Published online before print April 26, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
induction by calcineurin antagonists in activated human lymphocytesDepartment of Dermatology and Allergology, Hannover Medical School, Germany
1 Correspondence: Hannover Medical School, Department of Dermatology and Allergology, Ricklinger Str. 5, D-30449 Hannover, Germany. E-mail: Wittmann.Miriam{at}MH-Hannover.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(IFN-) by human T cells. This increased IFN- production is dependent on T cell receptor (TCR) and CD28 signaling as well as on the presence of IL-12. IL-27, which could mimic the effect of IL-12, was however less potent in inducing IFN- production in the presence of CsA and TCR stimulation. Other cytokines such as IL-23, IL-18, IL-2, or the Th2-related cytokine IL-4 are not able to support a calcineurin antagonist-dependent up-regulation of IFN-. CsA-dependent IFN- production is observable in therapeutic concentrations. The effect is independent of IL-10 or IL-4, as addition of these cytokines could not inhibit the CsA-induced IFN- production. The effect of calcineurin antagonists is associated with an increased c-fos expression and DNA-binding activity of the transcription factor activated protein-1 but not with increased DNA-binding activity of T-bet. Our study further supports the relevance of known calcineurin activities other than NFAT activation. The presented data may help to explain why concomitant infections (resulting in increased IL-12 expression) under therapy with calcineurin antagonists often have a negative impact on the activity of the underlying disease (e.g., autoimmune disease).
Key Words: T cells • cytokines • cell activation • signal transduction
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It was the goal of this study to better understand the mechanism of interferon- (IFN-) production by human T cells activated in the presence of calcineurin inhibitors (at concentrations seen in plasma/tissue under therapy). It has become clear that under the influence of calcineurin inhibitors, T cell cytokine production is not simply switched-off, but T cells activated in the presence of CsA display a differential pattern of gene expression [1
, 2
]. We focused on the cytokine IFN- because of its central role in disease maintenance and progression, as shown for chronic inflammatory skin diseases, autoimmune disorders, and GvHD/organ allograft survival.
IFN- is produced by CD4+ T helper cell type (Th) cells (defining the Th1 lineage), natural killer cells, and CD8+ T cells. IFN- is essential for mounting a cell-based immune response, promotes protection against a number of pathogens, but also plays a key role in the chronification of inflammatory responses (e.g., atopic dermatitis, rheumatoid arthritis, psoriasis). Of outstanding interest is to understand the interaction of signaling events leading to fine-tuned production of this key regulatory mediator.
Production of the Th1 cytokine IFN- is influenced by a number of transcription factors, especially T-bet, Ets-related molecule (ERM), and nuclear factor (NF)-
B [3
4
5
6
7
8
9
10
]. In addition, regulation of IFN- gene expression by NF of activated T cells (NFAT) is well-recognized [11
12
13
14
15
16
].
The activation of NFAT is inhibited by calcineurin antagonists such as CsA and tacrolimus, which act through ligation of distinct intracellular-binding protein targets (cyclophilin A and FK506-binding protein, respectively). These drug-binding protein complexes inhibit the function of the calcium-dependent phosphatase calcineurin, which activates NFAT. However, it is clear from a number of studies that NFAT is not the only target of calcineurin, which has been described to interact with other transcription factors such as Ets-like protein 1 (Elk1), cyclic adenosine monophosphate response element-binding protein, and myocyte enhancer-binding factor 2 [17 ].
In clinical care, failure to respond to calcineurin antagonists is a poorly understood problem. Many investigations have been undertaken to elucidate the factors contributing to T cell resistance to these drugs [18
19
20
21
22
]. Here, we demonstrate that T cells cannot only become resistant to the effect of calcineurin inhibitors but actually become activated differentially. In this study, we show that the calcineurin inhibitors CsA and tacrolimus can markedly up-regulate the production of IFN-. This increased IFN- production is dependent on a strong T cell receptor (TCR) signal, a costimulatory signal via CD28, and the presence of interleukin (IL)-12 (or IL-27). Of note, an increased IFN- production under the influence of CsA has been observed previously [2
, 23
, 24
]. However, IL-12 or IL-27 has not been identified as a critical mediator in those studies. Our observation improves the knowledge about the mechanisms leading to increased IFN- production under therapeutic levels of calcineurin inhibitors.
