In this study, we determined the respective roles of RelA and RelB NF-κB subunits in Epstein–Barr virus (EBV)-transformed B cells. Using different EBV-immortalized B-cell models, we showed that only RelA activation increased both survival and cell growth. RelB activity was induced secondarily to RelA activation and repressed RelA DNA binding by trapping the p50 subunit. Reciprocally, RelA activation repressed RelB activity by increasing expression of its inhibitor p100. To search for such reciprocal inhibition at the transcriptional level, we studied gene expression profiles of our RelA and RelB regulatable cellular models. Ten RelA-induced genes and one RelB-regulated gene, ARNTL2, were repressed by RelB and RelA, respectively. Apart from this gene, RelB signature was included in that of RelA Functional groups of RelA-regulated genes were for control of energy metabolism, genetic instability, protection against apoptosis, cell cycle and immune response. Additional functions coregulated by RelA and/or RelB were autophagy and plasma cell differentiation. Altogether, these results demonstrate a cross-inhibition between RelA and RelB and suggest that, in fine, RelB was subordinated to RelA. In the view of future drug development, RelA appeared to be pivotal in both classical and alternative activation pathways, at least in EBV-transformed B cells.
Epstein–Barr virus (EBV) is responsible for immunodeficiency-related diffuse large B-cell lymphomas (DLBCLs) of posttransplant or human immunodeficiency virus-infected patients. EBV is also associated with various other cancers, including DLBCLs of the elderly, Burkitt, Hodgkin’s or T-cell lymphomas and nasopharyngeal carcinomas.1 In vitro, EBV infects and transforms primary B cells, leading to the continuous proliferation of lymphoblastoid cell lines. This proliferation program, also called latency III program, is driven by the EBNA2 (Epstein–Barr Nuclear Antigen 2) protein, which regulates expression of the entire set of EBV latent genes, including the BNLF1 gene coding for the main EBV oncogene, the latent membrane protein 1 (LMP1). Other EBV latent proteins are EBNA1, required for episomal maintenance of the EBV genome, the EBNA3 proteins that modulate EBNA2 and regulate cell proliferation and LMP2A that mimics the B-cell receptor. Some EBV-associated B-cell lymphomas such as endemic Burkitt lymphomas or the rare primary effusion and plasmaplastic lymphomas express only the EBNA1 (so-called latency I). However, most EBV-associated tumors express LMP1, even in latency III or II (expression of EBNA1 and LMP proteins). LMP1 is a transmembrane protein acting as a constitutive active CD40 receptor, thereby continuously activating NF-κB.2 Previous results demonstrated that preserved NF-κB activity and protection against apoptosis would be the minimal prerequisite for all LMP1 natural mutated variants isolated from both normal and Reed-Sternberg cells from Hodgkin’s lymphomas.3
Two NF-κB activation pathways have been described, so-called classical (or canonical) and alternative (or non-canonical) (Vallabhapurapu and Karin4 for review). The classical pathway is induced in response to a variety of stimuli, such as CD40-Ligand, TNFα, IL-1, IL-6, bacterial lipopolysaccharide, as well as LMP1 of EBV. It involves RelA- or c-Rel- and p50-containing complexes. In resting cells, NF-κB dimers containing these subunits are retained in the cytoplasm by physical interaction with IκBα, β, ɛ or the p105 precursor of p50. Following activation of the classical NF-κB pathway, the IκBs and p105 are rapidly phosphorylated by the IκB kinase complex (IKK), containing the catalytically active kinases IKKα and IKKβ and the regulatory scaffold protein NEMO (NF-κB Essential Modulator or IKKγ).5 Phosphorylation of IκBs leads to their proteasomal degradation, releasing NF-κB dimers that translocate into the nucleus where they activate transcription of specific target genes.4 With much slower kinetics, the alternative NF-κB activation pathway is induced by a restricted subset of receptors such as the Lymphotoxin β receptor, B-cell-activating factor receptor or CD40, which then leads to NF-κB-inducing kinase activation.6 NF-κB-inducing kinase phosphorylates IKKα, which in turn mediates phosphorylation and proteolysis of p100, the precursor of p52. P100 acts as an IκB molecule specifically trapping RelB-containing complexes (i.e., RelB/p50 and RelB/p52 dimers). P100 proteolysis allows nuclear translocation of these complexes.7
Most reports addressing the question of NF-κB activation by LMP1 used reporter gene assays or studied the involvement of NEMO/IKKα/IKKβ or NF-κB-inducing kinase/IKKα, and most studies were performed in non-B-cell lineages.8, 9, 10, 11 Very few studies have addressed the question of the specific roles of RelA and RelB in EBV-transformed B cells. This question is of importance not only to understand the place of both subunits in B-cell transformation but also in the view of developing new drugs targeting NF-κB. Here, we addressed the question of the function, DNA-binding activity and gene expression signatures of RelA and RelB in EBV-immortalized B-cell line models.
Materials and methods
Patients and biopsies
For this study, 11 patients, three with EBV-positive posttransplant DLBCLs (EBV-DLBCL n° 1–3, following kidney transplantation for patients 1 and 2 and curative bone marrow transplantation for T-cell lymphoma for patient 3; patient 1 was under mycophenolate and tacrolimus and patients 2 and 3 were under cyclosporine A regimen. Patient 3 initially received one dose of antilymphocyte globulins), one with EBV positive human immunodeficiency virus-associated DLBCL (EBV-DLBCL n° 4), four with EBV-positive DLBCLs of the elderly (EBV-DLBCL n° 5–8), and three with nontumoral reactive lymph nodes (LN n° 1–3), were enrolled according to institutional regulations and after approval by the IRB of the university hospital of Limoges. EBERs were detected in all tumor cells. Tumor cells expressed CD20 and LMP1 confirmed by immunohistochemical analysis in all cases.
