(Received for publication, February 21, 1997, and in revised form, April 17, 1997)
From the The Kennedy Institute of Rheumatology, 1 Aspenlea Road, Hammersmith, London W6 8LH, United Kingdom
Interleukin-2 (IL-2) is a potent T cell mitogen.
However, the signaling pathways by which IL-2 mediates its mitogenic
effect are not fully understood. One of the members of the
mitogen-activated protein kinase (MAPK) family, p42/44MAPK (ERK2/1), is
known to be activated by IL-2. We have now investigated the response to IL-2 of two other members of the MAP kinase family, p54MAP kinase (stress-activated protein kinase (SAPK)/Jun-N-terminal kinase (JNK))
and p38MAP kinase (p38/Mpk2/CSBP/RK), which respond primarily to
stressful and inflammatory stimuli (e.g. tumor necrosis
factor-, IL-1, and lipopolysaccharide). Here we show that IL-2, and
another T cell growth factor, IL-7, activate both SAPK/JNK and p38MAP kinase. Furthermore, inhibition of p38MAP kinase activity with a
specific pyrinidyl imidazole inhibitor SB203580 that prevents activation of its downstream effector, MAPK-activating protein kinase-2, correlated with suppression of IL-2- and IL-7-driven T cell
proliferation. These data indicate that in T cells p38MAP kinase has a
role in transducing the mitogenic signal.
Interleukin-2 (IL-2)1 is a key factor
in driving the proliferation of activated T lymphocytes; a crucial
event in mounting an effective immune response (1). The high affinity
IL-2 receptor is a heterotrimeric complex composed of ,
, and
c subunits, the latter being shared with the receptors for IL-4,
IL-7, IL-9, and IL-15, other T cell growth factor cytokines (2).
Ligation of IL-2 to its receptor initiates the activation of several
intracellular enzymes including the tyrosine kinases: Jak1, Jak3 (3,
4), Syk (5) and p56lck (6); phosphatidylinositol 3-kinase (7,
8), and p70 S6 kinase (9-11). Furthermore, IL-2 activates ERK (12-14)
via a cascade of events involving the assembly of the
Shc·Grb-2·mSOS complex (15), the regulation of the GTPase
p21ras (16, 17), the activation of Raf-1 (18, 19), and by
inference MEK1/2. However, the contribution of all these events to the
proliferative activity of IL-2 is unclear. While the activation of Jak3
(4, 20, 21) and p70 S6 kinase (10, 22) are essential for the transduction of the mitogenic signal, a variety of studies have suggested that the activation of p56lck (6, 23), Syk (24), Jak1
(25), and the p21ras/ERK1/2 pathway are not required for the
proliferative response (14, 26, 27). Furthermore, while the activation
of Jak1 and Jak3 are common to the other T cell growth factors, IL-4
(28), IL-7 (29, 30), and IL-9 (31), activation of the other
IL-2-mediated events, with the exception of the IL-7-induced activation
of p56lck (32), has not been observed.
The apparent redundancy of ERK activation in T cell proliferation is in contrast to other cellular systems, where these kinases have been implicated in mitogenic responses to growth factors (33). This functional link is supported by the transforming potential of the proximal activators of this pathway, i.e. oncogenic Ras (reviewed in Ref. 34) and Raf (35) and a constitutively active form of MEK1 (36), as well as by studies with dominant-negative and antisense cDNA (37). The inhibition of fibroblast proliferation by a specific inhibitor of MEK1, PD098059 (38), further supports a role for this pathway in proliferation.
Recently two other subgroups of the MAP kinase family have been
characterized, SAPK/JNK and p38MAP kinase. These kinases respond to a
variety of physicochemical stresses (e.g. UV light,
translational inhibitors, hyperosmolarity), lipopolysaccharide, and
the pro-inflammatory cytokines TNF- and IL-1 (39-43).
Unlike ERK, these "stress kinases" have not been implicated previously in mitogenesis. The response of these enzymes to IL-2, other T cell growth factors, or other cytokines barring those mentioned above has not been investigated previously, although SAPK/JNK has been shown to be activated in T cells by co-stimulation through CD3 and CD28 (44). This study shows that both SAPK/JNK and p38MAP kinase are activated by IL-2. In addition (which is not the case for ERK) they are also activated by IL-7. The inhibition of p38MAP kinase activity by the specific inhibitor SB203580 resulted in suppression of T cell proliferation in response to IL-2 and IL-7, suggesting that rather than being solely involved in stress responses, in T cells at least, p38MAP kinase is required to transduce the mitogenic signal.
