| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
,
,,
*
T Cell Signal Transduction Laboratory, Hospital for Special Surgery, Weill Medical College of Cornell University, the
Department of Human Genetics and
Immunology Program, Memorial Sloan-Kettering Cancer Center, and the
Weill Graduate School of Medical Sciences of Cornell University, New York, NY 10021
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As a costimulatory receptor, CD28 functions by inducing the secretion
of multiple T cell cytokines, particularly IL-2, which following
binding to newly synthesized IL-2 receptors, drives T cells through the
cell cycle (5). Apparently, CD28 costimulation induces
both de novo IL-2 gene transcription and stabilization of newly
synthesized IL-2 mRNA (6, 7). Essential to TCR/CD28
induction of IL-2 is activation of the mitogen-activated protein
(MAP)4 kinases,
extracellular signal-regulated kinase (ERK), and c-jun
NH2-terminal kinase (JNK) (8, 9).
These serine/threonine kinases phosphorylate the c-fos and c-jun
proteins, respectively, which assemble to form the AP-1 transcription
factor (10). Together with NFAT and NF-
B transcription
factors, AP-1 binds to and turns on the IL-2 promoter (11, 12). Additionally, JNK has been shown to play an important role
in IL-2 mRNA stabilization (13). Both ERK and JNK are
activated synergistically by the TCR and CD28 and thus represent early
upstream points of signal integration in T cell costimulation
(14, 15). Another type of MAP kinase that is activated
synergistically by the TCR and CD28 is the p38 MAP kinase
(16). However, in contrast to ERK and JNK, p38 activation
is not necessary for the induction of IL-2 (17).
To understand further the mechanism of CD28 costimulation we used yeast-hybrid technology in an attempt to identify novel proteins that interact physically with the CD28 cytoplasmic tail (CD28 CYT). Previous studies have shown that the lipid/serine kinase, phosphatidylinositol 3-kinase (PI3-kinase), and the adaptor protein, Grb-2, bind CD28 CYT following phosphorylation of CD28 CYT on tyrosine residue 173, present in a YMNM motif (18). PI3-kinase and Grb-2 could potentially couple CD28 to MAP kinase and other intracellular signaling pathways (19, 20). However, the role of these signaling intermediates in CD28 costimulation is controversial, and other data indicate that distinct CD28 CYT interacting proteins may be involved in the relay of CD28 signals (21, 22, 23). Surprisingly, during our search for these proteins, one novel CD28 interactor that we identified was a previously undescribed member of the family of MAP kinase phosphatases (MKPs). MKPs are thought to down-modulate cellular responses by dephosphorylating and inactivating MAP kinases. Rather than functioning as a positive signaling element, we provide evidence that this phosphatase acts as a physiologic attenuator of CD28 costimulation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The yeast two-bait and modified two-bait interaction traps have
been described (24, 25). Screening of cDNA libraries
(contained in pJG4-5) was as before using TetR-fused human CD28 CYT as
bait1 (contained in pCWX200) and LCK as
bait2 (25). From the screen of a
Hela cell library, clone 14 was characterized as promoting uracil
(URA)-independent yeast growth regardless of LCK expression. Sequencing
of the 1.1-kb cDNA insert revealed an open reading frame coding for a
novel protein, MKP6, that by comparison with available sequences in the
database of expressed sequence tags (Unigene Cluster Hs. 91448; Ref.
26) was missing the first 18 amino terminal amino acids.
DNA covering the full-length coding region of human MKP6 was cloned by
PCR from Jurkat cell cDNA using a forward primer based immediately
upstream of the sequence shown in Fig. 1
A and a reverse
primer based immediately downstream of the MKP6 stop codon. Following
sequence verification, the exact full-length MKP6 coding region was
then subcloned into the EcoRI/XhoI sites of
pJG4-5. Further analyses of the CD28-MKP6 interaction in yeast used the
full-length form of MKP6 throughout. An additional prey protein used in
these studies was the human VHR phosphatase (27).
Additional bait proteins included the CYT of human TCR
(28), p58 KIR clone 6 (25), CD2
(29), and 
200, 199, and Y200F mutants of CD28 CYT
(30). DNA encoding these proteins was generated by PCR and
inserted into the EcoRI/XhoI sites of pJG4-5 or
the EcoRI/BamHI or
EcoRI/XhoI sites of pCWX200, respectively.
Assessment of bait-prey protein interaction by growth of transformed
yeast on URA dropout plates was as before, in the absence of
LCK.
|
For expression of amino-terminal GST-tagged MKP6 in Jurkat
cells, MKP6 DNA was subcloned into the BamHI/ClaI
sites of the eukaryotic GST expression vector, pEBG (a gift from Dr.
B. J. Mayer, Children’s Hospital, Boston, MA). For expression of
bacterial GST fusion proteins, MKP6, JNK2, CD28 CYT, CD28199 CYT,
and p58 KIR clone 6 CYT DNAs were inserted into the
BamHI/EcoRI sites of the pGEX3X expression vector
(Pharmacia, Piscataway, NJ). A pGEX3X-MKP6 (C111S) construct was
produced by site-directed mutagenesis of pGEX3X-MKP6 using a
QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA).
