Published online before print May 2, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Université Montpellier 1, Centre d’étude d’agents Pathogènes et Biotechnologies pour la Santé, CNRS UMR 5236, Montpellier, France; and
School of Biosciences, University of Birmingham, Edgbaston, Birmingham, United Kingdom
1Correspondence: UMR 5236, Université de Montpellier II, Place Eugene Bataillon, CC 100, 34095 Montpellier Cedex 05, France. E-mail: vlafont{at}crit.univ-montp2.fr
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
β T cells. We also showed that STAT6 is activated by IL-2 in NKT cells. Moreover, IL-2 induces the production of IFN-
and IL-4. The ability of IL-2 to induce pro- and anti-inflammatory cytokine production, in addition to proliferation, could open new therapeutic approaches for use in combination with molecules that activate NKT cells through TCR activation.
Key Words: nonconventional lymphocytes • STAT
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
chain composed of V24-J18 that is preferentially paired with a Vβ11 chain in humans. They have been classified in the nonconventional or innate-like lymphocyte family based on their mixed properties, where they have been shown to interact with the innate and adaptive immune systems (reviewed in ref. [1
]). The main distinguishing feature of NKT cells, compared with conventional T cells (named in this study β T cells), is their capacity to react with glycolipids in the context of the MHC class I-related molecule CD1d [2
]. For a long time, the identity and structure of endogenous and exogenous glycolipids presented by CD1d and recognized by NKT cells were unknown. The marine sponge-derived glycosphingolipid -galactosylceramide (-GalCer) was the first synthetic NKT ligand identified that specifically stimulated NKT cells in the context of CD1d [3
]. Although, NKT cells have been shown to recognize and respond to an endogenous lysosomal glycosphingolipid, isoglobotrihexosylceramide (iGb3) [4
, 5
], recent studies brought evidence that iGb3 is unlikely to be a physiologically relevant ligand of inducible NKT in vivo [6
, 7
]. On the other hand, several bacterial-derived glycolipids have also recently been shown to activate NKT cells (reviewed in ref. [8
]). This is the case of glycosphingolipids present in the wall of Sphingomonas or Ehrlichia bacteria [9
, 10
], specific diacylglycerols of Borrelia burgdorferi [2
] and phosphatidylinositol mannosides of Mycobacterium [11
].
Another hallmark of NKT cells is their capacity to rapidly produce large amounts of cytokines after TCR activation, including the typical Th1 cytokine IFN- and the Th2 cytokine IL-4. Moreover, in addition to their capacity to produce anti- or proinflammatory cytokines, NKT cells can exhibit cytotoxic activity through the existence of two mechanisms: the expression of Fas ligand and potential to trigger cell death via the Fas pathway and the ability to lyse target cells by secreting lytic granules that contain perforin and granulysin [1
, 12
].
As a result of their functional properties, NKT cells have been implicated in myriad immune responses, including responses to pathogens, tumors, tissue grafts, allergens, and autoantigens (reviewed in ref. [8 ]). However, although NKT cells can potentially act as effectors (cytotoxic activity), it is likely that regulatory function reflects their true physiologic role. As a result of the functional discrepancy of producing pro- and anti-inflammatory cytokines and the significance of this in therapeutic research, it becomes important to determine how NKT cells "decide" which way to go. To date, several possible explanations have been proposed, which are not mutually exclusive: an environmental model and a subset model. The environmental model refers to the environment in which NKT cells are found, which can be influenced by the type of cytokines, APCs, and the strength of antigen-mediated TCR stimulation. Thus, their environment determines the cytokine profile of NKT cells. The subset model consists of pre-existing subsets of NKT cells, which are programmed to produce different ratios of Th1/Th2 cytokines (reviewed in refs. [13 , 14 ]).
Furthermore, NKT cells are phenotypically similar to NK cells. They express stimulatory and inhibitory NK cell receptors such as CD161 (NKR-P1) and CD94/NKG2A, respectively, which regulate their activation state and influence their biological responses such as cytokine production. In mice, cross-linking of CD161 results in high IFN- production without IL-4 production [15
]. Blockage of the inhibitory receptor CD94/NKG2A increases recall and primary responses of NKT cells [16
].
In summary, NKT cells are similar to V9V
2 T cells, another nonconventional T lymphocyte subset. As all T cells, they are activated through their TCR/CD3 complex, but their level of activation and biological responses are regulated by the presence and recruitment of NK cell receptors. In addition to these properties, V9V2 T cells possess other mechanisms of regulation such as expression of cytokine receptors. More particularly, IL-2R recruitment triggers biological responses in V9V2 T cells through specific signaling pathways [17
]. Therefore, we were interested in determining whether IL-2R recruitment leads to specific cell signaling pathways and biological responses in NKT cells.