Our results may help to explain why concomitant infections [with increased endogenous IL-12 production and CD28 ligand (CD28L) expression] lead to activation of the underlying disease during treatment with CsA or tacolimus.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell isolation and culture
Peripheral blood mononuclear cells (PBMC) of healthy volunteers were separated by Ficoll-Hypaque density gradient centrifugation and resuspended in Iscove medium (Biochrom, Berlin, Germany) supplemented with 4% human AB serum (IAB). After 1–2 h (37°C, 5% CO2) on petri dishes (Heraeus, Hannover, Germany), the nonadherent cells were removed carefully. Where indicated, CD4+ T cells were isolated from nonadherent cells using a negative selection kit (Miltenyi Biotech, Bergisch Gladbach, Germany); purity of the resulting CD4+ T cells was 
95%, as verified by flow cytometry analysis of CD4+ T cells (mouse anti-human CD4 mAb, Coulter Immunotech, Hamburg, Germany). After separation, cells were resuspended in IAB. Where indicated, cells were labled with carboxyfluorescein succinimidyl ester (CFSE; MoBiTec, Göttingen, Germany) as described previously [25
]. For activated protein-1 (AP-1), electrophoretic mobility shift assay (EMSA) experiments, Jurkat T cells grown in RPMI, supplemented with 10% fetal calf serum (FCS), were used.
Stimulation procedure
Human nonadherent PBMC or purified CD4+ T cells were incubated with stimulating anti-CD3 (1 µg/ml)/anti-CD28 (0.2 µg/ml) mAb (referred to as "TCR stimulation" throughout the paper). As indicated, different cytokines were added along with the TCR stimulation. CsA (ranging from 1 µg/ml to 1 ng/ml) or tacrolimus (from 100 ng/ml to 10 pg/ml) was added.
Efficiency-controlled quantitative real-time fluorescence polymerase chain reaction (PCR)
Total RNA extraction from 1 x 105 CD4+ T cells and cDNA synthesis were performed using the High Pure RNA isolation kit and First Strand cDNA synthesis kit (Roche Molecular Biochemicals). Quantitative real-time fluorescence PCR was performed as described [26
]. FastStart PCR was performed (FastStart DNA Master SYBR Green I, Roche Molecular Biochemicals). For quantification of T-bet (forward 5'gatgtttgtggacgtggtcttg3', reverse 5'ctttccacactgcacccactt3'), IFN- (forward 5'gcatcgttttgggttctcttggctgttactgc3', reverse 5'ctcctttttcgcttccctgttttagctgctgg3'), Gata-3 (forward 5'gacgagaaagagtgcctcaag3', reverse 5'tccagagtgtggttgtggtg3'), c-fos (forward 5'tcaccctgcctctcctcaat3', reverse 5'gctgcatagaaggacccagatag3'), and ERM (forward 5'caatgctgaaacctctcaaagtgg3', reverse 5'ttcctctttctgtcaatcacaggc3'), an efficiency-adjusted, relative quantification was performed. Standard curves were created covering a range of six orders of magnitude by dilution series (dilutions from these standard curves were used as calibrators in each PCR run). These standard curves describing the PCR efficiencies of the target (T-bet, IFN-, Gata-3, c-fos) and the reference gene (ß-actin) allow an efficiency-corrected quantification using the Relative Quantification software (Roche Molecular Biochemicals).
Flow cytometric analysis of intracellular cytokines and membrane molecules
Five days after antigen stimulation, CFSE-labeled lymphocytes were stimulated with phorbol 12-myristate 13-acetate (10 ng/ml) and ionomycin (400 ng/ml; both from Sigma-Aldrich) in the presence of brefeldin A (3 µg/ml) for 2 h. Intracellular staining was performed using the Cytofix/Cytoperm kit (BD Biosciences, San Jose, CA). The allophycocyanin (APC)-conjugated mouse anti-human IFN- mAb and immunoglobulin G1 isotype control mAb (Becton Dickinson, Heidelberg, Germany) were used at final concentrations of 2 µg/ml for intracellular staining. Stained cells were measured by flow cytometry (FACSCalibur) and analyzed using CELLQuestProTM software (BD Biosciences). Expression of surface antigens was assessed using the following phycoerythrin- or fluorescein isothiocyanate-labeled mouse anti-human mAb: IL-18 receptor 
(IL-18R; R&D Systems), IL-12Rß2 (rat anti-human IL-12Rß2 mAb), and IL-12Rß1 (BD PharMingen, San Diego, CA); and CD54, CD40L, and CD25 (Coulter Immunotech). Appropriate isotype controls were used.