GST-ERK1 and GST-p38
bacterial expression constructs were gifts from
Dr. Z. Luo (Boston Unversity) and Dr. S. Arkinstall (Serono
Pharmaceuticals, Geneva, Switzerland) respectively. Bacterial GST
fusion proteins were produced in protease-negative
BL21(DE3)pLysS(Ion,ompT-) bacteria (Novagen,
Madison, WI) and purified from bacterial lysates with the use of
glutathione-agarose (Sigma, St. Louis, MO). Purified GST-ERK2 and
GST-p38–2 fusion proteins were obtained independently from Dr. M. Cobb
(University of Texas Southwestern Medical Center, Dallas, TX).
MKP6 expression
Mutiple tissue Northern blots (Clontech Laboratories, Palo Alto, CA) were probed with 32P-labeled MKP6 cDNA corresponding to the full-length MKP6 coding region. Blots were washed at high stringency before autoradiography.
PBMC were prepared by Ficoll separation of buffy coats from healthy
volunteers and depleted of monocytes and B cells by negative selection
using CD14 and CD19 Abs. Resultant T cells (>90% purity) were
resuspended in complete medium (RPMI 1640 supplemented with 10% FCS
and antibiotics) at 5 x 106 cells/ml and
stimulated or not with OKT3 (CD3; Ortho Pharmaceuticals, Raritan, NJ)
and CD28.2 (CD28; PharMingen, San Diego, CA) mAb (each at 0.3 µg/ml)
plus goat anti-mouse Ig (GAM, 1.5 µg/ml; Sigma) for varying times
as indicated. Total RNA was extracted from T cells with the use of
TRIzol reagent (Life Technologies, Grand Island, NY) and subject to
standard Northern analysis as above using the same MKP6 probe and a
32P-labeled control human 
actin probe.
To examine MKP6 protein expression, T cells (2 x 106 cells per condition) were lysed in Nonidet P-40 lysis buffer. Lysates were boiled in reducing SDS sample buffer, run on 10% SDS-PAGE gels, and transferred to polyvinylidene difluoride membranes (DuPont-NEN, Boston, MA). MKP6 protein was detected by enhanced chemiluminescence following the probing of membranes with a polyclonal MKP6 antiserum (1:1000 dilution) and donkey anti-rabbit Ig-HRP (DAR-HRP, 1:1000; Santa Cruz Biotechnology, Santa Cruz, CA). The MKP6 antiserum was raised by immunization of rabbits with an MKP6 peptide (72EYVKVPLADMPHAPIGLYFDTVADK96) coupled to keyhole limpet hemocyanin. The peptide sequence is unique among MKPs.
CD28-MKP6 association in vitro
MKP6 DNA was subcloned into the BamHI/XbaI
sites of pEF-FLAG (31 ; a gift of Dr. G. Koretzky,
University of Iowa, Iowa City, IA) to produce a construct encoding for
amino-terminal FLAG-tagged MKP6. COS-7 cells were transfected with
pEF-FLAG-MKP6 (20 µg/5 x 106 cells) by
electroporation (320 V/960 µF/0.4 cm gap cuvettes) and cultured in
complete medium for 36 h. Cells were lysed in Nonidet P-40 lysis
buffer, and 1 ml of lysate containing 0.5 mg of protein was incubated
with 10 µl of packed glutathione-agarose beads coated with 4 µg of
GST-CD28, CD28199, or p58 KIR clone 6 CYT fusion proteins for 30
min. Beads were washed three times in Nonidet P-40 lysis buffer, and
GST fusion proteins were eluted in 10 mM glutathione. Coeluted
FLAG-MKP6 was detected by Western blotting as above using anti-FLAG
M2 mAb (10 µg/ml; Sigma) and GAM-HRP (1:4000; Sigma).
CD28-MKP6 association in Jurkat cells
Jurkat TagC15 cells (32 ; a gift from Dr. J. Northrop, Affymax Research Institute, Palo Alto, CA) were transfected with pEBG-MKP6 or control pEBG by electroporation (20 µg/20 x 106 cells/condition as above) and cultured in complete medium for 36 h. Cells were washed and lysed in Nonidet P-40 lysis buffer, and lysates were precleared by rotation with Sephadex G-50 beads (Sigma) for 1 h. Precleared lysates (300 µl of 1 mg protein/ml/sample) were then incubated with 3 µg of 9.3 mAb (CD28, a gift from Dr. J. Ledbetter, Bristol-Myers Squibb, Seattle, WA) or control Leu 3a mAb (CD4; American Type Culture Collection, Manassas, VA) for 15 min, and immune complexes were rescued from lysates by rotation with 10 µl of packed protein A/G-coated agarose beads (Santa Cruz Biotechnology) for 1 h. Beads were washed five times in Nonidet P-40 lysis buffer and coimmunoprecipitated GST-MKP6 or GST alone was detected by Western blotting using a rabbit polyclonal anti-GST Ab (2 µg/ml; Santa Cruz Biotechnology) followed by protein A-HRP (1:10000; Zymed, San Francisco, CA).
In vitro phosphatase assays
Assays involving hydrolysis of p-nitrophenyl phosphate (p-NPP) were performed in 96-well plates. GST fusion proteins were incubated in 200-µl reaction volumes containing 20 mM p-NPP/50 mM imidazole (pH 7.5)/5 mM DTT for different times at 37°C. Hydrolysis of p-NPP was measured by absorbance at 405 nm.
