| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Medicine, Rheumatic and Autoimmune Diseases Section, and Center for Immunology, University of Minnesota Medical School, Minneapolis, MN 55455; and
Department of Molecular Pathology and Medicine, Cancer Immunotherapy and Gene Therapy Program, DIBIT/H.S. Raffaele, Milan, Italy
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and 
-chain phosphorylation, 2)
the hydrolysis of phosphatidylinositol 4,5-bisphosphate and production
of free inositol phosphates and diacylglycerol, 3) the mobilization of
calcium free ions, or 4) the dephosphorylation and nuclear
translocation of NFAT (2, 3, 4). Thus, intense TCR/CD3
stimulation in Th1 cells may alleviate any requirement for B7/CD28
costimulation in the TCR-dependent activation of select proximal and
distal signaling pathways. Nevertheless, IL-2 gene
expression (both steady-state mRNA levels and protein synthesis) as
well as AP-1-dependent transactivation in Th1 cells have been found to
be entirely suboptimal in the absence of costimulatory signals. Like the IL-2 gene itself, the activation of c-Jun N-terminal kinase (JNK)3 (but not extracellular signal-regulated kinase, ERK) in murine Th1 cells was found to be greatly augmented by Ab-mediated costimulation through the CD28 molecule (4). This finding is consistent with data originally obtained using the Jurkat T-leukemia cell line, where anti-CD28 mAb was shown to augment CD3-induced AP-1 transactivation through JNK-dependent phosphorylation of the c-Jun transactivation domain (5, 6, 7, 8, 9). Serine-phosphorylation of c-Jun on residues 63 and 73 by JNK has previously been shown to increase its capacity to transactivate at enhancer DNA (10). However, our result was somewhat paradoxical in that Th1 cells do not express detectable c-Jun protein early during activation when JNK activity is mostly up-regulated by CD3 and CD28 signals, and the AP-1 complexes isolated from Th1 cells do not contain c-Jun at a time when peak IL-2 transcription is observed (3, 11). Therefore, experiments shown here have examined the mechanism by which B7/CD28-mediated costimulatory signals and Ag-receptor occupancy integrate to regulate the expression and function of AP-1 proteins important to IL-2 gene transactivation in normal Th1 cells. Our results suggest that B7 costimulatory ligands on APC enhance JNK protein kinase activity leading to more vigorous Elk-1 transactivation at the c-fos gene and the increased expression of AP-1 proteins during Ag stimulation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The AE.7 and 16B.2 pigeon cytochrome c (PCC)-specific and I-Ek-restricted murine Th1 cells used in this study, and their cultivation, have been previously described (12, 13). The DO11 chicken OVA-specific and I-Ad-restricted murine Th1 cells were derived from CD4+ lymph node cells of DO11.10 TCR-transgenic mice (14) and maintained as short-term lines by intermittent stimulation with Ag and 3000 rad-irradiated BALB/c spleen cells. These DO11 T cells were used after 2–5 rounds of in vitro expansion with Ag and IL-2, and were rested 14 days before each experiment. The human T-leukemia cell line Jurkat (American Type Culture Collection (ATCC), Manassas, VA) was also examined in some experiments, for comparison with the normal Th1 cells. For Ab stimulations, Th1 cells (2–10 x 106 cells) were incubated with polystyrene-immobilized anti-CD3 (mAb 145-2C11) (15) and/or anti-CD28 (mAb 37.51) (16) as previously described (4). In some experiments, a goat anti-hamster IgG was used as a first-step reagent followed by immobilization of the anti-CD3 and/or -CD28 mAbs, with similar results. The calcineurin inhibitor cyclosporin A (CSA), the mitogen-activated protein kinase (MAPK)/ERK kinase 1 (MEK1) inhibitor PD 098059, and the stress-activated protein kinase (SAPK) inhibitor SB 202190 were obtained from Calbiochem (La Jolla, CA). CSA was dissolved in ethanol (final concentration 0.1%), whereas PD 098059 and SB 202190 were dissolved in DMSO (final concentration 0.1%); vehicle alone was used as the control. T cells were always pretreated with inhibitors for 1 h before activation and they remained present throughout the stimulation.
Ag-presenting cells
Animal use has been approved by the University of Minnesota
Institutional Animal Use and Care Committee and is in accordance with
National Institutes of Health guidelines. For the production of APC,
spleens were harvested from B10.A or BALB/c mice. Leukocytes were
incubated in culture dishes in complete medium for 2–3 h at 37°C
after which nonadherent cells were removed by vigorous rinsing with
PBS. The adherent spleen cells were treated with 10 µg/ml LPS (Sigma,
St. Louis, MO) at 37°C for 16 h in complete medium and then
removed from the plate using a cell scraper. Accessory molecule
expression on recovered, activated spleen cells was determined by
FACScan flow cytometry using CellQuest software (BD Biosciences,
Mountain View, CA). For Ag stimulations, LPS-activated splenic APC were
exposed to the Ag (either OVAp323–339 or
PCCp81–104) at 37°C for 2 h before mixing
with the responding Th1 cells (3–5 x 106)
at an APC-T cell ratio ranging from 1:2 to 1:25. To neutralize
potential costimulatory molecules on the APC, hybridomas producing
hamster IgG anti-murine B7-1 mAb 16-10A1 (provided by H. Reiser,
Boehringer Ingelheim Pharmaceutics, Ridgefield, CT)
(17), rat IgG2a anti-murine B7-2 mAb GL1 (ATCC)
(18), and rat IgG2a anti-murine ICAM-1 (ATCC)
(19) were obtained and their mAb purified using standard
protein G affinity chromatography techniques. Purified mAbs were used
at a final concentration of 5–20 µg/ml as indicated. The combination
of anti-B7-1 and -B7-2 mAbs was routinely capable of inhibiting
IL-2 production (assay as previously described in Ref. 20)
by >90% in Th1 cells responding to Ag and these APCs (see Fig. 1
A). Hamster IgG and rat IgG2a served as irrelevant Ab
controls (Pierce, Rockford, IL).
