Nuclear factor-KappaB modulation as a therapeutic approach in hematologic malignancies


  • Amit Panwalkar M.D.,

    1. Section of Developmental Therapeutics, Department of Leukemia, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Srdan Verstovsek M.D.,

    1. Section of Developmental Therapeutics, Department of Leukemia, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Francis Giles M.D.

    Corresponding author
    1. Section of Developmental Therapeutics, Department of Leukemia, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas
    • Department of Leukemia, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Box 428, Houston, TX 77030
    Search for more papers by this author
    • Fax: (713) 794-4297


Nuclear factor-kappaB (NF-κB) is a collective term that refers to a small class of dimeric transcription factors for a number of genes, including growth factors, angiogenesis modulators, cell-adhesion molecules, and antiapoptotic factors. Although most NF-κB proteins promote transcription, some act as inactivating or repressive complexes. The most common p50-RelA (p65) dimer known “specifically” as NF-κB, is relatively abundant, controls the expression of numerous genes, and exists as an inactive cytoplasmic complex bound to inhibitory proteins of the NF-κB inhibitor (IκB) family. The inactive NF-κB-IκB complex is activated by a variety of stimuli, including proinflammatory cytokines, mitogens, growth factors, and stress-inducing agents. The release of NF-κB facilitates its translocation to the nucleus, where it promotes cell survival by initiating the transcription of genes encoding stress-response enzymes, cell-adhesion molecules, proinflammatory cytokines, and antiapoptotic proteins. Constitutive activation of NF-κB in the nucleus is observed in some hematologic disorders. With the recent approval of bortezomib for patients with advanced multiple myeloma, NF-κB modulation is likely to be a therapeutic endeavor of increasing interest in coming years. Cancer 2004. © 2004 American Cancer Society.

The signal-transduction pathways involved in B-cell and T-cell activation in numerous immune and inflammatory responses have been well characterized.1, 2 Aberrant gene expression and dysregulation of these signaling pathways are implicated increasingly in neoplastic transformation and progression.1, 3, 4 Hence, signaling pathways and components of pathways that govern cell proliferation, survival, and oncogenesis are being studied thoroughly.5 Nuclear factor-kappaB (NF-κB) is a critical regulatory protein that activates the transcription of a number of genes, including growth factors, angiogenesis modifiers, cell-adhesion molecules, and antiapoptotic factors.6 The high level of interest in NF-κB is due to its broad role in inducing and controlling gene expression.1, 5, 7, 8


NF-κB, which was described first in 1986,9 was characterized initially from B-lymphocytes as a nuclear factor necessary for the transcription of the immunoglobulin κ light-chain gene. NF-κB is a collective term that refers to a small class of closely related dimeric transcription factors, which belong to a large family of proteins, the Rel (for reticuloendotheliosis) family. These proteins have been highly conserved from the fruit fly (Drosophila) to humans and are related through a DNA-binding, dimerization and inhibitor-binding region called the Rel homology domain (RHD).10 Members of the Rel/NF-κB family have been divided into two groups. In mammals, the first group comprising p105 and p100 has long C-terminal domains that contain multiple copies of ankyrin repeats, which act as inhibitors. By a process of proteolysis, the proteins in this group mature to give rise to p50 (NF-κB1) and p52 (NF-κB2), respectively, which are shorter, active, DNA-binding proteins.11–14 The second group is comprised of Rel (c-Rel), p65 (RelA), and RelB.15 The C-terminal halves of the latter group contain transcriptional-activation domains, whereas p100 and p105 have inhibitory domains.16 Nearly all Rel/NF-κB proteins can form homodimers and heterodimers, which are requisites for binding to the (κB) sites to influence gene expression. Although most of the NF-κB proteins are active transcriptionally, it is believed that some combinations act as inactive or repressive complexes. Thus, p50/p65, p50/c-rel, p65/p65, and p65/c-rel all are transcriptionally active, whereas p50 homodimer and p52 homodimer are transcriptionally repressive.17–21 The most common p50-RelA (p65) dimer, “specifically” known as NF-κB, is more abundant and controls the expression of more genes than any other heterodimers or homodimers.6 NF-κB exists as an inactive cytoplasmic complex, predominantly made up of p50-p65, bound to inhibitory proteins of the NF-κB inhibitor (IκB) family.22


