Interplay between the NF-κB and forkhead transcription factors

The NF-κB family of transcription factors plays central roles in cell proliferation and activation, and is critically required for multiple and diverse biological processes from developmental biology to carcinogenesis to inflammation (reviewed in refs^{1, 2}). Interestingly, their activity must be tightly regulated – indeed, actively inhibited – to ensure proper biological functions. The most well-described dampeners of the NF-κB system are the inhibitors of NF-κB (IκB) proteins, whose complex retains the NF-κB transcription factors in a transcriptionally inactive form in the cytoplasm (Figure 1). In the classically described pathway of NF-κB activation, inflammatory stimuli such as the cytokine TNF-α activate the IκB kinase (IKK) complex, which phosphorylates the IκB proteins, rendering them susceptible to ubiquitin-mediated proteosomal degradation and allowing the NF-κB proteins themselves to translocate to the nucleus and promote target gene transcription. The absence of such inhibitory components can be detrimental to the organism: IκBα deficiency in mice, for example, results in the spontaneous development of a severe, lethal inflammatory syndrome in the neonatal period.^{3, 4} Thus, quiescence of the NF-κB pathway must be absolutely maintained, at least until the advent of environmental signals that indicate that NF-κB-mediated responses are appropriate and adaptive.

Figure 1

A working model for the interactions between forkhead (Fox) and NF-κB. (a) In resting or inactive cells, NF-κB transcription factors (e.g. p50, RELA) are sequestered in the cytoplasm by the inhibitor of NF-κB (IκB) proteins, whose transcription is at least in part dependent on forkhead transcription factors like Foxj1 and Foxo3a. (b) Activating stimuli, such as Toll-like receptor ligands, activate the NF-κB pathway by resulting in the phosphorylation and activation of the IκB kinase (IKK) complex. This kinase machinery phosphorylates both Fox and IκB proteins, resulting in their ubiquitin-mediated proteosomal degradation. The NF-κB factors are now released from their cytoplasmic tethering, translocating to the nucleus and mediating target gene transcription (based on, in part, Figure 1 of ref¹⁹)

The inhibitory IκB genes appear to be regulated, at least in part, by the NF-κB proteins themselves, comprising a feedback loop by which the inhibition versus activation of the NF-κB system is intrinsically regulated and balanced.⁵ It has become increasingly clear, however, that signaling and genetic pathways beyond the classical NF-κB components can impact upon NF-κB activation: for instance, the NF-κB proteins and/or their IκB and IKK counterparts may be post-translationally modified, for example, by phosphorylation, acetylation, and sumoylation, etc., modulating the specific activity of the NF-κB transcriptional complexes.^{6, 7, 8, 9, 10, 11} Thus, a complex network of interdependence exists between the NF-κB transcription factors and their regulators, which include not only the IκB and IKK members of the NF-κB pathway itself, but a growing list of signaling intermediates like kinases, acetylases and sumoylases. Recent studies have demonstrated that the regulators of such modulators, that is the transcription factors that regulate the expression of IκB's, IKK's, etc., further contribute to this cross-regulatory network. Such transcription factors include members of another broad and diverse transcription factor family: the forkhead (Fox, ‘winged helix’) transcription factors (Figure 1).

The forkhead genes include an ever-growing family of at least 100 genes, which have generally been studied for their critical roles in developmental biology (reviewed in ref¹²). Although the specific mechanisms which regulate their cellular activities continue to be elucidated, the transcriptional activity of at least some Fox genes are known to be post-translationally regulated: for instance, members of the FoxO subfamily, which are homologs of the C. elegans longevity gene DAF-16, play key roles in cellular responses to stress, starvation and activation by normally being transcriptionally active in the nucleus in metabolically quiescent cells (reviewed in Tran et al¹³). Activating stimuli, particularly activators of phosphoinositide-3 kinase (PI3K) pathway, lead to the activation of Akt (protein kinase B, PKB), which phosphorylates the FoxO's, rendering them susceptible to 14-3-3-mediated export from the nucleus and halting their transcriptional activity. In this way, Fox transcription factors are potentially particularly well-suited to enforce quiescence and/or antagonize activation responses in resting or metabolically stressed cells.

