NF-κB as a Frequent Target for Immunosuppressive and Anti-Inflammatory Molecules*

Any immune or inflammatory reaction requires the de novo synthesis of special proteins that are usually not required for a day-to-day life of cells. These include cell adhesion molecules allowing transmigration of leukocytes and lymphocytes into inflammed tissue, chemokines attracting macrophages, and inflammatory cytokines that serve to amplify and spread the primary pathogenic signal. The predominant mechanism by which these proteins are newly synthesized involves an inducible transcriptional initiation of their respective genes. This is governed by preexisting transcription factors that, when activated, bind to regulatory regions of such genes and initiate a program of gene expression. Primary signals for this event can, for instance, be components of bacterial cell walls, such as lipopolysaccharide (LPS). The inflammatory cytokines TNF and IL-1 are then newly synthesized and trigger the same genetic program as LPS in cells that have never encountered the primary pathogenic signal. Eventually, this can lead to an avalanche of cytokines causing severe systemic effects, including toxic shock syndromes. Apart from these acute reactions, aberrant expression of inflammatory cytokines also plays a major causative role in chronic inflammatory autoimmune diseases such as rheumatoid arthritis (reviewed in Feldmann et al., 1996) and multiple sclerosis (reviewed in Steinman, 1996).

The transcription factor NF-κB has been recognized as a major regulator of pathogen- and inflammatory cytokine-inducible gene regulation (reviewed in Baeuerle and Henkel, 1994). This is evident, on one hand, when the endogenous and exogenous conditions that can trigger the activation of NF-κB are considered (Table I, Table II). On the other hand, it becomes apparent if the functions of genes are considered that were reported to be inducibly regulated by NF-κB binding motifs in promoters and enhancers (Table III). It is important to point out that not every stimulus will activate every target gene in a given cell type. On the contrary, there are pronounced cell type-specific differences with respect to inducing conditions that relate to the expression of receptors and other genetic factors. At the level of DNA, NF-κB synergizes with many additional transcription factors that may determine cell type-specific expression of genes and the input from other signal transduction pathways in their regulation. NF-κB usually plays the role of a powerful genetic switch that—although necessary—is rarely sufficient for inducible expression of a particular gene. Through the synergistic interaction of NF-κB with other factors in gene regulation a great degree of specificity is achieved without limiting the unique potential of NF-κB to coordinately induce a broad antipathogenic genomic response.

A hallmark of NF-κB is that it is kept in an inactive form in the cytoplasm waiting to be activated by a multitude of mainly adverse stimuli (Baeuerle and Baltimore, 1988a). The latency is achieved by binding of an inhibitory protein, called I κB, to a potentially DNA-binding dimer of NF-κB subunits (Baeuerle and Baltimore, 1988b; reviewed in Beg and Baldwin, 1993). Both NF-κB subunits and IκB proteins belong to larger protein families that constitute a system that currently consists of 10 known members (reviewed in Verma et al., 1995). Among the five NF-κB subunits, p65/RelA plays an outstanding role as a potent transcriptional activator that is most frequently found in inducible complexes of various immune and nonimmune cell types. c-Rel and RelB are likewise transcriptional activators that seem to have more prominent roles in immune cells. p50 and p52 have little or no transactivating capacity but form heterodimers of high DNA-binding affinity with the other three subunits. All NF-κB subunits share a 300-amino acid homology region (the Rel domain) that is sufficient for DNA binding and dimerization. The transactivating domains reside in the unique C-terminal portions of RelA, c-Rel, and RelB.

