| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Signaling1
,
* Benaroya Research Institute, Seattle, WA 98101;
Department of Immunology, University of Washington School of Medicine, Seattle, WA 98195; and
Fred Hutchinson Cancer Research Center, Seattle, WA 98109
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
opposes proliferative signaling by IL-2 through mechanisms that remain incompletely defined. In a well-characterized CD8+ T cell model using wild-type and mutated IL-2 receptors, we examined the effects of TGF- on distinct IL-2 signaling events in CD8+ T cells. IL-2 induces c-myc, cyclin D2, and cyclin E in a redundant manner through the Shc and STAT5 pathways. TGF- inhibited the ability of either the Shc or STAT5 pathway to induce these genes, as well as T cell proliferation. The inhibitory effects of TGF- were reversed by expression of a dominant-negative form of Smad3. TGF- did not impair proximal signaling by Shc or STAT5, and induction of some downstream genes, including cytokine-inducible Src homology-2-containing protein (CIS), bcl-xL, and bcl-2, was spared. Experiments with c-fos, cyclin D2, and CIS reporter genes revealed that promoter-proximal regulatory elements dictate the sensitivity of IL-2 target genes to inhibition by TGF-. By leaving the Shc and STAT5 pathways functional while inhibiting their target genes selectively, TGF- was found to uncouple the proliferative and antiapoptotic functions of IL-2. Thus, TGF- is not a simple antagonist of IL-2, but rather serves to qualitatively modify the IL-2 signal to create a unique pattern of gene expression that neither cytokine can induce independently. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
IL-2 promotes T cell proliferation by activating cell growth pathways and inducing genes such as c-myc, cyclin D2, and cyclin E (1). The proliferative effects of IL-2 are opposed by TGF-, which directly inhibits promitogenic gene induction and other cell cycle events (2, 3, 4, 5). Activated CD4+ Th cells release both IL-2 and TGF- (2), therefore responding T cells are expected to commonly receive both signals simultaneously (3, 4, 5). The balance between proliferative signals such as IL-2 and inhibitory signals such as TGF- is essential to maintain immune homeostasis, as defects in these pathways have severe consequences ranging from immune deficiency to fatal autoimmunity (1, 6, 7). Thus, the molecular cross-talk between the IL-2 and TGF- signaling pathways is a critical aspect of immune regulation, however, as discussed below, the underlying mechanisms remain poorly defined.
The major proliferative pathways downstream of the IL-2R are beginning to be elucidated (1, 8). Upon ligand binding, the IL-2R activates two tyrosine kinases, Janus kinase 1 (JAK1)3 and JAK3, which associate with the cytoplasmic domains of the IL-2R and 
c subunits (3). Subsequently, IL-2R undergoes tyrosine phosphorylation, creating docking sites for the adaptor protein Shc and the transcription factor STAT5. The IL-2 signal bifurcates at this point into Shc- and STAT5-mediated branches, which appear to function largely as independent signaling modules (9). Shc recruits several other adaptor proteins to form a complex that, at a minimum, activates the Ras/mitogen-activated protein kinase and phosphatidylinositol 3-kinase pathways (10, 11). In contrast, STAT5 is a transcription factor that undergoes tyrosine phosphorylation, followed by dimerization and translocation to the nucleus (12).
Importantly, both the Shc and STAT5 signaling modules can induce promitogenic gene expression and cell proliferation in an apparently redundant manner (9). That is, mutated forms of IL-2R that activate Shc but not STAT5 induce expression of promitogenic genes such as c-myc, cyclin D2, and cyclin E, as well as robust T cell proliferation (13, 14). Similarly, mutant receptors that activate STAT5 but not Shc can mediate these same effects (15). Finally, mutant receptors that activate JAK1 and JAK3 but fail to recruit Shc or STAT5 are absolutely nonmitogenic (9, 15). Thus, the cell cycle apparatus in T cells is dually regulated by the Shc and STAT5 pathways, which therefore represent potential targets for inhibition by TGF-.
TGF- triggers intracellular signaling events via two receptor subunits (TGF- type I receptor and type II receptor), which have intrinsic serine-threonine kinase domains (16). TGF- first binds to the type II receptor, which in turn transphosphorylates the type I receptor (17). The type I receptor then activates the Smad family of signal transducers (18, 19, 20), as well as Smad-independent pathways involving TGF-associated kinase/Ras, Daxx, and protein phosphatase type 2A (PP2A) (21, 22, 23, 24). Mice rendered genetically deficient in one Smad member, Smad3, exhibit immune dysregulation reminiscent of the phenotype seen with impaired TGF- responses (25, 26, 27, 28, 29), suggesting that the Smad pathway is an important mediator of TGF- signaling in the immune system.
