Nature Immunology
- 7, 1130 - 1132 (2006)
doi:10.1038/ni1106-1130
Cell cycle 'check points' T cell anergyChristopher E RuddChristopher E. Rudd is with the Cell Signalling Section, Department of Pathology, Cambridge University, Cambridge CB2 1QP, UK. cer51@cam.ac.uk A pathway has been defined linking cell cycle inhibitor p27Kip1 to the inhibition of cyclin-dependent kinase 2 and its phosphorylation of transcription factor Smad3 in the induction of in vivo tolerance.T cell tolerance is of chief importance to the homeostasis of the immune system and protection against autoimmunity. Although cell deletion and suppression contribute to tolerance, T cells can also become unresponsive or anergic after encountering antigen in the absence of signals from CD28 and interleukin 2 (IL-2) and thereby contribute to the tolerant state1. Manipulation of tolerance has potentially substantial therapeutic implications in the treatment of autoimmunity as well as graft and tumor rejection. In this issue of Nature Immunology, Li and coworkers document a pathway involving cell cycle inhibitor p27Kip1 in the inhibition of cyclin-dependent kinase 2 (Cdk2) and the maintenance of transcription factor Smad3 activity for the induction of in vivo tolerance2.
Anergy contributes to homeostasis in which many tissue cells present endogenous peptides without costimulation. Activation, proliferation and clonal expansion occur when T cells encounter foreign antigen in the presence of costimulatory signals. One longstanding issue has been the identification of the receptors and signaling processes that control anergy induction. Although models have postulated the existence of anergy-specific signals, the more tenable model is that unresponsiveness results from an incomplete array of activation signals (a subthreshold effect). In that context, many alterations in early signaling events have been associated with anergy induction. Those events include the 'preferential' activation of Src kinase p59Fyn and the GTP-binding protein Rap1, defective activation of the kinases p56Lck and the signaling protein Zap70, reduced activation of the GTPase Ras and the kinase Erk and decreased binding of the transcription factor AP-1 complex to the Il2 promoter. Advances have arrived with the reported development of autoimmunity in mice lacking various E3 ligases (such as Cbl-b, Itch and GRAIL) of the ubiquitin-mediated proteolytic pathway3. Calcium-dependent alteration of those ligases impairs receptor endocytosis and the degradation of key signaling proteins.
Of those various events, the most direct model proposes that the central gap is the failure to activate mediators of the cell cycle for progression from G1 to S phase. Anergy is an active process, as shown by the partial phosphorylation of substrates and its blockade by the inhibition of protein synthesis. Nevertheless, cells remain arrested in G1. In this model, CD28 and IL-2–IL-2 receptors act as surrogates that stimulate signals for cell cycle progression that are not provided by T cell receptor ligation. Rapamycin (which inhibits TOR protein kinases needed for cell division) and forced upregulation of the cell cycle inhibitor p27kip1 can induce an anergic-like state, and that can even occur in the presence of costimulation4,
5.
Cell cycling is a complex process with many participants that promote or inhibit different stages of the cycle (Fig. 1a). Central are the Cdks, whose activities are positively regulated by binding to cyclins. In resting cells, G1 Cdks are relatively inactive because of the abundance of cyclin-Cdk2 inhibitors such as p27Kip1, which ensure that cell cycling is not triggered without a proper signal. The inhibitor p27Kip1 binds and inhibits cyclin E–Cdk2 complexes in late G1 phase (Fig. 1a). T cell proliferation is accompanied by downregulation of p27Kip1, resulting in increased Cdk-mediated phosphorylation of transcription factors Smad2 and Smad3 and the retinoblastoma protein, leading to S-phase entry6. Phosphorylation of Smad2 and Smad3 inhibits their transcriptional activity and antiproliferative effects. That antiproliferative function is partially mediated by increased transcription of the Cdk inhibitor p15. Other inhibitory factors include p16CdkN2a of the INK4 family and p21Waf1/Cip1 of the Kip family6.
