, similar antigen presenting cells but with CD80, CD86 or CD275 have not been able to activate CD28 cells [30]. Thus, proliferation of CD28 cells appears to have a specific requirement for 4-1BBL. In our results, we demonstrated decrease in the frequency of CD28+ cells and showed decline in the intensity of the expression of this receptor within a few days after T cells were costimulated with CD80 or CD86 during anti-CD3 stimulation, confirming earlier observations [25]. Of these two B7-family ligands, greater decreases in CD28+ cell population consistently happened when cells were costimulated with CD86. This contrasts with a report by Lewis et al., in which CD86 costimulation of sub-lines of Jurkat T cell line, stimulated with PMA, did not cause significant decrease in CD28+ population [11]. We showed that, provision of 4-1BBL as a sole costimulation in conjunction with anti-CD3 stimulation, did not cause a significant change in CD28 expression even after one month culture. Nevertheless, combination of either CD80 or CD86 with 4-1BBL resulted in downregulation of CD28+ cell frequency, similar to that obtained using just CD80 or CD86 for costimulation. Thus, the greater mag-
nitude and longevity of T cell responses costimulated by the combined use of 4-1BBL with CD80 or CD86 does not appear to result from modulation of CD28 downregulation by 4-1BBL. As shown in Fig. 1 and reported before [22], CD4+ and CD8+ cell counts decline with CD80/86 costimulation after the second week of the culture; however switching of these non-responsive or anergic T cells to 41BBL costimulation condition resulted in reactivation and rapid proliferation within a few days post transfer [22]. We showed that combined costimulation using A549 cells expressing CD80 or CD86 with 4-1BBL allowed rapid and continued lymphocyte proliferation in response to a suboptimal concentration of anti-CD3 antibody, reaching up to 100-fold expansion of the lymphocyte population over 5 weeks without the use of exogenous cytokine or feeder cells. In recent report by Hippen et al. expression of 4-1BBL in conjunction with CD32 on the surface of K562 cells which then display membrane-bound anti-CD3 and anti-CD28 alongside the 4-1BBL, expansion levels of purified umbilical cord blood regulatory T cells exceeded 1250-fold [31]. Similarly, results of Suhoski et al. showed up to 104-fold expansion of CD8 T cells by artificial antigen presenting cells expressing CD80, 4-1BBL and anti-CD3 antibody [32]. These reports and others are in accord with our result regarding to the capability of 4-1BBL costimulation as a potent costimulation for T cell expansion. Lower T cell expansion in our results could be due to the weaker TCR signal from suboptimal, soluble anti-CD3 compared to the cell-attached antibody in other studies. Beside T cell proliferation and then resumption of growth by provision of 4-1BBL costimulation, in this report we showed reexpression of CD28 on the cells after they were transferred from CD80 or CD86 costimulation, and this was then sustained in cultures costimulated with 4-1BBL alone. Anergic T cells that were transferred to the condition of dual costimulation (i.e. 41BBL + CD80/86) also showed substantial subsequent proliferation, however only a transient increase in CD28 expression was observed before the receptor downregulated again, which emphasizes the dominant effect of CD80/CD86 on this process. Clearly, loss of CD28 is not an irreversible mechanism and CD28null cells may re-express this receptor if stimulated appropriately. Warrington et al. also have shown the re-expression of CD28 in CD4+CD28null lymphocytes after the combined action of IL-12 and anti-CD3 stimulation [33]. In their results CD28 expression was
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
45
Fig. 6. Frequency of CD4+ and CD8+ lymphocytes with 4-1BB expression in culture with anti-CD3 stimulation and different costimulation treatments. CD4+ (A) and CD8+ (B) lymphocytes were analyzed for expression of the 4-1BB at the time of culture establishment, and after 4 and 35 days of culture with anti-CD3 and different costimulation regimens. Portions of the cultures initially costimulated with CD80 or CD86 alone were transferred to conditions including 4-1BBL costimulation on day 21. By day 4, 4-1BB is upregulated, particularly with CD86 costimulation and on CD8+ cells. Transferring CD80/86 costimulated cells to regimens including 4-1BBL increased 4-1BB expression on CD8+ cells by day 35, relative to cultures with unchanged costimulation. Error bars show SDM. Statistically significant differences (p < 0.05) determined by two-tailed t test.
