An NFκB Activity Calculator to Delineate Signaling Crosstalk: Type I and II Interferons Enhance NFκB via Distinct Mechanisms

Abstract

Nuclear factor kappa B (NFκB) is a transcription factor that controls inflammation and cell survival. In clinical histology, elevated NFκB activity is a hallmark of poor prognosis in inflammatory disease and cancer, and may be the result of a combination of diverse micro-environmental constituents. While previous quantitative studies of NFκB focused on its signaling dynamics in single cells, we address here how multiple stimuli may combine to control tissue level NFκB activity. We present a novel, simplified model of NFκB (SiMoN) that functions as an NFκB activity calculator. We demonstrate its utility by exploring how type I and type II interferons modulate NFκB activity in macrophages. Whereas, type I IFNs potentiate NFκB activity by inhibiting translation of IκBα and by elevating viral RNA sensor (RIG-I) expression, type II IFN amplifies NFκB activity by increasing the degradation of free IκB through transcriptional induction of proteasomal cap components (PA28). Both cross-regulatory mechanisms amplify NFκB activation in response to weaker (viral) inducers, while responses to stronger (bacterial or cytokine) inducers remain largely unaffected. Our work demonstrates how the NFκB calculator can reveal distinct mechanisms of crosstalk on NFκB activity in interferon-containing microenvironments.

Keywords: NFκB; anti-viral response; immunoproteasome; interferon; mathematical model; signaling crosstalk; systems biology; translational inhibition.

Figure 1

A Simplified Model of NFκB Activity (SiMoN) can predict NFκB activity from 3 parameters. (A) Schematic of the key reactions controlling NFκB activity through IκB metabolism. The amount of free, transcriptionally active, NFκB (NFκB activity) is tightly controlled by the amount of IκB; therefore IκB synthesis (reaction T) and free IκB degradation (reaction P) may potentially offer alternative points of control. The primary, canonical activation pathway is through IKK (reaction K), however, interferons do not directly activate IKK. (B) Schematic of the Simplified Model of NFκB (SiMoN), which analytically calculates NFκB as a result of parameters T,P and K. (C) Modeled time-course concentrations of free NFκB (lower), in response to perturbed reaction rates obtained by multiplying the WT parameter value by the multiplier indicated (upper) utilizing the simplified model. (D) Steady-state free NFκB concentrations in response to: increased IKK activity and IκB translation inhibition (left) and increased IKK activity and free IκB degradation (right).

Figure 2

Interferons potentiate NFκB activation in response to the viral PAMP poly(I:C). (A) Electrophoretic mobility shift assay (EMSA) of nuclear NFκB activity in wild-type and ifnar^−/− BMDMs stimulated with LPS and poly(I:C). Quantitated activity is indicated below each band. (B) EMSA of nuclear NFκB activity in TEPMs cultured with or without IFNγ for 24 h prior to exposure to poly(I:C) or LPS. (A,B) show data representative of three biological replicates. Quantitations of phosphorimager data are relative to peak activity in controls which is set to 1. (C) Single-cell tracking of RelA-mVenus localization in 577 Poly(I:C) stimulated BMDMs cultured in the absence or presence (24 h) of IFNβ and IFNγ. Nuclear NFκB activity is indicated as nuclear:cytoplasmic ratio. The time-course response of each tracked cell is displayed as a row in the heatmap with brighter colors corresponding with increasing nuclear localization of NFκB. (D) The average nuclear NFκB activity of 577 tracked cells is shown for naïve and IFNβ- and IFNγ-primed conditions. (C,D) show data representative of two biological replicates.

Figure 3

Type I interferon signaling potentiates late NFκB activity by translational inhibition of IκBα (A) Experiments to determine IκBα translation rate in BMDMs. Top, schematic of the experimental design: ³⁵S-labeled Methionine pulsed at 0 h and following 8 h of Poly(I:C) stimulus. Middle, immunoprecipitates of IκBα following a ³⁵S-methionine pulse at either indicated timepoint. NFkB p65 immunoprecipitates are shown as normalization controls. Bottom, IκBα mRNA analysis using RNA protection assay. Ribosomal protein gene L32 is provided as a control. These data are representative of three biological replicates. Quantitations are relative to basal conditions which is set to 1. (B) Using SiMoN to determine whether the measured changes in the translation rate are sufficient to account for the NFkB activation defect in ifnar^−/− BMDMs. Left, timecourse simulation of NFκB activity in response to IKK activation following poly(I:C) stimulation with and without a 2-fold increase in IκBα translation measured in ifnar^−/− BMDMs (A). Right, bar graph of NFκB activity at the peak and 24 h time point as quantified from the simulation and experiment (Figure 2A). This indicates that the increase in translation rate measured in (A) is not sufficient to account for the decrease in NFkB activity observed in Figure 2A.