Like NF-κB, IκB is a member of a larger family of inhibitory molecules that includes IκBα, IκBβ, IκBϵ, IκBγ, and Bcl-3 in higher vertebrate cells. All these inhibitors contain multiple regions of homology known as the ankyrin-repeat motifs. The ankyrin repeats are regions of protein-protein interaction, and the specific interaction between ankyrin repeats and RHDs appears to be a crucial, evolutionarily conserved feature of the regulation of NF-κB proteins. Each IκB differs in the number of ankyrin repeats, and this number appears to influence the specificity with which IκB pairs with an Rel dimer. Proteins p100 and p105 also contain ankyrin repeats and occasionally are included in the IκB family.22 The interaction between IκB and NF-κB masks the nuclear localization sequence in the NF-κB complex, sequestering the factor in the cytoplasmic compartment.23

IκB Kinase

IκB kinase (IKK) is present in the cell cytoplasm as an enzyme with serine-protein-kinase activity that is responsible for IκBα phosphorylation and that links tumor necrosis factor (TNF)-induced and interleukin-1 (IL-1)-induced kinase cascades to NF-κB activation.24 It is a large, multisubunit complex with three known components. Two of these polypeptide components, IKKα and IKKβ, are catalytic subunits, whereas the third component, IKKα (NEMO), has a regulatory function.25 IKKα and IKKβ have partially similar primary structures (52% homology). IKK complexes exist in cells as IKKα-IKKβ heterodimers in association with an unknown number of IKKα units. Both catalytic subunits contain, from the N-terminal to the C-terminal, a kinase domain, a leucine zipper (LZ), and a helix-loop-helix (HLH) motif.26 The kinase activities of IKKα and IKKβ depend on LZ-mediated dimerization. LZ mutations that interfere with this process abolish kinase activity. IKKα or IKKβ activity also is destroyed by mutations within the HLH motif.27, 28 These mutations, however, do not interfere with dimerization or binding to IKKγ. IKK activation also requires an intact IKKγ subunit. IKK or NF-κB activity is absent in IKKγ-deficient cells that are treated with upstream activators.29 Activation of IKK depends on phosphorylation of its IKKβ subunit.30 IKKα phosphorylation, although it is concurrent with the phosphorylation of IKKβ, is not essential for the stimulation of IκB kinase activity. The inactive, unphosphorylated IKK complex is activated, in response to upstream stimuli, by IKK kinases. These IKK kinases are recruited to the complex through IKKγ. This results in phosphorylation of IKKβ and activation of IKK. Although, initially, only a small fraction of IKK is activated through direct phosphorylation by IKK kinases, the activated IKKβ subunit can phosphorylate the adjacent subunit by intramolecular transautophosphorylation. Because of the ability of IKKβ to propagate its active state by autophosphorylation at the activation loop, it is important to have modulation to reduce kinase activity and render it sensitive to inactivation by a phosphatase. Without this mechanism, prolonged IKK activation could result in prolonged NF-κB activation followed by increased production of both primary and secondary inflammatory mediators, causing further NF-κB activation.31 Constitutive IKK activation has been reported recently in Hodgkin disease (HD) cells.32 This aberration may cause NF-κB activation, thus protecting these cells from the induction of apoptosis.33 Increased NF-κB and/or IKK activity may protect numerous types of tumors from apoptosis-inducing therapies.34, 35 Thus, IKK offers a reasonable target for the development of new antileukemia approaches.

Activation of NF-κB and its Role in Cell Cycle Regulation

The inactive NF-κB-IκB complex is activated by a variety of stimuli, including proinflammatory cytokines, mitogens, growth factors, and stress-inducing agents.11 Different stimuli initiate different signal-transduction pathways that involve distinct scaffolding and signaling proteins. These include, but are not restricted to, NF-κB-inducing kinase (NIK), mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1 (MEKK1), IL-1 receptor-associated kinases (IRAKs), TNF-α receptor (TNFR)-associated factors (TRAFs), double-stranded (ds) RNA-dependent protein kinase (PKR), protein kinase C (PKC), and transforming growth factor-beta (TGF-β)-activated kinase (TAK1).36 A majority of these signaling pathways converge on the IKK complex, which plays a major role in NF-κB activation (Fig. 1).37

Figure 1.