|
The pNFATZH, pSXNeo/AP-1, and p15
CX/IL-2 plasmids were
provided by G.R. Crabtree (Stanford University, Palo Alto, CA). pNFATZH
contains three tandem repeats of the human IL-2 NFAT binding
sequence (-286 to -257 relative to the transcriptional start site of
the IL-2 gene) within the XhoI site of a minimal
IL-2 promoter (bp -319 to +47 of the IL-2 gene
containing an internal deletion and replacement of bp -296 to -72 by
XhoI linker DNA) fused to the lacZ gene. Plasmid
pAP-1ZH contains five tandem repeats of the sheep metallothionine
MT-II gene AP-1 site (TGACTCA) within the minimal
IL-2 promoter and was constructed as previously described
(3). Plasmid pIL-2ZH was constructed similarly by
inserting the entire human 5' IL-2 enhancer DNA sequence
(-326 to +47) contained within the XmnI-HindIII
fragment of p15CX/IL-2 into a pNFATZH vector that had its
corresponding XmnI-HindIII fragment deleted. The
resulting pIL-2ZH plasmid was identical with pNFATZH and pAP-1ZH except
that the intact 5' IL-2 gene sequence replaced the
3xNFAT/IL-2 minimal promoter or 5xAP-1/IL-2 minimal promoter sequences,
respectively. The generation of the NFATZH and AP-1ZH stable
transfectants of cloned 16B.2 Th1 cells was previously described
(3), and these transfectants have been maintained as pools
of cells to eliminate any effects of site of integration. IL-2ZH 16B.2
cells (stably transfected with pIL-2ZH) were produced and maintained by
the same method. Lysis of activated cells and assay for
-galactosidase activity was performed as previously described
(3).
EMSAs of DNA-binding complexes
EMSA was performed on nuclear extract proteins as previously
described (3) using the following T4 kinase-mediated
[-32P]ATP-labeled double-stranded
oligonucleotides: mouse NFAT,
5'-TCGAAAGAGGAAAATTTGTTTCATACAGAAGGCG-3'; mouse AP-1,
5'-TCGAGAAATTCCAGAGAGTCATCAGAAGA-3'.
Western blot detection of Fos and Jun proteins
c-Fos, FosB, and JunB protein expression were determined by
Western blot as previously described (3). T cells
(1–1.5 x 106) were solubilized in ice-cold
lysis buffer containing 20 mM Tris-HCl (pH 7.4), 1% Triton X-100, 10%
glycerol, 137 mM NaCl, 2 mM EDTA, 25 mM -glycero- phosphate, 1 mM
NaVO4, 2 mM sodium pyrophosphate, 1 mM PMSF, and
10 µg/ml leupeptin for 30 min. Then insoluble material was removed by
high-speed microcentrifugation in the cold, and samples were diluted
2:1 in 3x Laemmli sample buffer. Following separation by SDS-PAGE,
proteins were transferred to nitrocellulose. Equal protein loading was
confirmed at that time by staining of membranes with Ponceau S (data
not shown). Membranes were then probed with either a pan anti-Fos
Ab (peptide 128–152) or anti-JunB Ab (peptide 45–61) obtained
from Santa Cruz Biotechnology (Santa Cruz, CA). Detection was performed
using a goat anti-rabbit Ig-HRP conjugate (Bio-Rad, Hercules, CA)
and ECL chemiluminescence reagents (Amersham, Arlington Heights,
IL).
In vitro kinase assays and ex vivo detection of phosphoproteins
The solid-phase JNK protein kinase assay is according to Hibi et al. (21) and was previously described (4). In this assay, JNK is first immobilized on a GST-c-Jun fusion protein bound to glutathione Sepharose 4B beads (Amersham Pharmacia Biotech, Piscataway, NJ), and then an in vitro kinase reaction is conducted using the GST-c-Jun as substrate. The ERK immune complex kinase assay was also performed according to the method of Hibi et al. (21) and was conducted as previously described using anti-ERK2 polyclonal antisera (Santa Cruz Biotechnology) and myelin basic protein (Upstate Biotechnology, Lake Placid, NY) as substrate (4). Alternatively, ERK activity was assayed by solid-phase in vitro kinase reaction using GST-Elk-1 as both the ERK immobilizer and substrate. This GST-Elk-1 fusion protein contains aa 307–428 based on the human Elk-1 complementary DNA (plasmid encoding GST-Elk-1 provided by R.A. Hipskind, Institut de Genetique Moleculaire de Montpellier, Montpellier, France) and was expressed in Escherichia coli strain HB101. We have found that this GST-Elk-1 fusion protein binds only activated (dual phosphorylated) ERK, and does not bind either resting or activated JNK (data not shown). Incorporation of 32P into all substrates was found to be linear with time over the course of the in vitro kinase reactions and sensitive to differences in enzyme activity, as established by SDS-PAGE and autoradiography. Dual-specific phosphorylation of p38 indicative of its activation was detected using phospho-p38 MAPK (Thr180/Tyr182) Ab obtained from New England Biolabs (Beverly, MA) to probe a Western blot. The statistical comparisons indicated relied on Student’s paired t test.