Complementary DNAs for LMP1,12 Luciferase (Promega, Paris, France), IκBαS32,36A,12 RelA, RelB13 and p100/p5213 were cloned into the previously described pRT-1 doxycycline-inducible episomal vectors.14 The RelA insert was obtained by long distance amplification of the corresponding mRNA from an LCL cell line.
Cell lines, CD40-Ligand stimulation and cell transfection
Classical lymphoblastoid cell lines PRI, 1602, TSOC, LCL.4 and LCL.6 have been described in previous reports.3, 15 EREB2–5 cells are a nonclassical LCL cell line with an estradiol-inducible EBV-latency III proliferation program due to an estrogen receptor fused to the EBNA2 viral protein.16 Transfection, hygromycin selection and CD40-Ligand stimulation of EREB2–5 cells were performed as previously described.17, 18
Protein extracts and electrophoretic mobility shift assays (EMSAs)
Immunoprecipitation and western blot
Immunoprecipitation of p50 was performed as described in Supplementary Materials and Methods. Western blots were performed as previously described.20 The antibodies used are detailed in Supplementary Materials and Methods.
Analysis of apoptosis by flow cytometry
Cells were double stained with AnnexinV-FITC (BD Pharmingen, San Diego, CA, USA) and propidium iodide (Sigma-Aldrich, Saint-Louis, MO, USA) in cold PBS-CaCl2-MgCl2 (Invitrogen, Cergy-Pontoise, France) as described20 and analyzed by FACS Calibur cytometer (BD Pharmingen, Paris, France).
Transfected EREB2–5 cells were plated in 96-well plates (103/well), in 10% fetal calf serum medium containing or not doxycycline. Over 2 days, proliferation rates were measured using the CellTiter 96 AQueous One Solution Cell Proliferation Assay (Metabolism Tetrazolium Salt assay from Promega).
Cell sorting and RNA isolation
EREB2–5 cell line conditions used for gene expression profiling are detailed in Supplementary Materials and Methods. When indicated, NGFRt-expressing EREB2–5 transfected cells were purified with MACS microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) following previously published protocols.12 Isolation of the high-quality total RNA was performed from samples (cell lines and tissues) in the TRIzol Reagent (Invitrogen) using the RNeasy mini kit (Qiagen, Valencia, CA, USA). High quality (integrity and purity) of RNA was verified by the Agilent RNA 6000 Nano LabChip kit and the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).
Gene expression profiling
Amplification of RNAs and hybridization onto microarrays were performed on an Affymetrix Gene Atlas system with the Affymetrix Human Genome U219 Array. Data analysis was performed according to both the LIMMA and SAM methods.21, 22, 23 Biological functions of genes were studied using ‘Gene Set Enrichment Analysis’ (www.broadinstitute.org/gsea/index.jsp). Details are in Supplementary Materials and Methods.
Gene quantification with TaqMan low density array
cDNAs were reverse transcribed from total RNA samples using the High Capacity cDNA Archive Kit (Life Technologies, Carlsbad, CA, USA). PCR products were amplified from 200 ng of each cDNA sample using the TaqMan Universal PCR Master Mix and TaqMan Gene Expression Assays preloaded in each reaction well of TaqMan low density arrays (Life Technologies). Thereafter, TaqMan low density arrays were run on the 7900HT system for quantitative real-time PCR analysis. Details are in Supplementary Materials and Methods.
RelA and RelB differentially regulate proliferation and survival in EBV-immortalized B cells
To assess functional roles of RelA and RelB, we cloned the corresponding cDNAs and that of their respective inhibitor, IκBαS32,36A super-repressor and p100, into the doxycycline-regulatable pRT-1 vector.14 The same vector coding for Luciferase15 was used as control. After estradiol deprivation to induce arrest of the EBV-latency III program, EREB2–5 B cells were either pre-exposed or not to doxycyline for 24 h and then treated with estradiol for both 24 and 48 h. In this model, estradiol treatment of cells induces the EBV-latency III program with LMP1 expression that induces both NF-κB activation pathways.24
At 48 h, overexpression of each protein was checked by Luciferase assays or by western blot (Figure 1a). As expected, induction of the EBV-latency III program (E24hEBV and E48hEBV) reverted spontaneous apoptosis and induced cell growth in estradiol-starved E0hEBV cells in Luciferase-expressing cells (Figures 1b and c). RelA induction by doxycycline weakly over-increased protection against apoptosis and markedly over-increased cell proliferation when compared with Luciferase or uninduced EBV-latency III proliferating control cells, respectively (Figures 1b and c). Induction of the super-repressor form of IκBα, IκBαS32,36A, which inhibits the classical pathway,19, 25 abolished both EBV-induced protection against apoptosis and growth (Figures 1b and c). Surprisingly, induction of both RelB and p100 decreased EBV-latency III protection against apoptosis (Figure 1b). RelB also repressed EBV-induced proliferation of EREB2–5 cells, whereas p100 had no effect (Figure 1c). These results clearly suggest that RelA activation was associated with both proliferation and protection against apoptosis in EBV-latency III immortalized B cells. The role of RelB on apoptosis and proliferation would be more complex.