For in vitro MKP assays, GST-MAP kinases were incubated with
GST-MKP6 or MKP6 (C111S) or control GST alone in a 20-µl reaction
volume containing 50 mM imidazole (pH 7.5)/5 mM DTT. After 1 h at
37°C, reactions were stopped and activated T-X-Y dual-phosphorylated
GST-MAP kinases were detected by Western blotting using specific
anti-phospho-ERK or phospho-JNK mAb or an anti-phospho-p38
rabbit polyclonal Ab (all 1:1000; New England Biolabs, Beverly, MA)
plus GAM-HRP or protein A-HRP, respectively. To check that equivalent
quantities of GST-MAP kinases were analyzed, blots were stripped and
reprobed with anti-ERK or -JNK rabbit polyclonal Abs (2 µg/ml;
Santa Cruz Biotechnology) or an anti-p38 rabbit polyclonal Ab
(1:1000; New England Biolabs) followed by protein A-HRP.
Retroviral mediated gene transfer
DNA encoding FLAG-MKP6, FLAG-MKP6 (C111S), and human CD8-CD28
CYT or CD8-CD28199 CYT chimeric proteins was subcloned into the
NcoI/XhoI sites of pSFG (33).
CD8-CD28 fusion constructs have been described previously
(23). An internal ribosomal entry site of the
encephalomyocarditis virus together with DNA encoding for neomycin
phosphotransferase (G418R) was inserted 3' of
FLAG-MKP6 and FLAG-MKP6 (C111S) and also 3' of a control-inactive
mutant human low affinity nerve growth factor receptor (LNGFR) gene,
similarly cloned into pSFG (34, 35). Resulting constructs
contain di-cistronic genes comprised of FLAG-MKP6 or FLAG-MKP6 (C111S)
or LNGFR upstream of G418R.
Recombinant virions were pseudotyped with the gibbon-ape leukemia virus envelope using the PG13 packaging cell line (36). Retroviral gene transfer to PHA-activated human peripheral blood T cells (PBT) was performed by overnight infection in the presence of polybrene as previously described (37), except that in experiments involving transfer of CD8 chimeric receptors, PHA blasts were predepleted of CD8+ T cells by negative selection using the OKT8 mAb (CD8; American Type Culture Collection). For MKP6 gene transfer experiments, following infection, T cells were maintained in low dose IL-2 (10–20 U/ml; Chiron Therapeutics, Emeryville, CA) and selected in G418 (250 µg/ml; Sigma) for 5–7 days.
Efficiency of CD8 chimera and LNGFR gene transfer was determined by flow cytometry using CD4-FITC- and CD8-PE-coupled mAb (Becton Dickinson, Mountain View, CA) or an LNGFR mAb as appropriate. For LNGFR, this was found to be in the range of 40–60% postselection. To confirm FLAG-MKP6 and FLAG-MKP6 (C111S) gene transfer, 1 x 107 transduced and selected T cells were lysed in 300 µl of Nonidet P-40 lysis buffer, and lysates were rotated with 2 µg of anti-FLAG-M2 mAb plus 10 µl of packed protein A/G-coated agarose beads. Beads were washed five times in Nonidet P-40 lysis buffer and precipitated FLAG-tagged proteins were detected by Western blotting.
IL-2 induction
Washed transduced T cells were stimulated in 96-well flat-bottom plates (1 x 105 cells/well in a 200 µl total volume) with combinations of OKT3 or OKT8 mAb (both plate-bound by preincubation of plates with 1 µg/ml Ab), soluble CD28.2 mAb (1 µg/ml), soluble 9.6 mAb (CD2, 1 µg/ml; a gift from Dr. J. Ledbetter, Bristol-Myers Squibb) and PMA (varying concentrations as indicated) plus ionomycin (1 µM) as described (38). After 24 h, supernatants were harvested and assayed for IL-2 content with the use of IL-2 ELISA kits (R&D Systems, Minneapolis, MN).
MAP kinase activation in vivo
LNGFR- and FLAG-MKP6 (C111S)-transduced T cells were washed, and 4 x 106 T cells per sample were stimulated or not with 0.2 µg each of OKT3 and CD28.2 plus 2 µg of GAM in a total volume of 100 µl of RPMI 1640 at 37°C for the indicated times. Cells were lysed, and activation of MAP kinases was determined by Western blotting of whole cell lysates as indicated in In vitro phosphatase assays, except that phospho-JNK was detected using an anti-phospho-JNK rabbit polyclonal Ab (Promega, Madison, WI), and DAR-HRP was used in place of protein A-HRP. Equivalent loading of MAP kinases was demonstrated with the use of the same anti-MAP kinase Abs and DAR-HRP.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We used the yeast two-bait and modified two-bait interaction trap
systems to identify novel interactors of the CD28 CYT (24, 25). From library screens we isolated both PI3-kinase and Grb-2.