Construction of expression plasmids and reporter genes for transient transfection
Plasmid pEF-Gal4 was prepared by first excising a 1.7-kb
PvuII-SalI fragment containing the EF-1
promoter from the mammalian expression vector pEF (provided by G.
Koretsky, University of Iowa, Iowa City, IA) and then inserting it into
the polylinker of the pSP72 shuttle vector (Promega, Madison, WI). A
1.7-kb PvuII-BamHI fragment was then excised from
this EF-1 shuttle vector and used to replace a 0.4-kb
PvuII-BglII fragment containing the SV40 promoter
of the mammalian expression vector pBXG1 (provided by L. M. Staudt,
National Institutes of Health, Bethesda, MD) to create the plasmid
pEF-Gal4. This pEF-Gal4 vector contains an EF-1 promoter followed by
a cDNA encoding the Gal4 DNA-binding domain with a downstream
polylinker and poly(A) tail. To create pEF-Gal4Elk-1, a fragment
encoding the Elk-1 transactivation domain (aa 307–428) was generated
by PCR using a plasmid containing the human Elk-1 complementary DNA
(ATCC). The forward primer sequence (A)
5'-TAGGATCCATCTCCCAGCGCCAGAAG-3' contained a BamHI
restriction site, and the reverse primer (B)
5'-TGAAGATCTTGGCTTCTGGGGCCCTGG-3' contained a BglII site
to facilitate cloning. Following amplification, the Elk-1 PCR product
was cut with BamHI and BglII and inserted into
the BamHI site of pEF-Gal4, maintaining the Elk-1 sequences
in-frame with Gal4. To make plasmid pEF-Gal4-Elk-1(m) encoding a mutant
Gal4-Elk-1 fusion protein with alanine substitutions at serines 383 and
389, PCR mutagenesis was performed according to Xu et al.
(22). First, a mutated PCR fragment encoding aa 383–428
was generated from the Elk-1 complementary DNA using a
5'-phosphorylated forward primer
5'-GCTCCCATTGCGCCCCGTGCTC-3' (where the
alanine substitutions at residues 383 and 389 are indicated in bold)
and reverse primer (B). Second, a fragment encoding Elk-1 residues
307–382 was generated using forward primer (A) and a 5'-phosphorylated
reverse primer 5'-CAGGGTGCTCCAGAAGTG-3'. These two PCR products
were then directly ligated, restricted with BamHI and
BglII, and then cloned into the BamHI site of
pEF-Gal4 as before.
To make the 5xGal4luc reporter gene called p5GT109luc, a 
200-bp
BamHI-SalI fragment from pG5EC (provided by L. M.
Staudt), containing five tandem repeats of the Gal4 DNA binding
sequence and linker DNA, was cloned upstream of the thymidine kinase
promoter at identical restriction sites within the pT109luc plasmid
(ATCC). Unlike pEF-Gal4-Elk-1, cotransfection of pEF-Gal4-Elk-1(m) with
p5GT109luc did not result in luciferase reporter gene induction,
demonstrating the requirement for serines 383 and 389 for Gal4-Elk-1
transactivation (data not shown). pT109luc itself served as a control
luciferase reporter gene in these experiments, and it had no inducible
activity when cotransfected with pEF-Gal4-Elk-1 (data not shown). A
pIL-2luc reporter gene was also prepared by amplifying an
IL-2 gene fragment from pIL-2ZH (bp -314 to +40 relative to
the transcriptional start site) using a forward primer
5'-CGCAAGCTTCTTTTGTATCCCCACC-3' designed to contain a
HindIII site and a reverse primer
5'-GCGCAGATCTCAGGAGTTGAGGTTACT-3' containing a BglII site.
This PCR product was cut with HindIII-BglII and
then cloned into a pT81luc (ATCC) plasmid whose
HindIII-BglII fragment (containing the thymidine
kinase promoter) had been removed. The resulting pIL-2luc thus contains
the 5' IL-2 gene enhancer/promoter and a downstream
luciferase complementary DNA.