EMSAs
Nuclear extracts from isolated CD4+ T cells were prepared using the NE-PER nuclear and cytoplasmatic extraction reagents kit (Pierce, Bonn, Germany). Protein contents were determined using the Bio-Rad protein assay (Bio-Rad, Munich, Germany). EMSAs for T-bet and AP-1 were performed with equal amounts of protein. The binding reaction for the assays contained, in addition to the nuclear extracts, 1 µl Nonidet P-40, glycerol, MgCl2, poly (dI:dC), and 2 µl binding buffer. For detection of T-bet or AP-1-binding activity, double-stranded, biotin-labeled oligonucleotides (for T-bet: 5'-biotin-AAT TTC ACA CCT AGG TGT GAA ATT-3'; for AP-1 consensus oligonucleotide: 5'-biotin-CGC TTG ATG ACT CAG CCG GAA-3') were added to the reaction. For specificity control, nonlabeled oligonucleotides as well as mutant oligonucleotides (T-bet: 5'-biotin-AAT TTC ACG TCT AGG TAC GAA ATT-3'; AP-1 mutant oligonucleotide: 5'-biotin-CGC TTG ATG ACT TGG CCG GAA-3') were used. To resolve the complexes, native polyacrylamide gels were loaded in 0.5x Tris-boric acid-EDTA buffer for 1 h at 100 V. After electrophoresis, complexes were blotted onto a nylon membrane for 30 min at 20 V. With the help of ultraviolet light, the oligonucleotides were cross-linked to the membrane, which was then incubated in blocking buffer with horseradish peroxidase (HRP)-labeled streptavidin, washed, and incubated with luminol (Pierce LightShift® chemiluminescense EMSA kit). Chemiluminescense, as a measure for HRP bound to streptavidin-biotin-oligonucleotide complexes, was detected using the ChemiImagerTM 4400 (Biozym, Hess. Oldendorf, Germany). For the blocking experiments, nuclear extracts were preincubated with a 200x higher concentration of unlabeled oligonucleotides before biotin-labeled ones were added.
Enzyme-linked immunosorbent assay (ELISA)
Supernatants of cultured human PBMC or purified CD4+ T cells were harvested 48 h after addition of the stimuli. IFN- was detected by ELISA (Ready-Set-Go hIFN- ELISA, Biocarta, Hamburg, Germany).
Statistical analysis and densitometry
Comparative data were analyzed using the t-test (data are depicted in columns) or paired t-test (paired data are depicted). Data shown are representative for at least three independent experiments (mean±SEM is depicted). The ChemiImagerTM 4400 (Biozym) with the AlphaEase software Version 5.5. (AlphaInnotech, San Leandro, CA) was used to analyze the band intensities of the EMSA. In the figures, *, P < 0.05; **, P < 0.02; and ***, P < 0.01.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
production of human TCR-stimulated PBMC upon incubation with IL-12 in the ng/ml range within 48 h. In the presence of CsA, IFN- release into the supernatant was markedly inhibited in all concentrations tested (100–1 ng/ml; Fig. 1A
). However, in the presence of IL-12, CsA led to an increase of IFN- production (Fig. 1A)
. Timing of IL-12 addition to the culture system was irrelevant for the up-regulation of IFN- by CsA (24 h before or simultaneously with the TCR stimulation). A significant increase of IFN- (average: 2.5-fold) produced by aCD3/aCD28 and IL-12-treated T cells was consistently detectable in response to CsA in a dose range from 10 to 100 ng/ml. Even a high CsA concentration of 1 µg/ml in the cell culture could increase the IFN- secretion in 
60% of all individuals tested; however, some showed a decrease of IFN- production at this high CsA concentration as compared with the control (mean increase: 64%±11). Time kinetic experiments showed highest IFN- production between 48 and 72 h after stimulation and addition of CsA, whereas 24 h after stimulation, an increased IFN- secretion was not consistently observable. The absolute amount of secreted IFN- showed a mean increase of 270 ng/1 x 105 cells comparing cells stimulated by TCR + IL-12 with those stimulated alike in the presence of CsA 100 ng/ml. Tacrolimus could substitute for CsA, and similar result were obtained for 10 and 1 ng/ml tacrolimus (Fig. 1B)
, reflecting the known, higher potency of tacrolimus as compared with CsA. Of note, in two exemplary experiments, also lower concentrations of tacolimus (100 as well as 10 pg/ml) still resulted in increased IFN- production as compared with controls stimulated with aCD3/aCD28 + IL-12 in the absence of calcineurin inhibitors (not shown). Purified CD4+ T cells showed an even higher production of IFN- in the described setting than nonadherent PBMC (Fig. 1C)
.
