A final model of the degradation pathways controlling IκBα in basal and stimulated cells. In the resting cell, enough IκBα is synthesized that it can rebind any NF-κB released due to slow basal IKK activity. Free IκBα is degraded very rapidly, and only represents ∼15% of the total IκBα in the cell (transparent IκBα). When IκBα binds to NF-κB, it is stabilized, and must go through IKK-dependent phosphorylation and degradation (bold IκBα). Upon stimulation, the activity of IKK is increased such that most (if not all) IκBα is rapidly degraded and allows for NF-κB activation. Free IκBα must be continuously degraded to allow for this rapid and robust NF-κB activation.

IKK-independent IκB degradation is critical for signal responsiveness. (A) IκBαΔC288 accumulates as a free protein. Cells were incubated with an α-RelA antibody and immunoprecipitates and flow-through samples were analysed by WB. (^*) Nonspecific bands. (B) NF-κB activation is dampened due to stabilized free IκBα. EMSA for NF-κB activity in response to TNF stimulation in cells expressing WT IκBα, IκBαΔC288, and PESTA IκBα transgenes. (C) Quantitation of the EMSA in (B). Two separate experiments are graphed. WT IκBα is represented by black lines (solid and dashed), PESTA IκBα is represented by grey lines (solid and dashed), and IκBαΔC288 is represented by light grey lines (solid and dashed). (D) Stimulus-induced IKK-dependent phosphorylation of IκBαΔC288 and PESTA IκBα compared with WT IκBα. Phosphorylation of IκBαΔC288 is slightly slower than WT IκBα, whereas phosphorylation of PESTA IκBα is slightly faster than WT IκBα. Western blot showing phosphorylated IκBα after stimulation with 1 ng/ml TNF-α. (E) Stabilized IκBα degrades more slowly than WT IκBα. Left panel: IκBαΔC288 and WT IκBα transgenes in ikba^−/− were treated with 1 ng/ml TNF-α and whole cell extracts were run on SDS–PAGE, and analysed by WB. Right panel: quantification of the left panel. WT IκBα is represented by a black line, whereas IκBαΔC288 is represented by a light grey line.

IκB is an intrinsically unstable protein. (A) Schematic of IκBα primary sequence. The signal response domain (SRD) is followed by the ankryin repeat domain (ARD) and the PEST domain. IKK phosphorylation sites are indicated, as well as all lysines in IκBα. (B) IKK phosphorylation mutants do not inhibit free IκBα degradation. Left panel: western blot (WB) of WT and S32, 36A IκBα from extracts of nfκb^−/− cells treated with cycloheximide (CHX) for the indicated times. Endogenous IκBα and transgenic IκBα are denoted by arrows. Right panel: this is presented graphically with three separate experiments plotted with error bars signifying ±s.e.m. (standard error of the mean). (○) Transgenic WT IκBα and (⋄) S32, 36A IκBα. (C) Ubiquitination mutants do not slow free IκBα degradation. Left panel: WB of WT IκBα and different lysine mutants from cell extracts prepared as described in (B). Right panel: this is presented graphically with triplicate experiments plotted with error bars signifying±s.e.m. (○) Transgenic WT IκBα, (□) K21, 22R IκBα and (▵) KR9 IκBα. (D) Free IκBα is degraded by the proteasome. WB showing IκBα in the extracts of transgenic cells were treated with MG132. All protein levels increase over time, which show that the proteasome is involved in the degradation of free IκBα.