IL-7 and IL-2 were kindly provided by Dr. C. Faltynek (Sterling Winthrope, Malvern, PA) and Dr. P. Lomedico, (Roche Inc., Nutley, NJ), respectively. Rabbit antisera to SAPK/JNK were raised to the N-terminal peptide sequence, GVVKGQPSPSAQVQQ, and to p38MAP kinase as reported previously (45). Antibody to MAPKAP kinase-2 was from Upstate Biotechnology, Inc. (Lake Placid, NY), ERK was from Santa Cruz (Santa Cruz, CA) and c-Myc was generously provided by Dr. G. Evan (Imperial Cancer Research Fund, London). The p38MAPK inhibitor SB203580 was generously provided by Dr. J. Lee, SmithKline Beecham Pharmaceuticals (King of Prussia, PA). GST-Jun (2-89) and GST-ATF2 (19-96) were purified by standard techniques.
Cell Culture and Western ImmunoblottingThe murine cytokine-dependent T cell line, CT6 (kindly provided by Genentech, South San Francisco, CA) was maintained and proliferation assays and Western immunoblotting c-Myc and ERK performed as described previously (14, 46). Cell viability was assessed by the amount of merocyanine 540 (Sigma) incorporated into the cell membrane of live gated cells (47). Human peripheral blood T cell isolation and proliferation were as described previously (48). SB203580 was added to the cells 15 min prior to the addition of cytokine where indicated.
Immunoprecipitations and Affinity PurificationCells were
lysed at approximately 20 × 106/ml in 25 mM HEPES, pH 7.4, 50 mM -glycerophosphate,
150 mM NaCl, 1% Triton T-X100, 10 mM NaF, 10%
glycerol, 2 mM EGTA, 2 mM
Na3VO4, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin. Lysates were centrifuged
12,000 × g for 10 min to precipitate debris. Kinases
were immunoprecipitated for 3 h at 4 °C with constant agitation
with specific antibody/protein-G-Sepharose (Pharmacia Biotech Inc.,
Milton Keynes, Bucks, UK). Immunoprecipitates were washed twice with
lysis buffer prior to Western immunoblotting and a further two times
with kinase assay buffer prior to kinase assays (25 mM
HEPES, pH 7.4, 25 mM
-glycerophosphate, 25 mM MgCl2, 100 mM
Na3VO4, 2 mM dithiothreitol).
In vitro kinase assays
for SAPK/JNK were performed on immunoprecipitates resuspended in 50 µl of assay buffer and 30 µl of 0.1 mg/ml GST-ATF2-(19-96) or
GST-JUN-(2-89). Reactions were initiated by addition of 10 µl of 180 µM ATP containing 0.5 µCi [-32P]ATP
(Amersham International, Little Chalfont, Bucks, UK). After agitation
for 20 min at 25 °C, reactions were terminated by boiling with gel
sample buffer.
For the in vitro kinase assays for p38MAP kinase, Hsp27
kinase (which is similar, if not identical to MAPKAP kinase-2) and Hsp27 were purified from KB cells, and recombinant Hsp27 was prepared as described previously (41). To the immunoprecipitates were added 35 µl of kinase assay buffer, 30 µl of Hsp27 kinase (MAPKAP kinase-2),
which had been previously inactivated by treatment with protein
phosphatase 2A, and Hsp27 (1.5 µg/track); the reaction was initiated
and terminated as for the SAPK/JNK assays. All products were separated
by SDS-polyacrylamide gel electrophoresis, and dried gels were
autoradiographed at 70 °C.
For the in vitro MAPKAP kinase-2 assays, the activated
enzyme was immunoprecipitated from lysates of 5-10 × 106 cell equivalents for 3 h at 4 °C. The
immunoprecipitates washed as for the p38MAP kinase and SAPK/JNK assays,
resuspended in 50 µl of kinase assay buffer containing 30 µM final concentration of Hsp27 peptide (sequence
KKLNRTLSVA (49)). Reactions were initiated with 10 µl of 180 µM ATP containing 0.5 µCi of
[-32P]ATP. Following 20 min at 25 °C, reactions
were terminated by spotting the supernatants onto squares of P81 paper
and placing into 0.75% orthophosphoric acid. Following three washes in
acid and one in acetone, the squares were subjected to scintillation counting.