In addition, we identified one clone whose protein product bound CD28
CYT in yeast independent of CD28 CYT tyrosine phosphorylation. The 198
aa sequence of clone 14 is shown in Fig. 1
A. Clone 14 has not been
reported previously although it can be found in both the human and
murine divisions of the database of expressed sequence tags, where it
has been derived from a variety of tissue sources and cell types
including T cells (26). A similarity search performed with
the clone 14 sequence revealed significant homology to the family of
MKPs (22–32% similarity) of which there are at least ten members
(39). MKPs are dual specific phosphatases that
dephosphorylate MAP kinases upon threonine and tyrosine residues
present in T-X-Y sequences in the phosphorylation lip at the mouth of
the MAP kinase catalytic site. As such, MKPs reverse MAP kinase
phosphorylation mediated by upstream activating MEKs and inactivate MAP
kinases. Homology of clone 14 to this family extends throughout most of
the protein with the exception of the amino-terminal 20 and
carboxyl-terminal 30 aa. Of note is a centrally located
VHCXXGXSRSXTXXXAYLM catalytic core sequence that is conserved among
MKPs. Therefore, taken together with enzymatic data (see below), we
designate clone 14 as a novel member of the MKP family, MKP6.
The CD28 CYT-MKP6 interaction in yeast is specific and requires CD28 CYT residue tyrosine 200
The specificity of interaction between CD28 CYT and MKP6 in yeast
cells is illustrated in Fig. 1B. MKP6 bound to the CD28 CYT
but not to the CYT of a variety of other immunoreceptors including, as
shown, TCR, p58 KIR clone 6, and CD2. Conversely, CD28 CYT failed to
bind other MKPs apart from MKP6. Shown, for example, is the VHR MKP
(27), which did not bind CD28 CYT.
To determine which residues of CD28 CYT are necessary for MKP6 binding
we tested the effect of deleting increasing numbers of residues from
the carboxyl-terminal end of the tail. Essential results are depicted
in Fig. 1C, which shows that a CD28200 CYT, missing
carboxyl-terminal arginine 201 and serine 202, was able to interact
with MKP6, in contrast to a CD28199 CYT, additionally deleted of
tyrosine 200, which was unable to mediate this interaction. These
results show that arginine 201 and serine 202 are dispensable, whereas
tyrosine 200 is required for CD28 CYT-MKP6 binding. However, it should
be noted that tyrosine 200 can be mutated to phenylalanine (CD28 Y200F
CYT) without loss of binding activity (Fig. 1C).
Expression of MKP6 in PBT is induced by CD28 costimulation
Combined Northern blot (Fig. 2A) and RNA dot blot (data not
shown) analysis of different fetal and adult human tissues and a panel
of hemopoietic and nonhemopoietic cell lines confirmed that MKP6 was
expressed ubiquitously. Tissues with relatively strong MKP6 expression
included adult trachea, placenta, liver, heart, and thyroid, and fetal
thymus and lung. Somewhat lower levels of MKP6 were detected in other
adult and fetal tissues. However, we wondered whether MKP6 was induced
in T cells by CD28 costimulation. To address this we examined MKP6 mRNA
expression in purified populations of PBT that had been stimulated with
CD3 mAb (directed to the TCR complex) plus CD28 mAb for varying periods
in vitro. As shown in Fig. 2B, the combination of CD3 and
CD28 mAb induced strong expression of MKP6 mRNA in PBT. Induction was
apparent by 2.5 h and persisted up to 24 h poststimulation.
Little induction of MKP6 mRNA was noted when T cells were stimulated
with CD3 or CD28 mAb alone (Fig. 2B). In parallel with the
induction of MKP6 mRNA, CD3/CD28 stimulation of PBT also induced
expression of MKP6 protein as detected by Western blotting using an
MKP6-specific Ab (Fig. 2C). One other MKP, which has been
shown to be induced in T cells by mitogens, is the PAC-1 phosphatase
(40). However, PAC-1 and MKP6 differ in that PAC-1 is
relatively restricted to the nucleus, whereas MKP6 is expressed in both
nuclear and cytosolic compartments (as detected using green fluorescent
protein-MKP6 fusion protein constructs; data not shown).
|
We tested whether CD28 CYT and MKP6 interact in vitro and in T
cells. For in vitro studies, bacterially expressed GST-CD28 CYT or
control GST-p58 KIR clone 6 CYT or CD28199 CYT fusion proteins were
immobilized on glutathione-agarose beads and bathed in lysates of COS-7
cells that had been transiently transfected with FLAG-tagged MKP6.
Following bead washing, GST fusion proteins were eluted, and any bound
MKP6 was detected by Western blotting using an anti-FLAG mAb. Fig. 3A shows that FLAG-MKP6 bound
to GST-CD28 CYT but not to GST-p58 KIR clone 6 CYT or CD28199 CYT.
Therefore, CD28 CYT and MKP6 interact in vitro. Furthermore, as in
yeast cells, this interaction is prevented by deletion of tyrosine 200
from the CD28 CYT.
|
B,
GST-MKP6 was detected in CD28 immunoprecipitates but not in CD4
immunoprecipitates. Similarly, GST-MKP6 was detected in CD28
immunoprecipitates from "rested" serum-starved Jurkat cells, and
the amount of GST-MKP6 that was coimmunoprecipitated from these cells
was not increased by prior mAb-mediated cross-linking of CD28 (data not
shown). Therefore, in transfected Jurkat cells, CD28 and MKP6 interact
specifically and constitutively.