Transient cotransfection reporter system
Transient cotransfection of the Th1 cells was carried out by electroporation using a Genepulser (Bio-Rad) set to 960 µF and 340 V. Before transfection, Th1 cells (up to 8 x 107 cells in 0.4 ml of OPTI-MEM I medium per cuvette) were incubated at room temperature for 10 min with plasmids: 5 µg pGreen Lantern-1 (Life Technologies, Gaithersburg, MD); 10 µg of control luciferase reporter gene pT109luc, p5GT109luc, or pIL-2luc; 30 µg of control vector pEF-Gal4, pEF-Gal4-Elk-1, or pEF-Gal4-Elk-1(m). Cells were rested for 15 min at room temperature in the cuvettes after electroporation, and then incubated in complete medium for 2 days before stimulation. Luciferase activity was assayed in cell lysates using Reporter Lysis Buffer and Luciferase Assay System (Promega) according to the manufacturer’s instructions. Relative light units shown have been corrected for protein concentration in each lysate. The transfection efficiency for each electroporation sample was determined using the FACScan (FL1 detector) to detect pGreen Lantern-1 (green fluorescent protein) expression.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Using immobilized anti-CD3 mAb as a mimic of TCR/CD3 ligation in the absence of either Ag or APC, we had previously determined that normal Th1 cells require CD28 costimulation to maximally induce the transcription of a stably transfected AP-1 reporter gene (4), thus predicting a capacity for the CD28 molecule to regulate AP-1-dependent transactivation at a number of cis-acting DNA elements within the 5' IL-2 gene enhancer. Consistent with this, further experimentation with stably transfected Th1 cells confirmed that in addition to AP-1 elements, CD3 and CD28 mAbs also synergistically induce transactivation at NFAT DNA response elements, as well as at the 5' IL-2 gene enhancer itself (data not shown).
Based on these results, we investigated whether physiological
interactions between B7 costimulatory molecules and CD28 play an
important role in the up-regulation of AP-1-dependent transactivation
in Th1 cells responding to Ag and APC. Adherent spleen cells activated
with LPS served as the APC in these experiments. Use of such a
heterogeneous population of cells (containing activated dendritic
cells, macrophages, and some B cells) was intentional, and ensured the
presence of multiple potential costimulatory ligands including B7-1
(CD80), B7-2 (CD86), and ICAM-1 (CD54) on the APC (data not shown).
These APC were found to be potent stimulators of IL-2 synthesis in Th1
cells when Ag was present, and their expression of B7 molecules proved
essential to this response because the combination of anti-B7-1 and
B7-2 mAbs prevented the secretion of IL-2 (Fig. 1
A). Using these Ag-loaded APC as stimulator cells, the pair
of anti-B7 mAbs was observed to significantly reduce the activation
of the AP-1 and NFAT reporter genes in Th1 cells (Fig. 1B).
Transactivation of the 5' IL-2 gene enhancer reporter was
also significantly inhibited by the addition of anti-B7 mAbs, and
this effect appeared specific because the addition of an irrelevant IgG
failed to affect the response (Fig. 1, B and C).
Interestingly, an anti-ICAM-1 mAb demonstrated significantly less
inhibitory capacity than the B7 mAb pair. Thus, Ag- and APC-stimulated
5' IL-2 gene enhancer activation in Th1 cells was found to
be highly dependent on a synergy between the TCR and a unique set of
signals arising from B7-ligated CD28 molecules. Because the activation
of each of these three reporter genes can be expected to rely on the
binding of an AP-1 transcription factor complex to
cis-acting DNA elements, AP-1 proteins appeared to be the
targets of these synergistic signals in Th1 cells.
Induction of AP-1 complex formation during Ag presentation to Th1 cells depends on specific signals arising from the B7/CD28 interaction
The requirement for B7/CD28 interactions for the up-regulation of AP-1-dependent transcription during Ag stimulation is seemingly consistent with previous demonstrations of B7 and anti-CD28 mAb costimulation of JNK activity in T cells, and with a presumed role for JNK in the regulation of transactivation by Jun-containing AP-1 complexes (4, 5, 23, 24). However, at closer inspection these observations are more difficult to reconcile, as only a minority of AP-1 complexes in Th1 cells contain a c-Jun or JunD protein capable of regulation by direct JNK-dependent phosphorylation (3). Therefore, we examined whether physiological interactions between B7 costimulatory molecules and CD28 are important in the up-regulation of AP-1 complex formation in Th1 cells responding to Ag and APC. Using Ag-loaded APC to stimulate the activation of Th1 cells, T cell nuclear extracts were examined for the presence of AP-1 transcription factors using NFAT and AP-1 double-stranded oligonucleotide probes and an EMSA system. We and others have previously determined that the NFAT DNA-binding activity in Th1 cells consists of a trimeric protein complex containing both NFATp as well as a Fos and Jun (AP-1) heterodimer (3, 25, 26). We have already demonstrated in Th1 cells that CD28-mediated costimulatory signals are not required for the induction of increased intracellular calcium free ion (Ca2+) concentrations or for the dephosphorylation of NFATp that leads to its nuclear translocation (3); therefore, any requirement for B7 costimulatory signals during the Ag-stimulation of NFAT DNA-binding activity could be expected to be at the level of AP-1 complex assembly or its association with NFAT protein.