Activation of nuclear factor-kappaB (NF-κB) occurs through activation of the NF-κB inhibitor (IκB) kinase (IKK) through various pathways. Different stimuli initiate different signal-transduction pathways. These include, but are not restricted to, NF-κB-inducing kinase (NIK), mitogen-activated protein kinase/extracellular signal regulated kinase kinase 1 (MEKK1), interleukin-1 receptor (IL-1R)-associated kinases (IRAKs), tumor necrosis factor-alpha (TNFα) receptor (TNFR)-associated factors (TRAFs), double-stranded (ds) RNA dependent protein kinase (PKR), protein kinase C (PKC), and transforming growth factor-beta (TGFβ)-activated kinase (TAK1), which are activated by their respective ligands. NF-κB then translocates to the nucleus where, by binding to κB, it initiates transcription of various cytokines, adhesion molecules, etc., as depicted in this figure. These transcription products aid in further proliferation of cells. COX2: cyclooxygenase 2; ICAM: intracellular cell-adhesion molecule; iNOS: inducible nitric oxide synthase; LPS: lipopolysaccharide; VCAM: vascular cell-adhesion molecule.

IKK is a multisubunit complex. The predominant form is a IKKα-IKKβ heterodimer, associated with IKKγ.38 After stimulation, the NF-κB-IκB complex is activated by phosphorylation of the inhibitory protein. The IκBα subunit is phophorylated. This leads to attachment of ubiquitin (a small peptide) to IκBα by a stem cell factor-ubiquitin ligase. Multiple attachments of ubiquitin (a polyubiquitin chain) then flags the protein for degradation by the 26S proteasome, allowing the release of NF-κB. The 26S proteasome is a multicatalytic enzyme complex that is present in all eukaryotic cells. Its primary function is to degrade proteins.39 Recognition of proteins to be degraded is facilitated by attachment of ubiquitin chains to the target substrate. These substrate proteins may be tumor suppressors, cell cycle regulators, antiapoptotic proteins, etc.39 Thus, the release of NF-κB facilitates its translocation to the nucleus, where it promotes cell survival by initiating the transcription of genes encoding stress-response enzymes, cell-adhesion molecules, proinflammatory cytokines, and antiapoptotic proteins, such as Bcl-2, cIAP1, and cIAP2.40–42 More recently, it has been shown that NF-κB is involved in the production of vascular endothelial growth factor in human macrophages.43 In addition, it has been shown that other regulators of NF-κB transcriptional activity are active within the nucleus.44–46

A critical role for NF-κB in cell cycle progression is suggested by observations that NF-κB activity was elevated during the G0-to-G1 cell cycle transition in mouse fibroblasts.47 It also was found that the levels of NF-κB activation were linked to the signaling that controls cell cycle progression in HeLa cells and Jurkat T cells.48, 49 Inhibition of NF-κB caused impairment of cell cycle progression in human glioma cells50 and caused a retarded G1 to S-phase transition in HeLa cells.51 The identification of NF-κB binding sites in the promoter region of the cyclin D1 gene have provided direct evidence of the involvement of NF-κB in cell cycle regulation.52 Cyclin D1, in association with cyclin-dependent kinases (CDKs) CDK4 and CDK6, promotes G1 to S-phase transition through CDK dependent phosphorylation of pRb, thereby releasing the transcription factor E2F, which is required for the activation of S phase-specific genes.53–55

Dysregulation of NF-Kb and IκB Signaling and Oncogenesis

NF-κB in oncogenesis

The role of c-Rel.

