The Control of Apoptosis by NF-kB/Rel Transcription Factors in Immunity and Cancer

I. - Background Information
            In addition to coordinating immune and inflammatory responses, NF-kB/Rel transcription factors control cell survival. Activation of NF-kB antagonizes apoptosis or programmed cell death (PCD) induced by TNFa. The anti-apoptotic activity of NF-kB is also crucial to oncogenesis and cancer chemo- and radio-resistance, and indeed, serves a wide range of biological processes, including B lymphopoiesis and B- and T-cell costimulation. This activity involves induction of protective genes; however, its bases remain poorly understood. During the past few years, the efforts in my laboratory have focused on the identification and characterization of these protective genes. This is of interest since, in addition to gaining insights into mechanisms of immune activation and oncogenesis, NF-kB-controlled anti-apoptotic factors might provide new molecular targets for the treatment of chronic inflammatory conditions (e.g. rheumatoid arthritis and inflammatory bowel disease), as well as certain malignancies such as Hodgkin's lymphoma and multiple myeloma. NF-kB blockers are already part of the therapeutic regimen of these diseases. However, these blockers (e.g. glucocorticoids and proteasome inhibitors) can only achieve partial inhibition of NF-kB and exhibit considerable side effects, which limit their use in humans. Thus, a better therapeutic approach would be to block critical anti-apoptotic targets, rather than NF-kB itself.
            To identify these targets we devised a functional screening method known as the "death trap". This method provides substantial advantages over other methods also aimed at identifying these genes, as it ensures that isolated genes have anti-apoptotic functions. Foremost, it allows a functional assessment of candidate genes without any preconceived information about their sequences, and so represents an unbiased approach.
            Below, I summarize recent research projects, beginning with the generation of libraries enriched in anti-apoptotic cDNAs.

II. - Specific Research Projects

A - Identification and characterization of novel anti-apoptotic targets of NF-kB.

1) Generation and characterization of libraries enriched in anti-apoptotic cDNAs.
             To understand mechanisms by which TNF receptors (TNF-Rs) induce apoptosis and the basis for the protective activity of NF-kB, we have employed the "death trap" screen in NF-kB null cells. The approach consists of the functional complementation of the defects present in NF-kB/RelA null fibroblasts, which rapidly undergo apoptosis when treated with TNFa. Complementation is achieved by transfection of cDNA libraries derived from TNFa-activated, wild-type cells - which are instead refractory to cytokine-induced killing. Newly-generated libraries were transfected into relA-/- fibroblasts, apoptosis was induced with TNFa, and plasmids were recovered from surviving cells. After 4 cycles of selection, ~35% of the library cDNAs protected RelA null cells from killing by TNFa. Known inhibitors of TNF-R-triggered apoptosis, including RelA, cFLIP, and dominant negative (DN) forms of FADD, were enriched during selection, with RelA representing 3.6% of the newly-isolated library (~450-fold enrichment; De Smaele et al., Nature 2001).
             To analyze the selected library, we then adopted two distinct strategies: i) the testing of randomly chosen plasmids in cytoprotective assays; ii) the systematic examination of the library by the use of microarray technology. Both methods - described below in sections 2 and 3, respectively - were carried out successfully and lead to the discovery of novel pro-survival pathways controlled by NF-kB. They also resulted in the identification of Gadd45b and FHC as crucial mediators of two such pathways - namely the suppression of JNK signaling and oxygen radical accumulation, respectively, by NF-kB.

2) Analysis of cDNAs isolated randomly from the selected library.
             Among the plasmids that were chosen randomly from the selected library, 67 exhibited anti-apoptotic activity in NF-kB null cells, and 4 of these turned out to be under transcriptional control of NF-kB. They encoded a cell cycle regulator, 2 unknown products, and Gadd45b/Myd118 (described in detail below). 2 other genes isolated from this screen, but that are not targets of NF-kB, also appear of interest, since one has been reported to play a role in tumorigenesis and the other one in metastasis formation. None of these genes had been previously shown to have pro-survival activity, and thus, our findings may provide important new insights into mechanisms of oncogenesis.
By combining a genetic and a biochemical approach, we are now attempting to define physiologic functions and protective mechanisms by the NF-kB-regulated genes isolated from the above screen. Furthermore, we plan to explore possible roles of these genes in tumors that depend on NF-kB for their survival, and to confirm our findings in vivo, by exploiting transgenesis and gene targeting technologies in mice. Finally, we intend to take full advantage of our libraries and pursue the identification and characterization of additional anti-apoptotic targets of NF-kB.

3) Systematic analysis of the enriched library by the use of microarray technology.
             In parallel with the testing of random cDNAs, we have systematically analyzed our libraries by exploiting the gene array technology. Original and enriched libraries were transcribed in vitro to generate cRNA probes, which were then used to interrogate pairs of microchips (Affimetrix), each containing ~7,000 oligonucleotide sets derived, for the most part, from known genes. This system enabled us to examine the behavior of virtually all known genes during library selection, pointing out those capable of modulating apoptosis. Thus, herein microchips provided information relative to gene function, rather than gene expression. Indeed, this screening strategy combines the advantages of the "death trap" method, which yields functional information, and of microchip technology, which allows the simultaneous evaluation of a large number of genes. This method also enabled: i) clustering of isolated genes based upon their function, thereby unmasking potential cellular defects caused by loss of NF-kB; and ii) ranking isolated genes according to their "fold-enrichment" (FE) score, thereby providing a semi-quantitative indication of their protective efficacy.
             Using this method, we have identified 71 putative protective cDNAs. Validating our approach, RelA and DN-FADD ranked 7th and 1st, respectively, in terms of FE. Among the genes that were confirmed, so far, to have anti-apoptotic activity in relA-/- cells, 2 turned out to be under NF-kB control.

