RelAp43, a member of the NF-κB family involved in innate immune response against Lyssavirus infection

Abstract

NF-κB transcription factors are crucial for many cellular processes. NF-κB is activated by viral infections to induce expression of antiviral cytokines. Here, we identified a novel member of the human NF-κB family, denoted RelAp43, the nucleotide sequence of which contains several exons as well as an intron of the RelA gene. RelAp43 is expressed in all cell lines and tissues tested and exhibits all the properties of a NF-κB protein. Although its sequence does not include a transactivation domain, identifying it as a class I member of the NF-κB family, it is able to potentiate RelA-mediated transactivation and stabilize dimers comprising p50. Furthermore, RelAp43 stimulates the expression of HIAP1, IRF1, and IFN-β - three genes involved in cell immunity against viral infection. It is also targeted by the matrix protein of lyssaviruses, the agents of rabies, resulting in an inhibition of the NF-κB pathway. Taken together, our data provide the description of a novel functional member of the NF-κB family, which plays a key role in the induction of anti-viral innate immune response.

Figure 1. RelAp43 is a variant of RelA expressed in cell lines and tissues.

(A) Schematic representations of RelA, RelAp43 and of the exons and introns present in their corresponding mRNA. Exons (EX on the Figure) are represented by boxes numbered with arabic numbers, and introns are indicated by red lines with roman numbers. While RelA protein is encoded by a mRNA composed of all 11 exons, a part of the intron IX (figured as a red box in the picture), followed by a stop codon, is found in the mRNA transcript encoding RelAp43. Rel Homology domains (RHD) of RelA and RelAp43 are depicted here, such as the transactivation domain (TAD) of RelA. (B) Absolute quantification of RelA- (black bars) and RelAp43- (grey bars) encoding mRNA in the indicated cell line. (C) Absolute quantification of RelA (black bars) and RelAp43 (grey bars) transcription in human primary tissues. SC: spinal cord; CT: cerebral trunk; FC: frontal cortex; Ce: cerebellum; MC: peripheral blood mononucleated cells; SB: skin biopsy. Results are the mean mRNA level obtained after 3 independent experiments. (D) Expression of RelAp43 in HeLa cells transfected either with a control siRNA, a specific siRNA directed against RelAp43 (RelAp43 siRNA) or overexpressing FLAG-tagged RelAp43 (RelAp43 F). Western blot analysis was performed with an antibody targeting RelAp43 C-terminal part on 50 µg of total cell lysates of each condition. See also Fig. S2. (E) Immunoprecipitation of RelAp43 in HeLa and SK-N-SH (SK) cells transfected either with a control siRNA, a specific siRNA directed against RelAp43 (RelAp43 siRNA) or overexpressing FLAG-tagged RelAp43 (RelAp43 F) or non-transfected. Immunoprecipitations were performed with a pre-immune serum or with antibody directed against RelAp43 N-terminal part then analyzed by western blot revealed either with an antibody targeting RelAp43 C-terminal part or with an anti 3xFLAG antibody. Only 30% of the sample corresponding to HeLa cells overexpressing RelAp43F was loaded on the gel in comparison with the other samples that were loaded in totality after immunoprecipitation.

Figure 2. RelAp43 interacts with all human members of the NF-κB family and is supported by IκBα.

All experiments were performed three times independently. (A) IP using anti-FLAG antibody of FLAG-tagged RelAp43 (RelAp43 F) or RelA (RelA F) or CAT (CAT F) expressing cells or non-transfected cells (NT). The presence of FLAG-tagged protein or endogenous transcription factors of the NF-κB family was analyzed by western blot using specific antibodies in cell lysates either before (left panel) or after IP (right panel). (B) Cellular localization of transfected RelAp43 (green) analyzed by immunofluorescence and apotome imaging in absence (left panel), or in presence of transfected IκBα (red, right panel). Nuclei were visualized using DAPI staining. The scale bar corresponds to 2 µm. (C) Western Blot analysis of nuclear (N) and cytosolic fractions (C) of RelAp43 V5 and IκBα F co-transfected cells after nuclear-cytosolic fractionation. Nuclear and cytosolic fractions were controlled using anti histone H3 and anti actin β antibodies, respectively. See also Figure S2.