Smad3 is activated via serine phosphorylation of a C-terminal SSXS motif by the TGF- type I receptor (30). Smad3 then forms a secondary complex with Smad4 and enters the nucleus, where it interacts with transcription factors, coactivators, and corepressors to regulate gene expression (18, 19, 31). The consensus DNA-binding sequence for Smad complexes remains incompletely defined, apart from a CAGA motif found in some genes (18, 32). Smads regulate gene expression by a variety of mechanisms, including recruitment of transcriptional coactivators such as CREB binding protein (CBP)/p300 (33, 34, 35, 36) (37) as well as transcriptional corepressors, such as 5'TG3' interacting factor, Ski, and SnoN (38, 39, 40, 41, 42, 43). Recent studies have revealed that Smads can also regulate the stability of Smad nuclear interacting protein 1, a nuclear repressor of CBP/p300 (44, 45), and the cytoplasmic signaling protein human enhancer of filamentation 1 (46).
There are at least three mechanisms by which TGF- can inhibit cell proliferation. First, activated Smad3 has been shown to induce expression of cyclin-dependent kinase inhibitors such as p15, p21, and p27, which inhibit G1- to S-phase progression in lymphocytes and other cell types (47, 48, 49, 50). Second, in some cell types, TGF- can act through PP2A to inhibit p70S6 kinase (24), a serine/threonine kinase that regulates translation of specific mRNAs and is essential for G1/S progression (51). Third, TGF- can inhibit the induction of c-myc and other promitogenic genes by growth factors such as IL-2 (4, 52, 53).
Although this latter inhibitory mechanism is poorly understood at the molecular level, it could potentially be explained by reports that TGF- inhibits IL-2-induced tyrosine phosphorylation events in T cells. In an early study, TGF- was reported to inhibit IL-2R-mediated tyrosine phosphorylation of multiple substrates of unknown identity (54). More recently, several groups have reported that TGF-, through an unknown mechanism, inhibits the phosphorylation and activation of one or more components of the JAK/STAT pathway, including JAK1, JAK3, STAT5, and STAT3 (55, 56, 57). Based on this, it has been proposed that TGF- antagonizes IL-2R signaling by direct inhibition of proximal signaling events. Notably however, Sudarshan et al. (58) performed an extensive analysis of JAK1, JAK3, and STAT5 activation by IL-2 in the presence or absence of TGF- and, in direct contrast to earlier studies, saw no evidence of inhibition of these or any other IL-2R signaling components, despite seeing strong growth inhibition. In summary, inhibition of the JAK/STAT pathway is currently the only reported mechanism by which TGF- inhibits IL-2R signaling, yet this phenomenon is not consistently observed and hence remains controversial.
In the work reported here, we used IL-2R mutants to systematically investigate the effects of TGF- on the Shc and STAT5 branches of the IL-2R signal. To identify points of cross-talk between the TGF- and IL-2 receptors, we traced signaling events from the activated IL-2R complex at the cell surface through to the promoters of individual Shc and STAT5 target genes. Our results reveal that TGF- antagonizes IL-2 signaling primarily in the nucleus through the inhibitory activity of Smad3 on a subset of IL-2 target genes. TGF- can inhibit genes downstream of either the Shc or STAT5 pathways, and sensitivity to TGF- is dictated by regulatory elements within the genes themselves. By this mode of inhibition, TGF- uncouples the promitogenic and antiapoptotic effects of IL-2, thereby allowing T cells to remain viable without proliferating. Thus, TGF- does not simply oppose IL-2, but instead modifies the IL-2R signal to produce a pattern of gene expression that neither IL-2 nor TGF- can generate independently.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The murine IL-2-dependent T cell line CTLL-2 was obtained from the American Type Culture Collection (Manassas, VA) and maintained in RPMI 1640 supplemented with 10% FBS, 2 mM L-glutamine, 25 mM HEPES, 50 U/ml penicillin, 50 µg/ml streptomycin, 1 mM sodium pyruvate, 25 mM 2-ME, and 50 U/ml recombinant human (rh) IL-2 (Chiron, Emeryville, CA). Recombinant murine TGF- and recombinant human IL-15 were obtained from R&D Systems (Minneapolis, MN), and recombinant human GM-CSF was purchased from Amgen (Thousand Oaks, CA). Unless otherwise indicated, cytokines were used at the following concentrations for experiments: IL-2 (100 U/ml), TGF- (50 nM), IL-15 (10 ng/ml), and GM-CSF (100 ng/ml).