SAPK/JNK activation was
measured by phosphorylation of GST Jun-(2-89) (43) or
GST-ATF-2-(19-96) (50) as substrates, after immunoprecipitation from
human (Kit-225) and murine (CT6) T cells. The enzyme was activated when
precipitated from IL-2-treated cells, as compared with that from
unstimulated cells (Fig. 1, a and
b). Treatment of the responsive CT6 cells with IL-7 also
activated this kinase (Fig. 1b). SAPK/JNK activation was
dose-dependent for both cytokines (Fig. 1, c and
d) and was maximal at 2 ng/ml (100 pM). This
correlates with the proliferative response of CT6 to either cytokine
(46). As expected, activated SAPK/JNK was also immunoprecipitated from
T cells stressed with anisomycin (Fig. 1a) or exposed to
TNF- (Fig. 1b). No kinase activity was precipitated with
non-immune rabbit serum (Fig. 1, a and b), which confirmed the specificity of the assay.
IL-2 and IL-7 Cause the Activation of p38MAP Kinase
Generally, stressful and inflammatory stimuli that
activate SAPK/JNK also activate a related enzyme, p38MAP kinase. We
therefore investigated whether this second stress kinase would respond
to T cell mitogens. The activity of p38MAP kinase immunoprecipitated from CT6 cells was assayed by a kinase cascade involving Hsp27 kinase
(which is similar, if not identical to, MAPKAP kinase-2 (51)) and its
substrate (41). A substantial elevation in Hsp27 phosphorylation was
observed with p38MAP kinase immunoprecipitates from cytokine-treated
cells (Fig. 2a). Kinetic studies showed maximal activation by IL-7 was delayed when compared with IL-2 (Fig.
2b), although this has not been apparent in all experiments. TNF- also activated p38MAP kinase as expected, although it should be
noted that CT6 cells possess only the p75 TNF receptor (p75 TNFR) (52),
while previous studies on the TNF-
response have used cells that
also express the p55 TNFR. No kinase activity was precipitated by
non-immune rabbit serum (Fig. 2 a). Unlike SAPK/JNK, p38MAP
kinase has not been previously shown to be activated in T cells.
A p38MAP Kinase Inhibitor, SB203580, Suppresses IL-2 and IL-7-driven Proliferation of CT6 Cells and Activated PBMC but Not Cytokine-induced c-Myc Expression
Preincubation (15 min) of CT6
cells (Fig. 3a) with the specific p38MAP
kinase inhibitor, SB203580, equally inhibited proliferation induced by
either IL-2 or IL-7 (IC50 ~3 µM). A similar
IC50 is observed for inhibition of TNF- production by
monocytes2 and for collagen-induced
platelet aggregation (45). We next examined the effect of this
inhibitor on primary T cells isolated from human blood (Fig. 3,
b and c). These were treated with anti-CD3 and
then induced to proliferate to IL-2 and IL-7. SB203580 inhibited the
proliferation over a concentration range similar to that observed in
the T cell lines. The compound was not toxic at the maximum concentrations used as judged by merocyanine staining in conjunction with fluorescence-activated cell sorter analysis (47) (not shown), nor
did it inhibit all responses to the cytokines, since expression of
c-Myc induced by IL-2 or IL-7 (14) was unaffected by 10 µM SB203580 (Fig. 3d).
IL-2 Activation of MAPKAP Kinase-2 Is Sensitive to SB203580
The effect of SB203580 on T cell proliferation led us to
question how closely this correlated with inhibition of one known function of p38MAP kinase. MAPKAP kinase-2 is phosphorylated and activated by p38MAP kinase in a number of cell types (41, 51), and this
is inhibited by SB203580 (53). As expected MAPKAP kinase-2 was
activated by IL-2 (Fig. 4a) in a
dose-dependent manner as judged by assays of the enzyme
immunoprecipitated from CT6 cells. Activity was maximal at 20 ng/ml
IL-2; the proliferative response of the cells displayed a similar dose
dependence (not shown). The activation of the enzyme was inhibited by
SB203580 (Fig. 4b) in the concentration range 0.1-1
µM, in agreement with the in vitro inhibition
of p38MAP kinase (53) and previous studies on MAPKAP kinase-2 (54).
There was no MAPKAP kinase-2 activity in control immunoprecipitations
with nonspecific antisera (not shown). In contrast, studies of ERK
phosphorylation by gel retardation assay showed that the IL-2-induced
phosphorylation was unaffected by 10 µM SB203580 (Fig.
4c), as was the SAPK/JNK activation measured by GST-ATF-2
phosphorylation in specific immunoprecipitates (Fig. 4d),
indicating the specificity of the compound.
