Evidence for a phosphorylation-independent role for Ser 32 and 36 in proteasome inhibitor-resistant (PIR) IκBα degradation in B cells

Abstract

Constitutive NF-κB activity has emerged as an important cell survival regulator. Canonical inducible NF-κB activation involves IκB kinase (IKK)-dependent dual phosphorylation of Ser 32 and 36 of IκBα to cause its β-TrCP-dependent ubiquitylation and proteasomal degradation. We recently reported that constitutive NF-κB (p50/c-Rel) activity in WEHI231 B cells is maintained through proteasome inhibitor-resistant (PIR) IκBα degradation in a manner that requires Ser 32 and 36, without the requirement of a direct interaction with β-TrCP. Here we specifically examined whether dual phosphorylation of Ser 32 and 36 was required for PIR degradation. Through mutagenesis studies, we found that dual replacement of Ser 32 and 36 with Glu permitted β-TrCP and proteasome-dependent, but not PIR, degradation. Moreover, single replacement of either Ser residue with Leu permitted PIR degradation in WEHI231 B cells. These results indicate that PIR degradation occurs in the absence of dual phosphorylation, thereby explaining the β-TrCP-independent nature of the PIR pathway. Additionally, we found evidence that PIR IκBα degradation controls constitutive NF-κB activation in certain multiple myeloma cells. These results suggest that B lineage cells can differentiate between PIR and canonical IκBα degradation through the absence or presence of dually phosphorylated IκBα.

Introduction

The transcription factor NF-κB is associated with a number of physiological and pathological processes, including immune system development, inflammation, and the development of various cancers. NF-κB is typically found as a homo- or heterodimer of p50 (NFKB1), p52 (NFKB2), RelA (p65), RelB, or c-Rel. Regulation of NF-κB is mediated by a family of inhibitor molecules called IκB proteins, including IκBα, IκBβ, IκBɛ, IκBγ/p105, IκBδ/p100, or Bcl-3. Individual IκB proteins associate with NF-κB dimers to sequester them in the cytoplasm. Degradation of IκB proteins and release of free NF-κB leads to the regulation of gene expression involved in these different cellular processes [1].

Canonical inducible NF-κB activation through IκBα degradation has been studied extensively and is identified by several hallmark traits. Inactive NF-κB/IκBα complexes are found in the cytoplasm of most cell types. Upon stimulation with a variety of inducers, including tumor necrosis factor alpha (TNFα), bacterial lipopolysaccharide (LPS), and ionizing radiation, the IκB kinase (IKK) complex is activated. Active IKK can then phosphorylate IκBα on its N-terminal Ser 32 and 36 residues [2], [3]. Dually phosphorylated IκBα is recognized by the E3 ligase β-transducin repeat containing protein (β-TrCP) [4], [5], [6]. β-TrCP is known to recognize the specific sequence DpSGψXpS (where pS is phospho-Ser, ψ is a hydrophobic amino acid, and X is any amino acid) [7]. Upon binding, β-TrCP catalyzes poly-ubiquitin chain formation on Lys 21 and/or 22 of IκBα that leads to proteasome-dependent degradation of IκBα allowing the release of free NF-κB heterodimers to direct transcription in the nucleus [8]. Activation of NF-κB via canonical proteasome-dependent degradation of IκBα is shared by many known NF-κB activation pathways [1].

The susceptibility to IKK-dependent phosphorylation is an influential marker in the rate of canonical-inducible degradation of IκB proteins. The IKK complex is a Ser-Thr kinase comprised of two catalytic subunits, IKKα/IKK1 and IKKβ/IKK2, and a regulatory subunit, IKKγ/NEMO [1]. IκBα is a better IKK substrate than IκBβ, both in vitro and in vivo, which correlates with more rapid degradation of IκBα than IκBβ in vivo [9], [10]. Swapping of the N-terminal sequences between IκBα and IκBβ correspondingly altered the rate of degradation, suggesting that the susceptibility to IKK phosphorylation can indeed change the rate of inducible degradation in vivo [9]. Additionally, weaker inducers, such as genotoxic stress agents, activate the IKK complex more slowly and subsequently NF-κB is activated later [11], [12]. These studies together indicate that canonical inducible IκB degradation occurs by a linear process whose rate-limiting step is the phosphorylation of the IκB proteins.

Several studies have revealed that certain subtle mutations of Ser 32 and 36 can permit, while others inhibit, NF-κB activation. Single substitution of Ser 36 with Thr (S36T-IκBα) does not interfere with canonical IκBα degradation, suggesting that Thr at position 36 is permissive for both IKK-dependent phosphorylation and β-TrCP-dependent ubiquitylation [2]. In contrast, single substitution of Ser 32 with Thr (S32T-IκBα) does not allow for inducible degradation [2]. While it is clear that a double Ser to Thr substitution mutant (S32/36T-IκBα) is a much weaker substrate for IKK than wild-type IκBα in vitro [3], [13], [14], it is unclear whether the lack of degradation observed by S32T-IκBα is a consequence of a lack of IKK-dependent phosphorylation. In contrast, single substitution of either Ser 32 or 36 with Glu can still permit ubiquitylation of IκBα and inducible NF-κB activation, suggesting that this substitution has the capability of partially mimicking phosphoserines in vivo [8], [15]. Although a double Ser to Glu IκBα peptide could not compete with wild-type IκBα for β-TrCP recognition in vitro, the effects of double Ser to Glu or Asp mutants on IκBα degradation in vivo have not been clearly established [6], [16]. The differential susceptibilities of these Ser 32 and 36 mutants to canonical proteasome-dependent degradation reveal that the IKK and β-TrCP recognition specificity at these residues is still incompletely understood.

We have found that WEHI231 murine B cells (W231.Bcl-X_L) and normal murine splenic B cells maintain constitutive IκBα degradation and p50/c-Rel activity in a proteasome inhibitor-resistant (PIR) manner [17], [18]. The reason for proteasome inhibitor resistance was in part explained by the lack of a requirement for β-TrCP recognition in this pathway. Peculiarly, dual substitution mutagenesis of Ser 32 and 36 to Ala and certain IKK chemical inhibitors were each able to prevent PIR IκBα degradation in these B cells, suggesting the possible involvement of IKK-dependent dual phosphorylation of IκBα during PIR degradation. However, we were unable to detect dually phosphorylated IκBα species in this pathway. To determine whether dual phosphorylation of IκBα was required for PIR degradation, we now performed extensive mutagenesis studies and analyzed constitutive degradation of IκBα with mutations in Ser 32 and 36 in W231.Bcl-X_L cells. We also analyzed whether PIR degradation of IκBα also occurs in certain multiple myeloma cells. Our results indicate that Ser 32 and 36 function in a manner distinct from the canonical pathway to mediate PIR degradation in certain B cells and further suggest the possibility that this pathway may play a role in proteasome inhibitor resistance or unresponsiveness in certain multiple myeloma [19].