Regulation and function of nuclear IκBα in inflammation and cancer

Abstract

The nuclear translocation and accumulation of IκBα represents an important mechanism regulating transcription of NFκB-dependent pro-inflammatory and anti-apoptotic genes. The nuclear accumulation of IκBα can be induced by post-induction repression in stimulated cells, inhibition of the CRM1-dependent nuclear IκBα export by leptomycin B, and by the inhibition of the 26S proteasome. In addition, IκBα is constitutively localized in the nucleus of human neutrophils, likely contributing to the high rate of spontaneous apoptosis in these cells. In the nucleus, IκBα suppresses transcription of NFκB-dependent pro-inflammatory and anti-apoptotic genes, representing an attractive therapeutic target. However, the inhibition of NFκB-dependent genes by nuclear IκBα is promoter specific, and depends on the subunit composition of NFκB dimers and post-translational modifications of the recruited NFκB proteins. In addition, several recent studies have demonstrated an NFκB-independent role of the nuclear IκBα. In this review, we discuss the mechanisms leading to the nuclear accumulation of IκBα and its nuclear functions as potential targets for anti-inflammatory and anti-cancer therapies.

Introduction

IκBα is a critical regulator of the transcription factor NFκB, which induces expression of a wide range of genes involved in immune and inflammatory responses, cell proliferation and apoptosis [1-5]. Deregulation of IκBα cellular levels and localization results in a variety of diseases, including chronic inflammatory disorders and many types of cancer and leukemia [6-15]. Even though IκBα has been originally discovered as a cytoplasmic inhibitor of NFκB, it is now clear that it has important nuclear functions as well.

NFκB proteins form homodimers or heterodimers consisting of p65 (Rel-A), p50, p52, c-Rel, and Rel-B [16-20]. In the classical model of NFκB activation, IκBα inhibits NFκB activity by masking the nuclear localization signals (NLS) of NFκB dimers and retaining them in an inactive state in the cytoplasm. Following cell stimulation by extracellular stimuli, including inflammatory cytokines, bacterial and viral products, apoptotic signals, and other forms of cellular stress, IκBα is phosphorylated at serine residues 32 and 36 through a cascade of inducible protein kinases that involve IκB kinase (IKK), ubiquitinated, and selectively degraded by the 26S proteasome [21-26]. This results in unmasking of the NLS of the NFκB dimers, which then translocate to the nucleus and stimulate transcription of NFκB-dependent pro-inflammatory and anti-apoptotic genes. Studies have shown that individual NFκB dimers bind various κB sites with differential affinity, which is affected by differences in the affinity of each dimer for the κB site, the ability to interact with associated transcription factors and inhibitors, chromatin environment, and by the post-translational modifications of NFκB proteins [27-32].

One of the first genes induced following NFκB activation is IκBα, since IκBα promoter also contains the NFκB binding region [33-35]. This newly synthesized IκBα can then enter the nucleus, remove NFκB from gene promoters, and transport NFκB proteins back to the cytoplasm [36-39]. This feedback regulation by post-induction repression represents a crucial regulatory mechanism terminating NFκB activation during persistent stimulation, and limiting the NFκB response. Loss of this negative feedback regulation as well as increased degradation of IκBα have been associated with increased NFκB activation in inflammatory diseases as well as in numerous types of cancer and leukemia [6-10].

Regulation of IκBα nuclear transport and accumulation

Sequences determining the nuclear localization of IκBα

IκBα is the most abundant and best-characterized member of the IκB protein family that currently consists of nine IκB proteins: IκBα, IκBβ, IκBε, Bcl-3, IκBz, IκBNS, IκBh, and the precursor proteins p100 and p105. All IκBα proteins are characterized by ankyrin repeat domain (ARD), enabling IκB proteins to form complexes with NFκB dimers and bind other proteins. The IκBα molecule consists of three main regions: N-terminal region where the inducible phosphorylation and ubiquitination occur, the ARD, and an acidic C-terminal sequence that is important for basal degradation of free IκBα [4,5,20,40]. Even though IκBα does not contain the classical nuclear localization sequence (NLS; KK/RXK/R), and its small size (37 kD) would allow a simple diffusion through the nuclear pore complex (NPC), IκBα is transported to the nucleus by an active transport mediated by a nuclear import sequence localized within the ARD of IκBα [41-43].