This study shows that the previously termed "stress-activated kinases," p38MAP kinase and SAPK/JNK are activated in T cells by the mitogenic cytokines IL-2 and IL-7. Moreover evidence is presented that p38MAP kinase is involved in transducing these mitogenic responses. This is in contrast to ERK, which has been shown previously not to be necessary for T cell proliferation (14), opposing the paradigm in other cells that it is ERK that is required for proliferative effects and that p38MAP kinase is involved in stress responses.
The activation of p38MAP kinase by IL-2 and IL-7 is the first evidence
of the activation of this kinase in T cells. The recent identification
of highly specific pyrinidyl imidazole inhibitors of p38MAP kinase
allowed us to investigate its role in T cells responding to IL-2 or
IL-7. These inhibitors abrogate the synthesis of TNF- and IL-1 by
monocytes stimulated with lipopolysaccharide (42), and one, termed
SB203580, has been shown to be highly specific for p38MAP kinase (53).
The concentrations of drug that inhibited proliferation were
approximately one log higher than those required to inhibit the
activation of the immediate substrate MAPKAP kinase-2. This difference
in dose might be partly accounted for by the nature of the two
responses: one is a complex response and is measured 24 h after
stimulation, the other a single enzyme assayed after 20 min. The
possibility that SB203580 inhibits another protein kinase responsive to
IL-2 is unlikely, as it has no activity when tested on a large number
of other kinases at high concentration (53). However, the possible
existence of non-kinase targets cannot be absolutely discounted. These
results imply that p38MAP kinase has a novel role in mediating the
proliferative response of T cells to cytokines.
The nature of the signaling pathways leading to the activation of p38MAP kinase by IL-2 and IL-7 is an open question. The activation of MKK3/6 (55, 56), the G-proteins Rac and Cdc42 and p21-activated kinase have been shown or implicated in the activation of p38MAP kinase in a variety of cells (57, 58), however the response of these to IL-2 is unknown. p21ras, which is activated by IL-2 (16, 17), has also been suggested as a proximal activator for the kinase (59). However IL-7 does not appear to stimulate the Shc/p21ras/ERK1/2 cascade (14). As IL-2 and IL-7 both activate tyrosine kinases Jak1, Jak3, and p56lck (3, 4, 6, 29, 32), it is possible that these may be involved in initiating the signaling cascade to these kinases, and this is currently being investigated. Moreover, the events distal to the kinase are similarly unknown and warrant investigation. Our data have shown that the induction of c-Myc was not affected and that the pathway required for this event is still largely unknown. In other cells p38MAP kinase has been implicated in the translational control of TNF synthesis by unknown mechanisms (42). p38MAP kinase has also been shown to phosphorylate and activate the transcription factors ATF-1 and CREB (54), Elk-1 (60), CHOP (61), and ATF-2 (62). None of these are known to be activated by IL-2, and although Elk-1 is involved in promoting fos transcription induced by the cytokine, it is not required for IL-2-driven cell proliferation (27).
The role played by SAPK/JNK in IL-2 and IL-7 function could not be easily approached, as no specific inhibitor exists for this kinase. However, IL-2 does induce the synthesis of c-Jun, in which SAPK/JNK has been implicated (43). The activation of SAPK/JNK has been demonstrated previously in T cells (44) in response to antibody cross-linking of CD3 and CD28. This involves a calcium signal, a requirement unique to T lymphocytes. However, neither IL-2 nor IL-7 elevate intracellular calcium (2), and thus the mechanism of activation of SAPK/JNK in T cells by cytokines is distinct from that described previously. Nothing is known of the response of MKK4 (63, 64) or MEKK (59), which are proximal activators of SAPK/JNK, to IL-2 and IL-7. Furthermore despite its activation by some growth factors, e.g. epidermal growth factor on HeLa cells (58), cholecystokinin on pancreatic cells (65), and both endothelin and thrombin on airway smooth muscle cells (66), no role for SAPK/JNK in T cell mitogenesis has been defined.
In summary, we have shown that the previously termed stress-activated kinases p38MAP kinase and SAPK/JNK are activated in T cells by the mitogenic cytokines IL-2 and IL-7. Furthermore, using the specific inhibitor SB203580 a previously undescribed role for p38MAP kinase in mitogenesis has been observed. Moreover, p38MAP kinase is one of the few elements of IL-2 and IL-7 signaling pathways that can be ascribed a role in T cell proliferation. The involvement of these kinases in inflammatory responses has identified them as likely therapeutic targets. The observation that these kinases are also activated by T cell mitogenic cytokines and that at least p38MAP kinase has a role in transducing the proliferative response could have important consequences for any pharmacological modulation of these enzymes.