Multiple Sclerosis Resource Centre

Welcome to the Multiple Sclerosis Resource Centre. This website is intended for international healthcare professionals with an interest in Multiple Sclerosis. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

Molecular network of ChIP-Seq-based NF-κB p65 target genes involves diverse immune functions relevant to the immunopathogenesis of multiple sclerosis

Multiple Sclerosis and Related Disorders, 1, 3, pages 94 - 106

Abstract

Background

The transcription factor nuclear factor-kappa B (NF-κB) acts as a central regulator of immune response, stress response, cell proliferation, and apoptosis. Aberrant regulation of NF-κB function triggers development of cancers, metabolic diseases, and autoimmune diseases. We attempted to characterize a global picture of the NF-κB target gene network relevant to the immunopathogenesis of multiple sclerosis (MS).

Methods

We identified the comprehensive set of 918 NF-κB p65 binding sites on protein-coding genes from chromatin immunoprecipitation followed by deep sequencing (ChIP-Seq) dataset of TNFα-stimulated human B lymphoblastoid cells. The molecular network was studied by a battery of pathway analysis tools of bioinformatics.

Results

The GenomeJack genome viewer showed that NF-κB p65 binding sites were accumulated in promoter (35.5%) and intronic (54.9%) regions with an existence of the NF-κB consensus sequence motif. A set of 52 genes (5.7%) corresponded to known NF-κB targets by database search. KEGG, PANTHER, and Ingenuity Pathways Analysis (IPA) revealed that the NF-κB p65 target gene network is linked to regulation of immune functions and oncogenesis, including B cell receptor signaling, T cell activation pathway, Toll-like receptor signaling, and apoptosis signaling, and molecular mechanisms of cancers. KeyMolnet indicated an involvement of the complex crosstalk among core transcription factors in the NF-κB p65 target gene network. Furthermore, the set of NF-κB p65 target genes included 10 genes among 98 MS risk alleles and 49 molecules among 709 MS brain lesion-specific proteins.

Conclusions

These results suggest that aberrant regulation of NF-κB-mediated gene expression, by inducing dysfunction of diverse immune functions, is closely associated with development and progression of MS.

Highlights

 

  • We identified 918 ChIP-Seq NF-κB p65 binding sites on protein-coding genes.
  • The binding sites were accumulated in promoter and intronic regions.
  • Their molecular network was associated with immune functions relevant to MS.
  • They included 10 MS risk genes and 49 MS brain lesion-specific proteins.

Keywords: ChIP-Seq, GenomeJack, Molecular network, Multiple sclerosis, Next generation sequencing, NF-κB.

1. Introduction

The transcription factor nuclear factor-kappa B (NF-κB) acts as a central regulator of innate and adaptive immune response, stress response, cell proliferation, and apoptosis (Barnes and Karin, 1997 and Hayden et al, 2006). Deregulation of NF-κB function triggers development of cancers, metabolic diseases, and autoimmune diseases, such as multiple sclerosis (MS) and rheumatoid arthritis (RA) (Yan and Greer, 2008 and Gregersen et al, 2009). The NF-κB family proteins consist of five members, such as RelA (p65), RelB, c-Rel, NF-κB1 (p105), and NF-κB2 (p100) ( Gilmore, 2006 ). The latter two are proteolytically processed into p50 and p52, respectively. All the members share the Rel homology domain (RHD) acting for DNA binding and dimerization. The NF-κB family proteins constitute either homodimers or heterodimers, except for RelB that exclusively forms heterodimers. The p50-RelA heterodimer represents a predominant NF-κB dimer in various cell types. The NF-κB dimers interact with specific DNA sequences named the κB site located on promoters to activate or repress transcription of target genes. Only p65 and c-Rel act as a potent transcriptional activator, whereas p50 and p52 homodimers generally repress transcription ( Rothwarf and Karin, 1999 ). Optimal induction of NF-κB target genes requires phosphorylation of p65 within its transactivation domain in response to distinct stimuli by various kinases ( Viatour et al., 2005 ).

NF-κB activity is regulated tightly at multiple levels (Viatour et al, 2005 and Gilmore, 2006). In unstimulated cells, NF-κB proteins exist in an inactive state, being sequestrated in the cytoplasm via non-covalent interaction with the inhibitor of NF-κB (IκB) proteins, such as IκBα, IκBβ, IκBγ, and IκBε. Viral and bacterial products, cytokines, and oxidative stress activate the specific IκB kinase (IKK) complex, composed of two catalytic kinase subunits called IKKα and IKKβ and a regulatory subunit called NF-κB essential modulator (NEMO). IκB proteins, when phosphorylated by the IKK complex, are ubiquitinated, and processed for 26 S proteasome-mediated degradation, resulting in nuclear translocation of NF-κB dimers. The NF-κB signaling cascade is categorized into canonical and non-canonical pathways ( Oeckinghaus et al., 2011 ). The canonical pathway is activated by various proinflammatory cytokines, such as tumor necrosis factor-alpha tumor necrosis factor-alpha (TNFα) and interleukin-1 (IL-1), transduced by both IKKβ and NEMO that chiefly mediate phosphorylation of IκBα, followed by nuclear translocation of p65-containing NF-κB heterodimers. The non-canonical pathway, activated by CD40 ligand, B-cell activating factor (BAFF), and lymphotoxin-beta (LTβ), requires IKKα-mediated phosphorylation of p100 dimerized with RelB, which are processed to form the p52-RelB complex.

MS is an inflammatory demyelinating disease of the central nervous system (CNS) white matter, presenting with a relapsing-remitting (RR) and/or progressive clinical course. It is mediated by an autoimmune process triggered by a complex interplay between genetic and environmental factors, leading to development of autoreactive T helper type 1 (Th1) and type 17 (Th17) lymphocytes ( Comabella and Khoury, 2012 ). Several lines of evidence indicate that aberrant regulation of NF-κB signaling pathway plays a central role in acute relapse of MS. By gene expression profiling, we identified 43 differentially expressed genes in peripheral blood CD3+T cells between the peak of acute relapse and the complete remission of RRMS ( Satoh et al., 2008 ). We found that the molecular network of 43 genes showed the most significant relationship with transcriptional regulation by NF-κB. Our observations are supported by several studies that verified an aberrant expression of NF-κB signaling molecules in peripheral blood mononuclear cells (PMBC) during MS relapse (Achiron et al, 2007 and Lindsey et al, 2011). Intravenous methylprednisolone pulse (IVMP) immediately reduces the levels of activated p65 in PBMC of MS patients ( Eggert et al., 2008 ). Furthermore, interferon-gamma (IFNγ, a prototype Th1 cytokine, is identified as one of NF-κB target genes ( Sica et al., 1997 ), while interferon-beta (IFNβ, the first-line medication for RRMS, attenuates proinflammatory responses by inhibiting the NF-κB activity in lymphocytes ( Martín-Saavedra et al., 2007 ). Mucosa-associated lymphoid tissue lymphoma translocation gene 1 (MALT1), a key regulator of NF-κB activation, positively regulates the encephalitogenic potential of inflammatory Th17 cells ( Brüstle et al., 2012 ). To elucidate the precise role of NF-κB in MS relapse, it is highly important to thoroughly characterize NF-κB target genes involved in the immunopathogenesis of MS.