The nuclear export of IκBα is facilitated by two nuclear export signals (NES) located at the amino terminus (N-NES) [44-47] and carboxyl terminus (C-NES) [37,38]. The nuclear IκBα export is mediated by the NES receptor CRM1, also known as exportin 1, which belongs to the karyopherin β family and shares sequence homology in the Ran-GTP binding domain with members from this family [48,49]. In unstimulated cells, IκBα continuously shuttles between the nucleus and the cytoplasm [38,44]. However, in most cells, the nuclear export of IκBα is dominant over its import, resulting in the cytoplasmic localization of IκBα.

The nuclear translocation and accumulation of IκBα can be induced by three main mechanisms: by the post-induction repression in stimulated cells, by inhibition of the nuclear export of IκBα, and by the proteasome inhibition (Figure 1).

Figure 1.

Schematic representation of the main mechanisms inducing the nuclear translocation and accumulation of IκBα. The nuclear translocation and accumulation of IκBα can be induced by the post-induction repression in stimulated cells [36-39], by blocking the nuclear export of IκBα by CRM1 inhibition [54-61], and by the inhibition of the 26S proteasome [69-71].

Induction of nuclear IκBα accumulation by post-induction repression

In continuously stimulated cells, the newly synthesized IκBα translocates to the nucleus, dissociates NFκB dimers from gene promoters and transports them back to the cytoplasm, thus terminating transcription [36-39]. This feedback regulation by post-induction repression represents a crucial mechanism terminating NFκB activation during persistent stimulation. Impaired post-induction repression may result in a persistent activation of NFκB and increased cell survival. Blocking the nuclear export of IκBα by CRM1 inhibitors increases the nuclear IκBα accumulation, suppresses NFκB activity and induces apoptosis, representing an attractive therapeutic target.

Induction of nuclear IκBα accumulation by CRM1 inhibition

Leptomycin B (LMB) is a specific inhibitor of the nuclear protein export that interferes with the interaction between NES and CRM1 by covalently binding to a cysteine residue in the central domain of CRM1 [50-52]. It has been discovered in 1983 as a potent anti-fungal antibiotic produced by Streptomyces [53]. However, since then, LMB has been extensively used to study the nuclear-cytoplasmic shuttling of CRM1-binding proteins, including IκBα [54-57]. Studies from our laboratory have shown that in stimulated human leukocytes, LMB induces nuclear accumulation of IκBα by inhibiting IκBα nuclear export, resulting in the inhibition of NFκB activity, and increased leukocyte apoptosis [57-61]. Even though LMB possesses strong antitumor and anti-inflammatory properties [62,63], its toxicity prevents it from being clinically useful [64,65]. However, using high-content screening technologies and medicinal chemistry approaches based on modifying LMB, several recent studies identified novel selective nuclear export inhibitors (NEI) that maintain the high potency of LMB but are better tolerated [66-68]. These new NEIs have the potential to inhibit the constitutive activity of NFκB in cancer cells and chronic inflammatory disorders by increasing the nuclear levels of IκBα.