Dynamics of NF-κB complexes in EBV-immortalized B cells: reciprocal inhibition of RelA and RelB
To correlate functional assays with NF-κB DNA binding activities, we performed EMSA. Controls for specificity are given in Supplementary Figure S1. First, nuclear extracts from five EBV-latency III immortalized B cell lines, all with LMP1 expression, were isolated. NF-κB DNA binding patterns systematically corresponded to three bands, BI, BII and BIII (Figure 2a). Supershift experiments indicated that band B-I was likely to correspond to RelA/p50 dimers, band B-II to both c-Rel and RelB complexes and band B-III to p50 homodimers. Kinetics of EBV-latency III program induction in EREB2–5 cells showed that NF-κB activation paralleled LMP1 expression levels and reached maximum rates at 6 h (Figure 2b, lanes 1–5). After LMP1 expression or stimulation of its cell homolog CD40 for 24 h, the same NF-κB complexes with the same bands BI, BII and BIII were identified (Figure 2b, lanes 7–18). At the protein level, NF-κB subunits of the classical pathway, that is, RelA- and c-Rel-containing complexes, were rapidly translocated into the nucleus starting at 1 h after EBV-latency III induction (E1hEBV), whereas RelB complexes were activated later at 24 h (E24hEBV, Figure 2c). These results clearly show a rapid and strong activation of the classical NF-κB pathway and much delayed, slower and weaker activation of the alternative pathway by the EBV-latency III program.
Second, we analyzed NF-κB complexes in EREB2–5 cells transfected with RelA, RelB, IκBαS32,36A super-repressor and p100 after 48 h EBV-latency III program induction. Luciferase expression had no effect on NF-κB binding activity (Figure 3a, lane 2). RelA induction was associated with (i) an increase in band B-I containing the RelA/p50 and RelA/RelA dimers (Figure 3b, lanes 2–5) and (ii) a decrease in band B-II, now completely supershifted with c-Rel antibody (Figure 3b, lanes 2, 3 and 8). Thus, RelA overexpression repressed RelB but not c-Rel DNA binding activity. The expression of IκBαS32,36A led to a global decrease in NF-κB DNA binding activity (Figure 3c, lane 2). Over-expression of RelB deeply modified the NF-κB DNA binding activity, with two major bands: one corresponding to B-II and one new band, called B-IV (Figure 3d, lanes 2 and 3). Both bands B-II and B-IV were completely supershifted with RelB antibodies (Figure 3d, lane 6). As attested by the absence of supershift with RelA and c-Rel antibodies, RelB overexpression repressed NF-κB complexes activated by the classical pathway (Figure 3d, lanes 4 and 8). Moreover, band B-IV was completely supershifted with p50 antibodies, suggesting that p50 was the main partner of RelB (Figure 3d, lane 5). Finally, overexpression of p100 did not affect RelA DNA binding activity and specifically induced the loss of RelB binding activity in band B-II, such a loss was suggested by the complete extinction of band B-II in the presence of c-Rel antibody (Figure 3e, lane 8). This result can be interpreted as a specific inhibition of RelB DNA binding activity by an excess of p100 protein, which would trap RelB in the cytoplasm.
Mechanisms of RelA and RelB reciprocal inhibition
As p50 was the main partner of RelA and RelB by EMSA, we looked for the p50 partners by immunoprecipitation. P50 was indeed associated with RelA, RelB and p52 in Luciferase control cells (Figure 4, lane1). Induction of RelA was associated with a marked increased in immunoprecipitated RelA/p50 heterodimers and a slight increase in immunoprecipitated RelB/p50 complexes (Figure 4, lane 2). Furthermore, p50 was additionally associated with p100 in these RelA-overexpressing cells, which correlates with an increase in p100 expression in the input (Figure 4, lanes 1 and 5). These results led to the hypothesis that RelA overexpression could indirectly repress DNA binding of RelB/p50 complexes by increasing p100 expression, which in turn would trap RelB-containing complexes in the cytoplasm.
When RelB was overexpressed, RelB/p50 dimers were strongly increased and, concomitantly, RelA/p50 complexes were decreased when compared with Luciferase control cells (Figure 4, lane 3). In this condition, p100 expression and association with p50 were also increased (Figure 4, lanes 3 and 6). This p100 induction by RelB was likely to exert some inhibitory negative feedback on RelB but was certainly not sufficient to inhibit binding of all RelB/p50 heterodimers to DNA, as attested by the EMSA experiments presented in Figure 3.
Altogether, results presented in Figure 3 and 4 demonstrated a cross-talk between RelA and RelB so that RelA was likely to inhibit RelB DNA binding by overexpression of p100 and RelB repressed RelA DNA binding by competing for their association with p50.
RelA and RelB transciptomic signatures
To evidence cross-talk between RelA and RelB at the transcriptional level, that is, their cross-inhibition, we searched traces of such interplay using high-throughput gene expression profiling.
We first established the list of genes deregulated in EBV-transformed B cells. To avoid gene deregulation due to in vitro cell line artifacts, we also addressed genes truly deregulated in EBV-associated DLBCLs with the expression of LMP1 (three cases of EBV-related posttransplant DLBCL, one case with EBV-positive human immunodeficiency virus-associated DLBCL and four with EBV-positive DLBCLs of the elderly). Such variants of EBV-associated DLBCLs would also allow to prevent any bias due to a particular subtype of EBV-associated B-cell lymphomas.
On one hand, the different B cell lines were divided in two groups, those in which the EBV-latency III program was on or off. Cells expressing LMP1 alone were also included (Supplementary Figure S2). Selected genes yielded a cell line list of 1550 genes upregulated by the EBV-latency III program (Supplementary Table S1). On the other hand, EBV-DLBCL tumors were compared with three non-tumoral lymph nodes with benign follicular hyperplasia. This led to an EBV-related tumor list of 3415 genes (Supplementary Table S2). A total of 726 genes were in common between both lists, being thus bona fide EBV-induced genes in EBV-transformed B cells both in vivo and in vitro (Supplementary Figure S3 and Supplementary Table S3). Among these 726 genes, 612 (84%) were regulated by LMP1 alone and numerous genes were known targets of c-Myc, E2F1, Jun or NF-κB (Supplementary Figure S3 and Supplementary Table S3).