For stable transfections, linearized plasmids were introduced into CTLL-2 cells by electroporation, and stably transfected subclones were selected at limiting dilution for G418 resistance. Expression of dominant-negative Smad3 was assessed by Western blot (see Western blotting) with Abs to wild-type murine Smad3 and, as a loading control, STAT5 (SC-6033 and SC-835; Santa Cruz Biotechnology, Santa Cruz, CA). Chimeric receptor expression was assessed by flow cytometry with Abs to human GM-CSFR
or c (SC-458 and SC-457; Santa Cruz Biotechnology).
Thymidine incorporation assays were performed as described (15). Cell counts and percent viability were determined by trypan blue exclusion and light microscopy.
Northern blotting
CTLL-2 cells were washed three times in PBS and resuspended at 5 x 105 cells/ml. Cells were deprived of cytokine for 4 or 8 h, and then harvested either unstimulated or at serial time points after cytokine stimulation. Preparation of RNA and Northern blotting was performed as described (14).
Plasmid construction
The chimeric receptor chains wt, 325Shc
P, and 325-Y510 have been described (14, 15, 59). DN-Smad3 was generated by PCR using an oligonucleotide that converted serines 422, 423, and 425 of human Smad3 to alanine (60). The mutated cDNA was sequenced and cloned into the expression vector pNA' (59).
The reporter genes cytokine-inducible Src homology-2-containing protein (CIS)-luciferase (LUC) (61) and D2(1227)-LUC (formerly referred to as 1227-1168 LUC, Ref. 62) have been described. FOS-LUC was generated by inserting bases -701 to +80 of the human c-fos promoter into the XhoI and HindIII sites of pGL3-Basic (Promega, Madison, WI).
Western blotting
CTLL-2 cells were washed three times in PBS and resuspended at 5 x 105 cells/ml. Cells were deprived of cytokine for 4 or 8 h and then harvested either unstimulated or at serial time points after cytokine stimulation. Preparation of cell lysates and Western blotting has been described (14). The following polyclonal Abs were used: phospho-Jak1 (44–422; Biosource International, Camarillo, CA); phospho-Stat5 (9351S), phospho-AKT (9271S), AKT (9272), phospho-Erk (9101S), or extracellular signal-related kinase (Erk) (9102) (Cell Signaling Technology, Beverly, MA); and p70S6 kinase (SC-230; Santa Cruz Biotechnology).
Luciferase reporter gene assays
Reporter gene assays were performed as described using a subclone of CTLL-2 cells that is amenable to transient transfection (62). After overnight recovery, transfected cells were cultured for 4 h in complete medium lacking IL-2. Cells were either left unstimulated or stimulated with IL-2 alone (100 U/ml), TGF- alone (50 nM), or IL-2 + TGF-. After 5.5 h of stimulation, cells were lysed and assayed in triplicate for luciferase content (62). Fold induction was calculated by dividing the mean value for stimulated cells by that for unstimulated cells.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
inhibits IL-2-induced proliferation by a Smad3-dependent mechanism
Experiments were performed using the murine IL-2-dependent CD8+ T cell line CTLL-2, as these cells have well-characterized IL-2 signaling properties and are amenable to stable and transient transfection. As reported for primary T cells, the IL-2-induced proliferation of CTLL-2 cells was inhibited by 
50% at doses of TGF- ranging from 12.5 to 800 pM (Fig. 1A). Accordingly, TGF- also inhibited IL-2-induced expression of the promitogenic genes c-myc, cyclin D, and cyclin E (Fig. 1B).
|
fails to inhibit T cell proliferation in response to IL-15, a cytokine that is closely related to IL-2 (57). This seems at odds with the fact that the IL-15R uses the same and c subunits as the IL-2R and appears to activate the same downstream signaling pathways (1). However, this conclusion was based on studies of bulk human peripheral T cells and hence could potentially reflect differential sensitivities of heterogenous T cell subsets to IL-15 and/or TGF-. Therefore, we evaluated this issue in CTLL-2 cells, as these cells are clonal and proliferate equally well in response to IL-2 or IL-15. As expected, IL-15 induced a strong proliferative response that was equivalent to that induced by IL-2 (Fig. 1C). Importantly, TGF- inhibited the proliferative response to IL-15 by as much as 69% (Fig. 1C), and inhibition occurred at doses of TGF- as low as 1.6 nM (data not shown), similar to the results for IL-2 (Fig. 1A). Thus, the previously reported differential effect of TGF- on IL-2- vs IL-15-induced proliferation (57) is not applicable in all circumstances and hence does not reflect an invariant property of these receptor systems.