We additionally examined whether endogenous MKP6 could be detected in
CD28 mAb immunoprecipitates from PBT. However, we were unable to detect
MKP6 by Western blotting using our MKP6-specific antiserum (data not
shown). This could indicate that the physical interaction between CD28
and MKP6 is of low stoichiometry and/or that our MKP6-specific
antiserum is of low affinity. A low stoichiometric interaction between
CD28 and MKP6 is consistent with the observation that in in vitro
binding experiments (Fig. 3A), we estimate that GST-CD28 CYT
bound only one four-hundredth the total amount of FLAG-MKP6 that was
available.
MKP6 behaves as a general MKP in vitro
To show that MKP6 could function as a MKP we performed two
different types of in vitro assay. In the first assay we examined the
ability of purified GST-MAP kinases to activate MKP6 phosphatase
activity against the artificial substrate p-NPP. As shown in
Fig. 4A, all tested GST-MAP
kinases, including GST-ERK1, ERK2, JNK2, p38, and p38–2, were able
to induce the phosphatase activity of GST-MKP6. In contrast, control
GST alone did not induce phosphatase activity. Previous studies have
shown that the phosphatase activity of MKPs can be activated by MAP
kinases in a substrate-specific manner (41). Therefore,
these results imply that MKP6 can act as a general MKP in vitro.
|
, B–D). GST-MAP kinases were purified from bacteria in an
already dual-phosphorylated state, obviating a requirement for
incubation with MEKs in these experiments. Dual phosphorylation was
monitored by immunoblotting using specific anti-phospho-MAP kinase
Abs. As shown, GST-MKP6 was able to dephosphorylate each of GST-ERK1,
JNK2, and p38, thus confirming that MKP6 behaves as general MKP in
vitro. As expected, conversion of the predicted nucleophilic cysteine
residue to a serine residue (C111S) in the active site of MKP6
abrogated an ability to dephosphorylate these MAP kinases. Retroviral transfer of dominant-negative MKP6 (C111S) to primary human PBT reveals the function of MKP6 as a negative regulator of CD28 costimulatory signaling
Because MAP kinase activation is necessary for TCR/CD28-induced T cell cytokine secretion we considered the possibility that MKP6 may function as a negative regulator of CD28 costimulatory signaling. To test this, we transferred phosphatase-inactive FLAG-MKP6 (C111S) (see above) to primary human PBT with the use of retroviruses. FLAG-MKP6 (C111S) might behave as a dominant-negative mutant by competitively inhibiting endogenous MKP6, thereby resulting in augmented CD28 costimulatory signaling.
Retroviruses were constructed with di-cistronic genes encoding
FLAG-MKP6 (C111S) or a control null mutant LNGFR upstream of a G418
resistance gene. Following infection, T cells were then selected in
G418 to maximize the percentage of cells expressing transgenes. To
monitor transfer of FLAG-tagged proteins, G418-selected T cells were
lysed, and lysates were analyzed by Western blotting using an
anti-FLAG mAb (Fig. 5A).
As shown, FLAG-MKP6 (C111S) was readily detected in infected T cells.
No signals were detected in T cells infected with the LNGFR-encoding
retrovirus.
|
B). In control
LNGFR-transduced T cells, CD3 and CD28 mAb synergized to induce T cell
IL-2 secretion. However, in FLAG-MKP6 (C111S) transduced T cells CD3
and CD28 mAb induced considerably larger quantities of IL-2 under
identical stimulation conditions. We interpret these results as
evidence that in primary human T cells, MKP6 functions as a negative
regulator of CD28 costimulatory signaling. Specific enhancement of CD28 costimulatory signaling by dominant-negative MKP6 (C111S) and influence of wild-type MKP6
In additional experiments, we examined whether the
dominant-negative effect of FLAG-MKP6 (C111S) was specific to CD28
costimulation. Also, we examined the influence of FLAG-wild-type MKP6
upon CD28 costimulation, which might be expected to inhibit responses
by principle of overexpression. Gene transfer of FLAG-MKP6 (C111S),
FLAG-MKP6, and LNGFR to PBT was as before. However, although FLAG-MKP6
(C111S) was expressed well in transduced PBT, FLAG-MKP6 was expressed
at 
10-fold lower levels (Fig. 6A).
|
B, left).
Conversely, less IL-2 was secreted by FLAG-MKP6-transduced T cells,
although the inhibition was slight. Interestingly, the enhanced IL-2
secretion observed in FLAG-MKP6 (C111S)-transduced T cells was seen to
be specific to CD28 costimulation. Thus, the same FLAG-MKP6
(C111S)-transduced T cells did not secrete enhanced quantities of IL-2
when stimulated with the combination of CD3 mAb plus mAb against
another T cell costimulatory receptor, CD2 (42) (Fig. 6B, left).
We also examined IL-2 secretion in response to another control
stimulus, i.e., PMA and ionomycin (Fig. 6B,
right). Again, in contrast to results obtained with CD3 plus
CD28 mAb, the three different populations of T cells secreted similar
quantities of IL-2 in response to PMA (at 1 ng/ml) and ionomycin.
