Gene Expression Profiling in Conjunction with PhysiologicalRescues of IKK�-null Cells with Wild Type or Mutant IKK�Reveals Distinct Classes of IKK�/NF-�B-dependent Genes*
Received for publication, December 22, 2004Published, JBC Papers in Press, February 4, 2005, DOI 10.1074/jbc.M414401200
Paul E. Massa‡§¶, Xiang Li�, Adedayo Hanidu�, John Siamas§, Milena Pariali¶, Jessica Pareja**,Anne G. Savitt**, Katrina M. Catron�, Jun Li�‡‡, and Kenneth B. Marcu‡§¶**§§
From the ‡Genetics Graduate Program, the Departments of §Biochemistry and Cell Biology and **Microbiology, Institutefor Cell and Developmental Biology, State University of New York at Stony Brook, Stony Brook, New York 11794-5215,the �Department of Immunology and Inflammation, Boehringer Ingelheim Pharmaceuticals, Ridgefield, Connecticut06877-0368, and the ¶Center for Applied Biomedical Research, San Orsola Hospital, University of Bologna,Via Massarenti 9, Bologna 40138, Italy
Cellular responses to stress-like stimuli require theI�B kinase (IKK) signalsome (IKK�, IKK�, and NEMO/IKK�) to activate NF-�B-dependent genes. IKK� andNEMO/IKK� are required to release NF-�B p65/p50 het-erodimers from I�B�, resulting in their nuclear migra-tion and sequence-specific DNA binding; but IKK� wasfound to be dispensable for this initial phase of canoni-cal NF-�B activation. Nevertheless, IKK�(�/�) mouseembryonic fibroblasts (MEFs) fail to express NF-�B tar-gets in response to proinflammatory stimuli, uncoveringa nuclear role for IKK� in NF-�B activation. However, itremains unknown whether the global defect in NF-�B-dependent gene expression of IKK�(�/�) cells is causedby the absence of IKK� kinase activity. We show by geneexpression profiling that rescue of near physiologicallevels of wild type IKK� in IKK�(�/�) MEFs globallyrestores expression of their canonical NF-�B targetgenes. To prove that the kinase activity of IKK� wasrequired on a genomic scale, the same physiological res-cue was performed with a kinase-dead, ATP bindingdomain IKK� mutant (IKK�(K44M)). Remarkably, theIKK�(K44M) protein rescued �28% of these genes, albeitin a largely stimulus-independent manner with the no-table exception of several genes that also acquired tu-mor necrosis factor-� responsiveness. Thus the IKK�-containing signalsome unexpectedly functions in thepresence and absence of extracellular signals in bothkinase-dependent and -independent modes to differen-tially modulate the expression of five distinct classes ofIKK�/NF-�B-dependent genes.
The NF-�B pathway is important for a host of cellular pro-cesses including its central role in responses to stress-likestimuli, the antiapoptotic cascade, the initiation and mainte-nance of immune responses, embryonic and adult tissue devel-opment, and cell cycle progression (for review, see Refs. 1–11).In mammals the NF-�B family of transcription factors is com-prised of five subunits characterized by the presence of a con-served Rel homology DNA binding domain (for review, see Refs.2, 12, and 13). The p65(RelA), c-Rel, and RelB NF-�B subunitsare fully functional transcriptional activators, whereas the p50and p52 subunits lack a transcriptional activation domain (forreview, see Refs. 12 and 13). NF-�Bs function as specific hetero-or homodimers that bind to a GGGRNWTYCC consensus DNAsequence found in the promoters or enhancers of NF-�B targetgenes (for review, see Refs. 12 and 13). Transcriptional activat-ing NF-�B subunits are normally sequestered in the cytoplasmof unstimulated cells in a complex with one of the I�B familyproteins, which block their nuclear import and DNA bindingactivity. A large variety of extracellular activating stimuli in-duce the proteosomal dependent destruction of I�Bs therebydifferentially freeing NF-�Bs to bind DNA and activate thetranscription of their genomic targets (for review, see Ref. 14).Stress-like inducers of NF-�B (including proinflammatory cy-tokines such as TNF�1 and IL-1) function in a classicalmonophasic capacity to drive the canonical NF-�B activationpathway rapidly, which largely involves the activation ofp65(RelA)/p50 DNA binding activity and transcriptional com-petence (for review, see Refs. 1, 2, 8, and 15). More recently, adistinct class of NF-�B stimuli (exemplified by LT�, BAFF, andCD40 ligand), which also contribute to the implementation ofdifferentiation programs and the adaptive phase of immuneresponses, have been shown to function as biphasic activatorsinitially acting via the rapid canonical pathway and subse-quently feeding into a delayed noncanonical protein synthesis-dependent route characterized by the activation of RelB/p52heterodimers (for review, see Refs. 8, 9, 15, and 16).