Th1 cells stimulated with Ag and APC up-regulated both NFAT and AP-1
DNA-binding activities (Fig. 2). Similar to the IL-2 production and transactivation experiments
described above, the addition of a combination of neutralizing
anti-B7-1 and -B7-2 mAbs significantly inhibited the induction of
NFAT and AP-1 complexes. This was a specific effect of B7 molecule
neutralization during the incubation, as Ab against the ICAM-1 adhesion
molecule demonstrated considerably less inhibitory activity (Fig. 2A, and data not shown). Consistent with these results, we
also observed that immobilized anti-CD28 mAb synergized with
anti-CD3 in the induction of NFAT and AP-1 DNA-binding activities
in Th1 cells (data not shown). Thus, B7/CD28 interactions regulate AP-1
complex formation during Ag stimulation of Th1 cells.
|
Of the various Fos and Jun family members biochemically identified
within AP-1 complexes of Th1 cells during the induction of
IL-2 gene expression, only JunD protein exists at detectable
levels in resting cells; other members require T cell activation for
their expression (3). Therefore, we hypothesized a role
for B7- and CD28-mediated costimulatory signals in the regulation of
fos and/or junB gene expression. Stimulation of
Th1 cells with the combination of Ag and APC led to increased
expression of c-Fos and FosB protein (Fig. 3A). Furthermore, the addition of the anti-B7 mAbs during
the Ag stimulation resulted in a significant inhibition of this
response (42 ± 4% inhibition for c-Fos protein,
n = 4). Thus, this Western blot analysis was consistent
with the AP-1 and NFAT EMSA results, and suggested that the poor
association of AP-1 complexes with DNA observed following the addition
of neutralizing anti-B7 mAbs was the result of deficient Fos
protein expression. Also consistent with such a model, a combination of
immobilized anti-CD3 and -CD28 mAbs was found necessary to induce
the expression of c-Fos and FosB proteins in Th1 cells in the absence
of APC; stimulation with either anti-CD3 or anti-CD28 mAb alone
was insufficient (Fig. 3B). Even at saturating anti-CD3
mAb concentrations (
3 µg/ml mAb immobilized), CD3 stimulation
failed to induce more than a minimal c-Fos or FosB induction in the
absence of CD28 signaling. Finally, Western blots reprobed with Ab
against JunB demonstrated a similar requirement for CD28 costimulation
to induce the expression of this transcription factor following CD3
cross-linking (data not shown).
|
This requirement for B7/CD28 costimulation in the up-regulation of AP-1 protein expression following exposure of Th1 cells to Ag and APC was reminiscent of the CD28 dependence of JNK activation previously demonstrated by us and others (4, 5). We also observed that the calcineurin inhibitor CSA could blunt the induction of Fos protein in Th1 cells stimulated with either Ag plus APC or with the combination of anti-CD3 and -CD28 mAbs (data not shown), consistent with a role for a SAPK such as JNK or p38 in the regulation of fos gene regulation in these cells (7, 8, 27). Therefore, we hypothesized that a CD28-dependent MAPK mediates the costimulatory effects of B7/CD28 interactions on AP-1 protein synthesis in Th1 cells.
To address this issue, we examined whether the physiological activation
of ERK, JNK, or p38 could depend on the ligation of CD28 by B7
molecules. Whole cell extracts from Th1 cells stimulated with Ag and
APC either in the presence or absence of anti-B7 mAbs were assayed
for ERK and JNK protein kinase activities, and for activating
threonine-180 and tyrosine-182 phosphorylations of p38. Stimulation of
helper T cells with Ag rapidly induced both JNK and ERK protein kinase
activities, and p38 became phosphorylated consistent with its
activation. Interestingly, the addition of anti-B7 mAbs did
significantly limit the activation of JNK (58 ± 19% inhibition,
p 
0.01, n = 7) (Fig. 4). In contrast, ERK activation was little affected by the addition of
anti-B7 mAbs (5 ± 14% inhibition, p 0.89,
n = 6). Similarly, Ag- plus APC-induced phosphorylation
and activation of p38 demonstrated no significant inhibition following
the addition of the anti-B7 mAbs (9 ± 26% inhibition,
p 0.96, n = 5). The ERK and p38
activation events were also independent of ICAM-1/LFA-1 interactions,
given the lack of inhibition observed following the addition of
anti-ICAM-1 together with the anti-B7 mAbs. These data implied
that neither ERK nor p38 could mediate the unique costimulatory effects
of B7 molecules binding CD28 receptors that occurs during the response
to Ag-bearing APC. In contrast, the activation of JNK correlated well
with the B7-dependent expression of Fos protein.
|
This direct correlation observed between JNK activation and the
induction of AP-1 protein expression during Ag and APC stimulation in
Th1 cells suggested that JNK may play an important role as a downstream
mediator of B7/CD28 signals to the fos and junB
genes. Consistent with this notion, CSA proved incapable of inhibiting
the activation of either ERK or p38 during the response to Ag and APC,
whereas JNK activation (like Fos expression) was sensitive to the
effects of CSA (data not shown). c-fos gene transcriptional
regulation has been well studied in a number of cell types, and is
thought to depend mainly on the phosphorylation of the c-fos
gene regulator Elk-1 by a MAPK such as ERK (28). It has
generally been assumed that the activation of ERK in response to TCR
ligation would be sufficient to maximally induce Elk-1-dependent
c-fos gene expression in T cells (29). Less is
known about the regulation of FosB or JunB expression; however,
Ets-like transcription factors with sequence homology to Elk-1 and
similar regulation by MAPK such as ERK have been implicated in the
control of fosB and junB gene transcription
(30, 31, 32). Perhaps consistent with these published data, we
found that Ag-stimulated Fos protein expression in Th1 cells could not
be induced under conditions where ERK activation was blocked with the
MEK1 inhibitor PD 098059 (Fig. 3C). But based on this simple
paradigm, one should have expected little costimulatory signal
dependence of AP-1 protein induction and no sensitivity to CSA. Our
results instead suggested a more complicated regulation of c-fos,
fosB, and junB gene transcription that integrates
signals from both the TCR and CD28.