temporary, requiring the continued presence of IL-12 in the medium. Different members of TNF receptor superfamily, such as 4-1BB, have been proposed to act subsequent to CD28 downregulation, to compensate and regulate late T cell activation. This signal, at the peak of T cell expansion phase, maintains the survival of activated T cells when antigen becomes limiting [19,34–36]. This appears as one of the several bidirectional interactions among T cell costimulatory pathways. CD28 signal promotes the upregulation of CD40, OX40 and 4-1BB expression in T cells [37]. We also showed that CD80/86 costimulation are able to upregulate 4-1BB receptor, more effectively than 4-1BBL itself. However, growth arrested lymphocytes regained 4-1BB expression when they were costimulated with 4-1BBL alone or in combination with CD80/86. Therefore, as we have shown, there appears to be a regulatory role of 4-1BB for CD28 re-expression in scenario of T cell activation and there is a cross talk between these two costimulatory pathway to regulate their expression. It has been shown that signalling through CD137 and CD134, another member of the TNF receptor family, can upregulate telomerase activity of human CD8+ T cells. This is also an early function of CD28 during T cell activation [38]. Therefore, signalling through
4-1BB after its appearance on the surface of activated T cells, not only can provide costimulatory signal for their survival and further proliferation, but also according to our results, may cause reexpression of CD28 and restoration of the response to CD80/86 costimulation. In the context of aging of the immune system, reexpression of CD28 may be desirable to re-establish the state of immunocompetence. In contrast, induction of CD28 on CD28null T cells in autoimmune diseases may augment the autoimmune response. Mechanisms that restore CD28 expression are, therefore, potential therapeutic targets which need further investigations. Acknowledgment This work was funded by the Ministry of Health and Medical Education of Iran. References [1] Y. Zheng, Y. Zha, T.F. Gajewski, Molecular regulation of T-cell anergy, EMBO Rep. 9 (2008) 50–55. [2] P.G. Andres, K.C. Howland, D. Dresnek, S. Edmondson, A.K. Abbas, M.F. Krummel, CD28 signals in the immature immunological synapse, J. Immunol. 172 (2004) 5880–5886.
46
M. Habib-Agahi et al. / Cellular Immunology 256 (2009) 39–46
[3] S.K. Bromley, A. Iaboni, S.J. Davis, A. Whitty, J.M. Green, A.S. Shaw, A. Weiss, M.L. Dustin, The immunological synapse and CD28–CD80 interactions, Nat. Immunol. 2 (2001) 1159–1166. [4] S.A. Wetzel, T.W. McKeithan, D.C. Parker, Live-cell dynamics and the role of costimulation in immunological synapse formation, J. Immunol. 169 (2002) 6092–6101. [5] A. Viola, A. Lanzavecchia, T cell activation determined by T cell receptor number and tunable thresholds, Science 273 (1996) 104–106. [6] M.C. Miceli, M. Moran, C.D. Chung, V.P. Patel, T. Low, W. Zinnanti, Costimulation and counter-stimulation: lipid raft clustering controls TCR signaling and functional outcomes, Semin. Immunol. 