With the exceptions of UV radiation and the effects of some
* The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked“advertisement” in accordance with 18 U.S.C. Section 1734 solely toindicate this fact. This research was supported in part by a NationalInstitutes of Health grant (to K. B. M.) and by the Fondazione Cassa diRisparmio di Bologna. The costs of publication of this article weredefrayed in part by the payment of page charges. This article musttherefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.
‡‡ To whom correspondence may be addressed: Dept. of Immunologyand Inflammation, Boehringer Ingelheim Pharmaceuticals, 900 Ridge-bury Rd., P. O. Box 368, Ridgefield, CT 06877-0368. Tel.: 203-798-5714;Fax: 203-837-5714; E-mail: [email protected].
§§ Senior scholar of the Institute of Advanced Studies of the Univer-sity of Bologna. To whom correspondence may be addressed: Dept. ofBiochemistry and Cell Biology, SUNY at Stony Brook, 330 Life SciencesBldg., Stony Brook, NY 11794-5215. Tel.: 631-632-8553; Fax: 631-632-9730; E-mail: [email protected].
1 The abbreviations used are: TNF�, tumor necrosis factor �; CBP,CREB-binding protein; C/EBP, CCAAT/enhancer-binding protein;CREB, cAMP-response element-binding protein; EV, empty vector; IL,interleukin; MEF, mouse embryonic fibroblast; HA, hemagglutinin;M/CSF, macrophage colony-stimulating factor; MMP, matrix metallo-proteinase; RANTES, regulated on activation normal T cell expressedand secreted; RIP, receptor-interacting protein; TAD, transcriptionalactivation domain; VCAM1, vascular cell adhesion molecule 1; Wt., wildtype; 2T, 2-h TNF� stimulation; SMRT, silencing mediator of retinoidand thyroid hormone receptors; IKK, I�B kinase.
This paper is available on line at http://www.jbc.org 14057
DNA-damaging agents (17, 18), the release of NF-�Bs fromI�Bs is mediated by the cytoplasmic signalsome complex, whichconsists of two serine-threonine kinases (IKK�, IKK�) andNEMO/IKK�, a regulatory/docking protein (for review, seeRefs. 1, 7, 8, and 15). IKK� is essential for the phosphorylationof I�Bs on a pair of amino-terminal serines (residues 32 and 36in I�B�) thereby targeting I�B for ubiquitination and subse-quent proteosomal destruction (for review, see Refs. 1, 7, and8). In contrast, IKK� is not required for the phosphorylation ofI�Bs via the canonical NF-�B activation pathway in vivo, withthe exception of receptor activator of NF-�B (RANK) ligandsignaling in mammary epithelial cells (19). Rather, IKK� playsan essential role in epidermal keratinocyte differentiation in-dependent of both its kinase activity and NF-�B activation andhas also recently been found to play NF-�B-dependent and-independent roles in tooth development (20–22). With respectto its physiological role in NF-�B signaling pathways, IKK� isinstead essential for the activation of the noncanonical NF-�Bactivation pathway, which requires neither IKK� nor NEMO/IKK� (for review, see Refs. 8, 15, and 16). In this context, viaNF-�B-inducing kinase-dependent signaling, IKK� phospho-rylates multiple serines of the p100 precursor of the p52 sub-unit, thereby inducing its proteosome-dependent processinginto mature p52 subunits that are then freed to activate NF-�Btarget genes as RelB/p52 heterodimers (8, 9, 23, 24). We andother groups have also found that IKK� is required to activatethe transcription of canonical NF-�B target genes (25–29). Thelatter dependence on IKK� is independent of I�B� destructionand instead appears to involve one or more nuclear targetsperhaps including histone H3 (27, 28) and the SMRT transcrip-tional corepressor (29) resulting in the de-repression of NF-�Btarget genes.