Induction of IκBα nuclear accumulation by the proteasome inhibition

We have shown that in addition to the post-induction repression and by blocking the nuclear export of IκBα by the CRM1 inhibition, the nuclear IκBα accumulation can be induced by the proteasome inhibition (Figure 1) [69-71]. Bortezomib (Velcade, PS-341) and other 26S proteasome inhibitors have been developed to inhibit the cytoplasmic degradation of IκBα, thus inhibiting the NFκB signaling in cancer cells [72-75]. However, studies from our laboratory have shown that bortezomib, MG132, MG115 and other proteasome inhibitors inhibit NFκB activity by an additional mechanism that consists of inducing the translocation of IκBα from the cytoplasm to the nucleus in prostate and ovarian cancer cells, HeLa cells, leukemia HL-60 cells, monocytic cells and chronic T cell leukemia Hut-78 cells [69-71]. The proteasome inhibition-induced nuclear IκBα accumulation is dependent on de novo protein synthesis, since cycloheximide (CHX) completely blocks the proteasome-induced nuclear IκBα translocation [69]. This lack of IκBα nuclear translocation in response to the proteasome inhibition in CHX-treated cells could be explained by two mutually non-exclusive mechanisms. In the first model, treatment with CHX might prevent resynthesis of a protein that is otherwise necessary for the proteasome inhibition-induced nuclear translocation of IκBα, but has a short half-life; thus, treatment with CHX would significantly decrease its level. Alternatively, the proteasome inhibition-induced nuclear translocation of IκBα may require that the cellular (cytoplasmic) level of IκBα increases above certain threshold level. When cells are treated with CHX, de-novo synthesis of IκBα is inhibited, and IκBα never reaches this threshold level, even after the degradation of IκBα is blocked by the proteasome inhibition. Similar mechanism has been suggested to account for the proteasome inhibition induced nuclear accumulation of glucocorticoid receptor and the varicella-zoster virus DNA binding protein ORF29p [76-78]. This model is also supported by previous studies that used cells transfected with constructs expressing IκBα and demonstrated that when IκBα is overexpressed, it localizes in the nucleus [79-81]. Since bortezomib has been approved by FDA for the treatment of multiple myeloma and is being tested in clinical trials as a combination therapy to treat other cancers as well [82-85], understanding the mechanism how it induces the nuclear translocation and accumulation of IκBα may lead to the development of more specific and effective therapies in the future.

Induction of IκBα nuclear accumulation by UV light

Interestingly, a recent study by Tsuchiya et al suggested that IκBα translocates into the nucleus and associates with the nuclear IKKβ also in response to UV radiation and other types of oxidative stress [86]. However, in contrast to the proteasome inhibition or to the increased nuclear IκBα accumulation induced by the CRM1 inhibition, this UV light-induced nuclear IκBα translocation does not result in the nuclear IκBα accumulation and inhibition of NFκB activity. On the contrary, the UV light-induced nuclear translocation of IκBα is followed by IκBα degradation and activation of NFκB [86].

Constitutive nuclear localization of IκBα in human neutrophils

In most resting unstimulated cells, IκBα is localized in the cytoplasm and by binding to NFκB dimers, it inhibits their nuclear translocation [1-3,87]. In contrast, in human neutrophils (polymorphonuclear leukocytes), majority (more than 60%) of the total cellular IκBα is localized in the nucleus (Figure 2) [88]. Interestingly, neutrophils are cells that have one of the shortest live spans in the body. They circulate in the blood and in the absence of infection, they undergo apoptosis within 24 hours after the release from bone marrow [89,90]. Even though the NFκB subunits p50 and p65 are present in the nucleus of resting neutrophils as well [88,91], the nuclear IκBα prevents NFκB activation by binding to nuclear p65 NFκB [57]. In response to neutrophil stimulation with lipopolysaccharide (LPS) or pro-inflammatory cytokines, IκBα is phosphorylated by the enzymes of the IKK complex and degraded by the proteasome both in the cytoplasm and in the nucleus [57,92]. However, compared to macrophages and other inflammatory cells, the extent of NFκB activation in human neutrophils is considerably lower [93,94], and this is associated with a decreased production of NFκB-dependent pro-inflammatory cytokines [95,96]. Interestingly, this high nuclear accumulation of IκBα in resting cells is unique to human neutrophils, since in mouse neutrophils, IκBα is localized mainly in the cytoplasm (unpublished data).

Figure 2.

Confocal immunofluorescence microscopy of IκBα in human neutrophils. Resting human neutrophils were fixed and analyzed by confocal laser scanning microscopy using anti-IκBα antibody and FITC-conjugated secondary antibody (green fluorescence) as described [97]. DNA was visualized with propidium iodide (red fluorescence). The figure illustrates the nuclear localization of IκBα in human neutrophils and overlap of the DNA and FITC-IκBα staining (yellow).