 View larger version (15K): [in a new window] |
Figure 1. CsA induces IFN- production in TCR-stimulated human lymphocytes in the presence of IL-12. Nonadherent human PBMC (A, B) were stimulated with aCD3 (1 µg/ml) + aCD28 (0.2 µg/ml)-stimulating mAb. Where indicated, IL-12 (100 ng/ml in A; 10 ng/ml in B and C) was simultaneously given with the TCR trigger. Different concentrations of CsA or tacrolimus (Tac; B) were added along with the TCR stimulus. Supernatants were harvested 48 h after TCR stimulation and analyzed for IFN- by ELISA. Mean ± SEM of 16 independent experiments is shown (A). Comparison of CsA with tacrolimus was performed for six independent experiments. (C) For purified CD4+ T cells, five independent experiments are summarized, and normalization to the IFN- level produced by TCR + IL-12-stimulated lymphocytes was performed for each individual experiment. ns, non-stimulated.
|
production in the presence IL-27 but not upon incubation of T cells with IL-2, IL-18, IL-23, IL-4, or IL-13 in any IL-23 concentration tested (1–500 ng/ml) or in any CsA concentration tested. Upon TCR stimulation and addition of IL-2, IL-4, IL-13, IL-18, or IL-23, a significant reduction of IFN- secretion was observed by addition of CsA in low concentrations (Fig. 2A
). Down-regulation of IFN- by CsA was overcome for each cytokine if IL-12 was present in the culture system (Fig. 2B)
. IL-12, in combination with IL-18, has been shown to act (in a CsA-insensitive manner) synergistically on IFN- production, independent of a TCR trigger [5
]. In our setting, when T cells were stimulated with IL-12 and IL-18 (in different dose ranges from 10 to 500 ng/ml each), in the absence of TCR or costimulation, addition of CsA did not result in IFN- up-regulation but in a marked inhibition of IFN- production (not shown).
 View larger version (26K): [in a new window] |
Figure 2. Effect of CsA on IL-2-, IL-4-, IL-18-, IL-23-, and IL-27-induced IFN- production in the presence of TCR stimulation. IL-18 (50 ng/ml; n=10), IL-2 (10 U/ml; n=8), IL-4 (20 ng/ml; n=8), IL-23 (100 ng/ml; n=5), and IL-27 (10 ng/ml) were submitted to the same experimental setting as described in Figure 1
. Supernatants were harvested after 48 h for analysis by IFN- ELISA. Normalization to the IFN- level produced by TCR + IL-12-stimulated lymphocytes was performed for each individual experiment. Down-regulation of IFN- production by CsA was significant for all cytokines examined (n<0.05, as determined by t-test comparing TCR+cytokine vs. TCR+cytokine+CsA, 100 ng/ml), except for IL-27, which caused a significant increase in IFN- production in the presence of CsA (100 ng/ml). (B) Four independent experiments are depicted, showing that inhibition of IFN- production shown in A (by CsA, 100 ng/ml) is overcome by the presence of IL-12 (10 ng/ml) in the culture medium; no additive effect is seen with IL-27 + IL-12 in this setting. (C) IL-12 (10 ng/ml) and IL-27 (10 ng/ml) were compared in their capacity to induce IFN- in the presence of TCR stimulation and CsA (100 ng/ml), n = 6.
|
production under the influence of CsA. Compared with IL-12, the effect of IL-27 was significantly weaker (Fig. 2C)
and became only evident with higher doses of CsA (100 ng/ml; Fig. 2A
). If combined with IL-12 in the presence of TCR stimulation and CsA, IL-27 showed no significant, additive effect on the amount of IFN- produced (Fig. 2B)
.
CsA-induced IFN- was expressed in proliferating cells
We performed intracellular IFN- staining in CFSE-labeled lymphocytes to determine IFN- expression in the proliferating cells. As depicted in Figure 3
, highest expression of IFN- was found in the proliferating compartment (lower right quadrant) as compared with IFN- expression of nonproliferated cells (upper right quadrant). Induction of IFN- by CsA (10–100 ng/ml) was also evident by intracellular cytokine staining. Proliferation was not inhibited under the described stimulation scheme for CsA doses of 100 ng/ml or below. With higher doses (e.g., 1 µg/ml CsA), proliferation was markedly inhibited.