In this report we have investigated the physiological require-ment of the kinase activity of IKK� for the expression of NF-�B-dependent genes on a genomic scale in IKK�-null MEFs.Physiological expression of Wt. IKK� in IKK�(�/�) MEFs byretroviral transduction resulted in the rescue of specific NF-�B-dependent genes in the presence and absence of TNF�stimulation. Comparative microarray screens with NF-�B-compromised MEFs (p50(�/�) and Wt. � I�B�(S32A,S36A))revealed that the large majority of these IKK�-rescued genesare either dependent on basal or TNF�-inducible NF-�B, thusdemonstrating that 1) IKK� plays an essential role in control-ling the expression of both signal-induced and basal NF-�B-de-pendent genes, and 2) IKK� does not appear to influence theexpression of a large number of genes outside the NF-�B path-way. Comparable physiological rescue with a kinase-deadIKK� mutant protein (IKK�(K44M)) showed that most of thesegenes are dependent on IKK� kinase activity for their stimulus-dependent and -independent expression. However, the expres-sion of up to 28% of these NF-�B-dependent genes was alsosurprisingly rescued by the kinase-inactive IKK�(K44M) mu-tant. Furthermore both wild type and mutant IKK� are alsorequired for the basal levels of expression of specific NF-�B-de-pendent genes. Thus, our findings collectively reveal that thelevels of expression of different downstream NF-�B-dependentgenes are differentially codependent on catalytically activeIKK� in the presence or absence of extracellular stimuli.
EXPERIMENTAL PROCEDURES
Tissue Culture—Growth of IKK�(�/�) MEFs and their stimulationwith TNF� were performed as described previously. Wt. IKK�/CHUK-HA or IKK�(K44M)-HA (a kinase-inactive ATP binding domainmutant with lysine 44 mutated to methionine) with carboxyl-terminalHA epitope tags (30–32) was introduced into IKK�(�/�) MEFs bytransduction with a retroviral vector coexpressing a puromycin resist-ance gene followed by 6–8 days of puromycin selection (26, 33).
IKK�(�/�) cells harboring an empty retroviral vector (EV cells) weresimultaneously generated as a matched negative control.
RNA Preparation—Total cellular RNAs were extracted from celllysates with an RNeasy kit (Qiagen). Purified RNAs were converted todouble-stranded cDNA with a Super Script Double Stranded cDNAsynthesis kit (Invitrogen) and an oligo(dT) primer containing a T7 RNApolymerase promoter (GENSET). Biotin-labeled cRNAs were generatedfrom the cDNA samples by in vitro transcription with T7 RNA polym-erase (Enzo kit, Enzo Diagnostics). The labeled cRNAs were fragmentedto an average size of 35–200 bases by incubation at 94 °C for 35 min.Hybridization (16 h), washing, and staining protocols have been de-scribed (Affymetrix Gene Chip Expression Analysis Technical Manual;(34)).
DNA Microarrays and Clustering Analysis—We employed Affy-metrix MG-U74Av2 chips that include 12,400 genes. Chips werestained with streptavidin-phycoerythrin (Molecular Probes) andscanned with a Hewlett-Packard Gene Array Scanner. DNA microarraychip data analysis was performed using MAS5.1 software (Affymetrix)and as described previously (26). Levels of gene expression were quan-titated from the hybridization intensities of 16 pairs of perfectlymatched and mismatched control probes (35) (Affymetrix, Inc.). Theaverage of the differences (perfectly matched minus mismatched) foreach gene-specific probe family was calculated and expressed as signalvalues. The software computes how the expression level of each tran-script has changed between the base line and experimental samples(difference call/change call). A change call is a qualitative call thatdescribes whether a transcript in an experimental array has changedcompared with a base-line array. One array is designated as the exper-imental, and another array is designated as the base line. Wilcoxon’ssigned rank is used to generate a change p value. A change call isassigned based on analysis parameters. Change p values between 0.00and 0.0025 are given an Increase call. Change p values between 0.0025and 0.003 are given a marginal increase call. Change p values between0.997 and 0.998 are given a marginal decrease call. Change p valuesbetween 0.998 and 1.00 are given a decrease call (Affymetrix UserManual). For a comparative chip file (such as TNF�-stimulatedIKK�(�/�) MEF � Wt. IKK� versus IKK�(�/�) MEF � EV, the ex-perimental file (Wt. IKK� 2T) was compared with the base-line file (EV2T). We employed the following stringent selection criteria to identifysignificant changes in gene expression: 1) a change call of increase ormarginal increase in both samples, and 2) average -fold change valuesof 1.5 or greater (minimum of 1.3-fold each) in two independentlystimulated samples of IKK�(�/�) MEF � Wt. IKK� 2T versusIKK�(�/�) MEF � EV 2T. The following additional criteria were em-ployed to identify the spectrum of these genes that were also dependentupon NF-�B for their expression: 1) a change call of either increase ormarginal increase in Wt. MEF 2T versus Wt. MEF � I�B�SR-Ires-neomycin, or 2) a change call of either increase or marginal increase inWt. MEF 2T versus p50(�/�) MEF 2T, and 3) a valid presence call inWt. 2T screens. By this analysis we define NF-�B dependence to rep-resent genes whose expressions either directly (true direct targets ofNF-�B subunits) or indirectly (other downstream genes whose expres-sions are affected by the NF-�B pathway) require NF-�B. The TNF�inducibilities of genes rescued by either Wt. IKK� or IKK�(K44M) werebased on a combination of the following stringent criteria: 1) increasecalls in duplicate 2T versus US microarray screens of IKK�(�/�) MEFsrescued by Wt. IKK� or IKK�(K44M), and/or 2) TaqMan real time PCRanalysis performed at least in duplicate and an increase call in one oftwo screens. By these criteria, the vast majority of IKK�-rescued genesthat responded to TNF� did so with -fold change values of 2.0 andhigher in duplicate screens with a minimum unstimulated (US) average-fold change value of 1.7.