Then, EREB2–5 cells overexpressing Luciferase, RelA, RelB, IκBαS32,36A super-repressor and p100 were sorted after doxycycline induction (i.e., E.Luc, E.RelA, E.RelB, E.IκBαSS32,36A and E.p100 cells, respectively), and their gene expression profiles were analyzed to identify RelA and RelB target genes among the 726 EBV upregulated genes.
Forty-six genes were very likely to be upregulated by RelA and/or RelB in EBV-DLBCL-transformed B cells (Supplementary Materials and Methods for details and Figure 5). Of note, 44/46 genes were upregulated by LMP1 with a fold change ranging from 1.6–85 (Supplementary Table S4). As an independent experiment, we looked at the deregulation of these genes in the series published by Basso et al.26 (GEO Accession Number: GSE2350), which includes purified resting B cells from cord blood, lymphoblastoid cell lines and immunoblastic DLBCLs from both AIDS-related immunodeficient and immunocompetent patients. Indeed, these DLBCL variants are frequently associated with constitutive NF-κB activation.27, 28 Of the 46 RelA and/or RelB target genes, 34 could be retrieved from the gene list of Basso et al., and among them 31 were indeed deregulated either in lymphoblastoid cell lines or in immunoblastic DLBCLs (Supplementary Table S5). Deregulation of these genes was confirmed by quantitative PCR on cDNA from EBV-DLBCL tumors and LCL cells (Supplementary Figures S4 and S5).
Among the RelA and/or RelB target genes selected by our analysis, 45/46 were induced by RelA (Figure 5). Eleven of these 45 genes were also induced by RelB, that is, RelA- and RelB-co-regulated genes (Figure 5a). Twenty four of these 45 genes were apparently not regulated by RelB (so-called RelA-dominant genes) (Figure 5b). This can be interpreted as a simultaneous and overlapping RelA repression by RelB and RelB weak gene upregulation. The remaining 11 RelA-regulated genes were repressed by RelB, so-called RelA-only genes (Figure 5c). RelB induced the expression of one gene, ARNTL2 (Aryl hydrocarbon receptor nuclear translocator-like protein 2), whose expression was repressed by RelA (Figure 5c).
These results clearly suggest that RelB may repress transcription of some RelA target genes and vice versa. Additionally, it appeared that RelB signature was almost entirely included in the one of RelA (Figure 5d).
Functions of RelA- and/or RelB-regulated genes are shown in Supplementary Figure S6 and Supplementary Table S6. Genes dominantly regulated by RelA were associated with metabolism, genetic instability, protection against apoptosis, cell cycle and immune response. Additional functions of genes codominantly regulated by RelA and RelB were autophagy, transcription factors regulating cell cycle and plasma cell differentiation and NF-κB regulators. The only RelB-specific gene (i.e., repressed by RelA) was ARNTL2, a gene involved in the regulation of circadian rhythm and protection against hypoxia.
NF-κB pathways have been implicated in Hodgkin’s lymphomas, multiple myelomas and both EBV-negative and EBV-associated DLBCLs.29, 30, 31, 32, 33 Recently, it has been demonstrated in a mouse transgenic model that LMP1-expressing B cells are under permanent control of both T-cell-acquired and NK cell innate immune responses.34 We indeed demonstrated that LMP1 is able to promote autologous T cell targeting of EBV-infected B cells in humans.12 Despite this permanent immune control, LMP1 remains the main oncogenic agent of EBV, as demonstrated both in human EBV-related tumors and different mice models, including the model developed by Zhang et al.34 Indeed, almost all LMP1-positive EBV-associated B-cell lymphomas are associated with a more or less profound immunodeficiency status. The oncogenic potential of EBV is directly linked to continuous activation of NF-κB by LMP1.35 We thus worked on LMP1-expressing EBV-transformed B cells, raising the question of the respective roles of RelA and RelB. The model that emerged from our results is that RelA and RelB cross-talked together so that RelB was subordinated to RelA and would help RelA in reinforcing the regulation of some genes like those involved in autophagy or metabolism.
As a side conclusion of our work, we found that p50 was the main partner of RelA and RelB in EBV-immortalized B cells. Overexpressed RelB trapped most, if not all, available p50 subunits in the cell, inhibiting therefore DNA binding of both RelA/p50 and c-Rel/p50 complexes. Thus, in EBV-infected B cells, RelA and RelB were in competition for p50, explaining mechanistically how RelB exerted an inhibitory activity on RelA. Although poorly represented in our cell models, RelB/p52 complexes were also increased under RelB overexpression. Although processing of p105 to p50 is constitutive, p100 degradation or processing to p52 results from the activation of alternative signaling, allowing nuclear translocation of RelB-containing dimers associated with p50 or p52, respectively.36 The nfkb2/p100−/−-deficient mice model showed a functional overlap between p50 and p52 when bound to RelB.37 As RelB blocks RelA DNA binding by trapping its main partner, the p50 subunit, it can be put forward that the inhibitory effect of RelB overexpression on both proliferation arrest and increased apoptosis corresponds to its inhibitory effect on RelA. It is only because p100 overexpression was associated with both increased apoptosis and specific lack of RelB DNA binding activity that we can infer that RelB is likely to have a role in protection against apoptosis.
Several recent studies provided data that interconnections between the alternative and classical NF-κB pathways exist through expression control of NF-κB subunits.38 Regulatory regions of relb and nfkb2/p100 genes contain κB binding sites, and the expression of these genes is dependent on RelA.39, 40 In addition, experiments on rela−/− murine embryonic fibroblasts suggested that RelA-dependent expression of RelB is the main determinant of alternative pathway responses through Lymphotoxin β receptor stimulation.41 Consistently, as the overexpression of RelA increased p100 and RelB expression and RelB/p50 association together with decreased RelB DNA binding activity, we interpreted these results as the effect of RelB trapping by p100. However, another alternative would be that RelA and RelB association would directly inhibit RelB DNA binding activity, as reported in the murine embryonic fibroblast model.42 In any case, DNA binding of RelB was always weak in EBV-latency III immortalized B cells when compared with that of RelA or c-Rel or with RelB activity induced by CD40 stimulation. Kinetic experiments on induction of the EBV-latency III program demonstrated that RelA complexes were primarily translocated into the nucleus and paralleled LMP1 expression. Upon EBV stimulation, RelB activation was observed later at 24 h and was concomitant with TRAF3 degradation and p100 cleavage to p52 (not shown). Thus, not only RelB is secondarily induced by RelA but also it is negatively retro-controlled by RelA.