B - Gadd45b mediates the suppression of JNK signaling and apoptosis by NF-kB.

1) Identification of Gadd45b as an anti-apoptotic target of NF-kB.
             By using the above screen, we have identified Gadd45b/Myd118 - a member of the Gadd45 family of inducible factors - as a pivotal mediator of the NF-kB protective activity against TNFa cytotoxicity (De Smaele et al., Nature 2001). gadd45b is upregulated rapidly by the cytokine through a mechanism that requires NF-kB, is essential to antagonize TNFa-induced killing, and blocks PCD in NF-kB null cells. The protective activity of Gadd45B against TNFa-induced PCD involves the suppression of the c-Jun-N-terminal kinase (JNK) cascade, a pathway usually associated with cell death, and this suppression is central to the control of apoptosis by NF-kB. Thus, the NF-kB and JNK pathways - almost invariably co-activated by cytokine and stress - are intimately linked, and this link involves transcriptional up-regulation of Gadd45b by NF-kB. These findings define a novel protective mechanism that is mediated by NF-kB and establish a role for the persistent activation of JNK in the apoptotic response to TNFa. Of interest, Gadd45b may also contribute to NF-kB-dependent cancer chemoresistance, and thereby represent a potential therapeutic target (Franzoso et al., Cell Death Differ. 2003 ).

2) Regulation of the gadd45B promoter by NF-kB.
             The above study showed that NF-kB is required for gadd45b induction by stress and cytokines. Using an inducible system, we found that NF-kB is also sufficient to activate gadd45b expression. To determine the basis for the NF-kB control of this expression, we examined the gadd45b promoter. This analysis revealed the presence of 3 putative NF-kB-binding elements in the proximal region of this promoter. All 3 elements are evolutionarily conserved, suggesting that they play an important role in controlling gadd45b transcription. Indeed, each element has the ability to bind to NF-kB complexes, in vitro , and is required for optimal promoter transactivation, in vivo (Jin et al., DNA Cell Biol. 2002). These data establish the direct participation of NF-kB in the regulation of gadd45b, thereby providing mechanistic insights into the control of apoptosis by NF-kB.

3) Gadd45b mediates the protective effects of CD40 co-stimulation against Fas-induced apoptosis.
             In the immune system, a key homeostatic mechanism is the induction of apoptosis by Fas (CD95). In germinal center B cells, this apoptosis is counteracted by pro-survival signaling induced by CD40. Interaction between this receptor and its ligand on activated T cells controls various aspects of the T-dependent, humoral immune response. Pro-survival signaling by CD40 also plays a critical role in the pathogenesis of certain malignancies, including Hodgkin's Lymphoma. In a recent study, we found that Gadd45b mediates the protective effects of CD40 against Fas-induced apoptosis (Zazzeroni et al., Blood 2003). In B lymphocytes, Gadd45b is induced rapidly by CD40 through a mechanism that requires NF-kB, and this induction effectively suppresses the apoptotic response to Fas ligation. Importantly, Gadd45b up-regulation by CD40 precedes Fas-induced caspase activation, as well as up-regulation of other NF-kB-controlled inhibitors of apoptosis such as Bcl-xL and c-FLIPL. In the presence of Gadd45b, the Fas-induced apoptotic cascade is halted at mitochondria. However, in contrast to Bcl-xL, Gadd45b appears to block Fas cytotoxicity herein by suppressing a novel mitochondria-targeting mechanism activated by this receptor. These findings identify Gadd45b as a physiologic effector of CD40 pro-survival signaling and unravel a novel mechanism participating in Fas-induced PCD in B cells.

4) Gadd45b mediates the NF-kB suppression of JNK signaling by targeting MKK7 .
             Recent efforts in my laboratory focused on uncovering the precise mechanism(s) by which Gadd45b, and thereby NF-kB, blunt JNK signaling. We have now identified MKK7/JNKK2 - a specific and essential activator of JNK - as a target of Gadd45b, and in fact, of NF-kB itself (Papa et al., Nature Cell Biol. 2003). Gadd45b associates directly and tightly with MKK7, thereby inhibiting its catalytic activity, and this inhibition is central to the Gadd45b-mediated suppression of TNFa-induced apoptosis. Importantly, this targeting of MKK7 is critical for the suppression of apoptosis by NF-kB. Our data support a model whereby NF-kB activation induces expression of Gadd45b, which in turn inhibits MKK7, leading to the suppression of JNK signaling, and ultimately, apoptosis triggered by TNFa. These findings identify a crucial molecular link between the NF-kB and JNK pathways. They also establish a basis for the NF-kB control of JNK activation and identify the Gadd45b-MKK7 interaction as a potential target for anti-inflammatory and anti-cancer therapy.

III. - Future Research Interests
            Future efforts in my laboratory will continue to focus on a better understanding of the roles of NF-kB/Rel transcription factors in apoptosis, immunity, and cancer.