Rapid IκBα turnover is conferred by intrinsic C-terminal sequences. (A) Schematic representation of various IκBα constructs tested for their susceptibility to degradation as a free protein. (B) The N terminus of IκBα does not determine degradation of IκBα in the free form. Left panel: WB of IκBα in extracts of cells expressing WT and ΔN67 IκBα mutants in nfκb^−/− cells after treatment with cycloheximide (CHX). Right panel: triplicate experiments are represented graphically with error bars signifying ±s.e.m. Construction of stable cells and preparation of cell extracts were carried out as described in Figure 2. (•) Transgenic WT IκBα and (▴) ΔN67 IκBα. (C) The C terminus of IκBα controls the rapid degradation of free IκBα. Left panel: cells expressing transgenic IκBαΔC288 and IκBαΔC303 were treated with CHX for the indicated times and cell extracts were visualized by WB. Right panel: triplicate experiments representing WT IκBα and IκBαΔC303 are represented graphically. (•) Transgenic WT IκBα and (▪) IκBαΔC303. (D) The PEST domain of IκBα is responsible for high turnover rate of free IκBα. Left panels (top and bottom): cells expressing transgenic WT IκBα, IκBαΔC288 and PESTA IκBα were treated with CHX for the indicated times and cell extracts were visualized by WB. (^*) A nonspecific band, which can be used as a loading control. Right panel: triplicate experiments representing the relative degradation rates of WT IκBα (black line), IκBαΔC288 (light grey line) and PESTA IκBα (grey line) are represented graphically.

IKK phosphorylation of free IκBα does not require NF-κB, and has no functional consequence. (A) Computational simulations of basal NF-κB activity or of peak NF-κB activity after TNF stimulation. The IKK-dependent degradation rates of the free IκBαs were increased to match the IKK-dependent degradation rates of the NF-κB-bound IκBs. In the ‘wild-type' (WT) model, the IKK-dependent turnover rate is 40-fold more efficient for NF-κB-bound IκB than for free IκBα. This is decreased to 20-fold, and then finally to equal amounts of phosphorylation between free and bound IκBα. (B) Computational simulations of nuclear NF-κB levels in response to TNF for cells expressing WT IKK (black line) or a mutant IKK that does not discriminate between free IκBα and bound IκBα (grey line). (C) Computational simulations of total IκBα levels in response to TNF for cells expressing WT IKK (black line) or a mutant IKK that does not discriminate between free IκBα and bound IκBα (grey line). (D) Free IκBα is phosphorylated in response to TNF-α. Top panel: WB of P-IκB in response to TNF in ikba^−/− and in nfkb^−/− cells expressing WT IκBα. The amount of protein in each time-course set was adjusted to equalize the amount of total IκBα protein in the unstimulated cell extract. Bottom panel: the amount of phospho-IκBα (normalized to total IκBα levels) is compared between nfkb^−/− and diluted extracts from ikba^−/− cells after 10 min of stimulation. (E) Computational simulations of free IκBα levels in response to TNF for nfκb^−/− cells expressing either WT IκB (black line) or a virtual IκBαΔC288 mutant with a five-fold decrease in the IKK-independent degradation rate of free IκBα (grey line). (F) Free IκBαΔC288 is more responsive to TNF-α than WT IκBα. WB of IκBαΔC288 and WT IκBα after stimulation with 1 ng/ml TNF-α. (^*) A nonspecific band, which can be used as a loading control.

NF-κB determines the degradation pathway of IκBα. (A) NF-κB protects IκBα from proteasomal degradation in vitro. Top: purified 20S proteasome and IκBα were incubated at 37°C, with or without purified p65. (I) represents the proteasome inhibitor, MG132. (B) IκBα is highly stable in vivo in the presence of NF-κB. Left panel: WB showing WT, IKK phosphorylation and ubiquitination-defective mutants introduced into ikba^−/−, where all NF-κB subunits are present. Cells were treated with cycloheximide (CHX) for different lengths of time (up to 24 h) and the protein levels were visualized by WB. Right panel: this experiment was repeated twice and is represented graphically with error bars signifying ±s.e.m. (○) Transgenic WT IκBα, (□) K21, 22R IκBα, (▵) KR9 IκBα and (⋄) S32, 36A IκBα. (C) A model of NF-κB repression by IκBα in pre-stimulated cells. There are two processes that control IκBα degradation. In the resting cell, basal IKK activity phosphorylates bound IκBα and targets it for ubiquitin-dependent degradation. In addition, free IκBα is continuously synthesized and degraded in an IKK- and Ub-independent mechanism. This keeps NF-κB from being activated in the resting cell.