Usha Ponnappan

Department of Geriatrics, Medicine, Microbiology and Immunology, University of Arkansas for Medical Sciences, and GRECC, John L. McClellan Memorial VA hospital, VA Medical Research, GC143, 151/LR, 4300 West 7^th street, Little Rock, Arkansas 72205

Received 1/15/97 Accepted 1/29/98

4. SIGNALS AND SIGNALING PATHWAYS FOR NF KAPPA B

Activation of NF kappa B results from the degradation of I kappa B-alpha and subsequent translocation of the active NF kappa B dimer to the nucleus (86,87,116), figure 1. NF kappa B can be activated by a wide variety of physiological and non-physiological stimuli; these include cytokines, mitogens, viruses and viral products, oxidative stress, and chemical agents such as phosphatase inhibitors and ceramide. In T cells, almost any stimuli capable of activation results in NF kappa B induction including PHA, PMA, TNF-alpha, IL-1, IL-2, antigen receptor ligation, crosslinking of surface molecules such as CD2, CD3, CD28, and TCR, and infection with a wide variety of viruses such as HIV-1, HTLV, and HSV (86-89).Most signals tested so far target I kappa B-alpha as evidenced by its degradation in response to these activating stimuli, however the upstream signaling pathways preceding I kappa B-alpha degradation are very divergent and may be specific for a given activator (86,87, 127-130). I kappa B inactivation, without proteolytic degradation, has also been reported to occur as a consequence of tyrosine phosphorylation (131). Much research has been directed at elucidating the signal transduction pathways involved in regulating NF kappa B activation. Early studies implicated PKC as playing a central role in NF kappa B activation in response to many activators since phorbol esters, potent activators of PKC, were NF kappa B inducers and purified PKC released NF kappa B from I kappa B-alpha when incubated with cytosolic extracts (128-130,132,133). Furthermore, the activation of NF kappa B following stimulation of T cells with anti-CD3 has been demonstrated to be dependent upon PKC (134). Recent studies have revealed the existence of PKC independent pathways of NF kappa B activation including activation mediated by TNF- alpha, one of the most potent physiologic inducers of NF kappa B (135). The activation of NF kappa B by TNF-alpha has been shown to occur via the sphingomyelin pathway. In this pathway, phosphatidyl choline specific phospholipase C (PC-PLC) is activated which results in the generation of DAG which in turn leads to the activation of an acidic sphingomyelinase that hydrolyzes sphingomyelin to produce ceramide, in turn activating a ceramide dependent serine/threonine (ser/thr) kinase. Exogenous addition of PC-PLC, DAG, acidic sphingomyelinase, or ceramide can all activate NF kappa B in permeabilized cells (87,136). Although TNF-alpha mediated activation of NF kappa B has been shown to be independent of classical PKC activation, studies have demonstrated an important role for the nonclassical PKC isozyme PKC- zeta) (137-139). A dominant negative inhibitor of PKC-zeta inhibits NF kappa B activation by TNF-alpha, while a constitutively active mutant activates NF kappa B in NIH 3T3 cells (138).

Figure.1. Following treatment with a variety of activators upstream signal/s activate the NF kappa B inducible kinase (NIK) which in turn phosphorylates at least IKK1 and perhaps IKK2 (I kappa B kinase-alpha and I kappa B kinase-beta) in the I kappa B-kinase complex. Other signals such as PMA, UV or LPS etc., may directly activate the kinase complex or may first activate NIK. This kinase complex may also phosphorylate the p105/p65 complex, as well. IKK1/IKK2 then phosphorylates the NF kappa B-I kappa B-alpha complex on I kappa B-alpha series 32/36, which is followed by ubiquitination, proteasomal degradation and the nuclear translocation of NF kappa B. In the nucleus, NF kappa B induces the transcription of several immune response genes. ???? indicates possible defect in the proteasomal degradation pathway that likely results in the lowered nuclear translocation of NF kappa B in activated T cells from the elderly.

Given the convergence of all signaling pathways at or before the I kappa B-alpha target, existence of a common upstream effector has been postulated. Most of the studies have pointed to reactive oxygen intermediates (ROI's) as the common and critical effector molecule for various activating signals. This notion has been supported by the observations that most inducers of NF kappa B lead to the generation of ROI's and that activation of NF kappa B by many stimuli can be inhibited by antioxidants such as N-acetyl cysteine. Furthermore, treatment of cells with H₂O₂ can also activate NF kappa B (86,87,140). Recently published studies, however, have provided evidence against the role of oxidants in the induction of NF kappa B, leaving no proof of a direct functional role for ROI's in mediating the induction of NF kappa B (141-142).