However, when the combination of suboptimal PMA (0.1 ng/ml) and
ionomycin plus CD28 mAb was used as a stimulus, FLAG-wild-type
MKP6-transduced T cells secreted significantly less IL-2 (50%) and
FLAG-MKP6 (C111S)-transduced T cells again secreted severalfold more
IL-2 than LNGFR controls (Fig. 6B, right). Altogether, these
results provide strong evidence that MKP6 is involved specifically in
negative regulation of CD28 costimulation in T cells. Presumably, that
FLAG-MKP6 did not more strongly inhibit CD28 costimulatory signaling in
these experiments can be accounted for on the basis that FLAG-MKP6 was
only poorly expressed in the transduced T cells (Fig. 6A).
ERK and JNK but not p38 MAP kinases are hyperphosphorylated upon CD28 costimulation of dominant-negative MKP6 (C111S)-transduced PBT
Previous studies have shown that ERK and JNK rather than
p38 MAP kinases are involved in TCR/CD28 induction of IL-2. Therefore,
we asked whether ERK and JNK MAP kinases become hyperphosphorylated in
dominant-negative MKP6 (C111S)-transduced primary T cells following
TCR/CD28 stimulation, which could account for enhanced IL-2 secretion.
Fig. 7 shows that in both LNGFR- and
FLAG-MKP6 (C111S)-transduced T cells, ERK, JNK, and p38 MAP kinases
became transiently dual-phosphorylated in response to stimulation with
the combination of CD3 and CD28 mAb. However, the level of
dual-phosphorylation of ERK and JNK was considerably larger in
FLAG-MKP6 (C111S)-transduced T cells compared with LNGFR-transduced
controls. Hyperphosphorylation of ERK and JNK in these preactivated T
cells was evident by 15 min poststimulation. Unlike ERK and JNK, the
degree of dual-phosphorylation of p38 was comparable between LNGFR- and
FLAG-MKP6 (C111S)-transduced T cells. Thus, enhanced IL-2 secretion is
associated with hyperphosphorylation of ERK and JNK but not p38 MAP
kinases. Furthermore, these results demonstrate that, although in vitro
MKP6 behaves as a general MKP, in T cells MKP6 controls specifically
the activation of ERK and JNK during the course of CD28 costimulatory
signaling.
|
199 CYT hyperinduces IL-2 from PBT
Because a CD28-MKP6 physical association was difficult to detect
in primary T cells, we examined whether such an association occurs
functionally. We asked whether an MKP6 binding-deficient CD28 CYT would
hypercostimulate IL-2 from primary T cells. To test this, we
constructed chimeric receptors comprised of the extracellular plus
transmembrane domains of human CD8 linked to either the wild-type
CD28 CYT or a CD28199 CYT that was unable to bind MKP6 in yeast and
in vitro (Figs. 1 and 3). Chimeric receptors were then transferred to
CD8-depleted primary human PBT, again using retroviruses. As detected
by immunostaining with CD4 and CD8 mAb, CD8-CD28 CYT and CD8-CD28199
CYT chimeric receptors were expressed at comparably high levels on the
respective, predominantly CD4+, transduced T cell
populations (Fig. 8A). By
contrast, mock transduced T cells were not reactive with CD8
mAb.
|
B). As expected,
CD8-CD28 CYT-transduced T cells secreted IL-2 in response to CD3 plus
CD8 mAb, thus confirming that the CD28 CYT is sufficient to transduce
CD28 signals in primary T cells (43, 44). However,
importantly, CD8-CD28199 CYT-transduced T cells secreted severalfold
more IL-2 upon challenge with CD3 and CD8 mAb. Mock transduced T cells
did not produce any IL-2 when stimulated with CD3 and CD8 mAb
(similarly, chimeric receptors, which contained only the first five
amino acids of the CD28 tail or a CD28 tail with a Y173F mutation,
induced no IL-2 or much reduced IL-2, respectively, as determined in
separate experiments; data not shown). However, all three populations
secreted comparable amounts of IL-2 in response to stimulation with CD3
plus CD28 mAb. As in experiments with dominant-negative FLAG-MKP6
(C111S), enhanced IL-2 production mediated by the CD8-CD28199 CYT
chimeric receptor was associated with hyperphosphorylation of MAP
kinases (data not shown). Therefore, these results provide functional
evidence that MKP6 interacts physically with CD28 in primary human T
cells. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MKP6 was originally cloned by virtue of its ability to interact physically with the CD28 CYT in yeast. Moreover, this physical interaction can be demonstrated in vitro and in transfected Jurkat T leukemic cells. In these settings, MKP6 binding to CD28 is seen to be specific, independent of CD28 CYT tyrosine phosphorylation, but dependent upon tyrosine residue 200, located toward the distal end of the CD28 tail. Therefore, the mechanism of interaction with CD28 CYT is distinct from that seen with PI3-kinase and Grb-2, both of which bind CD28 CYT on phosphorylated tyrosine residue 173. However, it has been more difficult to demonstrate a physical association between endogenous MKP6 and CD28 in primary human PBT. Therefore, we sought independent functional evidence for this association in primary T cells. In this regard, a chimeric receptor with an MKP6 binding-deficient CD28 CYT (lacking the last three residues of CD28 CYT including tyrosine 200), in synergy with the TCR, induced severalfold more IL-2 from transduced primary human PBT than a similar receptor with a wild-type CD28 CYT. Excepting the possibility that the last three residues of CD28 CYT are involved in the activation of a distinct negative regulator of CD28 signaling, this finding is consistent with the idea that MKP6 and CD28 interact physically in primary T cells.