13 (2001) 115–128. [7] J.L. Bonnevier, D.L. Mueller, Cutting edge: B7/CD28 interactions regulate cell cycle progression independent of the strength of TCR signaling, J. Immunol. 169 (2002) 6659–6663. [8] L.J. Appleman, A.A. Van Puijenbroek, K.M. Shu, L.M. Nadler, V.A. Boussiotis, CD28 costimulation mediates down-regulation of p27kip1 and cell cycle progression by activation of the PI3K/PKB signaling pathway in primary human T cells, J. Immunol. 168 (2002) 2729–2736. [9] K. Okkenhaug, L. Wu, K.M. Garza, R.J. La, W. Khoo, B. Odermatt, T.W. Mak, P.S. Ohashi, R. Rottapel, A point mutation in CD28 distinguishes proliferative signals from survival signals, Nat. Immunol. 2 (2001) 325–332. [10] A.M. Dahl, C. Klein, P.G. Andres, C.A. London, M.P. Lodge, R.C. Mulligan, A.K. Abbas, Expression of bcl-X(L) restores cell survival, but not proliferation off effector differentiation, in CD28-deficient T lymphocytes, J. Exp. Med. 191 (2000) 2031–2038. [11] D.E. Lewis, M. Merched-Sauvage, J.J. Goronzy, C.M. Weyand, A.N. Vallejo, Tumor necrosis factor-alpha and CD80 modulate CD28 expression through a similar mechanism of T-cell receptor-independent inhibition of transcription, J. Biol. Chem. 279 (2004) 29130–29138. [12] A.N. Vallejo, CD28 extinction in human T cells: altered functions and the program of T-cell senescence, Immunol. Rev. 205 (2005) 158–169. [13] M. Schirmer, C. Goldberger, R. Wurzner, C. Duftner, K.P. Pfeiffer, J. Clausen, G. Neumayr, A. Falkenbach, Circulating cytotoxic CD8(+) CD28( ) T cells in ankylosing spondylitis, Arthritis Res. 4 (2002) 71–76. [14] E. Bryl, A.N. Vallejo, C.M. Weyand, J.J. Goronzy, Down-regulation of CD28 expression by TNF-alpha, J. Immunol. 167 (2001) 3231–3238. [15] A.N. Vallejo, E. Bryl, K. Klarskov, S. Naylor, C.M. Weyand, J.J. Goronzy, Molecular basis for the loss of CD28 expression in senescent T cells, J. Biol. Chem. 277 (2002) 46940–46949. [16] M. Fann, W.K. Chiu, W.H. Wood III, B.L. Levine, K.G. Becker, N.P. Weng, Gene expression characteristics of CD28null memory phenotype CD8+ T cells and its implication in T-cell aging, Immunol. Rev. 205 (2005) 190–206. [17] M.R. Alderson, C.A. Smith, T.W. Tough, T. vis-Smith, R.J. Armitage, B. Falk, E. Roux, E. Baker, G.R. Sutherland, W.S. Din, Molecular and biological characterization of human 4-1BB and its ligand, Eur. J. Immunol. 24 (1994) 2219–2227. [18] C. Takahashi, R.S. Mittler, A.T. Vella, Cutting edge: 4-1BB is a bona fide CD8 T cell survival signal, J. Immunol. 162 (1999) 5037–5040. [19] M. Croft, Co-stimulatory members of the TNFR family: keys to effective T-cell immunity?, Nat Rev. Immunol. 3 (2003) 609–620. [20] K.E. Pollok, Y.J. Kim, Z. Zhou, J. Hurtado, K.K. Kim, R.T. Pickard, B.S. Kwon, Inducible T cell antigen 4-1BB. Analysis of expression and function, J. Immunol. 150 (1993) 771–781. [21] M.G. Gilligan, P. Knox, S. Weedon, R. Barton, D.J. Kerr, P. Searle, L.S. Young, Adenoviral delivery of B7-1 (CD80) increases the immunogenicity of human ovarian and cervical carcinoma cells, Gene Ther. 5 (1998) 965–974.