The implication of c-Rel in neoplasia was suggested initially by the acute oncogenicity of its viral derivative, the v-Rel oncoprotein,56 which induced aggressive leukemias/lymphomas in chickens and transgenic mice. Elevated expression of this gene also was reported in a chicken B-cell lymphoma containing a proviral insertion in the c-rel promoter.57 Transformation of primary chicken lymphoid cells and embryo fibroblasts occurs due to overexpression of c-rel, albeit at a much lower frequency than v-rel, and occasionally gives rise to lymphoid tumors.58–61 Amplification of the human c-rel gene62, 63 was observed in 23% of diffuse lymphomas with a large-cell component (DLLC), which constitute approximately 50% of B-cell non-Hodgkin lymphomas.64–66 Amplification of the c-rel gene also is associated with extranodal presentation in 75–85% of DLLCs, identifying it as one of the most frequent genetic alterations in DLLC.65, 66 This amplification also was noted in primary mediastinal (thymic) B-cell lymphomas and in certain follicular large-cell lymphomas.65–68 Although it is amplified most commonly in human lymphomas, it also has been shown that the c-rel gene has undergone rearrangement in a few follicular lymphomas and in some DLLCs.68 A chromosomal rearrangement that created a hybrid protein (called c-Rel-Nrg), comprised of the N-terminal RH domain of c-Rel fused to unrelated gene sequences, was observed in a human diffuse lymphoma cell line.68 Knock-out mice with a homozygous deletion of the C-terminal transactivation domain of c-Rel were reported to develop lymphoid cell hyperplasia.69 The c-rel gene often is associated with advanced-stage disease.64–66 Together, these data suggest that loss of the C-terminal transactivation domain of c-Rel may provide a selective growth advantage.

The role of RelA.

RelA, the 65-kilodalton (kD) subunit of NF-κB, is encoded for by the human rela gene located on chromosome 11q13.63 Similar to c-Rel, RelA exhibits strong transcriptional activity, but to our knowledge chromosomal aberrations involving rela are relatively infrequent in tumors. The RelA protein interacts with factors that are critical for cellular gene expression and cell cycle progression,44, 48, 70, 71 and its inactivation in mice results in massive liver cell apoptosis.72 Consequently, many alterations in rela that impair cell growth or survival may be related to neoplastic change. There have been sporadic reports of rela gene translocation in non-Hodgkin lymphomas and multiple myelomas, and amplification was observed in a few DLLCs, but a larger survey failed to implicate rela in a majority of these tumors.73 However, it has been found that the persistent activation of NF-κB complexes that contain RelA in different tumor cell types enhances tumorigenicity.

The role of NF-κB1.

The p50/p105 subunit of NF-κB, NF-κB1, is encoded for by the nf-κb1 gene and is found on human chromosome 4q24.63 Few alterations in the structure and expression of nf-κb1 have been reported in leukemias and lymphomas. Possibly, structural changes in p50 and RelA may have an adverse effect on the growth or survival of tumor cells. However, there are some examples of nf-κb1 gene rearrangement in certain acute lymphoblastic leukemias.74 It is interesting to note that p105 was associated recently with LYL1, a bHLH protein involved in a translocation in certain human T-cell acute leukemias.75 Ectopic expression of LYL1 in T cells alters the ratio of p105 to the mature p50 protein, resulting in the repression of NF-κB-dependent gene activation by p50-p50 homodimers. The significant homology between LYL1 and TAL-1 also may be relevant, because TAL-1 activation is associated with large numbers of human T-cell acute lymphoblastic leukemias (ALL).

The role of NF-κB2.

Chromosomal translocations and deletions that affect the nf-κb2 locus on human chromosome 10q2463 are found in many lymphomas.56 Rearrangements of the nf-κb2 gene are present in small percentages of B-cell non-Hodgkin lymphomas, chronic lymphocytic leukemias (CLL), and multiple myelomas, are observed frequently in cutaneous T-cell lymphomas, and are associated with a poor prognosis.76–79 Chromosomal translocations or deletions in the nf-κb2 locus result in the production of truncated p100 proteins with a partially deleted C-terminal ankyrin domain that is fused in some cases to heterologous sequences.76 The functional consequence of these rearrangements is the constitutive nuclear localization of tumor-derived truncated p100 proteins and their binding to κB DNA motifs without further processing to a mature p52 form.77–81 Homozygous inactivation of the entire coding sequence of nf-κb2 is not tumorigenic.82, 83 This indicates that the simple loss of the IκB function of p100 is not sufficient to account for the oncogenicity of tumor-associated, truncated p100 proteins. However, it is interesting to note that mice carrying a homozygous deletion of the C-terminal ankyrin repeats of p100, but that still express a functional p52 protein, display gastric and lymph node hyperplasia.84 These data support a model in which tumor-derived p100 proteins function by altering NF-κB-mediated transcription. Consistent with this prediction, some truncated p100 proteins activate gene expression on their own, suggesting that tumor-derived NF-κB2 proteins may act as constitutive transactivators.80, 85 Overall, these results highlight the altered transcriptional activity of tumor-derived p100 proteins.