4.1 Modification and Degradation of I kappa B

Regardless of the upstream signaling events initiated by various activators the ultimate target is I kappa B-alpha. Previous studies demonstrated that modification of I kappa B-alpha rather than NF kappa B was important in the activation of NF kappa B (97,119). I kappa B-alpha was post-translationally modified in response to activating signals which ultimately led to its degradation (116,132). This induced degradation was usually rapid and in some cell types completed by 10 minutes. In vivo analysis of NF kappa B activation has shown that one of the first modifications of I kappa B-alpha is that of phosphorylation (128,131). Most NF kappa B activating stimuli result in hyperphosphorylation (86,87,143,144), which occurs on serine 32 and serine 36 of the I kappa B-alpha molecule (145,146). Phosphorylation of these residues are critical since mutations of these sites block phosphorylation by various activators and prevent degradation of I kappa B- alpha which in turn inhibits nuclear NF kappa B induction (145,146). Other agents such as the chymotrypsin inhibitor tosylphenylchloromethylketone (TPCK) which prevent the phosphorylation of I kappa B-alpha also prevent degradation and nuclear translocation of NF kappa B (130,147).

Phosphorylation of I kappa B-alpha is required for an additional modification of I kappa B-alpha, i.e. ubiquitination. The addition of ubiquitin, a 7kd protein occurs at lysine-21 and lysine-22 in the amino terminal region of the I kappa B-alpha molecule (148,149). Mutations in I kappa B-alpha that block phosphorylation have been shown to block ubiquitination in vitro (150). Mutations of lysine 21 and 22 block ubiquitination and degradation of I kappa B-alpha but not phosphorylation, demonstrating that phosphorylation precedes ubiquitination and ubiquitination of lysines are required for degradation of I kappa B-alpha (148,149). The addition of ubiquitin is proposed to target I kappa B-alpha for degradation by the 26S proteasome. It has been demonstrated that covalent attachment of polyubiquitin chains to I kappa B-alpha is required for proteasome mediated degradation of the protein following cellular activation (150-152). Thus, the signal dependent phosphorylation and ubiquitination of I kappa B-alpha targets the cytoplasmic inhibitor to the ubiquitin-proteasome pathway. Degradation is mediated by the 26S proteasome complex which is a multicatalytic protease that degrades multi-ubiquitinated proteins in an ATP-dependent manner. The 26S proteasome complex is involved in the turnover of short-lived and abnormal proteins and is also involved in processing of peptides for antigen presentation by Class I MHC (153). Proteasome activity has been demonstrated to be essential for the degradation of I-kappa B-alpha as pretreatment of cells with peptide aldehyde inhibitors of the proteasome such as N-acetyl leucinal-leucinal-norleucinal (LLnL) or lactacystin block I kappa B-alpha degradation and NF kappa B induction. Furthermore, in the presence of proteasomal inhibitors the phosphorylated form of I kappa B-alpha accumulates (143,144,154). Although many kinases have been reported to phosphorylate I kappa B-alpha in vitro and cause activation of NF kappa B, a kinase that specifically phosphorylates I kappa B-alpha on ser 32 and 36 has remained elusive, until now (89). Maniatis et. al., demonstrated a kinase complex that specifically phosphorylates I kappa B-alpha on ser 32 and ser 36 (155). In vitro reconstitution experiments demonstrated that the activity of this kinase required both ubiquitin and ubiquitin conjugating enzymes suggesting that ubiquitination of the I kappa B -kinase is required for its activation. Recent studies by three groups have identified I kappa-B kinase, which comprises of a previously identified serine/threonine kinase known as CHUK (conserved helix-loop-helix kinase) and a second kinase closely related to CHUK. These kinases have now been named as IKK1 and IKK2. These two kinases appear to occur in a high molecular mass complex (700kDa) (156-158). It appears that IKK1 and 2 act in concert. Activity of IKK1 is inducible, and is autophosphorylated. Several studies are now underway to determine the kinetics of activation by these two kinases.

4.2. Autoregulation of NF kappa B and I kappa B

Studies demonstrating that the degradation of I kappa B-alpha correlated with nuclear appearance of NF kappa B also demonstrated rapid induction of I kappa B-alpha mRNA within 20 minutes of stimulation and