Figure 3. RelAp43 binds on DNA κB sites and potentiates RelA-mediated transactivation of gene expression.

All experiments were performed three times independently. Error bars indicate standard deviations. *p<0.05. (A) EMSA analysis of nuclear extracts from FLAG tagged-CAT and FLAG tagged-RelAp43 transfected HeLa cells, using a radioactive κB probe. Cells were either left untreated (−, lanes 1 and 8), treated for 15 min with 10 ng/mL TNF-α (+, lanes 2 and 9) or treated for 15 min with 10 ng/mL TNF-α followed by chase periods of 30 min to 6 hours (lanes 3–7 and 10–14). Band A: RelA bound to κB-probe; band B: RelAp43 bound to κB-probe, as demonstrated in Figure 3B. (B) Supershift analysis of the complexes bound to the κB-probe. 100 ng of α-RelA antibody, α-FLAG antibody or pre-immune serum was incubated with the EMSA reaction mixture before gel electrophoresis. (C) Measurement of the luminescence of cells expressing luciferase under control of the Gal4 promoter and RelA or RelAp43 fused to the Gal4 DNA Binding domain (named as DB-RelA or DB-RelAp43 on the figure, respectively) or non transfected (NT). DB-RelA was arbitrary set to 1. (D) Luciferase assay of RelA transactivation properties in presence of CAT or RelAp43. RelAp43- or CAT-encoding plasmids were added to the transfection mix in the same conditions as in (C) (see also Figure S3). Their expression level was controlled by the western blot analysis showed on the right of the Figure. The level of luminescence obtained in the presence of DB-RelA and CAT encoding plasmids was arbitrary set to 1.

Figure 4. RelAp43 modifies the equilibrium of p50-comprising dimers and specifically potentiates the transcription of several NF-κB dependent genes.

Analysis of the stability of RelA and p50 using cycloheximide blockage of protein translation in HeLa cells overexpressing RelAp43 (A) or CAT (B). Quantification of the RelA and p50 expression, corrected by the actin controls, are given under each RelA and p50 bands. Expression levels in cells not treated with cycloheximide (time 0) were used as controls and set up to 100. (C) Modulation of transfected RelA-p50 dimer formation according to the amount of RelAp43. Dimer formation involving transfected RelA and p50 was analyzed by co-IP using anti-p50 antibody on protein extracts of HeLa cells cotransfected by FLAG-tagged RelA, V5-tagged p50 and FLAG-tagged RelAp43 or CAT as a control. The amount of FLAG-tagged RelA interacting with V5-tagged p50 was assessed using anti-FLAG antibodies. The quantification levels presented at the bottom of the lanes are the means and standard deviations (SD) of the intensity of the band corresponding to RelA obtained by substracting the intensity of the band observed with the pre-immune serum to the one obtained after p50 immunoprecipitation. These experiments (A, B and C) were repeated 3 times independently and results presented are representative of the three repetitions. (D) Modulation of endogenous RelA-p50 dimer formation according to the amount of endogenous RelAp43. Quantification of RelA obtained after co-IP using anti-p50 antibody (IP: p50) was corrected by that observed using a pre-immune serum (IP: pre-immune). Expression levels in cells transfected with control siRNA were used as controls and set up to 100. The experiments were repeated twice independently and the quantification levels presented at the bottom of the lanes are the means and standard deviations (SD) obtained. (E) Relative level of transcription of a set of apoptosis genes in HeLa cells transfected with RelAp43- or CAT-encoding plasmids (on the left) and with RelA- or CAT- encoding plasmids (on the right). The levels of RelAp43 and RelA mRNAs were normalized to the level of GAPDH mRNA and reported relatively to the level measured in CAT-expressing cells used as control (set to 1, not figured). Results presented here are the mean of three independent experiments. For one given repetition, all eight genes were studied on the same cDNAs using specific set of primers. The percentage of induction of HIAP1, IRF1 and IFN-β mediated by RelAp43 and relatively to that mediated by RelA is indicated on the corresponding bars (on the left). *p<0.05.

