[Frontiers in Bioscience 2, d49-60, February 15, 1997]

Table of Conents
 Previous Section   Next Section


Javier Gómez, David García-Domingo, Carlos Martínez-A.1 and Angelita Rebollo

Department of Immunology and Oncology, Centro Nacional de Biotecnología, Campus de Cantoblanco, E-28049 Madrid, Spain

Received 1/21/97; Accepted 1/30/97; On-line 2/15/97

6. A dual role for NF-kappaB gene family members in programmed cell death

A functional distinction has been suggested with regard to the involvement of different NF-kappaB monomers in apoptosis regulation; c-Rel promotes cell death, whereas RelA protects from it. These dual responses do not derive simply from the fact that these proteins are potential dimerization partners of a single complex. It has been proposed that a difference in the kappaB-specific motifs exists by which anti-apoptotic genes bind RelA complexes selectively, whereas c-Rel-containing dimers specifically bind to pro-apoptotic genes. Alternatively, c-Rel overexpression could prevent formation of RelA dimers required for the activation of anti-apoptotic genes. Even if any of these hypotheses is proven correct, none can account for the recent data that rule out such a simple scheme. First, apoptosis-promoting activity has also been reported for RelA. Thus, serum deprivation-induced apoptosis of 293 cells is characterized by an increase in RelA-containing NF-kappaB activity (89). Both effects, cell death and NF-kappaB activation, may be prevented by Bcl-2 or a dominant negative RelA mutant. c-Rel prevents apoptosis induced by either IgM crosslinking or a protease inhibitor in the WEHI 231 immature B cell lymphoma line, since microinjection of anti-c-Rel antibodies or an IkappaBalpha-GST fusion protein promotes cell death (90). Studies of normal and transformed murine B cells suggest that reduction in NF-kappaB DNA binding activity as a consequence of surface IgM ligation may be a determining event for the onset of apoptosis.

The pro- or anti-apoptotic properties of different NF-kappaB subunits is far from clear, and it is possible that a dual regulation may be exerted by the same proteins. Thus, studies of cellular response control by signaling mediators have led to the identification of molecules with multifunctional capabilities. Examples include proteins that act as cell cycle and cell death regulators, such as the tumor suppressor p53, the E2F transcription factors, the protooncogenes bcl-2 and c-myc (91-95). Other examples include the small G proteins of the Ras superfamily, of which both Rho-like and Ras-like proteins have been linked to cell proliferation (96-100) and to a dual control of apoptosis, either as suppressors or promoters of cell death (100-103; Gómez et al., submitted). Apoptosis is now envisioned as one of the options to be selected by the cell during cell cycle progression, and the switching from proliferation to cell death may be determined by a defective proliferative signal. If NF-kappaB activity is functionally linked to such mediators, it would not be surprising that this family of transcription factors would also be involved in driving diverse or even opposite cellular responses. There is evidence linking NF-kappaB activity with the functions of mediators implicated in cell proliferation. NF-kappaB activates p53, c-myc and c-H-ras genes. In addition, Ras has been proposed as the initiator of a signaling pathway that induces NF-kappaB through the sequential activation of the atypical protein kinase C (PKC) z isoform and a putative IkappaB kinase (43). Other reports have localized NF-kappaB in a pathway led by the Rho family proteins RhoA, Rac1 and Cdc42, that results in NF-kappaB activation through IkappaBalpha depletion. The same set of experiments showed involvement of RhoA and Cdc42, but not Rac1, in TNF-alpha-induced NF-kappaB activity (83). Relationships between NF-kappaB and several molecules involved in apoptosis control have been demonstrated. Furthermore, NF-kappaB activity increases following certain types of stimulation often associated with the onset of cell death in several systems, such as TNF-alpha, ultraviolet light, H2O2, calcium ionophores, phorbol esters or ceramides. Although in the case of TNF-alpha, as mentioned above, the apoptotic and NF-kappaB activation pathways seem to be divergent, the possible relevance of NF-kappaB in the cell death signals putatively triggered by other stimuli remains to be clarified. Finally, Fas receptor stimulation is followed by NF-kappaB DNA binding activity in some, but not all, cell types. In a cell line in which Fas triggers NF-kappaB activation, no apoptosis was detected when Fas ligation was accompanied by inhibition of NF-kappaB (88), suggesting that, in contrast to TNF-alpha-induced signaling, Fas-mediated cell death does not rely on NF-kappaB activity.

A few clinical implications of NF-kappaB function in apoptosis regulation are worth mentioning. TNF-alpha has been used as a therapeutic agent to trigger the killing of tumor and infected cells. However, the early expectations were not fulfilled, and transformed cells are often resistant to TNF-alpha-induced apoptosis. In the case of viral infection, this may be partially explained by the fact that some viruses express gene products promoting cellular NF-kappaB activity, making infected cells resistant to TNF-alpha-induced apoptosis. Moreover, NF-kappaB activation may also result in proviral transactivation, as is the case of HIV (104). There again, the dual role of NF-kappaB might cause the double effect of promoting both viral expression and the survival of infected cells. The pro-apoptotic effect of TNF-alpha, useful for the therapeutic removal of infected or tumorigenic cells, may therefore be productively enhanced by inhibition of NF-kappaB, either through the administration of suppressor drugs such as glucocorticoids, antioxidants (105) or Cu2+ (106), or through the genetic delivery of IkappaB proteins.