Functional IL-2R has been identified on NK cells, and all activated T cell subsets express IL-2R composed of three subunits: the chain (or CD25) responsible for IL-2 binding and the β and chains responsible for signal transduction (reviewed in refs. [18
19
20
]). IL-2 stimulates the proliferation of T and NK cells [21
]. However, in NK cells, IL-2 has an additional effect of increasing cytotoxic function and production of IFN-, emphasizing the differences in the signaling pathways induced through IL-2R in T and NK cells [22
, 23
]. In previous studies, we have shown that V9V2 T cells have specific and unique signaling pathways, which are a combination between NK and T cell signaling pathways. In V9V2 T cells, IL-2 activates not only the STAT3, STAT5, PI-3K, and ERK-2 MAPK pathways but also, STAT4 as in NK cells and the p38 MAPK pathway as in β T cells. Furthermore, IL-2 induces the production of IFN- in V9V2 T cells, as observed in NK cells [17
]. In this study, we demonstrate that activation of IL-2R on NKT cells leads to the induction of PI-3K, ERK, p38 MAPK, and STAT3, -4, -5, and -6. Moreover, IL-2 activation of NKT cells leads to the production of pro- and anti-inflammatory cytokines, IFN-, and IL-4, respectively. Finally, these results support a role for IL-2-induced cell signaling and specific biological responses in NKT cell subsets and open a new therapeutic approach.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
24, -pan TCR β, and -TCR V9 mAb and UCHT1 (anti-CD3 mAb) were purchased from Beckman Coulter (Brea, CA, USA).
Cell culture and stimulation
PBMC from healthy donors were prepared by density centrifugation on Ficoll-Paque (Eurobio, les Ulis, France). NKT cells, V9V2, and β T lymphocytes were purified by positive immunoselection using an anti-TCR V24, -TCR V9, and -pan TCR β mAb, respectively. This step was followed by the addition of magnetic beads coated with anti-mouse IgG (Dynal, Compiègne, France), according to the manufacturer. Following overnight incubation at 37°C, NKT cells, V9V2, and β T lymphocytes were spontaneously separated from the magnetic beads. NK cells were also purified from PBMC by positive immunoselection using magnetic beads coated with anti-CD56 antibody (Miltenyl Biotec, France). After this first step of purification, CD56+CD3+ cells were removed from CD56+ cells using an anti-CD3 mAb followed by the addition of magnetic beads coated with anti-mouse IgG. The high-affinity receptor for IL-2 is not present on primary T cells and is only weakly present on NK cells. To induce a high and identical level of CD25 expression on different cell populations, T lymphocytes were activated by TCR/CD3 recruitment in the presence of syngeneic monocytes. A nonspecific activator such as an anti-CD3 (2 µg/ml) was used for β T cells, but specific activators -GalCer (100 ng/ml) and HMB-PP (10 nM) were preferred for minor T populations, NKT and V9V2 T cells, respectively. Although anti-CD3 treatment could be used for all three T cell populations, we preferred using specific activators, as they activate T cells through the TCR/CD3 complex by identical mechanisms, leading to the expression of activation markers such as CD25 and CD69 [24
]. Additionally, we verified the levels of STATs to show that these levels do not vary before or after CD25 induction nor following TCR activation 1 week following (Supplementary Fig. 1). To increase CD25 expression on NK cells, cells were treated with PMA (10 ng/ml) for 24 h. After activation, all cell subsets were expanded and maintained in RPMI 1640 supplemented with 5% FCS, 5% human AB serum, 2 mM glutamine, and rIL-2 (20 ng/ml) at 37°C in a 5% CO2 humidified atmosphere during 1 or 2 weeks. After 1 week of expanded culture, V9V2 T, β T, NK, and NKT cells had a purity of >95%. Before stimulation, cells were quiescent by washing twice and being placed in RPMI 1640 with 10% FCS for 24 h in the absence of IL-2, and then, CD25 expression was analyzed. Unless otherwise specified, cells (20x106/ml) were stimulated with rIL-2 (20 ng/ml) or rIL-12 (20 ng/ml) at indicated times.
Flow cytometry
Cells (0.5 million) were incubated with human AB serum for 30 min to block nonspecific sites. Then, they were incubated with conjugated antibodies in PBS supplemented with 2% FCS and 0.02% NaN3 in a total volume of 50 µL for 30 min on ice. They were washed once between each step of staining, fixed in 1% paraformaldehyde, and analyzed by FACSCalibur (Becton Dickinson, San Jose, CA, USA) with CellQuest software. NKT cells were incubated with PE-conjugated anti-V24 and FITC-conjugated anti-β11 antibodies. The cell surface expression of CD25 was verified on different cell subsets by staining with a PE-conjugated anti-CD25 antibody. Concerning -GalCer-CD1d tetramer staining, a preliminary step to load -GalCer is necessary and realized as follows: biotinylated CD1d were incubated with -GalCer overnight at 37°C and therefore, with streptavidin-FITC. After loading, NKT were incubated with biotinylated CD1d tetramers -GalCer for 1 h at room temperature and then incubated with PE-conjugated anti-CD3 antibody during 30 min at 4°C before analyzing on a FACSCalibur (Becton Dickinson) with CellQuest software.