A number of previous studies identified hundreds of NF-κB target genes, including those involved in not only inflammatory and anti-apoptotic responses, but also anti-inflammatory and proapoptotic responses ( Pahl, 1999 ). Importantly, NF-κB target genes often activate NF-κB itself, providing a positive regulatory loop that amplifies and perpetuates inflammatory responses ( Barnes and Karin, 1997 ). However, it remains unclear how many of previously identified genes actually represent direct targets for NF-κB-mediated transcriptional activation.

Recently, the rapid progress in the next-generation sequencing (NGS) technology has revolutionized the field of genome research. As one of NGS applications, chromatin immunoprecipitation followed by deep sequencing (ChIP-Seq) provides a highly efficient method for genome-wide profiling of DNA-binding proteins, histone modifications, and nucleosomes ( Park, 2009 ). ChIP-Seq endowed with an advantage of higher resolution, less noise, and greater coverage of the genome, compared with the microarray-based ChIP-Chip method, serves as an innovative tool for studying the comprehensive gene regulatory networks. However, since the NGS analysis produces extremely high-throughput experimental data, it is often difficult to extract the meaningful biological implications. Recent advances in systems biology enable us to illustrate the cell-wide map of the complex molecular interactions by using the literature-based knowledgebase of molecular pathways ( Satoh, 2010 ). The logically arranged molecular networks construct the whole system characterized by robustness, which maintains the proper function of the system in the face of genetic and environmental perturbations. Therefore, the integration of high dimensional NGS data with underlying molecular networks offers a rational approach to characterize the network-based molecular mechanisms of gene regulation on the whole genome scale.

In the present study, to characterize a global picture of the NF-κB target gene network, we investigated the NF-κB p65 ChIP-Seq dataset of TNFα-stimulated human B lymphoblastoid cells. The dataset was retrieved from the public database of the Encyclopedia of DNA Elements (ENCODE) project ( encodeproject.org/ENCODE ) (Kasowski et al, 2010 and Gerstein et al, 2012).

2. Methods

2.1. ChIP-Seq dataset

To identify a comprehensive set of NF-κB p65-target genes, we studied a series of ChIP-Seq data retrieved from DDBJ Sequence Read Archive (DRA) under the accession number of SRP007993. The ChIP-Seq experiments were performed for the ENCODE project by researchers in Dr. Michel Snyder's Laboratory, Stanford University ( Kasowski et al., 2010 ). The data were derived from 10 distinct Epstein-Barr virus (EBV)-transformed human lymphoblastoid cell lines (LCL) numbered GM10847, GM12878, GM12891, GM12892, GM15510, GM18505, GM18526, GM18951, GM19099, and GM19193. In these experiments, the cells were exposed for 6 h to 25 ng/mL recombinant human TNFα (#14–8329-62; eBioscience), and then were fixed with formaldehyde to crosslink NF-κB–DNA complexes, and immunoprecipitated from sonicated nuclear lysates by using rabbit anti-NF-κB p65 antibody (sc-372; Santa Cruz Biotechnology) or normal rabbit IgG for input control. NGS libraries were constructed from 120 to 350 bp size-selected ChIP DNA fragments. They were processed for deep sequencing at a 28 bp read length on Genome Analyzer II (Illumina).

We converted SRALite-formatted files into FASTQ-formatted files, and combined 10 antibody-treated samples into the “test set” and 10 corresponding input controls into the “control set”. We mapped these data on the human genome reference sequence hg19 by using the Bowtie 0.12.7 program (bowtie-bio.sourceforge.net). Subsequently, we identified statistically significant peaks of mapped reads by using the MACS program (liulab.dfci.harvard.edu/MACS) under the stringent condition that satisfies the false discovery rate (FDR) ≤0.1% and fold enrichment ≥10 to avoid the detection of false positive binding sites if at all possible. Then, we identified the genomic location of MACS peaks by importing the processed data into GenomeJack v1.4, a novel genome viewer for NGS platforms developed by Mitsubishi Space Software ( www.mss.co.jp/businessfield/bioinformatics ). Based on RefSeq ID, MACS peaks were categorized into the following; the peaks located on protein-coding genes supplemented with NM-heading numbers, the peaks located on non-coding genes supplemented with NR-heading numbers, and the peaks located in intergenic regions with no relevant neighboring genes. Genomic locations of the peaks were further classified into the following; the promoter region defined by the location within a 5 kb upstream from the 5′ end of genes, the 5′ untranslated region (5′UTR), the exon, the intron, the 3′UTR, and intergenic regions outside these, as described previously ( Satoh and Tabunoki, in press ).

The consensus sequence motif was identified by importing a 400 bp-length sequence surrounding the summit of MACS peaks into the MEME-ChIP program ( meme.sdsc.edu/meme/cgi-bin/meme-chip.cgi ). The information on known NF-κB target genes was collected from web accessible databases constructed by Dr. Thomas Gilmore, Boston University ( www.bu.edu/nf-kb/gene-resources/target-genes ) and by Bonsai Bioinformatics, the Laboratoire d’Informatique Fondamentale de Lille (LIFL), Université Lille 1 (bioinfo.lifl.fr/NF-KB).

3. Molecular network analysis

To identify biologically relevant molecular networks and pathways, we imported Entrez Gene IDs of NF-κB p65 target genes into the Functional Annotation tool of Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7 ( david.abcc.ncifcrf.gov ) ( Huang et al., 2009 ). DAVID identifies relevant pathways constructed by Kyoto Encyclopedia of Genes and Genomes (KEGG) ( www.kegg.jp ) or by the Protein Analysis Through Evolutionary Relationships (PANTHER) classification system ( www.pantherdb.org ). They are comprised of the genes enriched in the given set with statistical significance evaluated by the modified Fisher's exact test corrected by Bonferroni's multiple comparison test. KEGG is a publicly accessible knowledgebase containing manually curated reference pathways that cover a wide range of metabolic, genetic, environmental, and cellular processes, and human diseases, currently composed of 207,161 pathways generated from 432 reference pathways. PANTHER includes the information on 175 signaling and metabolic pathways manually curated by expert biologists, expressed in a Systems Biology Markup Language (SBML) format. We also imported Entrez Gene IDs into Ingenuity Pathways Analysis (IPA) (Ingenuity Systems; www.ingenuity.com ) and KeyMolnet (Institute of Medicinal Molecular Design; www.immd.co.jp ), both of which are commercial tools for molecular network analysis.

IPA is a knowledgebase that contains approximately 2,500,000 biological and chemical interactions and functional annotations with definite scientific evidence. By uploading the list of Gene IDs and expression values, the network-generation algorithm identifies focused genes integrated in a global molecular network. IPA calculates the scorep-value that reflects the statistical significance of association between the genes and the networks by the Fisher's exact test.