It is interesting to speculate as to why MKP6 might require binding to CD28 to execute its function. We have not observed that MKP6 becomes activated upon interaction with CD28. Therefore, we favor the possibility that MKP6 might bind to CD28 to become localized to plasma membrane MAP kinase substrates. In quiescent cells, MAP kinases are known to be concentrated at the plasma membrane, at which site, or at least in early endosome compartments, MAP kinase activation is initiated (46, 47). As such, MKP6 binding to CD28 could allow localization to recently activated MAP kinases and subsequent inactivation before MAP kinase nuclear translocation.
Alternatively, in binding to CD28, MKP6 could inactivate MAP kinases that function proximally. Although MAP kinases are traditionally thought to act in the nucleus, a growing body of evidence indicates that this family of kinases performs additional roles at nonnuclear locations (48, 49). Perhaps most pertinent to costimulation is the recent observation that, in T cells, the ERK1 MAP kinase functions at a plasma membrane location by controlling the amplitude of TCR-generated signals (50). Soon after TCR triggering, ERK1 associates physically with the TCR complex (50, 51). Ostensibly, in its activated state this recruited ERK1 prevents docking to the TCR of the SHP-1 tyrosine phosphatase, which acts to down-regulate TCR signals, probably by dephosphorylating the TCR-associated protein tyrosine kinase, ZAP-70 (52). Interestingly, during the course of T cell activation initiated by natural TCR and CD28 ligands, and in Ag-specific T cell-APC conjugates, CD28 is known to translocate to the region of the TCR that becomes concentrated in the "central TCR supramolecular activation cluster" (Refs. 53, 54 ; and A. Kupfer, personal communication). Therefore, these findings suggest that MKP6, localized to the TCR by CD28, may function as a negative regulator by first inactivating TCR-associated ERK1, thereby resulting in more rapid SHP-1 recruitment. In turn, through the actions of SHP-1, MKP6 could ultimately down-regulate signals leading to the activation of other ERK isoforms and JNK MAP kinases. If true, then assumedly, at least some aspects of these signaling interactions are mimicked when CD3 plus CD28 mAb or the combination of PMA plus ionomycin and CD28 mAb are used to stimulate T cells in vitro.
Aside from the issue of physical interaction, one major and important finding in the current studies is that dominant-negative MKP6 (C111S)-enhanced CD28-mediated IL-2 secretion specifically. Similar enhancements of IL-2 secretion were not observed in response to PMA plus ionomycin alone or the combination of CD3 plus CD2 mAb. This indicates a specific involvement of MKP6 in CD28 costimulatory signaling. Furthermore, this finding raises the intriguing possibility that distinct MKPs may regulate different signaling pathways in T cells. With the exception of MKP6, described here, and PAC-1 (40), MKPs have been little studied in T cells. Like MKP6, PAC-1 expression is induced upon T cell activation. However, whether PAC-1 regulates specific signaling pathways has not been addressed. A full description of which MKPs are turned on in T cells and which types of signaling pathway each MKP regulates awaits.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 The sequence reported in this article is available in the Genbank database under accession number AF120032.
3 Address correspondence and reprint requests to Dr. Philip D. King, T Cell Signal Transduction Laboratory, Hospital for Special Surgery, Weill Medical College of Cornell University, 535 East 70th Street, New York, NY 10021.
4 Abbreviations used in this paper: MAP, mitogen-activated protein; CYT, cytoplasmic tail; DAR-HRP, donkey anti-rabbit Ig-HRP; ERK, extracellular-signal regulated kinase; GAM, goat anti-mouse Ig; JNK, c-jun NH2-terminal kinase; LNGFR, low affinity nerve growth factor receptor; MKP, MAP kinase phosphatase; PBT, peripheral blood T cell(s); PI3-kinase, phosphatidylinositol 3-kinase; p-NPP, p-nitrophenyl phosphate; URA, uracil.