restoration of I kappa B-alpha levels within an hour (87,127,128). Transfection of cells with transactivating NF kappa B subunits resulted in the production of high levels of I kappa B-alpha mRNA, indicating a regulatory pathway (127,128). Analysis of the I kappa B-alpha promoter has shown that it contains multiple NF kappa B binding sites and that these sites are functional in the upregulation of gene expression in response to inducers that activate NF kappa B (158,159). Thus, transactivating NF kappa B dimers can induce their own inhibitor, I kappa B-alpha, which then binds to cytoplasmic dimers to restore the inhibited state and reestablish cytoplasmic pools of NF kappa B/I kappa B complexes. The I kappa B-beta gene, however, is not upregulated by NF kappa B (125,130). The accumulation of newly synthesized I kappa B-alpha can also repress NF kappa B activity by entering the nucleus to inhibit previously activated NF kappa B once the stimulating agent is removed (129). This model is supported by the observation that I kappa B-alpha deficient cells exhibit high nuclear levels of NF kappa B for long times following induction with TNF-alpha (130). This built in feed back inhibition may assure a transient response once the initiating event fades preventing dysregulation of genes whose functions may be harmful if expression goes unchecked (87).

NF kappa B dimers are regulated not only at the level of cytoplasmic retention but also at the level of synthesis. The genes encoding p105, p100, and c-Rel all contain kappa B binding sites in their promoter regions and stimulation of cells leads to increased synthesis of these proteins. Synthesis of p65, however, the most potent activator of transcription, is not upregulated by NF kappa B dimers (87,159,160).

4.3. Role of NF kappa B in immune response

NF kappa B is well recognized for its critical role in regulating immune response genes. Extensive research has established a clear role for NF kappa B in the inducible regulation of a wide variety of genes involved in immune function and inflammatory responses including GM-CSF, IL-6, IL-8, IL-2, IL-2R-alpha, IFN-beta, cellular adhesion molecules such as VCAM-1, IFN-beta, and Class I MHC (86-88). In terms of T cell function NF kappa B plays a vital role in the regulation of both IL-2 and IL-2R-alpha genes. Mutations in the NF kappa B binding site in an IL-2 promoter inhibits promoter activation of the transcription unit (161). Furthermore, TNF-alpha mediated IL-2R-alpha expression following T cell activation is highly dependent upon NF kappa -B (162,163).

Gene knockout of Rel subunit and I kappa B-alpha genes in mice has been accomplished and has confirmed the important role of these proteins in immune function. There appears to be no redundancy within the Rel family proteins as data from knockout mice show that loss of a particular Rel protein cannot be compensated by another Rel protein (164-167). Table 1 summarizes the effect of subunit specific gene knockout on immune function.

Absence of the c-rel gene has a large impact on immune function. Both mature B and T cells exhibit proliferative defects in response to various activating agents. The defect in T cell proliferation correlated with a lack of IL-2 production as IL-2 levels in c-rel deficient T cells were 50 fold lower than that observed in wild type T cells. Production of other cytokines in response to stimulation was also affected as evidenced by 2-3 fold lower amounts of IL-3, IL-5, GM-CSF, TNF-alpha, IFN-gamma when compared to wild type T cells. Additionally, c-Rel deficient mice showed impaired T cell dependent humoral responses when antigenically challenged. Antigen-specific IgG1 levels were decreased by 50-100 fold (164,165). RelB knockout mice develop normally but also show defects in immune function. Loss of the RelB gene results in an absence of thymic dendritic cells indicating that RelB plays an important role in the development of dendritic cells. Furthermore, antigen presenting cells from the spleen of RelB -/- mice showed extremely poor stimulating capacity in mixed lymphocyte cultures providing evidence that RelB is important for antigen presenting cell function. Cell mediated immunity was also impaired in these mice as evidenced by poor delayed type hypersensitivity responses (166,167). Immune response defects are also observed in mice lacking the p50/p105 subunit. These mice cannot effectively clear the pathogens Listeria or Streptococcus. B cells from these mice are defective in their ability to proliferate and produce antibody in response to Lipopolysaccharide (LPS), while T cells proliferate poorly in response to TCR and CD28 stimulation. In addition, IgE levels were decreased by 40 fold suggesting that p50 plays an important role in heavy-chain class switching (168). Unlike c-Rel and RelB knockout mice, RelA knockout mice exhibit embryonic lethality due to widespread apoptosis within the liver, suggesting the importance of p65 in liver development. Embryonic fibroblasts from RelA deficient mice fail to induce mRNA for I kappa B-alpha and GM-CSF in response to TNF-alpha suggesting an essential role for RelA in the induction of these genes (169). I kappa B-alpha knockout mice develop normally but exhibit widespread dermatitis and die about 7 days after birth. In splenocytes and thymocytes of these mice, NF kappa B was found to be constitutively activated. Treatment of I kappa B-alpha -/- embryonic fibroblasts resulted in a prolonged and sustained NF kappa B DNA binding activity in the nucleus, indicating the importance of the I kappa B-alpha isoform and not other I kappa B isoforms in the termination of an NF kappa B response (126).