Hierarchical clustering was performed with the Cluster program(available at rana.lbl.gov/) as described previously (36). Genes that havedouble increase calls and are induced �1.5-fold (average -fold values) inthe duplicate primary Wt. IKK� versus EV-rescued IKK�(�/�) MEFs(see above description) were selected. The signal values (equivalent tothe quantities of mRNAs, see above) of the selected genes were median-centered by subtracting the median observed value and normalized bygenes to the magnitude (sum of the squares of the values) of a rowvector to 1.0. The normalized data were clustered by average linkageclustering analysis of the y axis (genes) using an uncentered correlationsimilarity metric, as described in the program Cluster. Signal values of50 or less were set to 50 before centering and normalization. Theclustered data were visualized with the Treeview program (available atrana.lbl.gov/).
TaqMan Real Time Quantitative PCR—TaqMan real time quantita-tive PCRs were performed as described previously (26, 37). Data from
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14058
TaqMan PCR analyses were normalized based on glyceraldehyde-3-phosphate dehydrogenase mRNA copy numbers using rodent glyceral-dehyde-3-phosphate dehydrogenase control reagents (Applied Biosys-tems). TaqMan probe sets were designed for the following genes usingeither Primer Express 1.5 or Bio-Rad Beacon Designer 2.0 software:ATF3, A20, ISG15, MyD118/GADD45�, SAA3, and VCAM1. MurineIL-6 and RANTES TaqMan reagents were obtained from ABI. TaqManreal time PCR experiments were performed in an ABI PRISM 7700sequence detector or in a Bio-Rad iCycler. DNA sequences for each ofthese probe sets are available upon request.
Western Blotting—Cell lysates were prepared in an isotonic lysis buffercontaining 1% Nonidet P-40 supplemented with protease inhibitors. SDS-PAGE (10% gel) transfer to polyvinylidene difluoride membranes wasperformed as described previously, and membranes were probed withprimary antibodies to either IKK� (Cell Signaling Technology) or NEMO(Santa Cruz Biotechnology) followed by an anti-rabbit-horseradish perox-idase-conjugated secondary antibody (Amersham Biosciences). Blots weredeveloped using a Lumi-Light Plus kit (Roche Applied Science).
Immunostaining—Prior to immunostaining cells were maintained in10-cm tissue culture-treated plates in their regular growth medium.Cells were trypsinized, resuspended in 12 ml of growth medium, andplated at 3 ml/well in 6-well plates containing 22-mm glass microscopecoverslips (VWR) precoated with poly-L-lysine (Sigma). Cells were in-cubated overnight at 37 °C with 5% CO2. The following day, cells werewashed and then fixed in 50% methanol and 50% acetone for 10 min.These and all subsequent washes were with phosphate-buffered saline.The fixed cells were rehydrated in phosphate-buffered saline, and thecoverslips were washed, blocked with phosphate-buffered saline con-taining 10% heat-denatured fetal bovine serum for 1 h, and washedagain. In situ expression of retrovirally transduced Wt. IKK�-HA andIKK�(K44M)-HA proteins in IKK�(�/�) cells were specifically detectedwith a primary anti-HA 12CA5 antibody (38). 12CA5 antibody wasdiluted in blocking solution and applied to the cells followed by a 1-hincubation at room temperature. After washing, alkaline phosphatase-conjugated goat anti-mouse IgG (Jackson) secondary antibody, diluted1:2,000 in block solution, was applied for a 1-h incubation at roomtemperature. The coverslips were washed and nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate developing substrate applied(Roche Applied Science). Cells were visualized on a Nikon Diaphotphase-contrast microscope and photographed using a Nikon D1X digitalcamera.