To test the role of Smad3 in the antiproliferative effects of TGF-, we introduced a dominant-negative form of Smad3 (DN-Smad3) into CTLL-2 cells and evaluated the effect on TGF- signaling. DN-Smad3 is a full-length version of murine Smad3 in which three C-terminal serine residues (Ser 422, 423, and 425) were point-mutated to alanine. This nonphosphorylable version of Smad3 competes with wild-type Smad3 for access to the TGF- receptor, which results in dominant inhibition of Smad3 signaling (60).
Multiple subclones of CTLL-2 cells that stably expressed high levels of DN-Smad3 were readily obtained. DN-Smad3 was expressed at vastly higher levels than endogenous Smad3, which could be seen in nontransfected cells only after prolonged exposure to film (Fig. 2A and data not shown). Expression of DN-Smad3 did not impair the proliferative response to IL-2 (data not shown). However, cells expressing DN-Smad3 were largely or fully protected from the antiproliferative effect of TGF- (Fig. 2B). Specifically, the magnitude of inhibition by TGF- was reduced from 40 to 50% for parental CTLL-2 cells to a mean value of 3% for DN-Smad3-expressing subclones. Accordingly, DN-Smad3 completely blocked the ability of TGF- to inhibit the IL-2-induced expression of c-myc, cyclin D2, and cyclin E (Fig. 2C). Thus, the antiproliferative effects of TGF- on IL-2R signaling are mediated largely, if not entirely, through Smad3.
|
does not inhibit proximal signaling events associated with the Shc or STAT5 pathways
The Shc and STAT5 pathways downstream of the IL-2R can each independently induce promitogenic gene expression and T cell proliferation, though to a lesser extent than the two pathways combined (14). The fact that TGF- inhibited proliferative signaling by only 40–50% suggested it might preferentially block either the Shc or STAT5 pathway while sparing the other. Alternatively, TGF- could inhibit the activity of both pathways equally, but incompletely. To distinguish these possibilities, we evaluated the effects of TGF- on biochemical events and target genes specific to the Shc vs STAT5 pathways.
Cytokine-deprived CTLL-2 cells were stimulated with IL-2 in the presence or absence of TGF-, and cytoplasmic fractions were subjected to immunoblotting with Abs specific for various IL-2R signaling components. RNA was prepared from parallel samples and analyzed by Northern blot, which confirmed that TGF- effectively inhibited promitogenic gene expression in these experiments (data not shown). TGF- had no effect on the tyrosine phosphorylation of JAK1, which is upstream of both the Shc and STAT5 pathways (Fig. 3). Similarly, TGF- did not alter tyrosine phosphorylation of cytoplasmic or nuclear STAT5 (Fig. 3 and data not shown). The Shc signaling pathway was evaluated with Abs specific for the serine phosphorylated forms of ERK and AKT and by assessing the apparent m.w. of p70S6 kinase, which increases upon phosphorylation. Similar to the results for JAK1 and STAT5, TGF- did not inhibit any of these phosphorylation events over the entire time course, with the exception of a slight decrease in the phosphorylation of p70S6 kinase in some experiments (Fig. 3). Thus, proximal events in the Shc and STAT5 signaling pathways appear to be unaffected by TGF-.
|
inhibits proliferative gene induction by both the Shc and STAT5 pathways
Although TGF- did not inhibit biochemical events in the Shc pathway, it did inhibit induction of the Shc target gene c-fos (Fig. 4), which was fully reversed by expression of DN-Smad3 (data not shown). Importantly, TGF- did not inhibit all Shc target genes, as it modestly enhanced the induction of c-jun (Fig. 4). As for STAT5 target genes, TGF- did not inhibit expression of either CIS or CD25 (Fig. 4), which initially suggested that STAT5 target genes might be spared from Smad3-mediated inhibition. To investigate this idea further, we again evaluated the effect of TGF- on c-myc and cyclin D2 expression, but this time using IL-2R mutants that selectively activate either the Shc or STAT5 pathways. Because Shc and STAT5 can each independently induce these genes, this provided a unique opportunity to test whether the mechanism of gene inhibition by TGF- was pathway-specific or gene-specific.
|
c fused to the extracellular domain of GM-CSFR (designated wt), and the cytoplasmic domain of IL-2R fused to the extracellular domain of GM-CSFR (designated wt). When ligated by GM-CSF, wt and wt generate an intracellular signal that is indistinguishable from that induced by IL-2 through the endogenous IL-2R (59). To selectively activate the Shc or STAT5 pathways, we used two mutated versions of called 325-ShcP and 325-Y510, which lack the STAT5 and Shc activation sites of IL-2R, respectively (13, 14, 15).