NF-κB dimers can potentially associate with five different IκB proteins, two of which represent precursor molecules for p50 and p52, called p105 and p100. The major players in inducible NF-κB activation are IκB-α, -β, and -ε, of which IκB-α is the best studied inhibitor. A structural characteristic of all IκB proteins is a cluster of five to seven ankyrin repeat motifs that, as an entity, are required for binding to the Rel homology domain of NF-κB subunits. In response to various signals, for instance, TNF, IL-1, and phorbol ester, IκB-α undergoes phosphorylation on serines 32 and 36 within its regulatory N terminus. This does not release IκB-α from NF-κB but turns the inhibitor into a substrate for ubiquitin-conjugating enzymes. Once ubiquitin is conjugated to lysines 21 and 22, IκB-α is rapidly degraded by the proteasome (reviewed in Baldwin, 1996). The liberated NF-κB is then transported into the nucleus and will initiate transcription by binding to regulatory κB motifs in target genes.

Although these downstream events are understood in some detail, it is still not known how TNF and IL-1 receptors trigger IκB phosphorylation. Neither has the IκB-α kinase been molecularly identified nor are the proteins and messengers known that regulate IκB kinase(s) (or counteracting IκB phosphatases). More progress has been made recently in understanding TNF and IL-1 signaling at the plasma membrane. In the case of TNF receptors, a number of adaptor molecules were identified, called TRADD and TRAFs (Hsu et al., 1995, Rothe et al., 1994, Hsu et al., 1996a), that become recruited to the receptor upon ligand binding and are necessary for NF-κB activation. Early signaling events in both TNF and IL-1 action appear to involve protein kinases, called RIP (Hsu et al., 1996b) and IRAK (Cao et al., 1996), that become associated with the respective receptors but do not directly phosphorylate IκB-α.

A role of NF-κB as an important genetic switch that rapidly turns on proinflammatory gene expression in response to mostly pathogenic conditions was proposed early on (Baeuerle and Baltimore, 1991). However, this assumption was mainly based on studies describing conditions that activate NF-κB in cell culture experiments and genes harboring exacting κB motifs in their enhancers. Only recently has more compelling evidence for the role of NF-κB in autoimmune and inflammatory diseases been obtained. One kind of experiment explored the state of NF-κB activation in a number of human diseases and animal disease models. Aberrant NF-κB activation was observed in smooth muscle cells and macrophages of atherosclerotic lesions (Brand et al., 1996), in synovial tissue of rheumatoid arthritis patients (Marok et al., 1996), in peripheral microglia of mice suffering from experimental autoimmune encephalomyelitis (Kaltschmidt et al., 1994), and in brain neurons adjacent to plaques of Alzheimer’s disease patients (Kaltschmidt et al., 1997). These studies were aided by immunocytochemical methods using antibodies specific for the nuclear form of the p65/RelA subunit that has been released from IκB (Kaltschmidt et al., 1995).

Direct functional evidence for a role of NF-κB in immune regulation has come from recent genetic studies using transgenic and gene knockout mice (reviewed in Baeuerle and Baltimore, 1996). Various forms and degrees of immunodeficiencies are observed when distinct members of the NF-κB family are inactivated by gene disruption. The full importance of the system may become apparent only after single gene knockout mice are crossed to generate double and multiple knockouts because individual NF-κB subunits are in part functionally redundant. However, already at this stage, the various mouse models strongly support a key role of the NF-κB/IκB system in an impressive number of immune regulatory processes.

The effect of adenovirus-mediated expression of IκB-α in endothelial cells supports the notion that the NF-κB transcription factor is a key regulator of proinflammatory genes (Wrighton et al., 1996). By this approach, activation of all known dimer combinations of NF-κB was prevented in a specific fashion. IκB overexpression resulted in a collective inhibition of LPS-induced NF-κB-regulated gene expression, including that of the VCAM-1, IL-1, IL-6, IL-8, and tissue factor genes. As a result, leukocyte adhesion to endothelial cells was fully suppressed. This provides a “proof of principle” that NF-κB activation is an attractive target for anti-inflammatory drugs. In the following sections, we will summarize studies showing that numerous molecules possessing immunosuppressive and anti-inflammatory activity are now being recognized to target mechanisms controlling the activity of transcription factor NF-κB, which may well explain, at least partially, their pharmacological behavior.