The mechanisms responsible for the high nuclear accumulation of IκBα in resting human neutrophils are not understood. There are two possible scenarios. In the first model, the high nuclear IκBα accumulation is the result of an increased nuclear import that is dominant over the nuclear export in human neutrophils. In the second model, the high nuclear IκBα level could be caused by the intranuclear binding of IκBα. This hypothesis is supported by our results indicating that the nuclear IκBα associates with the components of the nuclear matrix in human neutrophils [97]. Our study showed that a further increase of the nuclear accumulation of IκBα in the neutrophils increases caspase-3 activity and accelerates neutrophil apoptosis [57]. Thus, it seems likely that the high nuclear IκBα accumulation in human neutrophils represents one of the underlying mechanisms responsible for the high rate of spontaneous apoptosis in these cells (Figure 3). Since neutrophil apoptosis plays a critical role in the inflammatory response that characterizes sepsis, acute lung injury (ALI), and other inflammatory disorders [98-102], a better understanding of the mechanisms regulating the nuclear accumulation and function of IκBα in the neutrophils will contribute to the development of safer therapies for ALI, sepsis, and other neutrophil-mediated inflammatory disorders.

Figure 3.

Proposed model of NFκB regulation by nuclear IκBα in human neutrophils. A, In most unstimulated cells, IκBα is localized in the cytoplasm, and by binding to NFκB dimers, it prevents their translocation to the nucleus and NFκB-dependent transcription. B, In human neutrophils, IκBα is localized predominantly in the nucleus [57,88]. However, the nuclear IκBα in the neutrophils associates with NFκB p65 and p50 subunits in the nuclear matrix [97], thus suppressing transcription of NFκB-dependent genes and inducing neutrophil apoptosis.

Nuclear IκBα function

NFκB-dependent function of the nuclear IκBα

The subunit composition of NFκB dimers determines their affinity for IκBα. In vitro, IκBα preferentially binds to p50/65 heterodimer and p65 homodimer, while binding to p50 homodimer is substantially weaker [103-108]. However, even though IκBα can bind to 50 homodimers, it does not inhibit their in vitro DNA-binding activity. The in vitro interaction between IκBα and p65 has a very low dissociation rate resulting in an extremely high affinity and explaining the long half-life observed for the bound IκBα in vivo [109-112]. The precise mechanisms, by which the nuclear IκBα removes NFκB dimers from the target genes in vivo are insufficiently understood. Kinetic studies in living cells indicate a dynamic equilibrium between the promoter-bound and free NFκB dimers [113]. According to this model, the dissociating NFκB dimers may be immediately bound by the free IκBα present in the nucleus. In addition, a most recent study using stopped-flow fluorescence and NMR analysis indicates that the removal of NFκB from promoter DNA is a two-step process [111]. First, IκBα forms a ternary complex with NFκB-DNA, and subsequently, the negatively charged PEST domain of IκBα would displace DNA and dissociate NFκB from the promoter [111]. In vivo, several additional mechanisms are likely to be involved in the termination of NFκB activity. These mechanisms include termination of NFκB activation by p65 phosphorylation/dephosphorylation and acetylation, which regulate affinity for IκBα, nucleosome remodeling and the nuclear degradation of p65 NFκB by the associated proteasome [114-121].

Studies from our laboratory have demonstrated that the recruitment of IκBα to NFκB-dependent promoters is genes specific [60,61,70]. In LPS-stimulated human macrophages, the newly synthesized nuclear IκBα induced by post-induction repression is recruited to TNFα, IL-1β, and IL-6 gene promoters, resulting in the transcriptional suppression of these genes [60]. In contrast, the nuclear IκBα is not recruited to IL-8 promoter and the IL-8 expression is not inhibited by the LMB-induced nuclear IκBα [60]. In vivo, the IL-8 promoter is occupied predominantly by p65 NFκB homodimers phosphorylated on serine 536 [60]. Interestingly, this modification was shown to inhibit p65 binding to IκBα in vitro [119]. These studies indicate that the genes occupied by S536 phosphorylated p65 homodimers are not inhibited by the nuclear IκBα (Figure 4). IKKα, IKKβ and IKKε can phosphorylate p65 on serine 536 [122-127]. However, it is not clear at present whether this phosphorylation occurs before p65 binds to DNA or after, as a part of the preinitiation complex assembly. In this context, both IKKα and IKKβ were shown to be recruited to the promoters of NFκB-dependent as well as NFκB-independent genes [128-132], and could phosphorylate the promoter-bound p65, resulting in a prolonged transcription and decreased binding to the nuclear IκBα. Furthermore, the strength of the in vivo nuclear IκBα-p65 NFκB interaction might be influenced by the DNA sequence of κB response elements in the regulatory regions of NFκB-dependent genes. This would be consistent with studies demonstrating that a single nucleotide can influence the recruitment of specific NFκB dimers and the required cofactors for efficient gene transcription [133,134].