Total cell extract preparation and Western blot analysis
After stimulation, cells were lysed in buffer containing 50 mM HEPES (pH 7.4), 150 mM NaCl, 10 mM NaF, 10 mM iodoacetamide, 1% Nonidet P-40 (NP-40), 1 mM PMSF, and 1 mM Na2VO3 in the presence of 1 µg/ml of each protease inhibitor (leupeptin, aprotinin, chymostatin) for 20 min. Proteins were separated by 7.5% or 10% SDS-PAGE, depending on the proteins studied, and then transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, Beford, MA, USA) and detected with the indicated antibodies: antiphospho-STAT4 (1:5000), anti-STAT4 (1:1000), antiphospho-p38 MAPK (1:1000), anti-p38 MAPK (1:1000), antiphospho-p42/44 MAPK (1:1000), anti-ERK-2 (1:5000), antiphospho (ser 473)-PKB (1:1000), anti-PKB (1:1000), antiphospho-STAT6 (1/1000), or anti-STAT6 (1:500) antibodies. Corresponding HRP-conjugated secondary antibodies were used, and immunoreactive bands were visualized with the chemiluminescence Western blotting system (Amersham).
Affinity purification of DNA-binding proteins
After activation, whole cell extracts were prepared by lysis of 20 x 106 cells/ml in lysis buffer containing 50 mM Tris-HCl, pH 7.9, 1% NP-40, 150 mM NaCl, 0.1 mM EDTA, 10 mM NaF, 1 mM Na2VO3, 1 mM PMSF, 1 mM DTT, 1 µg/ml aprotinin, 1 µg/ml leupeptin, and 1 µg/ml chymostatin. The oligonucleotide sequences derived from the high-affinity serum-inducible element (SIE) of the c-fos gene (SIEM67) GTCGACATTTCCCGTAAATC and the FcR-IFN--activated sequence [IFN- response region (GRR)] GTATTTCCCAGAAAAGGAAC were used to affinity-purify DNA-binding proteins. STAT3 and STAT5 were purified with the SIEM oligonucleotide and STAT4 and STAT6 with GRR. DNA-binding proteins were isolated from whole cell extracts in the above buffer by incubation of cell lysates at 4°C for 2 h with 1 µg double-stranded, 5'-biotinylated oligonucleotide coupled to 30 µl of a 50% suspension of streptavidin agarose (Sigma Chemical Co., St. Louis, MO, USA). Complexes were washed twice in lysis buffer and eluted by boiling in reducing sample buffer. Affinity-purified proteins were further separated by SDS-PAGE and transferred to PVDF membranes (Millipore). Western blot analysis was performed with antibodies for STAT3 (1:1000), STAT4 (1:1000), STAT5 (1:500), and STAT6 (1:500), and corresponding HRP-conjugated secondary antibodies were used. Immunoreactive bands were visualized with the chemiluminescence Western blotting system (Amersham).
Measurement of IFN- and IL-4 cytokine production
Cells (2x106 cells/mL) were cultured in 48-well tissue-culture plates in RPMI 1640 supplemented with 10% FCS in a total volume of 0.25 mL per well. Cells were stimulated with rIL-2 or anti-CD3 antibody. Following different stimulation times, supernatants were collected and assayed for IFN- and IL-4 production using the IFN- kit (OptEIA set: human IFN-, BD PharMingen, San Diego, CA, USA) and IL-4 kit (Diaclone, Besançon, France), according to the manufacturer’s instructions. The mean of triplicate samples from the same experiment is shown for each data-point with their SEM and is representative of at least three experiments performed with separate human blood donors.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-GalCer and anti-CD3 antibody was performed. A representative staining of NKT cell populations is shown in Figure 1
, left panel. In Figure 1A
, right panel, the majority of expanded human NKT expresses a semi-invariant TCR composed of a V24 chain paired with a Vβ11 chain.
 View larger version (35K): [in a new window] |
Figure 1. Analysis of NKT cell phenotype and CD25 expression. (A) Phenotype characterization of purified NKT cells was performed with a PE-conjugated anti-CD3 mAb and CD1d tetramer:-GalCer (left panel) and with a PE-conjugated anti-V24 mAb and a FITC-conjugated anti-β11 mAb (right panel). Then, the stained NKT cells were analyzed by flow cytometry. (B) All cell subsets used in this study were stained with a PE-conjugated anti-CD25 mAb and analyzed by flow cytometry. These analyses were realized before each activation experiment.