KeyMolnet contains knowledge-based contents on 150,500 relationships among human genes and proteins, small molecules, diseases, pathways and drugs ( Satoh, 2010 ). They are categorized into the core contents collected from selected review articles with the highest reliability or the secondary contents extracted from abstracts of PubMed and Human Reference Protein database (HPRD). By importing the list of Gene ID and expression values, KeyMolnet automatically provides corresponding molecules as nodes on the network. The neighboring network-search algorithm selected one or more molecules as starting points to generate the network of all kinds of molecular interactions around starting molecules, including direct activation/inactivation, transcriptional activation/repression, and the complex formation within one path from starting points. The generated network was compared side by side with 484 human canonical pathways of the KeyMolnet library. The algorithm counting the number of overlapping molecular relations between the extracted network and the canonical pathway makes it possible to identify the canonical pathway showing the most significant contribution to the extracted network.

4. Results

4.1. Identification of 918 ChIP-Seq-based NF-κB p65 target genes

After mapping short reads on hg19, we identified totally 1630 stringent ChIP-Seq peaks that satisfied the criteria of both FDR≤0.1% and fold enrichment≥10. The genomic location of the peaks was determined by GenomeJack (Fig 1, Fig 2, and Fig 3, panel (a). After omitting the peaks located in non-coding genes (n=114), those in intergenic regions (n=502), and several redundant genes, we extracted 918 peaks located in protein-coding genes. They are tentatively designated as the set of ChIP-Seq-based NF-κB p65 target genes ( Supplementary Table 1 ). They included many genes important for regulation of innate and adaptive immune response and inflammation, such as CD22, CD69, CD70, CD83, CD86, CD209, IL1RN, IL2RG, IL3RA, IL7, IL18R1, IL21R, IL27RA, IL31RA, TGFB1, CXCR4, CX3CL1, TAP1, TYK2, IRAK2, NOD2, IRF1, IRF5, STAT1, GATA3, SMAD3, ICAM1, ITGAM, MMP9, FAS, BCL2, and TP53, in addition to NF-κB signaling molecules such as TRAF1, TRAF2, TRAF3, IKBKE, IKBKG, NFKB1, NFKB2, NFKBIA, NFKBIB, NFKBID, NFKBIE, REL, RELA, and RELB. The top 30 genes are listed in Table 1 . By searching NF-κB target gene databases, the set of 918 genes included 52 known targets (5.7%), such as ICAM1, IL1RN, IRF1, IRF2, CSF1, NFKBIA, NFKB1, NFKB2, RELB, TNFAIP3, MMP9, TP53, BCK2, and BCL3. The summits of the peaks were located in the promoter (n=326; 35.5%), 5′UTR (n=49; 5.3%), exon (n=23; 2.5%), intron (n=504; 54.9%), or 3′UTR (n=16; 1.7%) regions. Because we did not study NF-κB p65 ChIP-Seq data of TNFα-unstimulated cells, the possibility could not be excluded that a subset of ChIP-Seq peaks we identified are attributable to the constitutive binding of NF-κB p65 on target genes in the absence of TNFα stimulation.

gr1

Fig. 1 Location of NF-κB p65 ChIP-Seq peaks in the promoter region of target genes. From the dataset numbered SRP007993, we identified totally 1630 stringent ChIP-Seq peaks that satisfied the criteria of both FDR≤0.1% and fold enrichment≥10. The genomic location of the peaks was determined by importing the processed data into GenomeJack. An example of NF-κB p105 subunit (NFKB1) (Entrez Gene ID 4790 in Table 1 ), composed of two transcript variants NM_003998 and NM_001165412, is shown, where a MACS peak numbered 15207 is located in the promoter region of NFKB1 (panel a) with a NF-κB consensus sequence motif highlighted by an orange square (panel b). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

gr2

Fig. 2 Location of NF-κB p65 ChIP-Seq peaks in the intronic region of target genes. The genomic location of the peaks was determined by GenomeJack. An example of ADAM metallopeptidase domain 8 (ADAM8) (Entrez Gene ID 101 in Table 1 ), composed of three transcript variants NM_001109, NM_001164489, and NM_001164490, is shown, where a MACS peak numbered 2974 is located in the intronic region of ADAM8 (panel a) with a NF-κB consensus sequence motif highlighted by an orange square (panel b). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

gr3

Fig. 3 Location of NF-κB p65 ChIP-Seq peaks in the intergenic region. The genomic location of the peaks was determined by GenomeJack. An example of intergenic location of a MACS peak numbered 6239 is shown (panel a) with a NF-κB consensus sequence motif highlighted by an orange square (panel b). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 1 Top 30 ChIP-Seq-based NF-κB p65 target genes in human lymphoblastoid cells.

Chromosome Start End FE FDR (%) Location Gene ID Gene symbol Gene name
chr10 135087960 135090808 51.95 0 Intron 101 ADAM8 ADAM metallopeptidase domain 8
chr21 45376690 45379750 50.41 0 Intron 56894 AGPAT3 1-acylglycerol-3-phosphate O-acyltransferase 3
chr4 86477139 86479329 44.42 0 Intron 83478 ARHGAP24 Rho GTPase activating protein 24
chr6 138186536 138200020 39.23 0 Promoter 7128 TNFAIP3 Tumor necrosis factor, alpha-induced protein 3
chr19 10496255 10498250 37.53 0 Promoter 7297 TYK2 Tyrosine kinase 2
chr12 111866233 111869690 37.44 0 Intron 10019 SH2B3 SH2B adapter protein 3
chr4 103421036 103426349 36.81 0 Promoter 4790 NFKB1 Nuclear factor of kappa light polypeptide gene enhancer in B-cells 1
chr12 121997680 121999150 35.14 0 Intron 84678 KDM2B Lysine (K)-specific demethylase 2B
chr1 75196627 75200204 34.45 0 Intron 1429 CRYZ Crystallin, zeta (quinone reductase)
chr1 75196627 75200204 34.45 0 Promoter 127253 TYW3 tRNA-yW synthesizing protein 3 homolog (S. cerevisiae)
chr19 36389424 36392142 34.36 0 Promoter 10870 HCST Hematopoietic cell signal transducer
chr19 36389424 36392142 34.36 0 Intron 84807 NFKBID Nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, delta
chr16 10914391 10916414 33.92 0 Promoter 780776 FAM18A Family with sequence similarity 18, member A
chr7 155607589 155608757 32.79 0 Promoter 6469 SHH Sonic hedgehog homolog (Drosophila)
chr1 37938865 37946222 32.09 0 Intron 80149 ZC3H12A Zinc finger CCCH-type containing 12A
chr17 75429437 75430665 32.06 0 Intron 10801 SEPT9 Septin 9
chr5 150456868 150463839 31.6 0 5′UTR 10318 TNIP1 TNFAIP3 interacting protein 1
chrY 1396361 1397638 31.44 0 Promoter 3563 IL3RA Interleukin 3 receptor, alpha (low affinity)
chr17 1376790 1378860 31.19 0 Exon 4641 MYO1C Myosin IC
chr16 3012819 3014867 30.98 0 Promoter 79412 KREMEN2 Kringle containing transmembrane protein 2
chr16 3012819 3014867 30.98 0 Promoter 124222 PAQR4 Progestin and adipoQ receptor family member IV
chr22 50448174 50450362 30.14 0 Intron 400935 IL17REL Interleukin 17 receptor E-like
chr19 45503851 45505730 30 0 Promoter 5971 RELB v-rel reticuloendotheliosis viral oncogene homolog B
chr3 52344355 52345830 29.9 0 Promoter 25981 DNAH1 Dynein, axonemal, heavy chain 1
chr10 104153066 104156567 29.62 0 Promoter 4791 NFKB2 Nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 (p49/p100)
chr17 61772029 61780506 29.54 0 Intron 80774 LIMD2 LIM domain containing 2
chr14 35868073 35876918 29.08 0 Promoter 4792 NFKBIA Nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha
chr9 140130602 140131876 28.81 0 Promoter 10383 TUBB4B Tubulin, beta 2C
chr15 31803605 31804982 27.48 0 Intron 161725 OTUD7A OTU domain containing 7A
chr7 44787088 44791813 27.47 0 Intron 83637 ZMIZ2 Zinc finger, MIZ-type containing 2