[22] M. Habib-Agahi, T.T. Phan, P.F. Searle, Co-stimulation with 4-1BB ligand allows extended T-cell proliferation, synergizes with CD80/CD86 and can reactivate anergic T cells, Int. Immunol. 19 (2007) 1383–1394. [23] D.J. Lenschow, T.L. Walunas, J.A. Bluestone, CD28/B7 system of T cell costimulation, Annu. Rev. Immunol. 14 (1996) 233–258. [24] R.B. Effros, N. Boucher, V. Porter, X. Zhu, C. Spaulding, R.L. Walford, M. Kronenberg, D. Cohen, F. Schachter, Decline in CD28+ T cells in centenarians and in long-term T cell cultures: a possible cause for both in vivo and in vitro immunosenescence, Exp. Gerontol. 29 (1994) 601–609. [25] S.C. Eck, D. Chang, A.D. Wells, L.A. Turka, Differential down-regulation of CD28 by B7-1 and B7-2 engagement, Transplantation 64 (1997) 1497– 1499. [26] M. Berg, N. Zavazava, Regulation of CD28 expression on CD8+ T cells by CTLA4, J. Leukoc. Biol. 83 (2008). [27] N.J. Borthwick, M. Lowdell, M. Salmon, A.N. Akbar, Loss of CD28 expression on CD8(+) T cells is induced by IL-2 receptor gamma chain signalling cytokines and type I IFN, and increases susceptibility to activation-induced apoptosis, Int. Immunol. 12 (2000) 1005–1013. [28] H.L. Compton, J.P. Farrell, CD28 costimulation and parasite dose combine to influence the susceptibility of BALB/c mice to infection with Leishmania major, J. Immunol. 168 (2002) 1302–1308. [29] B. Salomon, J.A. Bluestone, Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation, Annu. Rev. Immunol. 19 (2001) 225–252. [30] E.C. Waller, N. McKinney, R. Hicks, A.J. Carmichael, J.G. Sissons, M.R. Wills, Differential costimulation through CD137 (4-1BB) restores proliferation of human virus-specific ‘‘effector memory” (CD28( ) CD45RA(HI)) CD8(+) T cells, Blood 110 (2007) 4360–4366. [31] K.L. Hippen, P. Harker-Murray, S.B. Porter, S.C. Merkel, A. Londer, D.K. Taylor, M. Bina, A. Panoskaltsis-Mortari, P. Rubinstein, R.N. Van, T.N. Golovina, M.M. Suhoski, J.S. Miller, J.E. Wagner, C.H. June, J.L. Riley, B.R. Blazar, Umbilical cord blood regulatory T-cell expansion and functional effects of tumor necrosis factor receptor family members OX40 and 4-1BB expressed on artificial antigen-presenting cells, Blood 112 (2008) 2847–2857. [32] M.M. Suhoski, T.N. Golovina, N.A. Aqui, V.C. Tai, A. Varela-Rohena, M.C. Milone, R.G. Carroll, J.L. Riley, C.H. June, Engineering artificial antigen-presenting cells to express a diverse array of co-stimulatory molecules, Mol. Ther. 15 (2007) 981–988. [33] K.J. Warrington, A.N. Vallejo, C.M. Weyand, J.J. Goronzy, CD28 loss in senescent CD4+ T cells: reversal by interleukin-12 stimulation, Blood 101 (2003) 3543– 3549. [34] P.R. Rogers, J. Song, I. Gramaglia, N. Killeen, M. Croft, OX40 promotes Bcl-xL and Bcl-2 expression and is essential for long-term survival of CD4 T cells, Immunity 15 (2001) 445–455. [35] M. Croft, Costimulation of T cells by OX40, 4-1BB, and CD27, Cytokine Growth Factor Rev. 14 (2003) 265–273. [36] G. Pulle, M. Vidric, T.H. Watts, IL-15-dependent induction of 4-1BB promotes antigen-independent CD8 memory T cell survival, J. Immunol. 176 (2006) 2739–2748. [37] A.H. Sharpe, G.J. Freeman, The B7-CD28 superfamily, Nat. Rev. Immunol. 2 (2002) 116–126. [38] F.J. Plunkett, O. Franzese, H.M. Finney, J.M. Fletcher, L.L. Belaramani, M. Salmon, I. Dokal, D. Webster, A.D. Lawson, A.N. Akbar, The loss of telomerase activity in highly differentiated CD8+CD28 , J. Immunol. 178 (2007) 7710– 7719.