The Elk-1 transcription factor is known to be a target of
phosphorylation by each of the MAPK (33, 34, 35, 36, 37), and we found
using an in-gel kinase reaction that the transactivation domain of
Elk-1 can indeed be phosphorylated by ERK, JNK, and p38 extracted from
activated Th1 cells (W. Li, unpublished observation). Based on this, we
undertook an investigation of the hypothesis that the Elk-1
transcription factor mediates the effect of B7 costimulation on
c-fos gene expression in Th1 cells. An Elk-1-dependent
transcription assay system consisting of a recombinant Gal4-Elk-1
fusion protein containing a Gal4 DNA-binding domain together with the
Elk-1 transactivation domain and a 5xGal4luc reporter gene containing
five tandem repeats of the Gal4 DNA-binding sequence was constructed
for these experiments. Transient coexpression of Gal4-Elk-1 together
with the 5xGal4luc reporter gene led to the accumulation of luciferase
activity when the transfected Th1 cells were activated with a
combination of CD3 plus CD28 mAbs, or with Ag plus APC (Fig. 5). Like the induction of the c-Fos protein, Elk-1 transactivation was
observed to be suboptimal in response to CD3 ligation alone; however,
the addition of immobilized anti-CD28 mAb led to an improved
response. Finally, Ag-induced Gal4-Elk-1 transactivation was found to
be significantly reduced when B7 molecules on the APC were neutralized
with Abs, consistent with the requirement for CD28 ligation by B7 that
leads to increased JNK signaling.
|
Similar to its inhibitory effects on Fos protein expression and on
5' IL-2 gene enhancer/promoter activity itself (Figs. 3C and 5C), blockade of MEK1 protein kinase
activity in the Th1 cells with PD 098059 effectively prevented
transactivation by Elk-1 in this reporter gene system (Fig. 5C). Despite this apparent requirement for ERK activation in
the elicitation of Elk-1-dependent c-fos gene transcription,
the intense activation of ERK that accompanies stimulation with CD3 mAb
alone or Ag stimulation in the presence of anti-B7 mAbs was clearly
insufficient to maximally induce Elk-1 activity (Figs. 4A
and 5, A and B; data not shown). The activation
of calcineurin also appeared necessary to stimulate maximal Elk-1
transactivation, based on its sensitivity to inhibition with CSA (data
not shown). Because CSA has no capacity to inhibit the activation of
ERK in Th1 cells (data not shown), pathways downstream of CD28 and
calcineurin and independent of ERK activation also appeared necessary
to fully activate this c-fos gene transcriptional
regulator.
The induction of Elk-1 transactivation and c-fos gene transcription in Th1 cells depends on the protein kinase activity of JNK
Based on the similarities observed between JNK activation and
Elk-1 transcriptional regulation in Th1 cells, we hypothesized that JNK
protein kinase activity mediates the B7- and CD28-dependent induction
of Elk-1 transactivation in cooperation with TCR-induced ERK activity.
To directly address these issues, we took advantage of the SAPK
inhibitor SB 202190 (38). Using SB 202190 at a low dose
capable of interrupting the kinase activity of only p38 (4 µM) (data
not shown), we found that the drug only ineffectively limits the
up-regulation of c-Fos and FosB protein expression (Fig. 3C). In contrast, at a higher dose (40 µM) capable of
interfering with the kinase activity of both p38 and JNK (data not
shown), SB 202190 proved to be a potent inhibitor of both c-Fos and
FosB protein expression (Fig. 3C). Consistent with such a
role for JNK in the transcriptional regulation of c-fos gene
expression, Elk-1 function and IL-2 gene transactivation
were similarly blocked in the presence of SB 202190 at concentrations
approaching 40 µM (Fig. 5D). SB 202190 (40 µM) appeared
to be neither toxic to Th1 cells nor nonspecific in its actions, as
this dose of drug did not interfere with the activation of ERK (data
not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There is currently considerable controversy surrounding the role of JNK in the regulation of IL-2 secretion and T cell proliferation. In naive CD4+ T cells, JNK is unlikely to mediate such a B7 costimulatory effect on AP-1-dependent IL-2 gene regulation. JNK levels are relatively low in these cells (39). Furthermore, our own investigation of JNK activation and c-Jun phosphorylation in naive CD4+ T cells stimulated with Ag in vivo indicates that JNK activation can occur even in the absence of CD28 signaling.4 A careful analysis of fresh splenic T cells in JNK2-deficient mice did in fact reveal a requirement for JNK2 signaling to achieve effective IL-2 production and proliferation in response to CD3 plus CD28 cross-linking (40); however, other independently derived JNK2-deficient mice demonstrated no apparent defect (41). Interestingly, naive CD4+ T cells developing in JNK1 and JNK2 double knockout mice, or in mice homozygous mutated for the JNK activator gene MAPK kinase 7 (MKK7), also demonstrated no defect in either IL-2 production or proliferation (42), implying that in naive T cells parallel signaling pathways exist to transmit costimulatory signals to the IL-2 gene. IL-2 production has not been examined specifically in untransformed Th1 cells genetically deficient in both JNK1 and JNK2 activity because differentiation to the Th1 phenotype is inhibited in the absence of JNK1 (43). Previous experiments using transformed Jurkat cells and transfection of a GST-c-Jun fusion protein to antagonize endogenous JNK function have consistently indicated that JNK protein kinase activity is essential for AP-1- and NFAT-dependent transactivation (5, 44). Furthermore, overexpression of dominant negative mutants of MKK4 or MKK7 (potential activators of JNK) leads to reductions in NFAT and 5' IL-2 gene enhancer DNA-dependent reporter gene activation in Jurkat T cells (7, 45). Data shown here strengthen the hypothesis that JNK is an important regulator of IL-2 production in Th1 cells, through its ability to sense the ligation of CD28 and as a result of its capacity to control Elk-1 transactivation and AP-1 complex assembly.
