Inhibitor of kappa B epsilon (IκBε) is a non-redundant regulator of c-Rel-dependent gene expression in murine T and B cells

Abstract

Inhibitors of kappa B (IκBs) -α, -β and -ε effect selective regulation of specific nuclear factor of kappa B (NF-κB) dimers according to cell lineage, differentiation state or stimulus, in a manner that is not yet precisely defined. Lymphocyte antigen receptor ligation leads to degradation of all three IκBs but activation only of subsets of NF-κB-dependent genes, including those regulated by c-Rel, such as anti-apoptotic CD40 and BAFF-R on B cells, and interleukin-2 (IL-2) in T cells. We report that pre-culture of a mouse T cell line with tumour necrosis factor-α (TNF) inhibits IL-2 gene expression at the level of transcription through suppressive effects on NF-κB, AP-1 and NFAT transcription factor expression and function. Selective upregulation of IκBε and suppressed nuclear translocation of c-Rel were very marked in TNF-treated, compared to control cells, whether activated via T cell receptor (TCR) pathway or TNF receptor. IκBε associated with newly synthesised c-Rel in activated cells and, in contrast to IκBα and -β, showed enhanced association with p65/c-Rel in TNF-treated cells relative to controls. Studies in IκBε-deficient mice revealed that basal nuclear expression and nuclear translocation of c-Rel at early time-points of receptor ligation were higher in IκBε-/- T and B cells, compared to wild-type. IκBε-/- mice exhibited increased lymph node cellularity and enhanced basal thymidine incorporation by lymphoid cells ex vivo. IκBε-/- T cell blasts were primed for IL-2 expression, relative to wild-type. IκBε-/- splenic B cells showed enhanced survival ex vivo, compared to wild-type, and survival correlated with basal expression of CD40 and induced expression of CD40 and BAFF-R. Enhanced basal nuclear translocation of c-Rel, and upregulation of BAFF-R and CD40 occurred despite increased IκBα expression in IκBε-/- B cells. The data imply that regulation of these c-Rel-dependent lymphoid responses is a non-redundant function of IκBε.

Figure 1. TNF inhibits IL-2 gene transcription and attenuates induction of AP-1 and NFAT.

11A2 cells were cultured with or without TNF 2.5 ng/ml for 8 days before (A) restimulation with plate-bound anti-CD3ε 5 µg/ml, PMA 10 ng/ml and ionomycin 50 ng/ml (P+I)_low, or PMA 50 ng/ml and ionomycin 100 ng/ml (P+I)_high for 24 hours. Secreted IL-2 protein was measured by immunoassay (mean +/− SEM, n = 5 experiments); (B) restimulation as above for 4 hours followed by RNA extraction, and analysis for IL-2 mRNA by RNase protection assay (mean +/− SD, n = 3 experiments). After restimulation with (P+I)_low, nuclear and cytoplasmic extracts were prepared, and equivalent amounts of protein assayed by immunoblot for (C) nuclear AP-1 and (D) nuclear NFAT2 (representative of 3 experiments for each condition). Each blot was probed then reprobed for the indicated proteins.

Figure 2. TNF pre-treatment inhibits subsequent induction and nuclear expression of NF-κB and upregulates IκBε.

11A2 cells were cultured with or without TNF 2.5 ng/ml for 8 days before restimulation with PMA 10 ng/ml and ionomycin 50 ng/ml or TNF 50 ng/ml for the time-points indicated. Nuclear and cytoplasmic extracts were prepared and equivalent amounts of protein assayed for the presence of NF-κB transcription factors (A, B) and IκB isoforms (C, D) by immunoblot. Each blot represents at least three similar experiments, probed then reprobed for the indicated proteins.

Figure 3. Transcription factor DNA binding and activity are suppressed by TNF.

Activation of pCD28RR partially restored by exogenous c-Rel. Control and TNF-treated 11A2 T cells were activated with (P+I)_high. Nuclear extracts were incubated with (A) ³²P-labelled NF-κB consensus oligonucleotide or cold competitor probe or (B) ³²P-labelled NF-κB oligonucleotide with or without NF-κB supershifting antibodies, before non-denaturing gel electrophoresis. Protein-bound oligonucleotide was visualised by both phosphorimaging and exposure of the dried gel to photographic film. Images are representative of three experiments. Control and TNF-treated cells were transfected with reporter plasmids (10 µg) expressing firefly luciferase under control of (C) pAP-1, (D) pNFAT/AP-1, (E) pNF-κB_consensus, (F) pCD28RR_{(NF-κB/AP-1)} promoters, and stimulated as above for 6 hours. Firefly luciferase expression relative to pSV40-driven renilla luciferase from a co-transfected plasmid (0.2 µg) is shown (mean +/− SEM for 9, 7, 4 and 5 experiments, respectively). (G) Summary of inducible promoter activities in TNF-treated cells as a percentage of induction in control cells, for all experiments. (H) Control and TNF-treated 11A2 cells were cotransfected with 2 µg of empty vector or c-Rel expression plasmid and pCD28RR reporter plus renilla luciferase plasmids prior to stimulation as for F. The dotted line represents stimulated expression in cells transfected with empty vector (mean +/− SEM, n = 4 experiments).