From the NF-κB p65 ChIP-Seq dataset numbered SRP007993, we identified 918 stringent peaks on protein-coding genes exhibiting false discovery rate (FDR) ≤0.1% and fold enrichment (FE) ≥10. Top 30 genes based on FE are listed with the chromosome, the position (start, end), FE, FDR, the location (promoter, 5′UTR, exon, intron, 3′UTR), Entrez gene ID, gene symbol, and gene name. Known NF-κB target genes by database search are in bold. The complete list of 918 genes is shown in Supplementary Table 1 .

By motif analysis with MEME-ChIP, both promoter and intronic regions of target genes contained the NF-κB consensus sequence motif, defined as 5′GGG(A/G)N(A/T)(C/T)(C/T)CC3′ where N accepts any nucleotide ( Gilmore, 2006 ) ( Fig. 4 , panels a and b), being consistent with the results of GenomeJack (Fig 1 and Fig 2, panel b). These results validated the specificity of mapping of NF-κB p65 ChIP-Seq short reads onto genomic regions containing the NF-κB consensus sequence motif.

gr4

Fig. 4 Identification of NF-κB consensus sequence motif located in promoter and intronic regions. The consensus sequence motif was identified by importing a 400 bp-length sequence surrounding the summit of MACS peaks of top 50 genes into the MEME-ChIP program, which identified a series of NF-κB consensus sequence motifs as top five most significant motifs in promoter regions (a) and intronic regions (b), typically defined as 5′GGG(A/G)N(A/T)(C/T)(C/T)CC3′, where N accepts any nucleotide.

4.2. Molecular network of ChIP-Seq-based NF-κB p65 target genes

Next, we studied the molecular network of 918 ChIP-Seq-based NF-κB p65 target genes by using four distinct pathway analysis tools operating on different computational algorithms. DAVID identified functionally associated gene ontology (GO) terms. They include “intracellular signaling cascade” (GO:0007242;p=0.0000008), “protein kinase cascade” (GO: 0007243;p=0.000004), and “positive regulation of biosynthetic process” (GO:0009891;p=0.00003) as the top three3 most significant GO terms ( Supplementary Table 2 ). These results suggest that ChIP-Seq-based NFκB p65 target genes play a role in a wide range of biological functions. KEGG showed close relationships with the “Neurotrophin signaling pathway” (hsa04722;p=0.0000004), “B cell receptor signaling pathway” (hsa04662;p=0.00008) ( Fig. 5 ), “Small cell lung cancer” (hsa05222;p=0.00009), “Apoptosis” (hsa04210;p=0.00016), “Pathways in cancer” (hsa05200;p=0.0007), “Leukocyte transendothelial migration” (hsa04670;p=0.0039), “Prostate cancer” (hsa05215;p=0.0042), “Pancreatic cancer” (hsa05212;p=0.0054), “Toll-like receptor signaling pathway” (hsa04620;p=0.0058), and “Focal adhesion” (hsa04510;p=0.0193) ( Table 2 ). Thus, ChIP-Seq-based NF-κB p65 target genes play a pivotal role in regulation of not only immune functions but also oncogenesis. PANTER indicated a significant relationship with the “Apoptosis signaling pathway” (P00006;p=0.00321), “B cell activation” (P00010;p=0.00326), “Toll receptor signaling pathway” (P00054;p=0.0058), “T-cell activation” (P00053;p=0.0104), and “Inflammation mediated by chemokine and cytokine signaling pathway” (P00031;p=0.0378) ( Supplementary Table 3 ). The core analysis tool of IPA extracted “CD40 signaling” (p=4.59E-15), “Molecular Mechanisms of Cancer” (p=8.38E-13), and “B Cell Receptor Signaling” (p=5.93E-12) as top 3 most significant canonical pathways associated with the set of 918 genes. All of these results support a predominant role of ChIP-Seq-based NF-κB p65 target genes in immune regulation and oncogenesis. Furthermore, IPA extracted the networks defined by “Gene Expression, Developmental Disorder, Hereditary Disorder” (p=1.00E-77), “Cellular Function and Maintenance, Cellular Growth and Proliferation, Hematological System Development and Function” (p=1.00E-64), and “Cellular Function and Maintenance, Immunological Disease, Cell Signaling” (p=1.00E-57) ( Fig. 6 ) as the top three most significant functional networks ( Supplementary Table 4 ).

gr5

Fig. 5 KEGG pathways of ChIP-Seq-based NF-κB p65 target genes. Entrez Gene IDs of 918 ChIP-Seq-based NF-κB p65 target genes were imported into DAVID. It identified KEGG pathways relevant to the set of imported genes ( Table 2 ). The second rank pathway termed “B cell receptor signaling pathway” (hsa04662) is shown, where NF-κB p65 target genes is colored orange. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 2 Top 10 KEGG pathways relevant to 918 ChIP-Seq-based NF-κB p65 genes.