Figure 5. Specific interaction of RelAp43 with the M protein of rabies virus and inhibition of NF-κB pathway.

The results presented here are representative of three independent experiments. Error bars indicate standard deviations. *p<0.05. (A) Determination of RelAp43 interaction with various lyssavirus M proteins in the yeast two-hybrid system. Yeast cells expressing Gal4-DB fused to M proteins from tested lyssaviruses were co-transformed to express Gal4-AD fused to RelAp43. Cells were plated on synthetic medium lacking histidine so that yeast growth is determined by M interaction with RelAp43 and activation of HIS3 reporter gene. (B) Western blot analysis of co-IP involving M proteins and RelA or RelAp43. Co-IP was performed using anti-FLAG antibody on cells co-transfected with on the one hand plasmid encoding RelA V5, RelAp43 V5 or the common part between them (RHDb) and on the other hand plasmid encoding FLAG-tagged M-Tha, M-SAD, CAT or non-transfected cells (NT). IP of FLAG-tagged proteins (not figured) and co-purification of V5-tagged protein were assessed (WB:V5 on the figure). Initial cell lysates were checked for V5- (cell lysate blot V5) and FLAG-tagged (cell lysate blot 3xFLAG) proteins expression. (C) Modulation of NF-κB activation in the presence of M protein from different lyssavirus strains after short-term stimulation. The NF-κB pathway was exogenously activated using 10 ng/mL TNF-α during 5 h (grey bars) or left untreated (black bars). The M protein of vaccinal strain SAD-B19 was arbitrary considered as a reference. The expression levels of M proteins in each condition were assessed by western blot (bottom of the figure). (D) Modulation of NF-κB activation in the presence of M-Tha (black bars), M-SAD (dark grey bars) or CAT (light grey bars) after extended stimulation. Increasing amounts of tri-phosphate RNA mimicking viral infection during 24 h were used. The expression levels of M-Tha, M-SAD and CAT in each condition were assessed by western blot (bottom of the figure). (E) Luciferase assay of RelA transactivation properties either without any other plasmid (light grey bars) or with a plasmid encoding RelAp43 (dark grey bars) plus a plasmid coding for the indicated M protein (SAD, Tha) or CAT or no additional plasmid (Mock).

Figure 6. RelAp43 is involved in IFN-β transcription during lyssavirus infection.

The results presented are the means of five (A) or three (B, C, D) independent experiments. Error bars indicate standard deviations. *p<0.05. (A) Luciferase assay of IFN-β promoter activation in presence of M protein. M-Tha (black bars), M-SAD (dark grey bars), or CAT (light grey bars) as a control were used. The measure obtained with M-SAD was arbitrary set to 1. The expression levels of M-Tha, M-SAD and CAT in each condition were assessed by western blot (bottom of the figure). (B) Infectious titres of Tha and SAD viruses from supernatant of Hela cells infected at an MOI of 1 at 24 and 48 hours p. i. in the presence of siRNA control (siRNAc) or anti-RelAp43 siRNA (siRelAp43). Titers are given in Fluorescent focus units per ml (FFU/ml). (C) Decrease of the number of RelAp43 mRNA copies in Hela cells infected with Tha or SAD virus and in non-infected cells (NI) in the presence of anti-RelAp43 siRNA (siRNA RelAp43) compared to siRNA control (siRNAc) and measured by quantitative RT-PCR. The level of transcription of RelAp43 in Hela cells measured 48 hours after infection with SAD and in the presence of control siRNA was arbitrary set to 1. (D) Modulation of IFN-β mRNA levels detected by quantitative RT-PCR in HeLa cells transfected by either a control siRNA or an anti RelAp43 siRNA, and infected with Tha or SAD virus compared to non-infected cells (NI).

Figure 7. Proposed model for the role of RelAp43 and of the M protein of lyssavirus in the regulation of the NF-κB dimer repertoire and downstream action on the transcription of different genes involved in innate immune response.

RelAp43 is designed as p43, M protein of lyssavirus wild isolates in orange; laboratory adapted and vaccine strains in green.