 View larger version (29K): [in a new window] |
Figure 3. CsA-induced IFN- in TCR + IL-12-stimulated lymphocytes is produced predominantly by proliferating cells. Nonadherent PBMC labeled with the fluorescent tracer dye CFSE were stimulated with aCD3/aCD28 + IL-12 (10 ng/ml) in the presence of CsA (10 or 100 ng/ml). After 5 days, intracellular cytokine staining for IFN- was performed. Lymphocytes were gated. Gates were set to discriminate nondivided (bright green), IFN--producing cells from divided. (A) Histograms with regard to IFN- staining are depicted for the gated proliferating and nonprofilerating cells in an exemplary experiment. For analysis of experiments summarized in B (n=8), quadrants (with regard to IFN- producting) were set according to the isotype control. Geometric mean fluorescence intensity (MFI) in the upper right (IFN- expression of nonproliferating cells) and lower right quadrants (IFN- expression of proliferating cells) was analyzed.
|
production was dose-dependent on IL-12, the strength of the TCR, as well as costimulatory signal-inducing effect of CsA + IL-12 was not observable. Therefore, strength of TCR engagement obviously plays a critical role in our experimental setting (Fig. 4
). For aCD28 antibodies, the threshold was approximately 20 ng/ml (we used 200 ng/ml). Thus, the IFN--inducing effect was also dependent on costimulation (Fig. 4)
. More than 1 ng/ml IL-12 was necessary to observe an IFN--inducing effect of CsA in the presence of a TCR trigger and costimulation (we routinely used 10 ng/ml).
 View larger version (12K): [in a new window] |
Figure 4. CsA-induced IFN- production is critically dependent on triggering of the TCR and CD28 pathway in the presence of IL-12. Human lymphocytes were stimulated with aCD3 mAb ranging from 0.01 to 2 µg/ml in the presence of aCD28 (200 ng/ml), IL-12 (100 ng/ml), and where indicated, CsA (10 ng/ml; left panel). Depicted in the right panel is the titration of stimulating aCD28 antibody (2–1000 ng/ml) in the presence of IL-12, aCD3 (1 µg/ml), and CsA (where indicated). Supernatants were harvested after 48 h and determined by ELISA for the presence of IFN-. Five independent experiments were performed.
|
were significantly down-regulated by CsA 100 ng/ml (Fig. 5
). Measurement was undertaken after 24 as well as 48 h. A significant down-regulation by CsA (1000 ng/ml) was also observable for IL-12Rß2 as determined 48 h after stimulation. For IL-12Rß1 and IL-12Rß2, we also examined the expression level in the absence of the ligand IL-12, but in the presence of aCD3/aCD28 triggering and CsA, up-regulation of these receptors could not be observed (not shown).
 View larger version (23K): [in a new window] |
Figure 5. Cell surface molecule expression was not up-regulated by CsA in combination with IL-12. Human lymphocytes were stimulated with aCD3/aCD28, IL-12, and indicated concentration of CsA. Two days after stimulation, cell surface expression of IL-18R, IL-12Rß2, CD25, and CD54 was analyzed by flow cytometry. Geometric mean values of cells stimulated in the absence of CsA were set to 100% in each individual experiment; corresponding results for CsA-stimulated cells from at least six independent experiments are given.
|
production was not dependent on decreased IL-10 or IL-4 levels production (Fig. 6
).
 View larger version (10K): [in a new window] |
Figure 6. Exogenous IL-10 or IL-4 did not counter-regulate the CsA-induced IFN- in the presence of IL-12 and TCR stimulation. Human lymphocytes were stimulated with aCD3/aCD28, IL-12 (gray circles), and CsA (10 ng/ml) (black circles). Increasing concentration of rhIL-10 or rhIL-4 was added along with the stimulation. Supernatants were harvested after 48 h and analyzed for IFN- release by ELISA. Twelve independent experiments were performed. Differences between the CsA-treated and nontreated samples were significant in all cytokine concentrations tested for IL-10 as well as IL-4, as determined by paired t-test of the original data (P<0.05). In the absence of CsA, incubation of the cells with IL-4, 10 and 100 ng/ml, resulted in a significant down-regulation of IFN- (as determined by paired t-test, P<0.01) in this experimental setting.
|
was up-regulated by addition of CsA to TCR + IL-12-stimulated T cells on the mRNA level with some time delay, resulting in a significant increase 24 h after stimulation, which was not observable 4 h after stimulation (Fig. 7
). Under the described experimental setting, the transcription factor T-bet showed no increased mRNA expression (Fig. 7)
or DNA-binding activity in EMSA experiments (not shown). ERM mRNA, another transcription factor, which has been reported to transduce cytokine signals to the IFN- promoter, was not induced either. In addition, Gata-3 (the "Th2" master transcription factor) was only faintly expressed in lymphocytes stimulated by TCR in the presence of IL-12 and CsA (data not shown).