As in prior studies, GM-CSF stimulation of CTLL-2 cells coexpressing wt and either 325-ShcP or 325-Y510 induced robust proliferation (data not shown) accompanied by induction of c-myc and cyclin D2 (Fig. 5A). TGF- inhibited the induction of c-myc and cyclin D2 irrespective of whether signaling was initiated by 325-ShcP or 325-Y510 (Fig. 5A). Similarly, for both receptor mutants, TGF- inhibited cell proliferation by 60–80% (Fig. 5B). Thus, TGF- can inhibit target gene induction and T cell proliferation mediated by either the Shc or STAT5 pathway. This implies that the mechanism of inhibition by TGF- is gene-specific rather than pathway-specific.
|
is determined by promoter-proximal regulatory elements
The above results suggested that regulatory elements within target genes themselves might dictate responsiveness to TGF-. This was investigated using reporter genes corresponding to genes that are inhibited by TGF- (c-fos and cyclin D2) or not inhibited (CIS). Notably, these reporter genes are also representative of both Shc and STAT5 target genes (c-fos and CIS, respectively). It is also noteworthy that the cyclin D2 reporter gene used in these experiments consisted of a 59-bp enhancer element that contains a single STAT5 site in tandem with an Sp1 site (62). Thus, even though the endogenous cyclin D2 gene is dually regulated by Shc and STAT5, the cyclin D2 reporter gene used in this study responds only to the STAT5 component of the IL-2R signal.
CTLL-2 cells were transiently transfected with the reporter genes FOS-LUC, D2(1227)-LUC, and CIS-LUC. After overnight recovery in complete medium containing IL-2, cells were deprived of IL-2 for 4 h and subdivided into four experimental groups. Cells were either left unstimulated or stimulated with IL-2 alone, TGF- alone, or IL-2 + TGF- for 5.5 h. Luciferase content was measured in triplicate for each sample. Because each of the four experimental groups was derived from the same transfected culture, it was not necessary to correct for transfection efficiency. Instead, results were normalized to the values of unstimulated samples. Experiments were performed three to five times to confirm that the observed reporter gene inductions were reproducible. IL-2 stimulation increased the expression of FOS-LUC by an average of 8.4-fold, CIS-LUC by 3.3-fold, and D2 (1227)-LUC by 3.1-fold (Fig. 6A). Stimulation with TGF- alone had little or no effect on expression of these reporter genes (Fig. 6A). However, TGF- inhibited the IL-2-induced expression of FOS-LUC and D2(1227)-LUC by an average of 59% and 44%, respectively (Fig. 6B). By contrast, TGF- had no effect on the induction of CIS-LUC, and this distinction was statistically significant (Fig. 6B; p < 0.01). Thus, the reporter genes recapitulated the Northern blot results for c-fos, cyclin D2, and CIS. These results indicate that TGF- and Smad3 inhibit IL-2 signaling by acting upon specific regulatory elements within the target genes themselves. This is especially evident when comparing D2(1227)-LUC and CIS-LUC, given that both reporter genes are induced by STAT5 (61, 62) but only D2(1227)-LUC is sensitive to TGF-.
|
uncouples IL-2R signals for cell viability vs cell proliferation
The foregoing studies indicate that TGF- neither globally antagonizes IL-2R signaling nor preferentially inhibits the Shc or STAT5 pathways. Instead, TGF- blocks a specific subset of target genes as dictated by regulatory elements contained within their promoters/enhancers. From a physiological perspective, one might expect TGF- to inhibit genes that have shared functions, as is the case for the proliferative genes c-myc, c-fos, cyclin D2, and cyclin E. Apart from proliferation, a second function of IL-2 is to promote T cell viability, therefore we investigated the effects of TGF- on this aspect of IL-2R signaling.