Rank Category Genes in the pathway p -Value FDR FE
1 hsa04722:Neurotrophin signaling pathway BCL2, BRAF, CALM1, CAMK4, CSK, FRS2, IRAK2, JUN, MAP3K5, NFKB1, NFKBIA, NFKBIB, NFKBIE, NTRK1, NTRK2, PIK3CA, PIK3CB, PIK3CD, PIK3R1, PLCG2, PRKCD, RAPGEF1, RELA, SH2B2, SH2B3, TP53, TP73 0.0000004 0.0000036 3.858
2 hsa04662:B cell receptor signaling pathway BCL10, CD22, CD81, IKBKG, JUN, LYN, NFKB1, NFKBIA, NFKBIB, NFKBIE, PIK3CA, PIK3CB, PIK3CD, PIK3R1, PLCG2, RASGRP3, RELA, VAV2 0.00008 0.00064 4.252
3 hsa05222:Small cell lung cancer BCL2, BIRC3, CDK6, COL4A2, IKBKG, LAMC2, NFKB1, NFKBIA, PIK3CA, PIK3CB, PIK3CD, PIK3R1, RELA, RXRA, TP53, TRAF1, TRAF2, TRAF3, TRAF4 0.00009 0.00072 4.008
4 hsa04210:Apoptosis BCL2, BIRC3, CAPN2, FAS, IKBKG, IL3RA, IRAK2, NFKB1, NFKBIA, NTRK1, PIK3CA, PIK3CB, PIK3CD, PIK3R1, PRKAR1B, RELA, TNFRSF10B, TP53, TRAF2 0.00016 0.0013 3.869
5 hsa05200:Pathways in cancer BCL2, BIRC3, BRAF, CCDC6, CDK6, COL4A2, CREBBP, CTNNB1, EGF, FAS, FGF2, IKBKG, JUN, KIT, LAMC2, MMP9, NFKB1, NFKB2, NFKBIA, NTRK1, PIK3CA, PIK3CB, PIK3CD, PIK3R1, PLCG2, RALGDS, RARA, RELA, RXRA, SHH, SMAD3, STAT1, TCF7, TGFB1, TP53, TRAF1, TRAF2, TRAF3, TRAF4, WNT10A 0.0007 0.0053 2.161
6 hsa04670:Leukocyte transendothelial migration ACTB, ACTG1, BCAR1, CLDN14, CTNNB1, CXCR4, GNAI2, ICAM1, ITGAM, ITK, MMP9, MYL2, MYL5, PIK3CA, PIK3CB, PIK3CD, PIK3R1, PLCG2, VAV2, VCL 0.0039 0.0304 3.003
7 hsa05215:Prostate cancer BCL2, BRAF, CREB1, CREBBP, CTNNB1, EGF, IKBKG, NFKB1, NFKBIA, PDGFD, PIK3CA, PIK3CB, PIK3CD, PIK3R1, RELA, TCF7, TP53 0.0042 0.0333 3.384
8 hsa05212:Pancreatic cancer BRAF, CDK6, EGF, IKBKG, NFKB1, PIK3CA, PIK3CB, PIK3CD, PIK3R1, RALGDS, RELA, SMAD3, STAT1, TGFB1, TP53 0.0054 0.0429 3.691
9 hsa04620:Toll-like receptor signaling pathway CD86, IKBKE, IKBKG, IRF5, JUN, MAP2K3, MAP3K8, NFKB1, NFKBIA, PIK3CA, PIK3CB, PIK3CD, PIK3R1, RELA, STAT1, TICAM1, TICAM2, TRAF3 0.0058 0.0455 3.158
10 hsa04510:Focal adhesion ACTB, ACTG1, BCAR1, BCL2, BIRC3, BRAF, CAPN2, COL1A1, COL4A2, CTNNB1, EGF, JUN, LAMC2, MYL2, MYL5, PARVB, PDGFD, PIK3CA, PIK3CB, PIK3CD, PIK3R1, RAPGEF1, SRC, TNR, VAV2, VCL 0.0193 0.1516 2.292

By importing Entrez gene IDs of 918 ChIP-Seq-based NF-κB p65 target genes into the Functional Annotation tool of DAVID, KEGG pathways showing significant relevance to the imported genes were identified. They are listed withp-value corrected by Bonferroni's multiple comparison test, false discovery rate (FDR), and fold enrichment (FE).

gr6

Fig. 6 IPA functional networks of ChIP-Seq-based NF-κB p65 target genes. Entrez Gene IDs of 918 ChIP-Seq-based NF-κB p65 target genes were imported into the core analysis tool of IPA. It extracted functional networks relevant to the set of imported genes ( Supplementary Table 4 ). The third rank network termed “Cellular Function and Maintenance, Immunological Disease, Cell Signaling” is shown, where NF-κB p65 target genes is colored by red. TRAF4 is highlighted by a blue ellipse. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

KeyMolnet by the neighboring network-search algorithm operating on the core contents extracted the highly complex molecular network composed of 3177 molecules and 5489 molecular relations. The network showed the most significant relationship with canonical pathways termed as “transcriptional regulation by p53” (p=2.00E-292), “transcriptional regulation by CREB” (p=1.44E-238), and “transcriptional regulation by NF-κB” (p=1.50E-169) ( Supplementary Fig. 1 ). These results suggest an involvement of the complex cross talk among core transcription factors p53, CREB, and NF-κB in the molecular network of 918 ChIP-Seq-based NF-κB p65 target genes.

4.3. ChIP-Seq-based NF-κB p65 target genes corresponding to MS risk alleles and MS lesion-specific proteins

Finally, we studied the relevance of ChIP-Seq-based NFκB p65 target genes to the immunopathogenesis of MS. The recent large-scale collaborative genome-wide association study (GWAS) involving 9772 cases of European origin discovered the collection of 102 MS risk SNPs outside the MHC region ( International Multiple Sclerosis Genetics Consortium et al., 2011 ). They validated 98 of the 102 SNPs overrepresented in MS patients versus the controls. Among the set of 98 genes, we found that CLEC16A, CD86, RGS14, ARHGEF3, TCF7, BATEF, EVI5, RNF213, ODF3B, and ZFP36L1 correspond to ChIP-Seq-based NF-κB p65 target genes.

A different study by a high-throughput proteomics technique comprehensively characterized the profile of MS lesion-specific proteome ( Han et al., 2008 ). They isolated proteins by laser-captured microdissection (LCM) from frozen brain samples of histologically validated acute plaques (AP), chronic active plaques (CAP), or chronic plaques (CP) of progressive MS. Peptide fragments were processed for mass spectrometric analysis. They identified 154, 405 and 231 MS lesion-specific proteins detected exclusively in AP, CAP and CP, respectively. The CAP proteome showed significant relationship with integrin-extracellular matrix interaction ( Satoh et al., 2009 ). We compared 709 MS lesion-specific proteins with 918 ChIP-Seq-based NF-κB p65 target genes. Totally, 49 MS lesion-specific proteins (6.2%) were classified into ChIP-Seq-based NF-κB p65 target genes ( Table 3 ). These results suggest that aberrant regulation of NF-κB-mediated gene expression, by inducing dysfunction of diverse immune functions, is actively involved in development of inflammatory demyelination in MS.

Table 3 ChIP-Seq-based NF-κB p65 target genes corresponding to MS lesion-specific proteome.