 | | Figure 1. The tolerogenic pathway involves inhibition of Cdk2 activity by p27Kip1 and an increase in Smad3 function. | | |  | (a) Regulation of the cell cycle. Cells progress from G0 or G1 to S (DNA synthesis), G2 and mitosis. Cyclin E–Cdk2 complexes drive the transition from G1 to S, whereas cyclin A–Cdk2 and cyclin A–cyclin B–Cdk2 complexes are involved in S and G2, respectively. The anaphase-promoting complex (APC) is required for anaphase and exit from mitosis. The inhibitor p27kip1 binds and inhibits cyclin E–Cdk2 kinase, whereas Cdk2 can phosphorylate the serine at position 212 of Smad3 (pS-212) and inhibit its transcription activity. Smad3 increases p15 (Cdk inhibitor 2B) and inhibits IL-2 production. The proteins p15, p16CdkN2a (p16; also called INK4a) and p21Waf1/Cip1 (p21) can inhibit Cdk4 needed for G1-phase progression. (b) In this model, anergy and in vivo tolerance is induced by p27Kip1 binding and the inhibition of cyclin E–Cdk2 complexes, which leads to less inhibitory phosphorylation of Smad3. That results in increased Smad3 activity, increased p15 expression and inhibition of IL-2 production. Conversely, proliferation and clonal expansion results from a loss of p27Kip1 expression and/or binding to cyclin E–Cdk2, leading to an increase in inhibitory phosphorylation of the serine at position 212 of Smad3 and a reduction in p15 and upregulation of IL-2 production.
 Full Figure and legend (48K) |
| |
Given that background, p27Kip1 and other inhibitors of the G1-to-S transition have been viewed as prime potential effectors of anergy. Suboptimal activation with costimulatory-deficient antigen-presenting cells concurrently fails to downregulate p27Kip1 and induces anergy. In fact, CD28 and IL-2 both downregulate p27Kip1, allowing for Cdk activation. Furthermore, p27Kip1-deficient T cells are essentially resistant to anergy induction7. Conversely, forced upregulation of p27Kip1 induces both in vitro anergy and in vivo tolerance5, that may operate 'preferentially' after the first cell division.
Despite that connection with the G1 cell cycle checkpoint, delineation of the pathway linking p27Kip1 to Cdk2 and in vivo tolerance has been missing. Although much information about the interconnections of components in cell cycling has been provided for other cell systems, that information has not been rigorously applied to the study of T cell anergy. Here, Li et al. initially show that T cells tolerized by in vivo blockade of CD28 and CD40 have impaired Cdk2 kinase activity and consequently fail to phosphorylate the serine at position 212 of Smad3 (ref. 2; Fig. 1b). They then show that T cells from DO11.10 T cell receptor–transgenic mice expressing a mutant form of p27Kip1 (p27 ) with a deleted Cdk-binding domain (and therefore unable to inhibit Cdks) retain Cdk2 activity and Smad3 phosphorylation and fail to upregulate the inhibitor p15. Consequently, those cells progress to G1-S despite tolerizing conditions and are resistant to in vivo tolerance induction2.
To causally link those steps, these investigators show that T cells expressing a mutant Smad3 that is resistant to Cdk2 phosphorylation become anergic in both priming and tolerizing conditions. Both conditions are associated with lower IL-2 production and proliferation. In addition, T cells with reduced Smad3 expression by means of RNA interference ('Smad3-knockdown' cells) are resistant to anergy induction. Those cells have less p15 expression, whereas phosphorylation-resistant cells expressing mutant Smad3 have more p15 expression in conditions producing anergy. Overall, the new aspect of this study is the definition of a pathway comprising p27Kip1-Cdk2-Smad3 in the G1-to-S transition during in vivo tolerance.
This study also opens the door to other issues. For example, the Smad3-knockdown DO11.10 p27 mice had much more T cell activation than did Smad3-knockdown DO11.10 mice, suggesting that the p27Kip1 Cdk-binding domain mediates additional inhibition independently of Smad3. The identity of that alternative pathway could be of great importance. Moreover, the true relevance of the higher p15 expression is unclear at present. It does confirm that the effect of Smad3 mutation reflects differences in transcription, an event not examined in this study. However, as p15 and p27Kip1 tend to coordinately inhibit Cdk4 and Cdk2 activities, their relative effects on in vivo anergy still need to be examined. The inhibitor p27Kip1 also inhibits Cdk4–cylin D complexes needed for G1 progression. Although not examined here, they may constitute a logical alternative pathway.