The role of Bcl-3.

The Bcl-3 protein is related structurally to the IκB family of Rel/NF-κB regulators. Bcl-3 is found in the nucleus of most cells, where it interacts with homodimers of p50 and p52 to alter their DNA-binding activities and to modulate gene expression.86 Bcl-3 facilitates NF-κB-mediated transcription by removing inactive p50 homodimers from DNA and acts as a potent coactivator for homodimers of p52. The human bcl-3 gene is located on chromosome 19q13.1 and is involved in a rare but recurrent t(14;19)(q32.3;q13.2) translocation in B-cell CLL.87 Translocations that affect bcl-3 in leukemias and lymphomas always result in overexpression of an intact Bcl-3 protein.87–89 Thus, chromosomal translocations involving bcl-3 very well may alter NF-κB dependent gene expression. Mice deficient in bcl-3 display a partial loss of B cells, suggesting the participation of Bcl-3 in B-cell survival that may be relevant to its oncogenic activity.90 Whereas constitutive expression of bcl-3 in thymocytes failed to induce tumors in mice,91 E-bcl-3 transgenic animals reportedly developed a lymphoproliferative disorder characterized by the accumulation of mature B cells.92 This phenotype is consistent with the notion that overexpression of Bcl-3 is a critical step in the multistep process that leads to cellular transformation.

Inactivation of IκB in oncogenesis

IκBα is an important inhibitor of NF-κB.93 Studies showing that overexpression of antisense iκbα transcripts transformed mouse NIH3T3 cells were to our knowledge the first to propose a role for the inactivation of IκB factors in oncogenesis.94 The death of neonatal mice resulting from the homozygous inactivation of iκbα has prevented the detection of a potential tumorigenic effect.95, 96 Recent analyses of many human tumors that demonstrate persistent nuclear NF-κB DNA binding and transactivation activity often have revealed defective IκB activity. Work on Hodgkin lymphomas unveiled inactivating mutations in iκbα and iκbϵ and revealed constitutive activation of IκB kinase, resulting in activation of NFκB.32, 97, 98 In other instances, oncogenesis was found to be correlated inversely with the levels of IκBα and IκBβ proteins and coincided with the activation of IκB kinases, which govern the destruction of IκB factors.32, 99, 100 These results are consistent with the fundamental role of IκB molecules in the regulation of Rel/NF-κB function. In this context, it has been proposed that IκBα acts as a tumor suppressor to control the oncogenic activation of NF-κB.97 The human iκbα gene is located on chromosome 14q13.101 Although this locus is not involved frequently in chromosomal rearrangements or deletions in HD,102 it is too early to rule out the possibility that subtle mutations may be involved in many instances.