Studies on one particular Fox gene, Foxj1, were initiated in T cells of the lymphoid system because microarray studies indicated its deficient expression in lupus-prone, but not nonautoimmune, lymphocytes, implicating this particular member as a regulator of lymphocyte tolerance and/or activation.¹⁴ Foxj1-deficiency resulted in a multisystemic cellular autoimmunity, associated with profound T-cell hyperproliferation and hyperactivity. Since NF-κB was widely known to play a major role in the regulation of immediate-early T-cell activation, a potential interaction between Foxj1 and the NF-κB system was sought: indeed, Foxj1-deficient T cells exhibited increased and spontaneous NF-κB activation, associated with a deficiency in the IκBβ subunit, and in vitro Foxj1 was capable of increasing IκBβ expression and suppressing NF-κB activity, as judged by luciferase reporter, nuclear translocation and gelshift studies. Similar findings were observed in animals deficient for Foxo3a, which were pursued because it too was observed to have NF-κB-suppressive activities in vitro.¹⁵ Foxo3a-deficient animals analogously developed a mild lymphoproliferative and multisystemic cellular autoimmune syndrome, also associated with T-cell hyperactivity. Like Foxj1-deficient T cells, Foxo3a-deficient T cells possessed increased and spontaneous NF-κB activation, but unlike Foxj1, Foxo3a deficiency interestingly resulted in a combined deficiency in both IκBβ and IκBɛ subunits. Such findings indicate that Foxj1 and Foxo3a play critical and nonredundant roles in the antagonism of NF-κB: they are actively required in quiescent cells, and in their absence, NF-κB activity is simply unrestrained due to the functional absence of the inhibitory IκB mechanisms.

Both Foxj1 and Foxo3a appear to regulate the IκB's at the transcriptional level, since both deficiencies result in significant reductions of the mRNAs corresponding to their respectively affected IκB subunits. In fact, Foxj1 appears to be capable of regulating the IκBβ promoter, and IκBβ's mRNA expression pattern strongly correlates with that of Foxj1, suggesting that Foxj1 is in fact a direct transcriptional regulator of IκBβ.¹⁴ On the other hand, Foxo3a does not appear to regulate the promoter of either IκBβ and IκBɛ, and whereas Foxj1-deficient cells contain normal levels of Foxo3a, Foxo3a-deficient cells are deficient in Foxj1, suggesting that Foxo3a may regulate at least IκBβ via Foxj1.¹⁵ Alternative explanations remain to be tested fully, including the possibility that Fox genes themselves might regulate the NF-κB pathway post-translationally, for example, by inhibiting the degradation of the IκB components – as might be suggested by prior studies demonstrating the ability of the Akt pathway to induce NF-κB activity by regulating the degradation of IκB.¹⁶ Regardless of the precise mechanism(s), though, such observations with Foxj1 and Foxo3a alone already interestingly imply the presence of a network of Fox genes, which coordinate the inhibition of other transcription factor families like NF-κB.

Interestingly, the interactions between the Fox and NF-κB families extends in the opposite direction as well: a recent study indicates that the IKK proteins may directly interact with, phosphorylate and inhibit the activity of Foxo3a, at least in part by targeting it to ubiquitin-dependent proteosome degradation.¹⁷ In this sense, Fox appears to be treated by the NF-κB pathway like an IκB protein, and may in fact participate in the autoregulatory aspects of the NF-κB pathway (Figure 1): perhaps NF-κB transcription factors regulate the transcription of Fox genes, and/or vice versa. As most of the interactions between the Fox and NF-κB systems have been limited to only a handful of published studies, it will be of exceptional future interest to determine if such interplays with NF-κB extend to other Fox genes, like Foxj1, and also whether or not this regulatory network extends, functionally, to other transcription factor families suggested to interact with Forkhead, such as NFAT.¹⁸

‘Note in proof : A recent study indicates that the forkhead member Foxp3 may inhibit both NF-κB and NFAT activities in T cells through direct protein–protein interactions (Bettelli et al. (2005) Proc. Natl. Acad. Sci. USA 102: 5138–5143).’