[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


Cells respond to intra- and extracellular signals by turning specific genes on or off and by modulating the extent of active gene transcription. Switching gene expression on and off is the responsibility of transcription factors, which operate singly or in association with other proteins. Transcription factors are usually organized in families, one of which is the NF-kappaB homo- or heterodimers. These proteins share the common property of being sequestered in the cytoplasm by the specific IkappaB proteins.

NF-kappaB was originally characterized as a lymphoid-specific protein that interacts with the immunoglobulin k light chain gene enhancer sequences (1). Five mammalian NF-kappaB/Rel proteins have been cloned and characterized, including c-Rel, NF-kappaB1 (p50/p105), NF-kappaB2 (p52/p100), RelA/p65, and RelB (2-9) (Fig 1A). These proteins have common domains, such as the Rel homology domain, which functions in DNA binding, dimerization, interaction with IkappaB. This domain encompasses the nuclear localization signals (NLS). NF-kappaB1 and NF-kappaB2 contain multiple copies of the ankyrin repeat at their C-termini. Processing of the precursors p105 and p100, which is signal-dependent, leads to the production of the p50 and p52 subunits. The cytoplasmic precursors p105 and p100 perform inhibitory functions by preventing their respective processed products binding to kappaB sites (10-13). NF-kappaB processing requires ATP and is mediated by a ubiquitin-dependent proteasome degradation pathway (14). Other members of the NF-kappaB/Rel protein family are not generated from precursor proteins and have a carboxy-terminal transactivation domain (15). The kappaB sites are present in the regulatory regions of genes involved in immune (IL-2, IL-2Ralpha) and inflammatory responses (IL-1, IL-6, TNFalpha, TNFß), as well as in genes of viruses, NF-kappaB/Rel members, IkappaB members, growth control proteins (p53, c-myc, Ras) and adhesion molecules (15-19). NF-kappaB/Rel protein functions are involved in cell transformation, tumor growth, apoptosis, embryonic liver development and functional differentiation of immune cells (20-23).

Figure 1.

A. Schematic representation of some members of the Rel/NF-kappaB and IkappaB families.

B. NF-kappaB regulation. In response to external signals, phosphorylated IkappaB binds to NF-kappaB factor. This phosphorylation targets IkappaB for ubiquitination. NF-kappaB bound to ubiquitinated and phosphorylated IkappaB is cleaved by a proteasome complex. Free NF-kappaB translocates to the nucleus and induces expression of certain genes.

In most cell types, NF-kappaB is associated with the inhibitor IkappaB in the cytoplasm. Following activation, a small fraction of NF-kappaB dissociates from the inhibitor and is translocated to the nucleus (24). The mechanism of NF-kappaB nuclear translocation following IkappaB degradation is speculated to include phosphorylation or dephosphorylation events that target IkappaB for ubiquitination and subsequent proteasome-dependent degradation (25, 26). There are at least nine mammalian IkappaB molecules with distinct and overlapping inhibitory specificities, including IkappaBalpha, IkappaBß, IkappaBgamma, IkappaBepsilon, Bcl-3, p105, p100, A238L and IkappaBR (12, 13, 27-31) (Fig 1A). The IkappaB family proteins share conserved motifs referred to as ankyrin repeats. These motifs are required for association of IkappaB proteins with NF-kappaB/Rel proteins (15, 29). In addition, the IkappaB proteins contain a acidic carboxy-terminus region (30) and a carboxy-terminus domain containing a Pro, Glu/Asp, Ser and Thr-rich PEST sequence, which are implicated in regulating protein half-life (32). Deletion of this sequence partially protects IkappaBa from degradation (33, 34). Phosphorylation of IkappaB is observed in many cell types following stimulation (35, 36). The kinase or kinases that phosphorylate IkappaBalpha upon stimulation recognize the same serine-32 and -36 residues. Recent studies demonstrate that the basal phosphorylation of IkappaBalpha occurs at the carboxy-terminal casein kinase II sites in the PEST region (37).

Prior to NF-kappaB activation, IkappaB undergoes complete degradation following stimulation, (12, 35, 38). IkappaB degradation is an efficient process that can be inhibited by serine protease inhibitors, suggesting that it is an obligatory step in NF-kappaB activation (30). Thus, the sequence of events leading to NF-kappaB activation may require phosphorylation of IkappaBalpha at serine residues, followed by phosphorylation-dependent multi-ubiquitination at lysine residues, degradation of IkappaBalpha by a ubiquitin-dependent proteasome and, finally, the release of free NF-kappaB transcription factor (Fig 1B).

Some kinases have been implicated in NF-kappaB activation. The best characterized kinase is probably the double-stranded RNA-activated kinase (PKR), which phosphorylates IkappaBalpha in vitro (39). Raf-1 has also been proposed to target IkappaB (40, 41), and experiments using dominant negative mutants have implicated zetaPKC as a regulator of NF-kappaB activation (42). Finally, epsilonPKC has also been suggested to induce NF-kappaB activation (42, 43). Several inducers trigger NF-kappaB activation, including TNF, IL-1, lipopolysaccharide (LPS), phorbol esters, okadaic acid, serum growth factors and nitric oxid (NO) (2, 3, 5, 15, 44, 45).

Phosphatases probably play an important role in NF-kappaB activation, either by regulating the kinase pathways or by direct dephosphorylation of IkappaB. Based on inhibition by cyclosporin A and transfection studies, the phosphatase calcineurin appears to be involved in NF-kappaB activation in T cells (46). Calcineurin increases NF-kappaB activity by increasing IkappaB phosphorylation and degradation, leading to an increase in the level of active nuclear NF-kappaB (47). Calcineurin is also involved in p105/NF-kappaB1 induction and in the decrease in p50 dimer levels in the nuclei of activated T cells (48). The result of these calcineurin-mediated processes would be to increase the ratio of active to inactive complexes, and thus to potentiate transactivation from the NF-kappaB site (48, 49). The serine/threonine phosphatase inhibitors PP1 and PP2A also activate NF-kappaB, suggesting the implication of a phosphatase in regulating aspects of the pathway.

Once in the nucleus, the NF-kappaB proteins not only bind to DNA, but also interact with many other proteins that may also participate in NF-kappaB regulation. c-Rel binds directly to the TATA binding protein and TFIIB (50, 51), and to the RNA polymerase II subunits. In some cases, a high-mobility-group protein [HMG-I(Y)] is required as coactivator for transcriptional activation (52). There are also activator proteins, such as the helix-loop-helix (HLH) family proteins or steroid receptors, which can interact with the NF-kappaB proteins to activate, repress, enhance or decrease transcription of various genes (15). Thus, NF-kappaB proteins, once in the nucleus, are subject to additional levels of control.