Although this study provides evidence that JNK mediates the unique B7/CD28-dependent costimulatory signal that promotes AP-1 complex formation during Ag stimulation by increasing the activity of the Elk-1 transcription factor, this activation of Elk-1 nevertheless also depends on MEK1. This indicates that Elk-1 may be an important point of signal integration between pathways highly sensitive to TCR signals (e.g., ERK activation) and pathways downstream of CD28 that promote the activation of JNK. Consistent with this, 5' IL-2 gene enhancer-driven transactivation, IL-2 mRNA accumulation, and IL-2 secretion have been shown to be consistently inhibited by the addition of the MEK1 inhibitor PD 098059 by overexpression of dominant negative mutants of MEK1 or by overexpression of the MAPK-specific phosphatase (MKP-1) (45, 46, 47).
We note that this signal integration observed in Th1 cells could not have been predicted based on results obtained using the Jurkat T-leukemia cell line alone, where significant Elk-1-dependent transactivation has been shown to be elicited with a PKC-activating phorbol ester alone (29). We have confirmed this Jurkat result; nevertheless, treatment of Th1 cells with PMA was insufficient to activate transcription in our Gal4-Elk-1 reporter gene system or to induce high levels of c-Fos protein expression (data not shown). These differences may relate to the differentiation state of the two T cell types, or alternatively may reflect the effects of cell transformation on signal transduction in the Jurkat T-leukemia line. Regardless, the costimulatory signal-dependent activation of Elk-1 in Th1 cells observed here appears to be essential for the up-regulation of immediate-early genes such as c-fos, the accumulation of AP-1 complexes at the 5' IL-2 gene enhancer, and the induction of IL-2 transcription in response to Ag stimulation.
The biochemical nature of the cooperativity that exists between JNK and MEK1 (or perhaps ERK) at the level of Elk-1 function remains unknown. Elk-1 is known to constitutively associate with serum response factor (SRF) homodimers at the serum response element (SRE) of the 5' c-fos gene DNA promoter/enhancer (reviewed by Cahill et al. in Ref. 28). Elk-1-dependent transactivation is regulated by phosphorylation at serine-383 and -389, and each of the MAPK has been shown to be competent to phosphorylate this nuclear factor at these residues in vitro (33, 34, 35, 36, 37). Perhaps the in vivo pattern of phosphorylation of Elk-1 in Th1 cells following activation of both JNK plus ERK is unique and different from that achieved by either kinase alone. Subtle differences in phosphorylation specificity have been observed with the various MAPKs, although the significance of these differences is not well understood. Unique phosphorylations might affect the ability of Elk-1 to transactivate, or to interact with the SRF dimer or some other transcriptional regulator.
Alternatively, an individual MAPK may have the capacity to directly regulate the function of some other signaling molecule that interacts with Elk-1 and influence its transactivation. The kinase suppressor of Ras (KSR), the adaptor protein Gab2, and calcineurin have each been identified as signaling proteins capable of interfering with the function of Elk-1 (48, 49, 50, 51). Perhaps these negative regulators are themselves turned off by costimulatory signals as a result of their phosphorylation by JNK. In contrast, the transcriptional coactivator CREB-binding protein (CBP) has been shown to physically interact with Elk-1, is capable of augmenting Elk-1 transactivation, and is itself a phosphorylation target for MAPK (52). CBP (as well as the related protein p300) acts as a bridge between the inducible transactivation factors and the basal transcription machinery at the promoter, and also regulates local chromatin structure through its histone acetyltransferase activity (53, 54, 55). Several groups have recently determined in T cells that CBP binds to NFAT and integrates signals from both Raf and the Rac1 GTPase to enhance NFAT-dependent transactivation (56, 57). Because Rac1 can also mediate the CD28-dependent activation of JNK and AP-1 in Jurkat T cells (7, 8, 29, 58), we speculate that TCR and CD28 signal integration at the c-fos gene in Th1 cells results from JNK and ERK cooperatively activating Elk-1/CBP complexes at the SRE.