|
chain (or CD25) responsible for IL-2 binding and the β and chains responsible for signal transduction (reviewed in refs. [18
19
20
]). The high-affinity IL-2R is not present on primary T cells and is only weakly present on NK cells, as CD25 expression is low or absent. As described in Materials and Methods, a strong expression of CD25 can be transiently triggered following TCR/CD3 activation in T cells [25
] and PMA treatment in NK cells [26
]. To explore the signaling pathways induced by IL-2 in NKT cells and to compare them with those of β T, V9V2 T, and NK cells, cells that express strong and comparable levels of the high-affinity chain of IL-2R were used (Fig. 1)
. In β T cells, strong CD25 expression was observed for 1 week following TCR activation. In NKT cells and V9V2 T cells, strong CD25 expression was observed during 2 weeks following TCR activation, and in NK cells, strong CD25 expression was observed for 1 week following PMA treatment (data not shown). Based on the CD25 expression analysis, we chose to study the effects of IL-2 1 week post-CD25 induction. At this time, CD25 is expressed by all cell subsets, and all effects as a result of CD25 induction are gone (data not shown). Thus, even if autocrine amounts of IL-2 are produced differently by all three T cell populations, interference with IL-2 stimulation is avoided.
IL-2 triggers phosphorylation of STAT4 in NKT cells
As IL-2 signaling has been shown to differ among NK, β T, and V9V2 T cells, particularly in STAT4 recruitment [17
], we first analyzed the phosphorylation of STAT4 induced by IL-2 in CD25+ NKT cells. Following incubation of the quiescent cells with IL-2, the tyrosine-phosphorylated form of STAT4 was observed (Fig. 2A
). IL-2 induced rapid phosphorylation of STAT4 in NKT cells, which was detectable at 5 min, reached a maximum at 10 min, and was sustained for at least 30 min, as described previously in NK and V9V2 T cells [17
]. IL-2-triggered STAT4 activation in NKT cells was identical to CD25 expression induced by an anti-CD3 antibody or -GalCer treatments (Supplementary Fig. 2).
 View larger version (15K): [in a new window] |
Figure 2. IL-2 triggers phosphorylation and DNA binding of STAT4 in NKT cells (A), which were activated by IL-2 at different times and then lysed. Total protein extracts were separated on a 7.5% SDS-PAGE and transferred to a PVDF membrane. An anti-(phospho or nonphospho) STAT4 antibody was used for immunoblotting (upper panel). Band intensities of phosphorylated proteins have been quantified and normalized by comparing band intensities of phosphorylated and unphosphorylated proteins and are shown in the histogram (lower panel). A.U., arbitrary units. (B) Various cell types were activated by IL-2 or IL-12 at different times and then lysed. Affinity purification of DNA-binding proteins was carried out with GRR oligonucleotides. Purified proteins were separated on a 7.5% SDS-PAGE, transferred to a PVDF membrane, and observed by Western blot analysis using an anti-STAT4 antibody. These experiments are representative of three experiments, and each experiment was performed with cells from different donors.
|
9V2 T, and NK cells but not in β T cellsβ T lymphocytes (Fig. 2B)
. In contrast, an intense band was detected in complexes from IL-12-activated β T lymphocytes (positive control) and IL-2-activated NKT, V9V2 T, and NK cells. For these last three populations, the kinetics of STAT4 binding is similar with rapid (5 min) and sustained binding, lasting at least 60 min. As demonstrated for V9V2 T [17
] and NK cells [28
], the rapid (5 min) phosphorylation and binding of STAT4 following IL-2 activation in NKT cells suggest that the activation of STAT4 is a direct effect of IL-2 and not a result of an indirect effect mediated through the autocrine secretion of STAT4-activating cytokines such as IL-12 or IFN-.
Other STAT proteins activated by IL-2 in NKT cells
STAT proteins such as STAT3 and STAT5 are recruited and tyrosine-phosphorylated upon IL-2 activation in all T cell subsets and NK cells [29
30
31
32
]. Phosphorylated forms of STAT3 and STAT5 can then dimerize, translocate to the nucleus, and bind DNA in the promoter region of STAT-regulated genes [33
, 34
]. In a similar way to that done for STAT4, we investigated DNA binding of STAT3 and STAT5 to biotinylated oligonucleotides with high-affinity binding sites for STAT3 and STAT5 (SIEM) to generate an affinity matrix to purify DNA-binding STAT complexes from cell lysates [30
, 35
]. Briefly, cells incubated with IL-2 were lysed, and the amount of protein present in extracts was normalized (Supplementary Fig. 3B). Then, proteins were purified by SIEM affinity-purified, separated by SDS-PAGE, and analyzed by Western blot with a specific anti-STAT3 or anti-STAT5 antibody. No STAT3 or STAT5 proteins were detected in SIEM complexes isolated from nonactivated cells of all cell populations used (Fig. 3
). In contrast, an intense band was rapidly (5 min) detected in complexes from IL-2-activated cells in all cell populations and lasted 60 min. In conclusion, IL-2 also activates STAT3 and STAT5 in NKT cells, and the kinetics of STAT activation in NKT cells is similar to that observed in β T, V9V2 T, and NK cells [17
, 28
, 30
, 35
].