MS plaque type (p65 targets/proteome) Entrez gene ID Gene symbol Gene name UniProt ID
AP (10/154) 375790 AGRN Agrin O00468
23092 ARHGAP26 Rho GTPase activating protein 26 Q9UNA1
527 ATP6V0C ATPase, H+ transporting, lysosomal 16kDa, V0 subunit c P27449
284001 CCDC57 Coiled-coil domain containing 57 Q2TAC2
1209 CLPTM1 Cleft lip and palate associated transmembrane protein 1 O96005
59269 HIVEP3 Human immunodeficiency virus type I enhancer binding protein 3 Q9BZS0
114783 LMTK3 Lemur tyrosine kinase 3 Q96Q04
4649 MYO9A Myosin IXA Q9UNJ2
5529 PPP2R5E Protein phosphatase 2, regulatory subunit B′, epsilon isoform Q16537
54434 SSH1 Slingshot homolog 1 (Drosophila) Q8WYL5
 
CAP (21/405) 3732 CD82 CD82 molecule P27701
1277 COL1A1 Collagen, type I, alpha 1 P02452
953 ENTPD1 Ectonucleoside triphosphate diphosphohydrolase 1 P49961
54932 EXD3 Exonuclease 3′-5′ domain containing 3 Q8N9H8
83856 FSD1L Fibronectin type III and SPRY domain containing 1-like Q9BXM9
2902 GRIN1 Glutamate receptor, ionotropic, N-methyl D-aspartate 1 Q05586
4151 MB Myoglobin P02144
4241 MFI2 Antigen p97 (melanoma associated) identified by monoclonal antibodies 133.2 and 96.5 P08582
4641 MYO1C Myosin IC O00159
100128731 OST4 Oligosaccharyltransferase 4 homolog (S. cerevisiae) P0C6T2
55690 PACS1 Phosphofurin acidic cluster sorting protein 1 Q6VY07
5142 PDE4B Phosphodiesterase 4B, cAMP-specific (phosphodiesterase E4 dunce homolog, Drosophila) Q07343
5581 PRKCE Protein kinase C, epsilon Q02156
10801 SEPT9 Septin 9 Q9UHD8
9644 SH3PXD2A SH3 and PX domains 2A Q5TCZ1
6714 SRC v-src sarcoma (Schmidt–Ruppin A-2) viral oncogene homolog (avian) P12931
10809 STARD10 StAR-related lipid transfer (START) domain containing 10 Q9Y365
6772 STAT1 Signal transducer and activator of transcription 1, 91kDa P42224
9144 SYNGR2 Synaptogyrin 2 O43760
5976 UPF1 UPF1 regulator of nonsense transcripts homolog (yeast) Q92900
152485 ZNF827 Zinc finger protein 827 Q17R98
 
CP (18/231) 83478 ARHGAP24 Rho GTPase activating protein 24 Q8N264
521 ATP5I ATP synthase, H+ transporting, mitochondrial F0 complex, subunit E P56385
8030 CCDC6 Coiled-coil domain containing 6 Q16204
1284 COL4A2 Collagen, type IV, alpha 2 P08572
51700 CYB5R2 Cytochrome b5 reductase 2 Q6BCY4
9732 DOCK4 Dedicator of cytokinesis 4 Q8N1I0
26088 GGA1 Golgi associated, gamma adaptin ear containing, ARF binding protein 1 Q9UJY5
2788 GNG7 Guanine nucleotide binding protein (G protein), gamma 7 O60262
4650 MYO9B Myosin IXB Q13459
9612 NCOR2 Nuclear receptor co-repressor 2 Q9Y618
23178 PASK PAS domain containing serine/threonine kinase Q96RG2
221692 PHACTR1 Phosphatase and actin regulator 1 Q9C0D0
5290 PIK3CA Phosphoinositide-3-kinase, catalytic, alpha polypeptide P42336
84687 PPP1R9B Protein phosphatase 1, regulatory (inhibitor) subunit 9B Q96B17
6146 RPL22 Ribosomal protein L22 P35268
6160 RPL31 Ribosomal protein L31 P62899
10044 SH2D3C SH2 domain containing 3C Q8N5H7
8027 STAM Signal transducing adaptor molecule (SH3 domain and ITAM motif) 1 Q92783

We identified a set of 918 ChIP-Seq-based NF-κB p65 target genes ( Suplementary Table 1 ). We compared these with 790 MS lesion-specific proteins identified by a high-throughput protemics technique ( Han et al., 2008 ). Overlapping genes between both are listed with Entrez Gene ID, gene symbol, gene name, and UniProt ID.

5. Discussion

In the present study, we identified 918 NF-κB p65 ChIP-Seq peaks on protein-coding genes from the dataset of TNFα-stimulated human B lymphoblastoid cells. They were located mainly in promoter and intronic regions of target genes with an existence of the NF-κB consensus sequence motif. Our observations are supported by a previous ChIP-Chip study showing that a substantial number of NF-κB-binding sites are located in intronic regions ( Martone et al., 2003 ). Unexpectedly, only 52 genes (5.7%) were known targets by database search, suggesting that both binding of NF-κB and recruitment of appropriate coactivators to responsive elements are crucial for the full-brown activation of target genes ( Ziesché et al., 2013 ). We studied the molecular network of 918 ChIP-Seq-based NF-κB p65 target genes by using four different pathway analysis tools of bioinformatics. KEGG, PANTHER, and IPA consistently showed that the molecular network has significant relationship with regulation of immune functions and oncogenesis, including B cell receptor signaling, T cell activation pathway, Toll-like receptor signaling, apoptosis signaling, and molecular mechanisms of cancers. We identified CD22, a negative regulator of B cell receptor signaling ( Collins et al., 2006 ) and CD81, a component of the CD19 complex pivotal for antibody production ( van Zelm et al., 2010 ), as two key genes of ChIP-Seq-based NF-κB p65 targets ( Fig. 5 ). Importantly, recent evidence indicated that B cells play a central role in MS pathogenesis ( Krumbholz et al., 2012 ). However, the possibility could not be excluded that these results are derived from a bias caused by the use of EBV-transformed human B lymphoblastoid cells stimulated with TNFα. It is worthy to note that the epidemiological association of EBV with a risk for development of MS is well established ( Levin et al., 2010 ). We also identified TRAF4, a negative regulator of IL-17 signaling ( Zepp et al., 2012 ) as one of ChIP-Seq-based NF-κB p65 target genes ( Fig. 6 ). KeyMolnet revealed an involvement of the complex crosstalk among core transcription factors, such as NF-κB, p53, and CREB in the NF-κB p65 target gene network, supporting previous observations (Park et al, 2005, Schneider et al, 2010, and Oeckinghaus et al, 2011).

Although ChIP-Seq serves as a highly efficient method for genome-wide profiling of DNA-binding proteins, it requires several technical considerations ( Landt et al., 2012 ). The specificity of the antibody, reproducibility of the results, sequencing depth, and the source of controls, along with cell types, developmental stages, and culture conditions capable of affecting epigenetic features, constitute critical factors. In general, DNA-binding of transcription factors is a highly dynamic process. However, the ChIP-Seq data reflect a snapshot of binding actions. Motif analysis of a defined set of high-quality peaks makes it possible to evaluate the antibody specificity to some extent ( Landt et al., 2012 ).