Figure 4

Type I interferon potentiates late NFκB activation by poly(I:C) by decreasing IκB translation and increasing bound IκB degradation via elevated RigI expression. (A) Immunoprecipitation kinase assay (kinase A) of IKK activity in WT and ifnar^−/− BMDMs in response to poly(I:C) and LPS. (B) Immunoblot of RIG-I expression after 8 h of poly(I:C) or LPS treatment in WT, ifnar^−/− and trif^−/− BMDMs; and rescue of trif^−/− cells with IFNβ. (C) EMSAs of NFκB activation by poly(I:C) and LPS in trif^−/− BMDMs with and without IFNβ co-treatment. (D) IKK activity in WT and ips^−/− BMDMs exposed to poly(I:C) and in ips^−/− cells with co-treatment with IFNβ. (E) IKK activity in trif^−/− BMDMs with and without IFNβ co-treatment. (F) EMSAs of NFκB activation by poly(I:C) in ips1^+/+ and ips1^−/− BMDMs. (A–D) show a dataset representative of at least three biological replicates, and (E,F) show a representative of two biological replicates (we gratefully acknowledge Zhijian James Chen for ips1^−/− bone marrow). Quantitations are relative to basal or peak activity, which is set to 1. (G) (Left) Simulated NFκB timecourse in response to IKK activation representative of poly(I:C) stimulation, with a 2-fold increase in IκBα translation (blue) or with both IκBα translation inhibition and 50% IKK activity reduction as seen in ifnar^−/− (green). (Right) Bar graph of NFκB activity at the peak and 24 h time point as quantified from simulations and experiments (Figure 2A). (H) Heatmap of NFκB activity calculated using SiMoN for 50 increasing IKK activity values and 50 increasing degrees of translation inhibition (2,500 total points). In both WT and ifnar^−/− poly(I:C) stimulation results in increased IKK activity during the early phase. Following this WT cells undergo 50% translation inhibition and IKK activity decreases. ifnar^−/− cells lack translation inhibition (horizontal dashed line, Figure 3), and have decreased late-phase IKK activity [vertical dashed line, this (A–F)].

Figure 5

Type II interferon amplifies weak NFκB activating stimuli by enhancing free IκBα degradation. (A) IκBα translational synthesis rates in naïve and IFNγ-conditioned TEPMs as revealed by ³⁵S-Met pulse assay. Average and standard deviation of three biological replicates are shown. (B) Immunoblot for p-IκBα in TEPMs exposed to either LPS or poly(I:C) with or without IFNγ priming. (C) Immunoblot of “free” IκBα compared to an actin control in MEFs deficient in canonical NFκB proteins RelA, cRel and p50 (termed “nfkb^−/−”) exposed to IFNγ. (D) Free IκBα levels in nfkb^−/− MEFs compared to an actin control. Immunoblot of lysates produced from MEFs exposed to 24 h priming with IFNγ or 4 h treatment with ribosomal inhibitor CHX, and followed by addition of proteasome inhibitor MG132. (E) Predictions from the Simplified Model of NFκB (SiMoN) with low (10% at peak) IKK activity, representative of poly(I:C) (blue), and high (40% at peak) IKK activity, representative of LPS. Values were calculated at 0, 0.5, 1, 2, and 4 h and fit with a smoothing spline for consistency with experimental time points. Free IκBα degradation was modulated from the default value (dashed lines) to 10-fold higher (solid lines) based on quantification of immunoblot in 5B. (F) Time course of NFκB induction (quantitated from EMSAs) in naïve or IFNγ-conditioned TEPMs stimulated with poly(I:C) and LPS. (G) Nuclear NFκB activity calculated using SiMoN as a function of bound IκBα degradation (IKK-activity, x-axis) and free IκBα degradation (colored lines). The blue and red arrows indicates the free IκBα degradation-dependent increase in NFκB activity for low and high IKK activities indicative of poly(I:C) and LPS, respectively. (H) Immunoblots of proteasome activator 28 (PA28) levels in TEPMs following exposure to IFNγ. (I) Immunoblots for IκBα and proteasome activator 28 (PA28) in nfkb^−/− MEFs. Both conditions were repeated following PA28 siRNA-mediated knockdown. (J) Immunoblot of IκBα and PA28 levels in nfkb^−/− MEFs transduced with retroviral transgenes. (K) Coomassie-stained SDS-PAGE showing free IκBα and PA28α/β levels following incubation with increasing amounts of purified 20S proteasome (upper panel) contrasted with GST-ubiquitin levels (lower panel), which serves as a negative control. (B–D) show a dataset representative of at least three biological replicates (H–K) show a dataset representative of two biological replicates. Quantitations are relative to basal or t = 0 activity, which is set to 1.

Figure 6

The mechanisms underlying interferon signaling crosstalk on NFκB. (A) Type I interferons reduce IκB expression and increase IKK activity through RIG-I and IPS-1. Type II interferons increase free IκB degradation through a PA28-dependent process. (B) Type I interferons reduce translation of IκBα and increase the expression of the cytosolic viral sensor to allow for enhanced IKK mediated degradation of NFκB-bound IκBα. Type II interferon increases the degradation rate of free IκBα. All mechanisms potentiate the NFκB response to weak signals emanating from viral PAMP sensors, but have little effect on bacterial-MyD88-mediated responses. (C) Three-dimensional heatmap of nuclear NFκB concentrations as a function of three biochemical reactions: IKK activity (reaction K), IκB translation efficiency (reaction T) and free IκB degradation (reaction P). The point in this parameter space reached following Poly(I:C) and LPS stimulus is marked with black circles. Signaling crosstalk by Type I and Type II interferons produce distinct trajectories through this three-dimensional parameter space (marked with white arrows to white circles).