Received for publication August 7, 2000. Accepted for publication October 2, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B transcription factor activation in T cells. Ann. NY Acad. Sci. 766:245.[Medline]
B/Rel transcription factors. Mol. Cell. Biol. 17:2605.[Abstract]
chain: distinction from the molecular CD3 complex. Proc. Natl. Acad. Sci. USA 85:9709. chain of the T cell receptor (TcR) with the SH2,3 domain of p56lck: differential regulation by TcR cross-linking. J. Biol. Chem. 271:10054.This article has been cited by other articles:
|
|
 |
|
  A. R. Raimondi, A. Molinolo, and J. S. Gutkind Rapamycin Prevents Early Onset of Tumorigenesis in an Oral-Specific K-ras and p53 Two-Hit Carcinogenesis Model Cancer Res., May 15, 2009; 69(10): 4159 - 4166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jurek, K. Amagasaki, A. Gembarska, C.-H. Heldin, and J. Lennartsson Negative and Positive Regulation of MAPK Phosphatase 3 Controls Platelet-derived Growth Factor-induced Erk Activation J. Biol. Chem., February 13, 2009; 284(7): 4626 - 4634. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Boutros, E. Chevet, and P. Metrakos Mitogen-Activated Protein (MAP) Kinase/MAP Kinase Phosphatase Regulation: Roles in Cell Growth, Death, and Cancer Pharmacol. Rev., September 1, 2008; 60(3): 261 - 310. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. Schuit and D. J. Drucker {beta}-Cell Replication by Loosening the Brakes of Glucagon-Like Peptide-1 Receptor Signaling Diabetes, March 1, 2008; 57(3): 529 - 531. [Full Text] [PDF]
|
|
|
|
|
|
S. Klinger, C. Poussin, M.-B. Debril, W. Dolci, P. A. Halban, and B. Thorens Increasing GLP-1-Induced {beta}-Cell Proliferation by Silencing the Negative Regulators of Signaling cAMP Response Element Modulator-{alpha} and DUSP14 Diabetes, March 1, 2008; 57(3): 584 - 593. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Salojin and T. Oravecz Regulation of innate immunity by MAPK dual-specificity phosphatases: knockout models reveal new tricks of old genes J. Leukoc. Biol., April 1, 2007; 81(4): 860 - 869. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lang, M. Hammer, and J. Mages DUSP Meet Immunology: Dual Specificity MAPK Phosphatases in Control of the Inflammatory Response J. Immunol., December 1, 2006; 177(11): 7497 - 7504. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Kowolik, M. S. Topp, S. Gonzalez, T. Pfeiffer, S. Olivares, N. Gonzalez, D. D. Smith, S. J. Forman, M. C. Jensen, and L. J.N. Cooper CD28 Costimulation Provided through a CD19-Specific Chimeric Antigen Receptor Enhances In vivo Persistence and Antitumor Efficacy of Adoptively Transferred T Cells. Cancer Res., November 15, 2006; 66(22): 10995 - 11004. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Graham, M. P. Bell, C. J. Huntoon, M. D. Griffin, X. Tai, A. Singer, and D. J. McKean CD28 Ligation Costimulates Cell Death but Not Maturation of Double-Positive Thymocytes due to Defective ERK MAPK Signaling J. Immunol., November 1, 2006; 177(9): 6098 - 6107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. O. Liu The Yins of T Cell Activation Sci. Signal., January 4, 2005; 2005(265): re1 - re1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-H. Chen, R. Luche, B. Wei, and N. K. Tonks Characterization of Two Distinct Dual Specificity Phosphatases Encoded in Alternative Open Reading Frames of a Single Gene Located on Human Chromosome 10q22.2 J. Biol. Chem., October 1, 2004; 279(40): 41404 - 41413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Sadra, T. Cinek, and J. B. Imboden Translocation of CD28 to lipid rafts and costimulation of IL-2 PNAS, August 3, 2004; 101(31): 11422 - 11427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Alonso, S. Narisawa, J. Bogetz, L. Tautz, R. Hadzic, H. Huynh, S. Williams, A. Gjorloff-Wingren, M. C. D. Bremer, L. J. Holsinger, et al. VHY, a Novel Myristoylated Testis-restricted Dual Specificity Protein Phosphatase Related to VHX J. Biol. Chem., July 30, 2004; 279(31): 32586 - 32591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Cook, L. Albacker, A. August, and A. J. Henderson CD28-dependent HIV-1 Transcription Is Associated with Vav, Rac, and NF-{kappa}B Activation J. Biol. Chem., September 12, 2003; 278(37): 35812 - 35818. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. V. Parry, C. A. Rumbley, L. H. Vandenberghe, C. H. June, and J. L. Riley CD28 and Inducible Costimulatory Protein Src Homology 2 Binding Domains Show Distinct Regulation of Phosphatidylinositol 3-Kinase, Bcl-xL, and IL-2 Expression in Primary Human CD4 T Lymphocytes J. Immunol., July 1, 2003; 171(1): 166 - 174. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-R. Chen, J. Han, R. Kori, A.-N. T. Kong, and T.-H. Tan Phenylethyl Isothiocyanate Induces Apoptotic Signaling via Suppressing Phosphatase Activity against c-Jun N-terminal Kinase J. Biol. Chem., October 11, 2002; 277(42): 39334 - 39342. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Zama, R. Aoki, T. Kamimoto, K. Inoue, Y. Ikeda, and M. Hagiwara A Novel Dual Specificity Phosphatase SKRP1 Interacts with the MAPK Kinase MKK7 and Inactivates the JNK MAPK Pathway. IMPLICATION FOR THE PRECISE REGULATION OF THE PARTICULAR MAPK PATHWAY J. Biol. Chem., June 21, 2002; 277(26): 23909 - 23918. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Zama, R. Aoki, T. Kamimoto, K. Inoue, Y. Ikeda, and M. Hagiwara Scaffold Role of a Mitogen-activated Protein Kinase Phosphatase, SKRP1, for the JNK Signaling Pathway J. Biol. Chem., June 21, 2002; 277(26): 23919 - 23926. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Masuda, H. Shima, M. Watanabe, and K. Kikuchi MKP-7, a Novel Mitogen-activated Protein Kinase Phosphatase, Functions as a Shuttle Protein J. Biol. Chem., October 12, 2001; 276(42): 39002 - 39011. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