Constitutive nuclear NF-κB Activity

Many human leukemias and lymphomas show constitutively active NF-κB in the nucleus. Deregulated NF-κB activation was reported initially in HD.103 It was shown that this persistent NF-κB is critical for Reed–Sternberg (RS) cell proliferation, resistance to apoptosis, and tumor formation when RS cells were placed in severe combined immunodeficient (SCID) mice.33 This suggested an important role for NF-κB in the pathogenesis of HD. In some HD cell lines and in biopsies from patients with recurrent HD, constitutive NF-κB activation appears to be a direct consequence of mutations in the gene encoding IκBα. These mutations invariably produce nonfunctional IκBα proteins that are missing various portions of the central ankyrin domain and the C-terminal region.32, 97, 98 Although IκBα appears to be the most frequent target of mutations in HD cells, a mutant allele of IκBϵ was detected recently in an HD-derived cell line that also contained a mutation in the iκbα gene.97 It is important to note that mutations affecting iκb genes are not the only means by which constitutive NF-κB activity is achieved in HD. For instance, it has been shown that activated NF-κB complexes in RS cells result from persistent activation of the signal-transduction pathway, which leads to the proteolysis of IκBα. This appears to occur through an autocrine activation of the RS cell by several cytokines secreted by the RS cell.32, 104, 105 The instability of IκBα in these cells has been correlated with constitutive IκB kinase activity. Numerous other mechanisms of NfκB activation, including Epstein–Barr virus (EBV) infection, have been found in the RS cells in Hodgkin lymphomas.106 This deregulated NF-κB activity is detected in all patients with HD, and its susceptibility to inhibition by recombinant IκBα molecules has important applications for therapy. Persistent nuclear NF-κB activity also is now recognized as a common characteristic of childhood ALL; activated NF-κB complexes were detected in 93% of the patients examined.107 Inhibition of the proteasome in primary ALL cultures resulted in hyperphosphorylated forms of IκBα. Similar to the scenario described above for HD, these results point to the constitutive activation of upstream IKK kinases and suggest a crucial role for NF-κB in leukemia cell survival. NF-κB is necessary for bone marrow cell transformation and tumorigenesis by Bcr-Abl, a chimeric oncoprotein that is involved in ALL and in chronic myelogenous leukemia (CML).108 Bcr-Abl expression leads to increased nuclear translocation of NF-κB and enhances the transactivation function of RelA. This effect is dependent on the tyrosine kinase activity of Bcr-Abl and also partially requires Ras function. However, the exact role of NF-κB in Bcr-Abl-induced leukemias remains to be clarified, because the antiapoptotic activity of NF-κB is not necessary for Bcr-Abl-mediated cell survival.108 There is ample evidence indicating that the Tax protein of human T-cell leukemia virus Type I (HTLV-I) leads to persistent NF-κB activity through the activation of IKK kinases.109 This also was demonstrated in some patients with acute myelogenous leukemia (AML).110 However, continuous NF-κB activation in primary adult T-cell leukemia (ATL) cells was described recently that also occurred through a mechanism that was independent of HTLV-I Tax and that likely involved increased degradation of IκBα. Despite the lack of any Tax protein expression in TL-Om1 cells, persistent nuclear expression of NF-κB DNA-binding complexes composed of RelA/p50 was observed.111 These complexes are distinct from the major c-Rel/p50 DNA-binding complexes found in HTLV-I-infected T cells that constitutively express Tax. Thus, it appears that different NF-κB subunit activation can occur in primary leukemic cells derived from patients with ATL that do not express significant levels of Tax.111 To our knowledge, the contribution of these complexes to leukemogenesis has yet to be addressed. In patients with diffuse, large-cell, B-cell lymphoma (DLBCL), high expression levels of the target genes of NF-κB reportedly have been associated with a poor prognosis. Two cell lines of DLBCL also showed constitutive activation of IκB kinase activity. Furthermore, NF-κB inhibition caused cell death, showing that this pathway is necessary for cell growth.112 Primitive AML cells that demonstrated aberrant expression of NF-κB were subjected to action of proteosome inhibitors and resulted in rapid cell death in one experiment with leukemic stem cells.113 High levels of NF-κB activity also were demonstrated in CLL B-cells.114 In EBV-infected cells, it has been shown that the EBV-encoded latent membrane protein 1 acts like a constitutively active receptor of the TNF receptor family, causing TRAF2-mediated NF-κB activation and resulting in sustained proliferation, which leads to lymphoma.115 Primary CML blasts have shown constitutive p65/RelA NF-κB/Rel DNA-binding activity, thus representing a potential target for molecular therapies in patients with CML.116 Data from studies of NF-κB activation in two non-Hodgkin lymphoma cell lines derived from mantle cell lymphoma (MCL) samples suggest that constitutive NF-κB expression plays a key role in the growth and survival of MCL cells and that the proteasome inhibitor PS-341 and BAY-11, which prevents the phosphorylation of IκB, may be useful therapeutic agents for patients with MCL, a lymphoma that is refractory to most current chemotherapy regimens.117 Patients with B-CLL have constitutive high NF-κB activity, which is modulated by a number of cytokines.118 Ohshima et al recently reported on the role of Bcl10 in mucosa-associated lymphoid tissue (MALT) lymphomas, which usually involve extranodal sites, especially the stomach, lung, and salivary glands.119 The Bcl10 gene, which has been isolated from the breakpoint region of t(1;14) (p22;q32) in MALT lymphomas, is considered to be an apoptosis-associated gene and involves a caspase recruitment domain-containing protein that activates NF-κB.120 Current data suggest that activation of Bcl10 and NF-κB may be important in MALT lymphomagenesis and that nuclear localization of Bcl10 may be important in the progression of MALT.119–121