Figure 4. Association of IκBε with newly synthesised c-Rel and p65/c-Rel complexes in TNF-treated cells.

Control and TNF-treated 11A2 cells were restimulated with (A) (P+I)_high, (B) P+I_low, for the indicated times. Nuclear and cytosolic extracts were prepared. (A) IκBs α, β and ε were immunoprecipitated from 100 µg cytosolic protein and immunoblotted for associated c-Rel and p65. (B) p65-containing complexes were immunoprecipitated from 100 µg cytosolic protein and immunoblotted for associated IκB. Blots were probed and reprobed for the indicated proteins and those shown are representative of more than 3 experiments at both concentrations of P+I.

Figure 5. Increased basal nuclear c-Rel, priming for IL-2 expression, and enhanced sensitivity to TCR stimulation in IκBε−/− T cells.

Splenocytes and LNCs from WT C57BL/6J (IκBε+/+, filled circles) mice, or mice hetero- (+/−, half-filled circles) or homozygous (−/−, open circles) for IκBε deletion, were cultured with anti-CD3 for 48 hours, washed, then stimulated with IL-2 at day 2 and day 5. On day 7, cells were washed and restimulated, or not, with plate-bound anti-CD3 for up to 4 hours. Extracts from resting and stimulated cells were analysed for (A) nuclear and cytosolic NF-κB and IκB – a four-hour time-course is shown – (B) nuclear NF-κB – a two hour time-course representative of three experiments is shown, with nuclear c-Rel and p65 normalised to actin and expressed relative to values for resting IκBε+/+ T cells as indicated. (C) Relative nuclear expression of c-Rel (mean +/− SD, n = 3 experiments). (D) Cytosolic IκB for experiment shown in (B), with IκBα and -β normalised to actin and expressed relative to values for resting IκBε+/+ T cells as indicated. (E) T cells blasts were stimulated, or not, for 6 hours with PMA and ionomycin in the presence of brefeldin A, labelled with anti-CD4-PECy7, fixed, permeabilised, labelled with anti-IL-2-FITC and analysed by flow cytometry. Shown are% IL-2+CD4+ in the live cell gate (mean +/− SD, n = 3 experiments; one mouse per genotype per experiment). (F) LNC were cultured in triplicate with soluble anti-CD3 at the concentrations indicated. At 48 hours, 1 µCi 3H-thymidine was added to each well and culture continued for 18 hours prior to analysis of 3H-thymidine uptake. Mean +/− SEM, n = 5 experiments (one mouse per genotype per experiment); p values refer to IκBe−/− vs IκBε+/+, other P values were not significant.

Figure 6. Increased basal thymidine uptake in splenocytes and LNC, increased LN cellularity and B cell numbers in IκBε−/− mice.

(A) Basal thymidine uptake by splenocytes and LNCs in IκBε+/+, +/− and −/− mice (median, n = 10 experiments). (B) Enlarged axillary lymph node from an IκBε−/− mouse compared to those from IκBε+/+ and IκBε+/−. (C) Total lymph node cell numbers, one mouse per IκBε genotype per experiment (mean +/− SEM, n = 10). (D) LNC were labelled with anti-CD4-PECy7, anti-CD8-PE, anti-B220-FITC and anti-CD11b-APC before flow cytometric analysis. The percentage of each cell subset was multiplied by the total LNC number to give numbers for each cell subset (mean, n = 6 experiments).

Figure 7. Increased basal and BCR-inducible nuclear c-Rel in IκBε−/− splenic B cells.

Splenic B cells were isolated by CD19 positive selection from three each of IκBε+/+ and IκBε−/− mice. (A) Cells were rested for 2 hours then stimulated with anti-IgM F(ab')₂ fragment 10 µg/ml for the times indicated at 10⁷ B cells/ml per condition in a 12-well plate. There were insufficient cells from IκBε−/− mouse 2, following selection, to carry out stimulation of its B cells. Nuclear and cytoplasmic extracts from resting and stimulated cells were analysed for nuclear NF-κB and IκB by immunoblot. Nuclear and cytoplasmic c-Rel were normalised to actin and expressed relative to values for B cells of IκBε+/+ mouse 1 (mouse numbers allocated randomly), as indicated.

Figure 8. Enhanced survival, increased basal expression of CD40 and increased upregulation of BAFF-R ex vivo in IκBε−/− splenic B cells.

Splenic B cells were isolated by CD19 positive selection from three each of IκBε+/+ and IκBε−/− mice, plated at 10⁶/ml in 6-well plates without stimulation for 48 hours, and analysed by flow cytometry for (A) viability, as% B220+ cells in the live cell gate (FSc/SSc), (B) median fluorescent intensity (MFI APC) due to CD40 expression on CD40+ live B cells, (C) MFI FITC due to BAFF-R expression on BAFF-R+ live cells. Histograms are for one set of randomly paired IκBε+/+ and IκB−/− B cells, representative of three for each genotype. In a different experiment, splenic B cells were isolated and plated as above, then stimulated with anti-IgM F(ab')₂ fragment 10 µg/ml and anti-IL-4 50 ng/ml (∼250 u/ml) for 72 hours. Splenic B cells were analysed for (D)% viability (FSc/SSc), (E) CD40 expression (MFI PE), and (F) BAFF-R (MFI APC) by flow cytometry.