To further search for functional traces of RelA and RelB interactions in EBV-related DLBCLs, we established gene expression profiles of both our B-cell line models and EBV-associated DLBCL tumors. A total of 726 six genes were upregulated in both tumors and cell lines, most of them being regulated by the LMP1 oncogene of EBV. On the basis of RelA and RelB inhibition or activation experiments, 46 genes were found to be regulated either by RelA or RelB. The number of RelA- and/or RelB-regulated genes seemed to be low, when compared with the number of genes induced by LMP1 alone, and probably not all NF-κB target genes were selected following our methodology. However, LMP1 is a very strong activator, and overexpression of IκBαS32,36A or p100 could have not completely abolished NF-κB DNA binding activity in EBV-immortalized B cells. Moreover, as tissue control for EBV-related DLBCLs, we took lymph nodes with benign follicular hyperplasia, that is, lymph nodes with current intense secondary B-cell immune response. Thus, it is not excluded that some NF-κB activation would exist in these control tissues. We can hypothesize that we detected the most NF-κB-sensitive genes deregulated in DLBCLs when compared with benign lymph nodes submitted to an intense immune response.
The expression of ARNTL2 gene was induced by RelB and repressed by RelA overactivation (so-called RelB only). Reciprocally, 10 RelA-regulated genes were repressed by RelB overexpression (so-called RelA only). These results are a strong indication that RelA and RelB reciprocal inhibition may occur at the transcriptional level in transformed B cells. Genome-wide gene expression analysis derived from Lymphotoxin β receptor-stimulated murine embryonic fibroblasts revealed that the majority of the induced genes require both RelA- and RelB-containing dimers.43 Here, RelB gene expression profile was almost included in that of RelA. As RelB activity was secondarily induced by RelA in EBV-transformed B cells and genes repressed by RelB in our experimental in vitro conditions were in fact overexpressed both in vitro and in EBV-related DLBCLs, it is very likely that the optimal equilibrium between RelA and RelB would exist in EBV-transformed B cells so that RelA-specific functions remained and RelA/RelB co-regulated functions were secondarily reinforced.
Taking both RelB and RelA together, one of the main NF-κB signatures was that of metabolism with regulation of genes involved, for example, in cholesterol synthesis, fatty acid degradation or the pentose phosphate pathway. Recent publications have identified NF-κB as a major regulator of energy homeostasis via the regulation of metabolic functions.44 Here, ACSL1 (acyl-CoA synthetase long chain family, member 1) codes for an enzyme that produces acetyl-CoA, the requisite building block for lipid biosynthesis, and also contributes to energy production through the β-oxidation into mitochondria.45, 46 Tumor cells reactivate de novo lipid synthesis such as that of fatty acids, which are important substrates for energy production, essential components of all biological lipid membranes and participate in several signaling pathways that regulate in turn metabolism, proliferation, survival and invasive properties.47
Consistently with functional results, the RelA signature indicated that it is very likely to participate in strong proliferation of EBV-transformed B cells. Indeed, other RelA-regulated genes revealed that RelA was in the center of key functions associated with cell transformation, including regulation of genes involved in cell cycle progression such as BATF (basic leucine zipper transcription factor, ATF-like) and CCNE1 (cyclin E1). The B-cell-specific transcription factor, BATF, promotes latent EBV infection with transforming potential and contributes to in vivo progression of EBV-associated lymphomagenesis.48, 49, 50 G1/S checkpoint alteration by cyclin E1 (CCNE1) overexpression is associated with high cell proliferation and chromosomal instability in many human tumors and related to a poor clinical outcome in DLBCLs.51, 52 RelA could also participate in genetic instability through upregulation of APOBEC3.53 RelA-regulated genes could also modulate the tumor microenvironment through regulatory molecules with pro- or anti-tumor effects such as CD58/LFA3 or IL1R2, respectively.54, 55
In addition to RelA, RelB could contribute to increased energy metabolism by targeting genes directly involved in cholesterol or sorbitol pathways, as well as by increasing the expression of DRAM1 (damage-regulated autophagy modulator) autophagy regulator. Autophagy may promote the survival of tumor cells within the tumor microenvironment by providing an alternate route of nutriment acquisition when tumor cells are deprived of blood flow. In addition, both RelA and RelB regulated i) the well-known NF-κB tumor markers of EBV-DLBCLs such as NFKBIA/IKBA, TNFRSF8/CD30 and TRAF132 and ii) ZBTB32 (zinc-finger and BTB domain-containing protein 32), involved in plasma cell differentiation.56 ZBTB32 is a transcriptional partner of Blimp-1 (B-lymphocyte-induced maturation protein-1), a tumor marker related to apoptosis protection of tumor cells engaged in terminal B-cell differentiation such as in EBV-DLBCL tumors.48, 56
ARNTL2 was found to be regulated by RelB only. With CLOCK, ARNTL1-2 regulates the circadian rhythm of cells and is a cancer risk factor.57 ARNTL2 expression is increased in B leukemia and colorectal cancer, which is related to tumor invasiveness and aggressiveness.58, 59 ARNTL2 is also a partner of HIF1α (Hypoxia-Inducible Factors 1 alpha) suggesting a potential physiological link between circadian rhythm and cellular oxygen status.60
In conclusion, our results demonstrated a cross-talk between RelA and RelB so that RelB appeared to be subordinated to RelA and modulated RelA activity. In that context, NF-κB dependent transformation of B cells could occur in two steps. The first step would be induction of the entire set of RelA target genes by RelA alone, including RelB and its inhibitor p100. The second step would be attenuation of some RelA-specific genes by RelB and long-term reinforced expression of RelA and RelB co-regulated genes. RelA alone would act directly on tumor development and progression through direct regulation of genes associated with energy metabolism, genetic instability, survival and cell cycle progression but also anti-tumor immunity. The presence of RelB would favor long-term maintenance of EBV-associated tumors, reinforcing the energy metabolism subversion, protection against apoptosis and promoting resistance to hypoxic stress. RelA would be the right NF-κB subunit to target in the view of new drug development, at least for EBV-transformed B cells.