 View larger version (14K): [in a new window] |
Figure 7. CsA increases IFN- but not T-bet mRNA expression. Purified CD4+ T cells were stimulated with aCD3/aCD28 and CSA (10ng/ml) in the absence or presence of IL-12. After 4 or 24 h, cells were lysed, and IFN- and T-bet mRNA expression was analyzed by quantitative real-time PCR. The T-bet/ß-actin and IFN-/ß-actin mRNA ratios are expressed as the percentage of the ratio in the absence of CsA (100%). This figure summarizes four independent experiments.
|
expression/production and c-fos mRNA expression. Increased IFN- production was only consistently seen from 48 h onward, and IFN- mRNA levels were increased after 24 h but not after 4 h.
 View larger version (17K): [in a new window] |
Figure 8. Binding activity for AP-1 is increased under the influence of CsA. AP-1 protein DNA activity was detected by EMSA in Jurkat T cells, which were stimulated with aCD3/aCD28 and IL-12 (10 ng/ml) in the presence or absence of CsA (100 ng/ml). Cell lysis and nuclear extraction were performed 6 h after stimulation. For the blocking experiments, unlabeled oligonucleotides, containing the same binding site for AP-1 as the labeled ones, were used in a 200-fold higher concentration; mutant oligonucleotides were used in the same concentration as those for AP-1. One representative experiment is depicted (left panel). Densitometry was performed, and four independent experiments are summarized in the middle panel. c-fos mRNA expression was determined by quantitative real-time PCR in purified human CD4+ T cells 4 h after stimulation (right panel).
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
by activated T cells. Requirements for this IFN- induction by CsA are the cooperation of three different pathways: TCR, CD28, and IL-12 (IL-27)-dependent signaling. The critical involvement of IL-12 (IL-27) for an increased IFN- production in the presence of CsA has not been described before. Analyzing the signaling events, we found an increased expression of c-fos mRNA and increased DNA AP-1-binding activity induced with some time delay in the presence of CsA. Thus, our study contributes to the understanding of how the integration of different signal transduction pathways regulates the IFN- production in human lymphocytes. Furthermore, our findings improve the understanding of "CsA resistance," especially in situations of endogenous IL-12 production (e.g., concomitant infection).
Calcineurinantagonists and IFN- production
Data about IFN- expression regulated by CsA seem conflicting. A number of studies describe a down-regulation of IFN- expression by calcineurin inhibitors [18
, 31
32
33
34
35
]. In line with this, we show a marked and consistent down-regulation of IFN- by calcineurin inhibitors when the cell culture was devoid of IL-12/IL-27.
In contrast to that, few studies performed with human cells found increased IFN- expression in vitro [2
, 23
] and in vivo [24
] under the influence of CsA [23
] or tacrolimus [2
]. In our hands, an increased IFN- production in the presence of calcineurin inhibitors is critically dependent on IL-12/IL-27. Different experimental settings have to be taken into consideration, especially different stimulation procedures, cell origin, and media.
In the two studies, which have shown increased IFN- production of human lymphocytes in the presence of calcineurin inhibitors [2
, 23
], 10% FCS has been used. In these studies, the IFN- induction has not been found to be dependent on the presence of IL-12 in the culture. However, presence of a low amount of IL-12 has not been excluded. IL-12 might have been present, especially as the type of media used may have an important impact on endogenous IL-12 production. For example, it has been shown that dendritic cells (DC), differentiated in media containing FCS, produced much higher IL-12 levels upon stimulation as compared with cells cultured in medium containing human serum [36
]. Use of different culture media has also been shown to result in a different surface protein expression and cytokine profile [37
]. It is interesting that it has also been shown that serum components, such as fibronectin, can induce AP-1 activity [38
]. As we have used 4% human serum, this could explain in part the differences seen, next to differences in the stimulation procedure and time-point of IFN- detection.
CsA-induced IFN- production is independent of IL-4 and IL-10
Rafiq et al. [23
] {working with mitomycin-treated, CD80-transfected P815 cells; aCD3 (mAb UCHT1) at 2 µg/ml, CsA at 400 ng/ml; culture medium RPMI+10% iron-supplemented BCS} as well as Dumont et al. [2
] {PBMC (or purified CD3+ T cells), cultured in RPMI 10% FCS supplemented with 2-mercaptoethanol, aCD3, and aCD28 mAb, were used at 400 ng/ml; tacrolimus at 1.2 and 12.5 nM} have shown decreased production of IL-4 and IL-10 in their experimental settings. Rafiq et al. [23
] reasoned this decrease to be responsible for the observed IFN- up-regulation. However, in the presence of IL-12, IL-4 nor IL-10 seems to be critically involved in the up-regulation of IFN- in our hands. We observed CsA-induced IFN- production also in the presence of high exogenous levels of IL-4 or IL-10 (e.g., 100 ng/ml). A relative lack of these cytokines might reinforce the effect in vivo, but their absence seems not responsible for the IFN- production seen. Main producers of IL-10 are monocytes or DC; however, we found even higher IFN- production in purified CD4+ T cells as compared with PBMC. Furthermore, CsA might support IL-12 production in certain situations [39
] and down-regulate the IL-12 "counterplayers" IL-10 and IL-4.