 View larger version (42K): [in a new window] |
Figure 3. Activation of STAT3 and STAT5 proteins by IL-2 in NKT cells. Various cell types were activated by IL-2 at different times and then lysed. Affinity purification of DNA-binding protein was achieved with SIEM-specific oligonucleotides. Purified proteins were separated on 7.5% SDS-PAGE, transferred to a PVDF membrane, and observed by Western blot analysis using anti-STAT3 and -STAT5 antibodies. These experiments are representative of three experiments, and each one was performed with cells from different donors.
|
β T, V9V2 T, and NK cells [17
], showing a rapid (5–10 min) and sustained (60 min) phosphorylation.
 View larger version (51K): [in a new window] |
Figure 4. Activation of PI-3K, ERK-2, and p38 MAPK pathways by IL-2 in NKT cells, which were activated at different times with IL-2 and then lysed. Total protein extracts were collected and separated on a 10% SDS-PAGE, transferred to a PVDF membrane, and observed by Western blot analysis using antiphospho-PKB followed by anti-PKB antibodies (A), antiphospho-p42/44 MAPK followed by anti-ERK-2 antibodies (B), or antiphospho-p38 followed by anti-p38 antibodies (C). Band intensities of phosphorylated proteins have been quantified, normalized by comparing band intensities of phosphorylated and unphosphorylated proteins, and are shown in the histogram. These experiments are representative of at least three experiments, and each one was performed with cells from different donors.
|
β T, V9V2 T, and NK cells [17
], rapid (10 min) and sustained (30–60 min). On the other hand, results for p38 MAPK were different from one cell type to another. Previously, it was shown that IL-2 activates the p38 kinase pathway in β T cells [42
] but not in NK cells [41
]. More recently, we have shown that p38 was also phosphorylated and activated by IL-2 in V9V2 T cells [17
]. Using a specific antibody, which only interacts with the phosphorylated form of p38 kinase, we confirmed that p38 was phosphorylated and activated by IL-2 in V9V2 T and β T cells and not in NK cells. Moreover, we showed that p38 is also recruited and activated in NKT cells (Fig. 4C)
. Nevertheless, kinetics of phosphorylation and activation differs from one subset to another. A rapid (10 min) and sustained (60 min) phosphorylation of p38 kinase was observed in NKT cells, whereas the phosphorylation of p38 kinase is delayed and only visible from 30 min in V9V2 T and βT cells. These results show that mechanisms leading to activation of p38 are different. The rapid activation of p38 suggests a direct recruitment of the p38 pathway in NKT cells, whereas in V9V2 T and β T cells, p38 activation could be a result of an indirect mechanism and could result in different regulation of biological responses.
IL-2 induces IFN- and IL-4 production by NKT cells
IL-2 is a pleiotropic cytokine required for proliferation and activation of many cell types, including T and NK cells [20
, 21
]. However, in NK and V9V2 T cells, IL-2 has the additional effect of increasing cytotoxic function and IFN- production [22
, 23
]. STAT4 plays a role in IFN- production by regulating multiple components of IFN--inducing signaling pathways. As STAT4 is activated by IL-2 in NKT cells, we wondered if IFN- production could be induced by IL-2, as is the case in NK and V9V2 T cells. We investigated IFN- production in IL-2-stimulated NKT cells by comparing the kinetics over 24 h of NK, β T, and V9V2 T cells. IFN- production in NKT cells is detected, starting from 6 h, and reaches a plateau at 18 h (Fig. 5A
). IFN- production is detected at 18 h and 24 h in NK and V9V2 T cells; no IFN- production was detected in supernatants from IL-2-activated β T cells. Therefore, in terms of IFN- production, NKT cells behave more like NK cells than like β T cells, as they can produce IFN- upon IL-2 activation without the requirement of cosignaling, sharing yet another characteristic of V9V2 T cells [17
]. Moreover, as NKT cells also have the ability to produce IL-4, we determined whether IL-2 induced IL-4 production and compared NKT to β T cells. NK and blood-expanded V9V2 T cells do not produce IL-4 and were excluded from the study. IL-4 production is observed in IL-2-activated NKT cells (at 18 h and 24 h points) but not in β T cells (Fig. 5B)
. Thus, pro- and anti-inflammatory cytokines are produced by NKT cells following IL-2 stimulation. However, IFN- is produced before IL-4 production and reaches a plateau at 18 h in the presence of IL-2. Moreover, we have analyzed and compared IL-4 and IFN- production induced by IL-2 and TCR/CD3 activation in NKT cells. IL-4 is more strongly induced by IL-2 than by anti-CD3, and IFN- is less induced by IL-2 than by anti-CD3 (Fig. 5C)
. This suggests that IL-2 could be considered to be a weak or strong activator depending on the biological response measured.