Figure 4.

Schematic representation of the regulation of IL-8 transcription by S536 p65 and nuclear IκBα in LPS-stimulated human macrophages. In LPS-stimulated human macrophages, the IL-8 promoter is occupied predominantly by S536 phosphorylated p65 homodimers, which do not bind to IκBα. Consequently, the IL-8 expression is not inhibited by the LMB-induced nuclear IκBα [60]. In contrast, the gene promoters of IL-1β and IL-6 are occupied by p65/p50 heterodimers, and their transcription is repressed by the LMB-induced nuclear IκBα [60].

In addition, the regulation of NFκB-dependent transcription by the nuclear IκBα depends on the subunit composition of NFκB complexes. Our studies indicate that in the chronic T cell leukemia Hut-78 cells, the expression of NFκB-dependent anti-apoptotic genes cIAP1 and cIAP2 is inhibited by the bortezomib-induced nuclear IκBα, while expression of Bcl-2 is not suppressed [70]. Analysis of the in vivo binding of NFκB proteins to cIAP and Bcl-2 promoters by chromatin immunoprecipitation showed that NFκB p65 and p50 subunits are recruited to cIAP1 and cIAP2 promoters, whereas the Bcl-2 promoter is occupied only by NFκB p50. Thus, these data suggest that cIAP1 and cIAP2 promoters associate with NFκB p65/50 heterodimers and this binding and transcription are inhibited by the bortezomib-induced nuclear IκBα. In contrast, Bcl-2 promoter is occupied predominantly by NFκB p50/50 homodimers and its transcription is not inhibited by IκBα.

NFκB-independent function of the nuclear IκBα

The nuclear IκBα not only regulates NFκB binding to NFκB-responsive promoters and NFκB-dependent transcription, but it also physically interacts with different repression elements including nuclear co-repressors, and histone acetyltransferases and deacetylases (HDACs), resulting in transcriptional repression [132,135]. In resting cells, IκBα together with HDACs are recruited to the promoters of Notch target genes correlating with transcriptional repression, whereas in response to NFκB activation, IκBα is released from the chromatin, resulting in Notch-dependent transcriptional activation [136,137]. In addition, IκBα negatively regulates HIV-1 expression by directly binding to the HIV-encoded Tat protein, resulting in the nuclear export and cytoplasmic sequestration of the HIV transactivator [138]. According to this study, IκBα acts as a potent repressor of HIV-1 transcription by inhibiting both NFκB and Tat transacting factors, which are major players in the transcriptional activation and elongation of HIV-1 transcripts [138].

Conclusion

The studies carried out within the last decade clearly demonstrated that in addition to the cytoplasmic retention of NFκB dimers in unstimulated cells, IκBα has important functions in the nucleus as well. Nuclear IκBα is involved in the regulation of numerous pro-inflammatory and anti-apoptotic NFκB-dependent genes as well as NFκB-independent genes through its interactions with HDACs and other transcriptional coregulators. The nuclear translocation and accumulation of IκBα can be induced by the post-induction repression in stimulated cells, by blocking the nuclear export of IκBα by CRM1 inhibitors, and by the proteasome inhibition. A better understanding of the mechanisms regulating the nuclear shuttling of IκBα in stimulated cells, IκBα nuclear translocation and accumulation in response to the proteasome inhibition and the nuclear IκBα accumulation in resting human neutrophils could lead to the development of new therapies aimed at the inhibition of NFκB activity by increased nuclear localization of IκBα. In addition, an important future goal will be to analyze the in vivo NFκB post-translational modifications, and DNA and NFκB subunit preferences of the nuclear IκBα, since they might hold a key to more specific anti-inflammatory and anti-cancer therapies.