IL-2 promotes T cell viability by several mechanisms, chief among these being activation the phosphatidylinositol 3-kinase/AKT pathway and induction of the antiapoptotic genes bcl-xL and bcl-2 (63). In Fig. 3A, we showed that phosphorylation of AKT was not impaired by TGF-. Moreover, as shown in Fig. 7A, IL-2-induced expression of bcl-xL and bcl-2 proceeded normally in the presence of TGF-. Like c-myc and other promitogenic genes, bcl-xL and bcl-2 are dually regulated by the Shc and STAT5 pathways, therefore these results reinforce the notion that the target genes of TGF- are related by function rather than by their mechanism of regulation with respect to the IL-2R.
|
affected the ability of IL-2 to sustain T cell viability in culture. CTLL-2 cells were washed free of all cytokines and cultured for 1 day with graded doses of IL-2 in the presence or absence of a saturating amount of TGF- (50 nM). Even at limiting doses of IL-2, at which any inhibitory effect of TGF- ought to be most pronounced, T cell viability was unaffected by TGF- (Fig. 7B). For example, cells cultured at a limiting dose of IL-2 (8.3 U/ml) were 60% viable after 1 day irrespective of whether TGF- was present or absent (Fig. 7B). Notably, the antiproliferative effect of TGF- was still evident in these experiments (data not shown), indicating that TGF- uncouples the antiapoptotic and promitogenic components of the IL-2R signal through its differential effects on IL-2R target genes. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
inhibits IL-2R signaling in CD8+ T cells. A priori, there were several possible mechanisms of antagonism to consider, including global inhibition of all aspects of IL-2R signaling (e.g., through proximal inhibition of JAK activity), inhibition of specific downstream pathways (e.g., the Shc or STAT5 pathways), or inhibition of individual target genes. The differences between these models are not only of biochemical interest, but also speak to the functional relationship between IL-2 and TGF-. Depending on the mechanism used, TGF- could either cause general attenuation of the IL-2 signal, or it could inhibit specific aspects of the signal while sparing others, thereby generating a signal that is qualitatively distinct from that produced by IL-2 alone.
Our findings support the latter model. First, we provide the first evidence that TGF- antagonizes IL-2R signaling largely, if not entirely, through the transcription factor Smad3. Specifically, a dominant-negative version of Smad3 blocked the ability of TGF- to inhibit individual IL-2 target genes, including c-myc, c-fos, cyclin D2, and cyclin E, as well as IL-2-induced T cell proliferation. In contrast to some prior reports, TGF- did not inhibit IL-2-induced activation of JAK1 or STAT5, and we further showed that biochemical events downstream of Shc, including phosphorylation of ERK, p70S6 kinase and AKT, were similarly unaffected by TGF-. Consistent with these biochemical analyses, TGF- did not inhibit all target genes of the Shc or STAT5 pathways, but instead inhibited individual target genes downstream of either pathway. These latter target genes are all functionally related to cell proliferation. By contrast, the induction of genes involved in cell viability, including bcl-xL and bcl-2, was not inhibited by TGF-, and accordingly TGF- did not impair the ability of IL-2 to maintain T cell viability in culture. Thus, the combination of TGF- and IL-2 can generate a unique signal that promotes cell viability but with minimal induction of proliferation.
There are prior reports that TGF- inhibits the IL-2-induced tyrosine phosphorylation and activation of multiple proteins in T cells, including JAK1, JAK3, STAT5, and/or STAT3, with some differences noted between studies (54, 55, 56, 57). Thus, it has been proposed that TGF- antagonizes IL-2R signaling by direct inhibition of proximal signaling events. If JAK1 and JAK3 were inhibited by TGF-, one would expect most if not all events associated with the Shc and STAT5 pathways to be similarly inhibited, as these pathways are critically dependent on JAK kinase activity (1). However, we did not observe decreased phosphorylation of JAK1, STAT5, or the Shc pathway components AKT and ERK in the presence of TGF-. Therefore, our results are inconsistent with an inhibitory effect of TGF- at the level of JAKs, or proximal components of the Shc or STAT5 pathways. Furthermore, we failed to see any differential sensitivity to TGF- when comparing IL-2- and IL-15-induced T cell proliferation, in contrast to one of these prior studies (57). The discrepancy between our results and prior studies may reflect the use of different T cell types, stimulation conditions, or other experimental variables. Indeed, as discussed further below, a recurring theme with TGF- signaling is that the biochemical and biological consequences are often context-dependent. Notably, our results are consistent with those of Sudarshan et al. (58) who performed an extensive analysis of TGF- signaling in PHA-activated human PBMC and failed to see any effects of on IL-2-induced phosphorylation of JAK1, JAK3, or STAT5A. Thus, even if TGF- does inhibit JAK or STAT activity under some circumstances, our results and those of Sudarshan et al. (58) show that this need not occur for TGF- to exert an antiproliferative effect.