 View larger version (22K): [in a new window] |
Figure 5. IFN- and IL-4 production induced by IL-2. (A) Different cell subsets (2x106 cells/mL) were incubated with IL-2 ( ) or not (•). At indicated times, supernatants were removed and tested for the presence of IFN-. (B) NKT and β T cells (2x106 cells/mL) were incubated with IL-2 () or not (•). At indicated times, supernatants were removed and tested for the presence of IL-4. (C) NKT cells (2x106 cells/mL) were incubated with IL-2 or coated anti-CD3 antibody or not (NS) for 18 h. Then, supernatants were removed and tested for the presence of IFN- (upper panel) and IL-4 (lower panel). These experiments are representative of three experiments, and each one was performed with cells from different donors.
|
 View larger version (31K): [in a new window] |
Figure 6. STAT6 pathway recruitment in NKT cells (A), which were activated with IL-2 at different times and then lysed. Total protein extracts were resolved on 7.5% SDS-PAGE, transferred to a PVDF membrane, and observed by Western blot analysis using antiphospho-STAT6 followed by anti-STAT6 antibodies (upper panel). Band intensities of phosphorylated proteins have been quantified and normalized by comparing band intensities of phosphorylated and unphosphorylated proteins and are shown on the histogram (lower panel). (B) NKT cells were activated with IL-2 at different times and then lysed. Affinity purification of DNA-binding proteins was carried out with GRR oligonucleotides. Purified proteins were then separated on a 7.5% SDS-PAGE, transferred to a PVDF membrane, and observed by Western blot analysis using an anti-STAT6 antibody. This experiment is representative of three experiments, and each experiment was performed with cells from different donors.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Therefore, we studied IL-2-induced regulation of NKT cells by determining signaling pathways and biological responses triggered following the recruitment of IL-2R by IL-2, which can lead to distinct effects, depending on the cell population. IL-2 can induce a variety of intracellular events in T and NK cells, some of which overlap (such as those controlling proliferation and survival) [20
, 21
] and some of which are distinct (such as those controlling cytotoxicity and cytokine production) [22
, 23
]. We have previously shown that the pattern of IL-2-induced signaling in V9V2 T cells is a combination of NK and β T cell profiles and that IL-2 triggers IFN- production [17
]. Hence, signaling pathways and biological responses induced by IL-2 in NKT cells could also have a mixed pattern. Therefore, we studied IL-2-triggered signaling in NKT cells and have shown that the pattern of IL-2-induced signaling is specific and leads to pro- and anti-inflammatory cytokine production.
One of the major signaling pathways triggered by cytokines or growth factors is the Jak/STAT pathway. The STAT family of transcription factors appears to play an important role in mediating gene activation by most hematopoietic and many nonhematopoietic cytokines, growth factors, and neurotrophic factors [33
]. Some members, such as STAT1, STAT3, and STAT5, are activated in numerous cell types by a wide variety of cytokines. These proteins may be involved in mediating more general types of signals, such as those for cell growth and survival [47
]. Others, including STAT4 and STAT6, are more restricted in their expression and are activated in response to a small number of factors, reflecting a more specialized role [48
]. Notably, studies using STAT4-deficient mice have demonstrated that STAT4 is essential for IL-12 signaling and that Th1 development is severely impaired in these mice [49
], and other studies have shown that STAT6 is involved in IL-4 signaling and Th2 development [50
]. Wang and co-workers [28
] have reported that STAT4 is activated by IL-2 in NK cells and proposed that activation of STAT4 may control cytotoxic activity induced by IL-2. More recently, we have shown that IL-2 induced transcriptional activity of STAT4 in V9V2 T cells and that STAT4 could be involved in IL-2-induced IFN- production [17
]. These cell-type specificities of STAT activation are an attractive model for increasing the diversity of cellular responses induced by a given cytokine.
In the present paper, we showed that IL-2 induces tyrosine phosphorylation and DNA binding of STAT4 in NKT cells. These data suggest that IL-2-induced STAT4 activation is specific to the NK cell lineage and nonconventional T cells. As in NK cells and V9V2 T cells, the rapid phosphorylation and transcriptional activity of STAT4 (5 min, Fig. 2
) suggest that the IL-2-triggered STAT4 activation in NKT cells is a direct mechanism and not the consequence of production and release of cytokines such as IL-12 and/or IFN- (two cytokines triggering STAT4 activation) following IL-2 treatment. Although DNA-binding and transcriptional activity of STAT depend on tyrosine phosphorylation, serine phosphorylation is required to attain full transcriptional activity. This has already been described for STAT1, STAT3, STAT5 [51
, 52
], and more recently, for STAT4 [53
]. In Figure 2
, we can observe two bands with distinct electrophoretic mobility corresponding to tyrosine-phosphorylated STAT4 proteins. The existence of these two tyrosine-phosphorylated forms has been reported in a previous study, where it was shown that the more rapidly migrating form is only phosphorylated on a tyrosine residue (tyrosine 693), and the more slowly migrating form is phosphorylated on tyrosine 693 and serine 721 residues [27
]. Therefore, the slowly migrating form in our study suggests that IL-2 induces serine and tyrosine phosphorylation of STAT4 in NKT cells. As serine phosphorylation of STAT4 is critical for IL-12-induced IFN- production [53
], it may also be involved in IL-2-induced IFN- production. Moreover, we have compared the levels of STAT4 activation induced by IL-2 or IL-12 in NKT cells and found that IL-2 was as strong an activator of STAT4 as IL-12.