There exists consense in different experimental settings, that IL-4 is sensitive to the effect of calcineurin antagonists. Thus, in all studies, a significant down-regulation of IL-4 expression is observable (e.g., refs. [2
, 18
, 23
]), which is, however, not true for other Th2 cytokines such as IL-13 and IL-5 [2
, 40
]. IL-13/IL-5 production has not been analyzed in this study. However, in the presence of IL-12 (IL-27) and IFN-, IL-13 as well as IL-5 (if produced under these Th1-polarizing conditions) might not account for marked effects on T cells (which do not respond to IL-13), as IL-12 and IFN- probably superimpose their action or counteract IL-13/IL-5 {shown on the level of transcription factors Gata-3 and signal transducer and activator of transctiption 4 (Stat4) by Ouyang et al. [41
]}.
CsA-induced IFN- production is independent of increased IL-12Rß2 expression
van Rietschoten et al. [42
] have described that CsA can up-regulate the high-affinity receptor for IL-12 (IL-12Rß2) on naive T cells on the mRNA level. We could not detect increased surface expression of the IL-12Rß2, as determined by flow cytometry. This may point to post-translational regulation, resulting in the observed difference between mRNA and protein data or to differences in stimulation schemes and cell origins in the study by van Rietschoten et al. [42
] and our study.
CsA "resistance" and costimulation
Of note, in all studies dealing with CsA resistance (including our data), there is consensus about the critical importance of costimulation via CD28. Two important signaling events might be crucial to explain the CD28 dependence: activation of Elk1 and inhibition of glycogen synthase kinase 3 (GSK-3) [43
], which is a NFAT nuclear export kinase; thus, its inhibition results in decreased nuclear export and accumulation of NFAT in the nucleus.
The c-fos promoter contains a composite binding site also recognized by Elk1 [44 ]. The activity of Elk1 is stimulated by all three mitogen-activated protein kinase (MAPK) cascades, which are activated by CD28 (as well as by TCR) triggering. It was shown that c-fos promoter activity correlates temporally with the phosphorylation state of Elk1 [45 ]. Calcineurin has been identified as a physiologically relevant Elk1 phosphatase [46 , 47 ], which means that Elk1 transcriptional activity is regulated negatively by calcineurin. Thus, in the presence of calcineurin inhibitors, Elk1 activation by CD28-induced MAPK activation is enhanced further by loss of the calcineurin-dependent Elk1 "inhibition." As a result, c-fos activity is enhanced. Our data support an important role of Elk1/c-fos activity in integrating pathways activated by calcineurin inhibition in addition to CD28, as we found increased c-fos mRNA expression as well as increased AP-1 DNA-binding activity. Our finding of AP-1 DNA-binding activity under the influence of CsA is in line with data shown by Boulougouris et al. [48 ] in another experimental system, showing that AP-1 responses were stimulated by CD28 in a CsA-resistant manner.
Transcription factors that regulate IFN-
In our hands, calcineurin-induced IFN- production occurs on integration of signaling pathways triggered by IL-12, TCR, and CD28. Two of them, the IL-12 [5
, 7
, 8
, 49
] as well as the CD28/AP-1 [50
, 51
] signaling pathways, are recognized to be insensitive to calcineurin inhibitors. Among the transcription factors activated by induction of these pathways, T-bet and AP-1, respectively, seem central for IFN- induction [15
, 52
]. IL-12 and CD28 in combination have been described to stabilize IFN- mRNA [53
]. IL-12 is of central importance for the polarization of T cells along the Th1 lineage. IL-12-dependent activation of T-bet [3
] might play an important role in opening the IFN- locus [54
55
56
] to allow access of transcription factors to the promoter/enhancer region. In T cells, IL-27 signaling induces phosphorylation of Stat1 and Stat3 but also of Stat4 and Stat5 [27
, 28
]. Stat1 signaling has been linked to the induction of T-bet and IL-12Rß2 expression [30
].
T-bet was not activated by CsA in our system at the time-points measured. Although CsA seems not to up-regulate T-bet DNA-binding activity, this does not exclude an important role of T-bet for increased IFN- expression, e.g., by accessibility of the IFN- locus for other transcription factors.