Our results are the first to implicate Smad3 as mediating the inhibitory effect of TGF- on IL-2R signaling. That being said, it should be noted that the dominant-negative Smad3 strategy used in this study could potentially affect signaling by functionally related Smad family members such as Smad2. Nevertheless, our results clearly demonstrate that intact Smad signaling is required for TGF- to inhibit all IL-2 target genes analyzed in this study, including c-fos, c-myc, cyclin D2, and cyclin E. We further showed, using reporter genes, that gene-specific regulatory sites dictate the sensitivity of target genes to TGF-. TGF--regulated inhibitory sites have been identified in several genes (38, 40, 64, 65, 66, 67, 68). Some of these sites are bound directly by Smad3, which can recruit transcriptional corepressors, such as 5'TG3' interacting factor, Ski, and SnoN (38, 39, 40, 41, 42, 43). One recent report suggests that, in naive T cells, Smad3 and Smad4 bind to the IL-2 promoter and inhibit expression by recruiting the transcriptional repressor transducer of ErbB-2, which is down-regulated only in the presence of CD28 costimulatory signals (69). Other possible mechanisms of inhibition include competition between IL-2-activated transcription factors and Smad3 for the transcriptional coactivator CBP/p300 (70, 71) or Smad3-mediated degradation of unidentified signaling proteins or transcriptional regulators downstream of the IL-2R, as shown for Ski, SnoN, Smad nuclear interacting protein 1, and human enhancer of filamentation 1 (39, 42, 45, 46, 72). The reporter genes described in this study should allow these various possibilities to be tested in the context of IL-2R signaling.
The c-myc gene has recently been shown to contain a TGF- inhibitory element within its promoter-proximal region (67, 68). This site is adjacent to a well-characterized E2F binding site and, upon TGF- stimulation of the HaCaT cell line, is bound by a complex containing Smad3, Smad4, E2F4/5, DP1, and p107 (67). E2F4/5 in association with p107 is a transcriptional repressor in the context of other genes and therefore is proposed to mediate repression of c-myc transcription in response to TGF-. Although the formation of a Smad3, Smad4, E2F4/5, DP1, and p107 complex has not yet been demonstrated in T cells, such a mechanism is consistent with our findings that 1) DN-Smad3 reverses the inhibitory effects of TGF- on c-myc, and 2) TGF- inhibits induction of c-myc by either the Shc or STAT5 pathways, each of which would theoretically be blocked by a repressive complex at the promoter-proximal E2F site.
In mammary epithelial cells, TGF- activates PP2A, which in turn inactivates p70S6 kinase, thereby inhibiting cell proliferation (24). We observed a very modest inhibition of p70S6 kinase phosphorylation in the presence of TGF-, suggesting this mechanism may operate at a low level in T cells. This could also explain the slight residual reduction in thymidine incorporation seen after TGF- treatment in some CTLL-2 subclones expressing DN-Smad3 (Fig. 2). However, this mechanism cannot account for the effects of TGF- on promitogenic gene expression, as complete blockade of p70S6 kinase by rapamycin treatment has no effect on transcription of the c-myc, c-fos, or cyclin D2 genes in CTLL-2 cells (B. H. Nelson, unpublished observations).
We have shown that the combination of TGF- and IL-2 generates a signal that neither TGF- nor IL-2 can generate on their own. In other words, TGF- does not inhibit the IL-2 signal so much as edit it. Although these results were obtained in vitro using a T cell line and a highly circumscribed set of stimuli, the general principle that Smad3 inhibits specific target genes rather than entire signaling pathways may have profound implications for primary lymphocytes, which integrate a multitude of signals in vivo and select among a variety of genetic programs underlying proliferation, survival, differentiation, and other physiological behaviors. Indeed, this principle may explain the often contradictory effects that have been ascribed to TGF- under different physiological circumstances. For example, depending on the T cell subset and context of stimulation, TGF- can either promote (73) or inhibit (74, 75, 76, 77) T cell apoptosis, or have a neutral effect, as demonstrated here in the case of IL-2R signaling. If TGF- is viewed as a qualitative modifier of other signals, rather than a simple antagonist, then its observed activity on a cell will depend largely on the other signals being received concurrently. This implies that the study of cytokine cross-talk will become increasingly challenging as more stimuli are taken into consideration. However, this line of investigation may ultimately reveal a level of information processing in cells that matches the physiological complexities we currently observe but cannot yet explain in molecular terms.
|
Acknowledgments |
|---|
; Warren Leonard for the CD25 probe; and Bryan Carson, Richard Tempero, and Brad Stone for helpful comments. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Brad H. Nelson at the current address: British Columbia Cancer Agency, Vancouver Island Center, 2410 Lee Avenue, Victoria, British Columbia, V8R 6V5, Canada. E-mail address: bnelson{at}bccancer.bc.ca
3 Abbreviations used in this paper: JAK, Janus kinase; PP2A, protein phosphatase type 2A; CBP, CREB binding protein; CIS, cytokine-inducible Src homology-2-containing protein; rh, recombinant human; ERK, extracellular signal-related kinase; LUC, luciferase.