In regards to other STAT proteins, IL-2 also induces tyrosine phosphorylation and activation of STAT3 and STAT5 in T and NK cell lineage [28
, 54
]. Hence, we confirmed that STAT3 and STAT5 are activated by IL-2 in NKT cells, and the kinetics of STAT activation was found to be similar to β T, V9V2 T, and NK cells [17
, 28
, 30
, 35
]. Previously published data concerning STAT1 recruitment upon IL-2 activation have yielded conflicting results. Some investigators have reported STAT1 activation in NK cells [47
] but not in T cells [34
, 48
]. In addition, in the cases where STAT1 activation was reported, activation is weak in comparison with that induced by IFN-. The difference in the data could be explained by a difference in the sensitivity of experimental protocols and/or cells used. In previous studies, we have shown that although a weak phosphorylation of STAT1 can be detected occasionally, no transcriptional activity of STAT1 induced by IL-2 was associated with phosphorylation status in T cells. We have studied the activation status of STAT1 in IL-2-stimulated NKT cells, but no transcriptional activity of STAT1 has been observed (Supplementary Fig. 4). This suggests that IFN- production induced by IL-2 in NKT cells cannot be involved in the STAT1 pathway. In summary, the IL-2-induced STAT pathway shows that NKT cells act similarly to NK and V9V2 T cells in terms of STAT3, STAT4, and STAT5 activation.
In addition to the activation of the STAT pathway, the recruitment of IL-2R triggers a series of intracellular signaling pathways including the PI-3K and MAPK pathways. It has been reported that the activation of PI-3K is necessary for IL-2-induced growth and differentiation [18
, 36
]. Also, it has been established that IL-2 activates PKB (also called Akt) by a PI-3K-dependent pathway [39
] and that PKB can regulate survival responses in a variety of cell types [37
, 38
]. PI-3K, through other effectors, such as E2F, couples IL-2R to cell-cycle regulation [39
]. In NKT cells, we have shown that IL-2 activates the PI-3K pathway and that the kinetic profile was similar to that observed in β T, V9V2 T, and NK cells (Fig. 4A)
. Thus, the activation of this pathway may influence the regulation of cell-cycle progression leading to proliferation and survival in NKT cells.
In activated, primary T cells and T cell lines, ERK and p38 MAPK pathways are activated by IL-2 [18
, 36
]. Yu et al. [41
] have demonstrated that in NK cells, IL-2 activates the ERK-2 pathway but not the p38 kinase pathway. Considering that V9V2 T cells resemble NK cells more than β T cells in terms of the IL-2-induced STAT pathway, we studied the implication of MAPK activation in V9V2 T cells and showed that the p38 and ERK-2 kinase pathways are triggered by IL-2. These findings and data from other published reports [40
, 41
, 55
] suggest that the ERK-2 pathway is activated in all types of IL-2-responder cells, but activation of the p38 pathway is limited to cells of the T cell lineage. In the present study, we showed that ERK-2 and p38 MAPK pathways are recruited by IL-2 in NKT cells. Activation kinetics of ERK-2 is similar to that observed in β T, V9V2 T, and NK cells (Fig. 4B)
, and our data showed that the kinetics of phosphorylation and activation of p38 MAPK induced by IL-2 vary from one cell type to another (Fig. 4B)
. The phosphorylation of p38 MAPK is rapid (10 min) and sustained (60 min) in NKT cells and delayed (30 min) in β T and V9V2 T cells. Differences in the activation kinetics could reflect variations in p38 kinase-regulated biological responses, as the p38 pathway has been associated with various biological responses such as proliferation and cytokine production [56
]. Notably, p38 kinase is involved in IL-2 induction of the TNF-β gene in β T cells [57
] and is required for IFN- and TNF- production by V9V2 T cells [58
, 59
]. Other reports have shown that p38 is also implicated in Th2 cytokine gene regulation such as IL-4. Together, this suggests that IL-2-induced p38 pathway activation could be involved in production by NKT cells of IFN- and IL-4, two cytokines with contrasting effects, pro- and anti-inflammatory, respectively. Early activation of p38 kinase could be involved in IL-4 regulation, and late activation could be involved in IFN- regulation.