Finally, the discovery of unique B7/CD28 costimulatory signals integrating with TCR signals at the level of Elk-1 transactivation does not preclude the possibility of other costimulatory signals also acting elsewhere within the c-fos and IL-2 genes to influence transcription. The p38 downstream effector molecule MAPK-activated protein kinase-2 is responsive to CD28 signals in T cells and has been shown to be a SRF kinase (59, 60). In addition, the c-fos gene enhancer contains a cis-acting cAMP response element just downstream of the SRE, and transactivation by phosphorylated CREB appears to be sensitive to costimulatory signals in T cells (61). Consistent with this, CREB phosphorylation can be induced by the p38 (and ERK) downstream effector molecule mitogen- and stress-activated protein kinase-1 (62). Of note, the CBP/p300 coactivators bind both CREB and SRF, in addition to Elk-1, NFAT, Rel, and AP-1 proteins (52, 53, 54, 55, 56, 57, 63, 64, 65, 66). It remains to be determined whether costimulatory signals play a general role in the regulation of CBP-mediated transcriptional coactivation at immediate-early genes as well as at the 5' IL-2 gene enhancer itself.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Daniel L. Mueller, Center for Immunology, University of Minnesota Medical School, 312 Church Street SE, 6-120 Basic Science and Biomedical Engineering Building, Mayo Mail Code 334, Minneapolis, MN 55455. E-mail address: dmuell{at}lenti.med.umn.edu
3 Abbreviations used in this paper: JNK, c-Jun N-terminal kinase; CBP, CREB-binding protein; CSA, cyclosporin A; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase; MEK1, MAPK/ERK kinase 1; PCC, pigeon cytochrome c; SRE, serum response element; SRF, serum response factor; SAPK, stress-activated protein kinase; MKK, MAPK kinase.
4 T. Zell, A. Khoruts, E. Ingulli, J. L. Bonnevier, D. L. Mueller, and M. K. Jenkins. Analysis of jun N-terminal kinase and p38 mitogen-activated protein kinase phosphorylation in CD4 T cell stimulated by Ag in vivo. Submitted for publication.
Received for publication December 18, 2000. Accepted for publication May 8, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
to activate JNK and the IL-2 promoter in T lymphocytes. EMBO J. 17:3101.[Medline]
B by cyclin-dependent kinases associated with the p300 coactivator. Science 275:523.This article has been cited by other articles:
|
|
 |
|
  K. Praveen, Y. Zheng, F. Rivas, and T. F. Gajewski Protein Kinase C{theta} Focusing at the cSMAC Is a Consequence rather than Cause of TCR Signaling and Is Dependent on the MEK/ERK Pathway J. Immunol., May 15, 2009; 182(10): 6022 - 6030. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Morton, B. McManus, P. Garside, A. McI. Mowat, and M. M. Harnett Inverse Rap1 and Phospho-ERK Expression Discriminate the Maintenance Phase of Tolerance and Priming of Antigen-Specific CD4+ T Cells In Vitro and In Vivo J. Immunol., December 15, 2007; 179(12): 8026 - 8034. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Nandiwada, W. Li, R. Zhang, and D. L. Mueller p300/Cyclic AMP-Responsive Element Binding-Binding Protein Mediates Transcriptional Coactivation by the CD28 T Cell Costimulatory Receptor J. Immunol., July 1, 2006; 177(1): 401 - 413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jimenez-Perianez, G. Ojeda, G. Criado, A. Sanchez, E. Pini, J. Madrenas, J. M. Rojo, and P. Portoles Complement regulatory protein Crry/p65-mediated signaling in T lymphocytes: role of its cytoplasmic domain and partitioning into lipid rafts J. Leukoc. Biol., December 1, 2005; 78(6): 1386 - 1396. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Takeshita, K. Suzuki, S. Sasaki, N. Ishii, D. M. Klinman, and K. J. Ishii Transcriptional Regulation of the Human TLR9 Gene J. Immunol., August 15, 2004; 173(4): 2552 - 2561. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Adams, A. M. Grierson, A. M. Mowat, M. M. Harnett, and P. Garside Differences in the Kinetics, Amplitude, and Localization of ERK Activation in Anergy and Priming Revealed at the Level of Individual Primary T Cells by Laser Scanning Cytometry J. Immunol., August 1, 2004; 173(3): 1579 - 1586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Bonnevier and D. L. Mueller Cutting Edge: B7/CD28 Interactions Regulate Cell Cycle Progression Independent of the Strength of TCR Signaling J. Immunol., December 15, 2002; 169(12): 6659 - 6663. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or presence (
) of
anti-CD28 mAb (10 µg/ml) as indicated. Relative light units are
shown to indicate luciferase reporter gene activity. B,
Th1 cells were cotransfected with Gal4-Elk-1 and 5xGal4luc DNA, and
then were incubated overnight with syngeneic LPS-treated whole spleen
cells. Groups of cells were either unstimulated or were stimulated with
the indicated concentration of Ag OVAp. Cultures also received either
isotype control Abs (
). Th1 cells were then incubated
overnight either without stimulus or with the combination of
anti-CD3 and anti-CD28 immobilized mAbs (1 µg/ml each).
Cultures also received PD 098059 at the concentration indicated.
Luciferase reporter gene activity was determined for each of the
groups, and is expressed as the mean ± SEM difference between
duplicate mAb-stimulated T cell groups and an unstimulated control
group. D, Th1 cells were transfected and stimulated
overnight as in C, in either the absence or the presence
of the indicated concentration of SB 202190. Results are expressed as
in C. Each of the results shown is representative of at
least three experiments.