IKK Inhibition and Modulation of NF-κB Activity

IKK is a convergence point for a broad spectrum of inflammatory mediators, which cause NF-κB activation.122 Numerous IKK inhibitors have been identified in the past few years. This opens new arenas for modulation of NF-κB activity and, thus, cancer chemotherapies. Many agents that inhibit NF-κB activity have antiinflammatory activity. IKK kinase activity is stimulated when cells are exposed to the cytokine TNFα or by overexpression of the cellular kinases MEKK1 and NIK. It has been shown that the antiinflammatory agents aspirin and sodium salicylate specifically inhibit IKKβ activity in vitro and in vivo. The mechanism of this inhibition is the binding of these agents to IKKβ to reduce adenosine triphosphate binding.122 Curcumin (diferuloylmethane), the active ingredient in turmeric, blocks the disappearance of IκBα and inhibits the phosphorylation of IκBα.123 It also inhibits the IκB kinase 1 (IKK1) and IκB kinase 2 (IKK2) activities induced by lipopolysaccharide. Tetrahydrocurcumin, hexahydrocurcumin, and octahydrocurcumin are less active in this respect. These data suggest that curcumin may exert its antiinflammatory and anticarcinogenic properties by suppressing the activation of NF-κB through inhibition of IKK activity.123 More recently, it was reported that curcumin suppressed constitutive IκBα phosphorylation through the inhibition of IKK activity in human myeloma cells, leading to cell apoptosis.124 Curcumin also down-regulated the expression of NF-κB-regulated gene products, including IκBα, Bcl-2, Bcl-x(L), cyclin D1, and IL-6, leading to the suppression of proliferation and arrest of cells at the G1 to S-phase transition of the cell cycle. At equimolar concentrations, curcumin inhibits DNA synthesis, as demonstrated in five leukemia cell lines, three nontransformed hematopoietic progenitor cell populations, and four nontransformed fibroblastic cell lines in a concentration dependent manner.125 Helenalin is a sesquiterpene lactone derived from the plant Arnica.126 Helenalin inhibits NF-κB activation in response to different stimuli in T cells, B cells, and epithelial cells and abrogates κB-driven gene expression.126 This inhibition was not due to a direct modification of the active NF-κB heterodimers but to modification of the NF-κB/IκB complex, thus preventing the release of IκB. More recent studies have revealed that cytotoxic sesquiterpene lactones, including ambrosin, alantolactone, hymenin, and helenalin, induce apoptosis in Jurkat leukemia T cells.127, 128 Sulfasalazine is a direct inhibitor of IKKα and IKKβ, but it did not appear to inhibit NF-κB/Rel activation or TRAF1 transcript levels in B-CLL lymphocytes.129, 130 Therefore, NF-κB/Rel activation in B-CLL appears to be independent of IKK activation. Arsenic trioxide is an active agent in the treatment of patients with acute promyelocytic leukemia.131 Arsenic rapidly down-regulates constitutive IKK as well as NF-κB activity and induces apoptosis in Hodgkin-RS cell lines that contain functional IκB proteins. Treatment of nonobese diabetic/SCID mice with arsenic trioxide induces significant reduction of xenotransplanted L540Cy Hodgkin tumors concomitant with NF-κB inhibition.132 In vitro, it has been shown that tetrathiomolybdate, a specific copper chelator, inhibits the development of metastasis and angiogenesis through suppression of NF-κB.133