Increasing evidence suggests a central role of aberrant NF-κB activation in the immunopathogenesis of MS. Pathologically, RelA (p65), c-Rel, and p50 subunits of NF-κB are overexpressed in macrophages in active demyelinating lesions of MS ( Gveric et al., 1998 ). RelA (p65) is activated in oligodendrocytes surviving in the lesion edge ( Bonetti et al., 1999 ). Genetically, a predisposing allele in the NFKBIL gene is associated with development of RRMS, while a protective allele in the promoter of the NFKBIA (IκBα) gene is found in the patients with primary progressive MS, suggesting that the NF-κB cascade contributes certainly to susceptibility to MS ( Miterski et al., 2002 ). Targeted disruption of the NFKB1 (p105) gene confers resistance to development of experimental autoimmune encephalomyelitis (EAE), an animal model of MS ( Hilliard et al., 1999 ). In vivo administration of selective inhibitors of NF-κB protects mice from EAE ( Pahan and Schmid, 2000 ). Furthermore, the CNS-restricted inactivation of NF-κB ameliorates EAE, accompanied by suppression of activation of proinflammatory genes in astrocytes ( van Loo et al., 2006 ). We found that 10 genes among 98 MS risk alleles (10.2%) and 49 proteins among 709 MS lesion-specific proteins (6.2%) have met with ChIP-Seq-based NF-κB p65 target genes, although it is difficult to evaluate the statistical significance of enrichment of MS-associated genes.

NF-κB acts as a central regulator of various cellular processes (Barnes and Karin, 1997 and Pahl, 1999). Actually, molecular network of ChIP-Seq-based NF-κB p65 target genes not only involves immune responses relevant to MS and other autoimmune diseases, such as rheumatoid arthritis (RA) ( Myouzen et al., 2012 ) and systemic lupus erythematosus (SLE) ( Zhang et al., 2012 ), but also regulates oncogenic processes of various cancers. Therefore, we consider that the involvement of molecular network of NF-κB p65 target genes in immune regulation is not a MS-specific phenomenon. However, our results would suggest that deregulation of NF-κB might be actively involved in development and progression of MS, and that drug development targeted to fine-tuning of NF-κB function in autoreactive T and B cells and CNS resident cells could provide a promising approach to suppress the clinical activity of MS ( Yan and Greer, 2008 ).

6. Conclusion

We identified the comprehensive set of 918 stringent NF-κB p65 binding sites on protein-coding genes from the ChIP-Seq dataset of TNFα-stimulated human B lymphoblastoid cells. They were located mainly in promoter and intronic regions with an existence of the NF-κB consensus sequence motif. Pathway analysis by KEGG, PANTHER, IPA, and KeyMolnet showed that the NF-κB p65 target gene network is closely associated with the network involved in regulation of diverse immune functions relevant to the immunopathogenesis of MS.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgments

This work was supported by grants from the Research on Intractable Diseases (Grant nos. H21-Nanchi-Ippan-201 and H22-Nanchi-Ippan-136), the Ministry of Health, Labour and Welfare (MHLW), Japan, and the High-Tech Research Center (HRC) Project (Grant no. S0801043), the Genome Research Center (GRC) Project, and the Grant-in-Aid (Grant nos. C22500322 and C25430054), the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

Appendix A. Supplementary materials

 

Download file

Supplementary Fig. 1 KeyMolnet networks of ChIP-Seq-based NF-κB p65 target genes. Entrez Gene IDs of 918 ChIP-Seq-based NF-κB p65 target genes were imported into KeyMolnet. The neighboring network-search algorithm extracted the highly complex molecular network composed of 3177 molecules and 5489 molecular relations. Red nodes represent NF-κB p65 target genes, whereas white nodes exhibit additional nodes extracted automatically from the core contents of KeyMolnet to establish molecular connections. The molecular relation is indicated by the solid line with arrow (direct binding or activation), solid line with arrow and stop (direct inactivation), solid line without arrow (complex formation), dashed line with arrow (transcriptional activation), and dashed line with arrow and stop (transcriptional repression). The molecular network termed “transcriptional regulation by NF-κB” is highlighted in orange.

 

Download file

Supplementary Material

 

Download file

Supplementary Material

 

Download file

Supplementary Material

 