Received for publication January 23, 2003. Accepted for publication March 31, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Annu. Rev. Immunol. 16:137.[Medline]
in T-cell biology. Nat. Rev. Immunol. 2:46.[Medline]
on IL-2- and IL-4-dependent T cell-cycle progression. J. Immunol. 144:1767.[Abstract]
by human T lymphocytes and its potential role in the regulation of T cell growth. J. Exp. Med. 163:1037. signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12:171.[Medline]
-chain tyrosines couple to at least two signaling pathways and synergistically mediate interleukin 2-induced proliferation. Proc. Natl. Acad. Sci. USA 93:2077. signals through a heteromeric protein kinase receptor complex. Cell 71:1003.[Medline]
signalling. Trends Genet. 12:493.[Medline]
signal transduction. Annu. Rev. Biochem. 67:753.[Medline]
signal transduction. Science 270:2008. signal transduction. Science 272:1179.[Abstract]
-induced apoptosis is mediated by the adapter protein Daxx that facilitates JNK activation. Nat. Cell Biol. 3:708.[Medline]
inhibits p70S6 kinase via protein phosphatase 2A to induce G1 arrest. Genes Dev. 14:3093.-mediated signal transduction. Mol. Cell. Biol. 19:2495.. EMBO J. 18:1280.[Medline]
1-deficient mice. J. Immunol. 153:1936.[Abstract]
1-null mouse is lymphocyte mediated. Proc. Natl. Acad. Sci. USA 92:12215.1-deficiency in mice is dependent on MHC class II antigen expression. J. Clin. Invest. 98:2109.[Medline]
receptor. Cell 95:779.[Medline]
receptor kinase activation to transcriptional control. Genes Dev. 16:1950.-inducible elements in the promoter of human plasminogen activator inhibitor-type 1 gene. EMBO J. 17:3091.[Medline]
. Genes Cells 3:613.[Abstract]
transcriptional responses in endothelial cells. Proc. Natl. Acad. Sci. USA 95:9506.-induced transcriptional activation. Genes Dev. 12:2153.-stimulated cooperation of smad proteins with the coactivators CBP/p300. Genes Dev. 12:2114. signaling. Cytokine Growth Factor Rev. 12:1.[Medline]
signaling through interaction with smads. J. Biol. Chem. 274:35269. signaling. Mol. Cell 4:499.[Medline]
signaling by the SnoN oncoprotein. Science 286:771. family signaling. Curr. Top. Microbiol. Immunol. 254:145.[Medline]
signal transduction. Genes Dev. 14:1605.. EMBO J. 19:5178.[Medline]
-mediated p15INK4B induction and growth inhibition in astrocytes is SMAD3-dependent and a pathway prominently altered in human glioma cell lines. J. Biol. Chem. 274:35053.1 induces the cyclin-dependent kinase inhibitor p27Kip1 mRNA and protein in murine B cells. J. Immunol. 160:770. inhibition of endothelial cell proliferation: alteration of EGF binding and EGF-induced growth-regulatory (competence) gene expression. Cell 49:415.[Medline]
. Mol. Cell. Biol. 8:3088. on early and late activation events in human T cells. J. Immunol. 150:3109.[Abstract]
inhibits IL-2-induced tyrosine phosphorylation and activation of Jak-1 and Stat 5 in T lymphocytes. J. Immunol. 159:175.[Abstract]
, secreted from a new type of CD4+ suppressor T cell, NY4.2, in the prevention of autoimmune IDDM in NOD mice. J. Autoimmun. 10:299.[Medline]
is reversed by IL-15. J. Immunol. 167:553. does not inhibit IL-12- and IL-2-induced activation of Janus kinases and STATs. J. Immunol. 162:2974. and chains mediate the signal for T-cell proliferation. Nature 369:333.[Medline]
-indu
alanine) in 12 stably transfected CTLL-2 subclones (numbered 1–12) compared with untransfected cells (P), assessed by Western blot of cell lysates with Smad3-specific antiserum. Reprobing with antiserum to STAT5 demonstrated similar protein loading in all lanes. B, Thymidine incorporation assay showing diminished or abrogated antiproliferative signaling by TGF-