Resistance to Chemotherapy Due to NF-κB Activation

There is evidence to suggest that overexpression of NF-κB may lead to chemoresistence and that inhibition of this pathway may lead to successful therapy with older chemotherapeutic agents. This has been shown by inhibition of NF-κB by PS-341, which markedly increased the sensitivity of myeloma cells to chemotherapeutic agents.134 Ectopic expression of mutated K-ras or N-ras in the IL-6 dependent ANBL6 multiple myeloma cell line induced cytokine independent growth. Hu et al. recently reported on the signaling pathways activated by oncogenic ras that may stimulate IL-6 independent growth.135 On depletion of IL-6, two mutated, ras-containing myeloma lines demonstrated constitutive activation of mitogen-activated extracellular kinase 2(MEK)/extracellular signal-regulated kinase, phosphatidylinositol-3 kinase (PI3-kinase)/AKT, mammalian target of rapamycin (mTOR)/p70S6-kinase, and NF-κB pathways. The mTOR inhibitors rapamycin and CCI-779, the PI3-kinase inhibitor LY294002, and the MEK inhibitor PD98059 all reduced the growth rate of these mutant ras-containing cells. These results indicate that several pathways that contribute to stimulation of cytokine independent growth are activated downstream of oncogenic ras in myeloma cells. Data supporting the combinations of targeted approached are being generated rapidly. The coadministration of bortezomib and suberoylanilide hydroxamic acid resulted in diminished NF-κB activation, raising the possibility that combined proteasome and histone deacetylase inhibition may represent a novel strategy in leukemia, particularly for apoptosis-resistant, Bcr-Abl positive hematologic malignancies.136

An important issue for the development of NF-κB modulation agents is the impact of such modulation on normal cells. Although to our knowledge the contribution of NF-κB modulation to the clinical activity of the proteasome inhibitor bortezomib is not fully defined to date, the adverse event profile of this agent may be instructive.137 On initial Phase I clinical trials of bortezomib, dose-limiting toxicities (DLT) included diarrhea and sensory neurotoxicity, with the latter usually more pronounced in patients with antecedent neuropathy. In a Phase I study in patients with advanced hematologic malignancies that included a relatively intensive schedule in which bortezomib was dosed twice weekly for 4 consecutive weeks over a 6-week cycle, DLTs included thrombocytopenia, hyponatremia, hypokalemia, fatigue, malaise, and peripheral neuropathy. Other less severe toxicities that were reported commonly included gastrointestinal symptoms.137

Although IKK inhibitors appear to be a particularly promising in terms of NF-κB modulation, various other approaches are being developed. Bortezomib (PS-341) has been approved by the U.S. Food and Drug Administration for the treatment of patients with advanced multiple myeloma.137 It is not clear to date whether the therapeutic activity of bortezomib is due predominantly to its effects on of NF-κB modulation or to other consequences of its activity as a proteasome inhibitor. It has been shown that agents developed as inhibitors of the 26S proteasome inhibit IKK degradation and NF-κB nuclear translocation. These agents are receiving relatively more attention than small-molecule inhibitors, which may interfere directly with binding of NF-κB to DNA. However, this occurs over a large interaction surface, making it unlikely that a small, nonpolar blocking molecule can be developed. These same issues discourage somewhat the development of inhibitors of the dimerization of NF-κB proteins.

Other classes of agents in development include 1) agents that regulate NF-κB protein expression and binding to DNA, including antisense compounds, NF-κB-decoy compounds, and RNA inhibitors; 2) agents that interfere with IKK complex formation (e.g., the NEMO binding-domain peptide); 3) agents that block IKKβ activation process (e.g., the toll-IL-1 receptor adapter protein peptide); 4) heat-shock protein 90 inhibitors; 5) proteasome inhibitors; and 6) inhibitors of NF-κB transcriptional activity, including glucocorticoids.138 There also is ongoing exploration of nonsteroidal antiinflammatory drugs as NF-κB modulators. Although many of the relevant agents are cyclooxygenase (COX) inhibitors, the effects of these drugs on NF-κB is reported to be independent of COX inhibition.122 Increasing data documenting the activity of thalidomide and its analogs in various tumors have focused in part on the ability of these agents to inhibit NF-κB activation.139 Wide varieties of selective IKK inhibitors currently are in development, including SPC-839,140 PS-1145,141 BMS-34551,142 and SC-514.143 Curcumin, the relevant in vitro properties of were discussed earlier, is an agent currently under study in patients with malignancies, with a focus on antioxidant, antiangiogenic, and PKC-inhibitory activities.144


NF-κB modulation is likely to be a therapeutic endeavor of increasing interest in coming years. The recent approval of bortezomib for patients with advanced myeloma is likely to accelerate the focus in developmental therapeutics on this target.136 An understanding of the relatively specific sensitivity of myeloma cells to bortezomib will help define important mechanisms of resistance to proteasome inhibition. We can expect to see accelerated studies of combinations of targeted therapies as the various cellular pathways in which NF-κB plays a pivotal role are elucidated.136, 145, 146