Download file

Supplementary Material

References

  • Achiron et al., 2007 A. Achiron, A. Feldman, M. Mandel, M. Gurevich. Impaired expression of peripheral blood apoptotic-related gene transcripts in acute multiple sclerosis relapse. Annals of the New York Academy of Sciences. 2007;1107:155-167
  • Barnes and Karin, 1997 P.J. Barnes, M. Karin. Nuclear factor-κB. A pivotal transcription factor in chronic inflammatory diseases. New England Journal of Medicine. 1997;336:1066-1071
  • Bonetti et al., 1999 B. Bonetti, C. Stegagno, B. Cannella, N. Rizzuto, G. Moretto, C.S. Raine. Activation of NF-κB and c-jun transcription factors in multiple sclerosis lesions. Implications for oligodendrocyte pathology. American Journal of Pathology. 1999;155:1433-1438
  • Brüstle et al., 2012 A. Brüstle, D. Brenner, C.B. Knobbe, et al. The NF-κB regulator MALT1 determines the encephalitogenic potential of Th17 cells. Journal of Clinical Investigation. 2012;122:4698-4709
  • Collins et al., 2006 B.E. Collins, B.A. Smith, P. Bengtson, J.C. Paulson. Ablation of CD22 in ligand-deficient mice restores B cell receptor signaling. Nature Immunology. 2006;7:199-206
  • Comabella and Khoury, 2012 M. Comabella, S.J. Khoury. Immunopathogenesis of multiple sclerosis. Clinical Immunology. 2012;142:2-8
  • Eggert et al., 2008 M. Eggert, R. Goertsches, U. Seeck, S. Dilk, G. Neeck, U.K. Zettl. Changes in the activation level of NF-kappa B in lymphocytes of MS patients during glucocorticoid pulse therapy. Journal of the Neurological Sciences. 2008;264:145-150
  • Gerstein et al., 2012 M.B. Gerstein, A. Kundaje, M. Hariharan, et al. Architecture of the human regulatory network derived from ENCODE data. Nature. 2012;489:91-100
  • Gilmore, 2006 T.D. Gilmore. Introduction to NF-κB: players, pathways, perspectives. Oncogene. 2006;25:6680-6684
  • Gregersen et al., 2009 P.K. Gregersen, C.I. Amos, A.T. Lee, et al. REL, encoding a member of the NF-κB family of transcription factors, is a newly defined risk locus for rheumatoid arthritis. Nature Genetics. 2009;41:820-823
  • Gveric et al., 1998 D. Gveric, C. Kaltschmidt, M.L. Cuzner, J. Newcombe. Transcription factor NF-κB and inhibitor IκBα are localized in macrophages in active multiple sclerosis lesions. Journal of Neuropathology and Experimental Neurology. 1998;57:168-178
  • Han et al., 2008 M.H. Han, S.I. Hwang, D.B. Roy, et al. Proteomic analysis of active multiple sclerosis lesions reveals therapeutic targets. Nature. 2008;451:1076-1081
  • Hayden et al., 2006 M.S. Hayden, A.P. West, S. Ghosh. NF-κB and the immune response. Oncogene. 2006;25:6758-6780
  • Hilliard et al., 1999 B. Hilliard, E.B. Samoilova, T.S. Liu, A. Rostami, Y. Chen. Experimental autoimmune encephalomyelitis in NF-κB-deficient mice: roles of NF-κB in the activation and differentiation of autoreactive T cells. Journal of Immunology. 1999;163:2937-2943
  • Huang et al., 2009 da W. Huang, B.T. Sherman, R.A. Lempicki. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols. 2009;4:44-57
  • International Multiple Sclerosis Genetics Consortium et al., 2011 International Multiple Sclerosis Genetics Consortium, Wellcome Trust Case Control Consortium 2, S. Sawcer, et al. Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature. 2011;476:214-219
  • Kasowski et al., 2010 M. Kasowski, F. Grubert, C. Heffelfinger, et al. Variation in transcription factor binding among humans. Science. 2010;328:232-235
  • Krumbholz et al., 2012 M. Krumbholz, T. Derfuss, R. Hohlfeld, E. Meinl. B cells and antibodies in multiple sclerosis pathogenesis and therapy. Nature Reviews Neurology. 2012;8:613-623
  • Landt et al., 2012 S.G. Landt, G.K. Marinov, A. Kundaje, et al. ChIP-Seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Research. 2012;22:1813-1831
  • Levin et al., 2010 L.I. Levin, K.L. Munger, E.J. O'Reilly, K.I. Falk, A. Ascherio. Primary infection with the Epstein-Barr virus and risk of multiple sclerosis. Annals of Neurology. 2010;67:824-830
  • Lindsey et al., 2011 J.W. Lindsey, S.K. Agarwal, F.K. Tan. Gene expression changes in multiple sclerosis relapse suggest activation of T and non-T cells. Molecular Medicine. 2011;17:95-102
  • Martone et al., 2003 R. Martone, G. Euskirchen, P. Bertone, et al. Distribution of NF-κB-binding sites across human chromosome 22. Proceedings of the National Academy of Sciences of the United States of America. 2003;100:12247-12252
  • Martín-Saavedra et al., 2007 F.M. Martín-Saavedra, N. Flores, B. Dorado, et al. Beta-interferon unbalances the peripheral T cell proinflammatory response in experimental autoimmune encephalomyelitis. Molecular Immunology. 2007;44:3597-3607
  • Miterski et al., 2002 B. Miterski, S. Böhringer, W. Klein, et al. Inhibitors in the NFκB cascade comprise prime candidate genes predisposing to multiple sclerosis, especially in selected combinations. Genes and Immunity. 2002;3:211-219
  • Myouzen et al., 2012 K. Myouzen, Y. Kochi, Y. Okada, et al. Functional variants in NFKBIE and RTKN2 involved in activation of the NF-κB pathway are associated with rheumatoid arthritis in Japanese. PLoS Genetics. 2012;8:e1002949
  • Oeckinghaus et al., 2011 A. Oeckinghaus, M.S. Hayden, S. Ghosh. Crosstalk in NF-κB signaling pathways. Nature Immunology. 2011;12:695-708
  • Pahan and Schmid, 2000 K. Pahan, M. Schmid. Activation of nuclear factor-κB in the spinal cord of experimental allergic encephalomyelitis. Neuroscience Letters. 2000;287:17-20
  • Pahl, 1999 H.L. Pahl. Activators and target genes of Rel/NF-κB transcription factors. Oncogene. 1999;18:6853-6866
  • Park et al., 2005 J.M. Park, F.R. Greten, A. Wong, et al. Signaling pathways and genes that inhibit pathogen-induced macrophage apoptosis. CREB and NF-κB as key regulators. Immunity. 2005;23:319-329
  • Park, 2009 P.J. Park. ChIP-Seq: advantages and challenges of a maturing technology. Nature Reviews Genetics. 2009;10:669-680
  • Rothwarf and Karin, 1999 D.M. Rothwarf, M. Karin. The NF-κB activation pathway: a paradigm in information transfer from membrane to nucleus. Science's STKE. 1999;1999 RE1
  • Satoh et al., 2008 J. Satoh, T. Misawa, H. Tabunoki, T. Yamamura. Molecular network analysis of T-cell transcriptome suggests aberrant regulation of gene expression by NF-κB as a biomarker for relapse of multiple sclerosis. Disease Markers. 2008;25:27-35
  • Satoh and Tabunoki, 2013 Satoh J, Tabunoki H. Molecular network of ChIP-Seq-based vitamin D receptor target genes. Multiple Sclerosis; 2013 [in press]. PMID: 23401126. DOI: 10.1177/1352458512471873.
  • Satoh, 2010 J. Satoh. Bioinformatics approach to identifying molecular biomarkers and networks in multiple sclerosis. Clinical and Experimental Neuroimmunology. 2010;1:127-140
  • Satoh et al., 2009 J.I. Satoh, H. Tabunoki, T. Yamamura. Molecular network of the comprehensive multiple sclerosis brain-lesion proteome. Multiple Sclerosis. 2009;15:531-541
  • Schneider et al., 2010 G. Schneider, A. Henrich, G. Greiner, et al. Cross talk between stimulated NF-κB and the tumor suppressor p53. Oncogene. 2010;29:2795-2806
  • Sica et al., 1997 A. Sica, L. Dorman, V. Viggiano, et al. Interaction of NF-κB and NFAT with the interferon-γ promoter. Journal of Biological Chemistry. 1997;272:30412-30420
  • Viatour et al., 2005 P. Viatour, M.P. Merville, V. Bours. Chariot A. Phosphorylation of NF-κB and IκB proteins: implications in cancer and inflammation. Trends in Biochemical Sciences. 2005;30:43-52
  • Yan and Greer, 2008 J. Yan, J.M. Greer. NF-κBa potential therapeutic target for the treatment of multiple sclerosis. CNS & Neurological Disorders: Drug Targets. 2008;7:536-557
  • van Loo et al., 2006 G. van Loo, R. De Lorenzi, H. Schmidt, et al. Inhibition of transcription factor NF-κB in the central nervous system ameliorates autoimmune encephalomyelitis in mice. Nature Immunology. 2006;7:954-961
  • van Zelm et al., 2010 M.C. van Zelm, J. Smet, B. Adams, et al. CD81 gene defect in humans disrupts CD19 complex formation and leads to antibody deficiency. Journal of Clinical Investigation. 2010;120:1265-1274
  • Zepp et al., 2012 J.A. Zepp, C. Liu, W. Qian, et al. Cutting edge: TNF receptor-associated factor 4 restricts IL-17-mediated pathology and signaling processes. Journal of Immunology. 2012;189:33-37
  • Zhang et al., 2012 W. Zhang, Q. Shi, X. Xu, et al. Aberrant CD40-induced NF-κB activation in human lupus B lymphocytes. PLoS One. 2012;7:e41644
  • Ziesché et al., 2012 Ziesché E, Kettner-Buhrow D, Weber A, et al. The coactivator role of histone deacetylase 3 in IL-1-signaling involves deacetylation of p65 NF-κB. Nucleic Acids Research; 2013;41:90–109. 10.1093/nar/gks916 .

Footnotes

Department of Bioinformatics and Molecular Neuropathology, Meiji Pharmaceutical University, 2-522-1 Noshio, Kiyose, Tokyo 204-8588, Japan

lowast Tel./fax: +81 42 495 8678.