Another issue concerns the central function of Smad3 in anergy. Is its involvement limited to direct effects on the cell cycle? Challenging that idea is the finding that Smad proteins can inhibit Il2 transcription by binding to the Il2 promoter8. Consistent with that result, the Smad3 mutant that is nonphosphorylated suppressed IL-2 production even in normal priming conditions. Although CD28 and/or IL-2 influence Smad3 function by means of their downregulation of p27Kip1, Smad3 can directly inhibit IL-2 production. That finding emphasizes the complexity of mediator function in cell cycle arrest. In that vein, in certain cases, the ability to overcome the G1 cell cycle block for proliferation is insufficient to restore antigen responsiveness9,
10. Those studies indicate the requirement for another rapamycin-sensitive, IL-2-dependent signal in the reversal of clonal anergy that is distinct from the signal that drives proliferation. Given that the Smad3-knockdown DO11.10 p27 mutant allowed for antigen responsiveness, the 'derepressed' effects of Smad3 on transcription may provide that additional signal. Smad family members repress or activate many genes. In keeping with that complexity, although T cells from Smad3-deficient mice produce more IL-2, the mice do not have an autoimmune phenotype11.
Another key issue concerns the possible connection between p27Kip1-Smad3 and other 'upstream' mediators of anergy. The simplest scenario would be that defects in membrane adaptor Lat and Ras signaling fail to downregulate p27Kip1 and/or Smad3 function. That idea is possible, as Smad proteins function as 'nodes' in integrating Ras-Erk signaling. In contrast, despite the fact that p27Kip1 and Smad3 are ubiquitinated, it is unlikely that the autoimmunity of mice deficient in Cbl-b, GRAIL and Itch is due to targeting of those components by E3 ligase. The loss of E3 ligases would increase p27Kip1 and Smad3, the very factors that inhibit cell cycle progression. Another scenario would be the downregulation of 'upstream' components, such as Vav1 and phosphatidylinositol-3-OH kinase, that are needed for the p27Kip1 and Smad3 pathway or, more likely, that there are additional pathways that 'tip the balance' between alterations in anergy induction and the development of autoimmunity. Facilitated cell cycling is necessary but insufficient for the onset of autoimmunity.
Finally, the involvement of Smad indicates a common possible connection between anergy induced by the absence of costimulation and the suppressive cytokine TGF- . Smad2 and Smad4 (and p15) are prominent in TGF--induced suppression, and Smad3 is essential for the TGF- suppression of IL-2 production and T cell receptor–induced proliferation11. If antigen receptor–induced and TGF--induced anergy share common mediators, the targeting of Smad proteins in immunotherapy could have particularly potent and beneficial effects. Similarly, p27Kip1 and Smad3 seem to serve as a 'checkpoint' in the development of leukemia. The dual loss of p27Kip1 and Smad3 expression promotes T cell leukemogenesis in mice, whereas p16CdkN2a and p15 are often inactivated in T cell acute lymphoblastic leukemia12. It would be an exciting prospect if the same inhibitory mechanisms control anergy induction and tumor progression.
REFERENCES
-
Schwartz, R.H.
Annu. Rev. Immunol. 21, 305–334 (2003). | Article | PubMed | ISI | ChemPort |
-
Li, L.
,
Iwamoto, Y.
,
Berezovskaya, A.
&
Boussiotis, V.A.
Nat. Immunol. 7, 1157–1165 (2006). | Article |
-
Liu, Y.C.
Annu. Rev. Immunol. 22, 81–127 (2004). | Article | PubMed | ISI | ChemPort |
-
Powell, J.D.
,
Lerner, C.G.
&
Schwartz, R.H.
J. Immunol. 162, 2775–2784 (1999). | PubMed | ISI | ChemPort |
-
Boussiotis, V.A.
et al. Nat. Med. 6, 290–297 (2000). | Article | PubMed | ISI | ChemPort |
-
Sherr, C.J.
Science 274, 1672–1677 (1996). | Article | PubMed | ISI | ChemPort |
-
Rowell, E.A.
et al. J. Immunol. 175, 3359–3368 (2005). | PubMed | ChemPort |
-
Tzachanis, D.
et al. Nat. Immunol. 2, 1174–1182 (2001). | Article | PubMed | ISI | ChemPort |
-
Appleman, L.J.
,
Berezovskaya, A.
,
Grass, I.
&
Boussiotis, V.A.
J. Immunol. 164, 144–151 (2000). | PubMed | ISI | ChemPort |
-
Colombetti, S.
,
Benigni, F.
,
Basso, V.
&
Mondino, A.
J. Immunol. 169, 6178–6186 (2002). | PubMed | ChemPort |
-
McKarns, S.C.
,
Schwartz, R.H.
&
Kaminski, N.E.
J. Immunol. 172, 4275–4284 (2004). | PubMed | ChemPort |
-
Wolfraim, L.A.
et al. N. Engl. J. Med. 351, 552–559 (2004). | Article | PubMed | ISI | ChemPort |
|