IL-2 is not the only cytokine that activates NKT independently from TCR engagement [60
]. A combination of IL-12 and IL-18 has been shown to activate antitumor activity of NKT cells [61
]. Thus, it is possible that IL-12 could be involved in the effects of IL-2 on STAT activation and cytokine (IL-4 and/or IFN-) production. IL-12 is produced by dendritic cells, macrophages, and human B lymphoblastoid cells but not by T lymphoid cells as NKT cells. In our study of STAT activation and IL-2-induced cytokine production in NKT cells, no APCs are present and thus, no cells that produce IL-12. Therefore, it is unlikely that endogenous IL-12 is implicated in the effects of IL-2. Also, it has been reported that IL-4 induces production of IFN- by NKT cells [62
]. We have shown that IL-2-induced NKT cells produce IFN- before IL-4. Therefore, IFN- production is not a result of an indirect effect mediated by IL-4. Moreover, we have analyzed and compared IL-4 and IFN- production induced by IL-2 and TCR/CD3 activation in NKT cells. IL-4 production induced by IL-2 is higher than that induced by anti-CD3, and IFN- production induced by IL-2 is lower than at induced by anti-CD3. This suggests that the production of cytokines is differently regulated and specific to the activation pathway.
Functional IL-2R is identical in all subsets of cells and composed of three subunits: the chain (or CD25) responsible for IL-2 binding and the β and chains responsible for transduction signals, but its recruitment by IL-2 triggers various signaling pathways and biological responses depending on the cell type. Although these results appear inconsistent, they could be explained by several mechanisms. CD25, the high-affinity chain of IL-2, is located in the microdomains of the plasma membrane [63
]. It is possible that the proteins present or close to these structures could be different depending on the cell types. Proteins present in the microdomains could be receptors or molecules responsible for signal transduction. In particular, other cytokine receptors, which share one or more chains with IL-2R, could be present and would allow IL-2 to trigger at least some of the signaling events induced by other cytokines (IL-12, IFN-, IL-4). In NKT cells, this is particularly relevant for IL-4, in which the receptor shares the chain with IL-2R and surprisingly, induces IFN- [62
]. Additionally, there are other mechanisms responsible for the recruitment of STAT to IL-2R complexes. For example, novel docking proteins such as signal transducing adaptor molecule, in which the expression is dependent on cell type, might allow the recruitment of various Jak/STAT complexes to the IL-2R. The IL-2 signaling pathway, which differs from the established paradigm of IL-2R signaling, has not only been reported for nonconventional T cells [17
] but also for CD4+CD25+ regulatory T cells [64
], in which IL-2 induces an intact Jak/STAT kinase pathway but does not activate downstream targets of PI-3K, and as a result, IL-2 has a hypoproliferative effect on these cells. In this case, the authors demonstrated that negative regulation of the PI-3K signaling pathway is inversely associated with expression of the lipid phosphatase and tensin homologue that is deleted on chromosome 10.
In this study, we have shown that IL-2 induces not only the proliferation of NKT cells but also IFN- and IL-4 production. As a result of these effects, IL-2 could be used to potentiate various types of NKT cell therapies also used to amplify therapeutic effects of molecules such as -GalCer, which activate NKT cells though TCR activation. As has been investigated by Parekh et al. [65
], in mice, IL-2 could maintain NKT cells in an active and productive state, particularly to compensate for their nonreactive state, which has been described in vivo after -GalCer injection. Potential therapeutic compounds such as -GalCer induce activation of NKT cells, production of cytokines, and down-modulation of TCR leading to a TCR-nonresponding state. Moreover, use of -GalCer analogs characterized for their Th1 or Th2 cytokine induction in the presence of IL-2 could open new therapeutic approaches for direct use.
Received October 4, 2007; revised March 20, 2008; accepted March 26, 2008.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
14 NKT cells by glycosylceramides Science 278,1626-1629 9V 2 T cells by NKG2D J. Immunol. 175,2144-215114 NKT cells in innate and acquired immune response Annu. Rev. Immunol. 21,483-513[CrossRef][Medline] production by natural killer (NK) cells and NK1.1+ T cells upon NKR-P1 cross-linking J. Exp. Med. 183,2391-2396}9V{}2 T cells use a combination of mechanisms to limit the spread of the pathogenic bacteria Brucella J. Leukoc. Biol. 77,652-660 9V 2 T cells: an amalgamation of NK and β T cell signaling J. Immunol. 171,5225-5232 T lymphocytes without inducing down-modulation of T cell antigen receptor J. Biol. Chem. 276,15961-15967 production in a manner distinct from LPS activation of monocytes J. Immunol. 162,659-668, but not the T cell antigen receptor, in human T lymphocytes EMBO J. 13,5605-5615[Medline] T cells expressing NK cell receptors J. Immunol. 159,3723-3730[Abstract] T lymphocytes involved in intravesical bacillus Calmette-Guerin immunotherapy of bladder cancer? Biochem. Soc. Trans. 25,363S[Medline] production but not for cell proliferation Proc. Natl. Acad. Sci. USA 99,12281-12286 by human V 9V 2 T cells via engagement of Fc RIIIA, the low affinity type 3 receptor for the Fc portion of IgG, expressed upon TCR activation by nonpeptidic antigen J. Immunol. 166,7190-7199 by NK and NKT cells J. Immunol. 176,5299-5305 subunit of the interleukin-2 receptor (CD25) to influence interleukin-2 receptor signal transduction Immunol. Cell Biol. 80,351-357[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