Yang et al. [5
] distinguished in mouse CD4+ T cells between a CsA-sensitive and a non-CsA-sensitive IFN--induction pathway, the latter being induced by cooperation of the cytokines IL-12 and IL-18 (involvement of ERM and NF-B). In our study, we failed to detect increased mRNA expression of ERM. Of note, ERM has also been described to play an important role in ICAM-1 expression/up-regulation [57
]. Thus, the observed, decreased cell surface expression of ICAM-1 (Fig. 5)
supports the notion that ERM might not be critically involved in human IFN- induction in our experimental setting.
NFAT translocation at therapeutic CsA levels
TCR stimulation results in activation of the calcium/calcineurin pathway mounting in nuclear translocation and activation of NFAT. It has been demonstrated that even in the presence of therapeutic levels of CsA, some NFAT is still translocated to the nucleus. Batiuk et al. [58
] showed that the CsA effect on calcineurin activity is unlikely to achieve more than 50% inhibition in the in vivo situation. Of note, patients treated for severe atopic dermatitis or psoriasis usually receive much lower CsA doses as compared with transplant recipients. Batiuk et al. [58
] suggested that CsA is much less available in vivo in body fluids than it is for isolated cells in vitro.
Hypothesis
Based on the findings and data described above, we draw up the following hypothesis. For IFN- gene transcription, Fos seems to play an important role in the NFAT1:AP-1 (Fos/Jun)-DNA complex, once the locus is opened (e.g., via IL-12/IL-27
T-bet), and counter-regulatory Th2 pathways (e.g., Gata-3) are blocked. Relatively low amounts of NFAT in the presence of excess AP-1 lead to increased IFN- mRNA accumulation (also as an effect of increased stability [53
, 59
]) and consequently, to increased protein secretion. In this situation, AP-1 activity results from activation of the (CsA-insensitive) CD28 pathway in addition to disinhibition of Elk1 as a result of reduced calcineurin activity. CD28 signaling also results in inhibition of GSK-3 [43
], resulting in decreased nuclear export and consequently, accumulation of NFAT in the nucleus. As remodeling of the IFN- locus, NFAT accumulation, as a result of decreased nuclear export, and increased IFN- mRNA stability need some time to establish, this may explain the time delay of the increased IFN- production observed.
Conclusion
Our findings point to the fact that CsA therapy could be detrimental for the patient in a situation of increased endogenous IL-12 expression. In a situation of concomitant infection, professional APCs are activated to produce a high amount of IL-12 and up-regulate their costimulatory capacitiy (CD28L), which will, upon interaction with T cells, result in the scenario described here. A CsA-dependent increase in IFN- production thus explains why concomitant infections quite often worsen the underlying disease (e.g., autoimmune diseases, GvHD, psoriasis).
Received June 18, 2005; revised February 15, 2006; accepted March 6, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
production in CD4 and CD8 T cells Science 295,338-342 production in Th1 CD4+ T cells: evidence for two distinct pathways for promoter activation Eur. J. Immunol. 29,548-555[CrossRef][Medline] production by natural killer cell stimulatory factor: characterization of the responder cells and synergy with other inducers J. Exp. Med. 173,869-879B and NFAT with the interferon- promoter J. Biol. Chem. 272,30412-30420 gene expression by nuclear factor of activated T cells Blood 98,1480-1488 promoter that is a potential NFAT binding site Transplantation 61,933-939[CrossRef][Medline] (IFN-) locus revealed by genome sequence comparison J. Biol. Chem. 279,4802-4810 production by T cells when co-stimulated through CD28 Eur. J. Immunol. 28,1481-1491[CrossRef][Medline] by cyclosporin A in atopic patients: a discrepancy between in vitro and in vivo events Acta Derm. Venereol. 74,20-21[Medline] interferon production in the presence of interleukin Transplantation 45,967-972[CrossRef][Medline] production by downregulation of interleukin-18 in human mixed lymphocyte reactions Acta Med. Okayama 54,201-209[Medline] production J. Immunol. 159,3247-3256[Abstract] promoter Int. Immunol. 15,1149-1160 J. Immunol. 162,5894-5901 mRNA in a cytolytic T cell clone: Ca(2+)-induced transcription followed by mRNA stabilization through activation of protein kinase C or increase in cAMP Eur. J. Immunol. 25,889-895[Medline]This article has been cited by other articles:
 |
|
 |
|
  H. Wang, B. Grzywacz, D. Sukovich, V. McCullar, Q. Cao, A. B. Lee, B. R. Blazar, D. N. Cornfield, J. S. Miller, and M. R. Verneris The unexpected effect of cyclosporin A on CD56+CD16 and CD56+CD16+ natural killer cell subpopulations Blood, September 1, 2007; 110(5): 1530 - 1539. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