Young LS, Rickinson AB . Epstein-Barr virus: 40 years on. Nat Rev Cancer 2004; 4: 757–768.
Mosialos G, Birkenbach M, Yalamanchili R, VanArsdale T, Ware C, Kieff E . The Epstein-Barr virus transforming protein LMP1 engages signaling proteins for the tumor necrosis factor receptor family. Cell 1995; 80: 389–399.
Faumont N, Chanut A, Benard A, Cogne N, Delsol G, Feuillard J et al. Comparative analysis of oncogenic properties and nuclear factor-kappaB activity of latent membrane protein 1 natural variants from Hodgkin’s lymphoma’s Reed-Sternberg cells and normal B-lymphocytes. Haematologica 2009; 94: 355–363.
Vallabhapurapu S, Karin M . Regulation and function of NF-kappaB transcription factors in the immune system. Annu Rev Immunol 2009; 27: 693–733.
Israël A . The IKK complex, a central regulator of NF-kappaB activation. Cold Spring Harb Perspect Biol 2010; 2: a000158.
Scheidereit C . IkappaB kinase complexes: gateways to NF-kappaB activation and transcription. Oncogene 2006; 25: 6685–6705.
Dejardin E . The alternative NF-kappaB pathway from biochemistry to biology: pitfalls and promises for future drug development. Biochem Pharmacol 2006; 72: 1161–1179.
Luftig M, Yasui T, Soni V, Kang M-S, Jacobson N, Cahir-McFarland E et al. Epstein-Barr virus latent infection membrane protein 1 TRAF-binding site induces NIK/IKK alpha-dependent noncanonical NF-kappaB activation. Proc Natl Acad Sci USA 2004; 101: 141–146.
Atkinson PGP, Coope HJ, Rowe M, Ley SC . Latent membrane protein 1 of Epstein-Barr virus stimulates processing of NF-kappa B2 p100 to p52. J Biol Chem 2003; 278: 51134–51142.
Eliopoulos AG, Caamano JH, Flavell J, Reynolds GM, Murray PG, Poyet J-L et al. Epstein-Barr virus-encoded latent infection membrane protein 1 regulates the processing of p100 NF-kappaB2 to p52 via an IKKgamma/NEMO-independent signalling pathway. Oncogene 2003; 22: 7557–7569.
Saito N, Courtois G, Chiba A, Yamamoto N, Nitta T, Hironaka N et al. Two carboxyl-terminal activation regions of Epstein-Barr virus latent membrane protein 1 activate NF-kappaB through distinct signaling pathways in fibroblast cell lines. J Biol Chem 2003; 278: 46565–46575.
Le Clorennec C, Youlyouz-Marfak I, Adriaenssens E, Coll J, Bornkamm GW, Feuillard J . EBV latency III immortalization program sensitizes B cells to induction of CD95-mediated apoptosis via LMP1: role of NF-kappaB, STAT1, and p53. Blood 2006; 107: 2070–2078.
Ferreira V, Sidénius N, Tarantino N, Hubert P, Chatenoud L, Blasi F et al. In vivo inhibition of NF-kappa B in T-lineage cells leads to a dramatic decrease in cell proliferation and cytokine production and to increased cell apoptosis in response to mitogenic stimuli, but not to abnormal thymopoiesis. J Immunol 1999; 162: 6442–6450.
Bornkamm GW, Berens C, Kuklik-Roos C, Bechet J-M, Laux G, Bachl J et al. Stringent doxycycline-dependent control of gene activities using an episomal one-vector system. Nucleic Acids Res 2005; 33: e137.
Le Clorennec C, Ouk T-S, Youlyouz-Marfak I, Panteix S, Martin C-C, Rastelli J et al. Molecular basis of cytotoxicity of Epstein-Barr virus (EBV) latent membrane protein 1 (LMP1) in EBV latency III B cells: LMP1 induces type II ligand-independent autoactivation of CD95/Fas with caspase 8-mediated apoptosis. J Virol 2008; 82: 6721–6733.
Kempkes B, Spitkovsky D, Jansen-Dürr P, Ellwart JW, Kremmer E, Delecluse HJ et al. B-cell proliferation and induction of early G1-regulating proteins by Epstein-Barr virus mutants conditional for EBNA2. EMBO J 1995; 14: 88–96.
Caux C, Massacrier C, Vanbervliet B, Dubois B, Van Kooten C, Durand I et al. Activation of human dendritic cells through CD40 cross-linking. J Exp Med 1994; 180: 1263–1272.
Feuillard J, Schuhmacher M, Kohanna S, Asso-Bonnet M, Ledeur F, Joubert-Caron R et al. Inducible loss of NF-kappaB activity is associated with apoptosis and Bcl-2 down-regulation in Epstein-Barr virus-transformed B lymphocytes. Blood 2000; 95: 2068–2075.
