Novel regulation of CD80 and CD86 signaling by Notch1 in dendritic cells
MetadataShow full item record
Modulation of T cell immune responses by dendritic cells (DC) is importantly dependent upon the engagement of CD28 on T cells by CD80 and CD86 expressed on DC. While CD28 signaling in T cell activation has been well characterized, it has only recently been shown that CD80 and CD86, which do not have demonstrated signaling motifs in their cytoplasmic tails, are also capable of signaling to the DC. Functionally, CD80/CD86 engagement results in DC production of IL-6 that is necessary for full T cell activation. Despite this significant immunological role, how CD80 and CD86 signal remains poorly understood. We have now found that crosslinking CD80 and CD86 with the soluble chimeric fusion protein CD28-Ig activates the PI3K/Akt pathway. This results in phosphorylation and inactivation of its downstream target, the transcriptional regulator FoxO3a, which subsequently derepresses IL-6 expression. A second event downstream of Akt phosphorylation is activation of the canonical NF-κB pathway, which actively upregulates IL-6 expression. During these studies, we unexpectedly found that CD80/CD86-induced PI3K signaling is regulated by a previously unrecognized crosstalk with Notch1 signaling. This crosstalk is facilitated by Notch-mediated upregulation of the enzymatic activity of casein kinase II (CK II), which phosphorylates and inactivates PTEN - a central negative regulator of PI3K signaling. This results in full activation of PI3K and its downstream signaling upon crosslinking CD80/CD86. In contrast to T cell activation, ligation of CD80 and CD86 on DC by the CD28 family member CTLA4, expressed on activated T cells or T regulatory cells, mediates immune suppression. It has recently been shown that CTLA4 (or the soluble chimeric receptor CTLA4-Ig) can induce signaling downstream of CD80 and CD86, resulting in the production of indoleamine 2, 3 dioxygenase (IDO) by DC. IDO catabolizes the essential amino acid tryptophan in the microenvironment to L-kynurenine, and this depletion blocks T cell activation and creates a localized immunosuppressive environment. As for IL-6 production, the molecular events involved in induction of IDO expression by activated CD80 and CD86 are incompletely characterized. We have found that IDO production by DC is regulated distinctly at early and late time points. Shortly after ligation of CD80 and CD86, the PI3K-Akt-NF-κB pathway induces IL-6 production. Additionally, the PI3K pathway is regulated by Notch1 signaling via casein kinase II similar to our observations in CD28-Ig mediated induction of DC IL-6. At late time points, activation of CD80 and CD86 causes accumulation of FoxO3a (which we propose is dependent upon NF-κB binding to the FoxO3a promoter), resulting in the sustained production of IDO. Unexpectedly, we have found that production of IL-6 versus IDO by DC is not dependent upon whether it is CD28-Ig or CTLA4-Ig that ligates CD80 and CD86, as previously proposed. Rather, it appears to be dependent upon the expression of SOCS3 by DC, which plays the determinative role in inhibiting IDO production downstream of CD80 and CD86 activation. In summary, here we have shown that ligation of CD80 and CD86 by CD28-Ig or CTLA4-Ig can induce activation of the PI3K-Akt pathway downstream of CD80 and CD86 in DC. This induces IL-6 as well as IDO production (which is critically dependent upon exogenous addition of IFN-γ) at distinct time points. Additionally, we have made the novel observation that Notch signaling regulates IL-6 and IDO production by modulating the activation of the PI3K pathway via phosphorylation of PTEN by CK II. Altogether, our data suggest a unified model of CD80/CD86 activation in DC. Upon T cell interaction with DC, CD80 and CD86 engage CD28, and provide a costimulatory signal to the T cell. In the presence of concurrent Notch signaling, which is initiated upon ligation of Notch1 on DC by Jagged2 on T cells, PI3K signaling is activated in the DC. Activation of PI3K results in downstream activation of Akt and the canonical NF-κB signaling, which in turn phosphorylates/inactivates FoxO3a. This inactivation derepresses IL-6 production and suppresses Foxo3a-induced IDO production. In the first 24 hours, IL-6 production driven by NF-κB results in full T cell activation. At late time points, NF-êB counters the initial IL-6 induction by increasing FOxO3a expression that induces IDO production/activity, which is dependent upon IFN-γ production by the fully activated T cell. Our model suggests that T cell activation by DC is self-restraining and is in fact temporally regulated.