Faumont N, Le Clorennec C, Teira P, Goormachtigh G, Coll J, Canitrot Y et al. Regulation of DNA polymerase beta by the LMP1 oncoprotein of EBV through the nuclear factor-kappaB pathway. Cancer Res 2009; 69: 5177–5185.
Baran-Marszak F, Feuillard J, Najjar I, Le Clorennec C, Béchet J-M, Dusanter-Fourt I et al. Differential roles of STAT1alpha and STAT1beta in fludarabine-induced cell cycle arrest and apoptosis in human B cells. Blood 2004; 104: 2475–2483.
Smyth GK, Michaud J, Scott HS . Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinforma Oxf Engl 2005; 21: 2067–2075.
Reiner-Benaim A . FDR control by the BH procedure for two-sided correlated tests with implications to gene expression data analysis. Biom J Biom Z 2007; 49: 107–126.
Tusher VG, Tibshirani R, Chu G . Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001; 98: 5116–5121.
Graham JP, Arcipowski KM, Bishop GA . Differential B-lymphocyte regulation by CD40 and its viral mimic, latent membrane protein 1. Immunol Rev 2010; 237: 226–248.
Traenckner EB, Wilk S, Baeuerle PA . A proteasome inhibitor prevents activation of NF-kappa B and stabilizes a newly phosphorylated form of I kappa B-alpha that is still bound to NF-kappa B. EMBO J 1994; 13: 5433–5441.
Basso K, Margolin AA, Stolovitzky G, Klein U, Dalla-Favera R, Califano A . Reverse engineering of regulatory networks in human B cells. Nat Genet 2005; 37: 382–390.
Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000; 403: 503–511.
Camara DA, Stefanoff CG, Pires ARC, Soares F, Biasoli I, Zalcberg I et al. Immunoblastic morphology in diffuse large B-cell lymphoma is associated with a nongerminal center immunophenotypic profile. Leuk Lymphoma 2007; 48: 892–896.
Ranuncolo SM, Pittaluga S, Evbuomwan MO, Jaffe ES, Lewis BA . Hodgkin lymphoma requires stabilized NIK and constitutive RelB expression for survival. Blood 2012; 120: 3756–3763.
Demchenko YN, Kuehl WM . A critical role for the NFkB pathway in multiple myeloma. Oncotarget 2010; 1: 59–68.
Cormier F, Monjanel H, Fabre C, Billot K, Sapharikas E, Chereau F et al. Frequent engagement of RelB activation is critical for cell survival in multiple myeloma. PloS One 2013; 8: e59127.
Chao C, Silverberg MJ, Martínez-Maza O, Chi M, Abrams DI, Haque R et al. Epstein-Barr virus infection and expression of B-cell oncogenic markers in HIV-related diffuse large B-cell Lymphoma. Clin Cancer Res Off J Am Assoc Cancer Res 2012; 18: 4702–4712.
Compagno M, Lim WK, Grunn A, Nandula SV, Brahmachary M, Shen Q et al. Mutations of multiple genes cause deregulation of NF-kappaB in diffuse large B-cell lymphoma. Nature 2009; 459: 717–721.
Zhang B, Kracker S, Yasuda T, Casola S, Vanneman M, Hömig-Hölzel C et al. Immune surveillance and therapy of lymphomas driven by Epstein-Barr virus protein LMP1 in a mouse model. Cell 2012; 148: 739–751.
Shair KHY, Bendt KM, Edwards RH, Bedford EC, Nielsen JN, Raab-Traub N . EBV latent membrane protein 1 activates Akt, NFkappaB, and Stat3 in B cell lymphomas. PLoS Pathog 2007; 3: e166.
Derudder E, Dejardin E, Pritchard LL, Green DR, Korner M, Baud V . RelB/p50 dimers are differentially regulated by tumor necrosis factor-alpha and lymphotoxin-beta receptor activation: critical roles for p100. J Biol Chem 2003; 278: 23278–23284.
Lo JC, Basak S, James ES, Quiambo RS, Kinsella MC, Alegre M-L et al. Coordination between NF-kappaB family members p50 and p52 is essential for mediating LTbetaR signals in the development and organization of secondary lymphoid tissues. Blood 2006; 107: 1048–1055.
Shih VF-S, Tsui R, Caldwell A, Hoffmann A . A single NFκB system for both canonical and non-canonical signaling. Cell Res 2011; 21: 86–102.
Liptay S, Schmid RM, Nabel EG, Nabel GJ . Transcriptional regulation of NF-kappa B2: evidence for kappa B-mediated positive and negative autoregulation. Mol Cell Biol 1994; 14: 7695–7703.
Bren GD, Solan NJ, Miyoshi H, Pennington KN, Pobst LJ, Paya CV . Transcription of the RelB gene is regulated by NF-kappaB. Oncogene 2001; 20: 7722–7733.
Basak S, Shih VF-S, Hoffmann A . Generation and activation of multiple dimeric transcription factors within the NF-kappaB signaling system. Mol Cell Biol 2008; 28: 3139–3150.
Jacque E, Tchenio T, Piton G, Romeo P-H, Baud V . RelA repression of RelB activity induces selective gene activation downstream of TNF receptors. Proc Natl Acad Sci USA 2005; 102: 14635–14640.
Lovas A, Radke D, Albrecht D, Yilmaz ZB, Möller U, Habenicht AJR et al. Differential RelA- and RelB-dependent gene transcription in LTbetaR-stimulated mouse embryonic fibroblasts. BMC Genomics 2008; 9: 606.
Moretti M, Bennett J, Tornatore L, Thotakura AK, Franzoso G . Cancer: NF-κB regulates energy metabolism. Int J Biochem Cell Biol 2012; 44: 2238–2243.
Samudio I, Harmancey R, Fiegl M, Kantarjian H, Konopleva M, Korchin B et al. Pharmacologic inhibition of fatty acid oxidation sensitizes human leukemia cells to apoptosis induction. J Clin Invest 2010; 120: 142–156.
Pike LS, Smift AL, Croteau NJ, Ferrick DA, Wu M . Inhibition of fatty acid oxidation by etomoxir impairs NADPH production and increases reactive oxygen species resulting in ATP depletion and cell death in human glioblastoma cells. Biochim Biophys Acta 2011; 1807: 726–734.
Menendez JA, Lupu R . Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis. Nat Rev Cancer 2007; 7: 763–777.
Care MA, Barrans S, Worrillow L, Jack A, Westhead DR, Tooze RM . A microarray platform-independent classification tool for cell of origin class allows comparative analysis of gene expression in diffuse large B-cell lymphoma. PLoS One 2013; 8: e55895.
Johansen LM, Deppmann CD, Erickson KD, Coffin WF 3rd, Thornton TM, Humphrey SE et al. EBNA2 and activated Notch induce expression of BATF. J Virol 2003; 77: 6029–6040.
Farrell CJ, Lee JM, Shin E-C, Cebrat M, Cole PA, Hayward SD . Inhibition of Epstein-Barr virus-induced growth proliferation by a nuclear antigen EBNA2-TAT peptide. Proc Natl Acad Sci USA 2004; 101: 4625–4630.
Hwang HC, Clurman BE . Cyclin E in normal and neoplastic cell cycles. Oncogene 2005; 24: 2776–2786.
Tzankov A, Gschwendtner A, Augustin F, Fiegl M, Obermann EC, Dirnhofer S et al. Diffuse large B-cell lymphoma with overexpression of cyclin e substantiates poor standard treatment response and inferior outcome. Clin Cancer Res 2006; 12: 2125–2132.
Shinohara M, Io K, Shindo K, Matsui M, Sakamoto T, Tada K et al. APOBEC3B can impair genomic stability by inducing base substitutions in genomic DNA in human cells. Sci Reports 2012; 2: 806.
Challa-Malladi M, Lieu YK, Califano O, Holmes AB, Bhagat G, Murty VV et al. Combined genetic inactivation of β2-Microglobulin and CD58 reveals frequent escape from immune recognition in diffuse large B cell lymphoma. Cancer Cell 2011; 20: 728–740.
Ma Y, Visser L, Roelofsen H, de Vries M, Diepstra A, van Imhoff G et al. Proteomics analysis of Hodgkin lymphoma: identification of new players involved in the cross-talk between HRS cells and infiltrating lymphocytes. Blood 2008; 111: 2339–2346.
Yoon HS, Scharer CD, Majumder P, Davis CW, Butler R, Zinzow-Kramer W et al. ZBTB32 is an early repressor of the CIITA and MHC class II gene expression during B cell differentiation to plasma cells. J Immunol 2012; 189: 2393–2403.
Wille JJ Jr . Circadian rhythm of tumor promotion in the two-stage model of mouse tumorigenesis. Cancer Lett 2003; 190: 143–149.
Mazzoccoli G, Pazienza V, Panza A, Valvano MR, Benegiamo G, Vinciguerra M et al. ARNTL2 and SERPINE1: potential biomarkers for tumor aggressiveness in colorectal cancer. J Cancer Res Clin Oncol 2012; 138: 501–511.
Charfi C, Voisin V, Levros L-C, Edouard E, Rassart E . Gene profiling of Graffi murine leukemia virus-induced lymphoid leukemias: identification of leukemia markers and Fmn2 as a potential oncogene. Blood 2011; 117: 1899–1910.
Hogenesch JB, Gu YZ, Moran SM, Shimomura K, Radcliffe LA, Takahashi JS et al. The basic helix-loop-helix-PAS protein MOP9 is a brain-specific heterodimeric partner of circadian and hypoxia factors. J Neurosci 2000; 20: RC83.
This work was supported by Institut National du Cancer, Cancéropôle Grand-Sud-Ouest, Comité Orientation Recherche Cancer du Limousin, Conseil Régional du Limousin, Comités Limousin de la Ligue contre le Cancer and Association pour la Recherche sur le Cancer. We thank the tumor banks of the University hospitals of Bordeaux and Limoges for providing biopsies and Véronique Pantesco (Microarray Core Facility of the Institute for Research in Biotherapy, Montpellier, France) for Affymetrix analyses. We thank Dr Jeanne Cook-Moreau, CNRS-UMR-7276 Limoges, for English corrections and Mr Lionel Forestier platform GENOLIM, University of Limoges, France. N Faumont was supported by Comité Orientation Recherche Cancer du Limousin, A Chanut and S Durand-Panteix by Association pour la Recherche sur le Cancer. A Chanut was also supported by Société Française d’Hématologie.
The authors declare no conflict of interest.
Supplementary Information accompanies this paper on the Leukemia website
About this article
Cite this article
Chanut, A., Duguet, F., Marfak, A. et al. RelA and RelB cross-talk and function in Epstein–Barr virus transformed B cells. Leukemia 28, 871–879 (2014). https://doi.org/10.1038/leu.2013.274
- cell cycle
- B-cell lymphoma.
Apoptosis- and survival-related gene mRNA profile in peripheral blood leukocytes in children with acute EBV infectious mononucleosis
Russian Journal of Infection and Immunity (2020)
Ferruginol induced apoptosis on SK-Mel-28 human malignant melanoma cells mediated through P-p38 and NF-κB
Human & Experimental Toxicology (2019)
Identification of a Nine-Gene Signature and Establishment of a Prognostic Nomogram Predicting Overall Survival of Pancreatic Cancer
Frontiers in Oncology (2019)
Critical Reviews in Clinical Laboratory Sciences (2018)
PLOS Pathogens (2017)