NF-�B DNA Binding—Activation of NF-�B p65-dependent DNAbinding activity was quantitated with an enzyme-linked immunosor-bent assay-based kit (Active Motif, Inc.) (39). Nuclear extracts wereprepared from unstimulated or 2-h TNF�-stimulated WT. IKK�-nulland Wt. IKK� or IKK�(K44M)-rescued IKK�(�/�) MEFs and appliedto 96-well plates containing an immobilized NF-�B DNA binding con-sensus oligonucleotide according to the manufacturer’s instructions(Active Motif, Inc.). DNA-bound NF-�B was detected with a p65-specificprimary antibody followed by the addition of a secondary antibodyconjugated to horseradish peroxidase and absorbance quantitated at450 nm with a microplate spectrophotometer. The specificity of NF-�BDNA binding was confirmed by competitions with wild type and mutantNF-�B binding sequences. Negative controls for stimulus-dependentNF-�B nuclear localization and DNA binding activity also includednuclear extracts prepared from NEMO(�/�) and p65/p50(�/�) MEFs.
RESULTS
Physiological Rescue of IKK�(�/�) MEFs with Wt. IKK�IKK�(K44M) Proteins Does Not Interfere with the Induction ofNF-�B DNA Binding Activity—By employing DNA microarraychip technology, we previously reported that the IKK� proteinwas as essential as the IKK� and NEMO/IKK� signalsomesubunits for the genomic NF-�B-dependent transcriptional re-sponse induced by TNF� or IL-1 stimulation. Our finding of astrict requirement for IKK� in the regulation of NF-�B-depend-ent transcription in MEFs was controversial because earlierstudies had shown that cells derived from IKK�-null miceexhibited no significant defect in the stimulus-dependent in-duction of NF-�B nuclear localization and DNA binding activ-ity. However, in agreement with our observations, two otherstudies had shown that IKK� was required for the TNF�- andIL-1-dependent transcriptional induction of IL-6 gene expres-sion (25, 40). Subsequent to these reports, chromatin immuno-precipitation experiments showed that the role of IKK� in
engendering DNA-bound NF-�B with transcriptional compe-tence was associated with its TNF�-dependent binding to theI�B� gene promoter and with its ability to directly phosphoryl-ate histone H3 on serine 10 in vitro (27, 28) and more recentlyalso to phosphorylate and thereby facilitate the release of theSMRT corepressor from specific NF-�B target gene promoters(29).
To determine whether the intrinsic defect of IKK�(�/�)MEFs to express NF-�B-dependent genes on a genomic scalewas solely caused by the absence of a functional IKK� kinase,we employed retroviral transduction to rescue physiologicallevels of Wt. IKK� expression in a large population ofIKK�(�/�) MEFs. To determine whether IKK� kinase activitywas required to rescue the expression of their NF-�B-depend-ent genes, we also derived a similar population of IKK�(�/�)MEFs expressing near physiological levels of a kinase-deadIKK� mutant (IKK�(K44M)) (31, 32). To rule out any effects ofstable retroviral transduction, we also generated a matchednegative control population of IKK�(�/�) MEFs harboring theempty retroviral vector (IKK�(�/�)-EV cells). Western blottingrevealed that the levels of Wt. IKK� and IKK�(K44M) expres-sion in IKK�(�/�)-rescued MEFs were similar to the expres-sion of endogenous IKK� in wild type NF-�B competent MEFs(see Fig. 1). Importantly, as shown in Fig. 2, in situ immuno-staining also revealed uniform expression of either the Wt.IKK�-HA or mutant IKK�(K44M)-HA proteins in their respec-tive stably transduced cell populations. These stable popula-tions of IKK�(�/�) cells expressing physiologically comparablelevels of either Wt. IKK�-HA or IKK�(K44M)-HA were used inall subsequent experiments.
As discussed above, studies with IKK�-null mice haveproven that the loss of IKK� has no effect on the induction ofNF-�B DNA binding activity by proinflammatory cytokines(40–42). Consequently, before proceeding to compare the ef-fects of exogenous wild type and kinase-dead IKK� on theTNF�-induced NF-�B-dependent transcriptional response ofIKK�-null MEFs, it was necessary for us to verify that ourretrovirally derived cell populations exhibited TNF�-inducedNF-�B DNA binding activity comparable with that of theirIKK�(�/�) counterparts and Wt. MEFs. To this end, we em-ployed a quantitative enzyme-linked immunosorbent assay-based DNA binding assay to directly compare levels of NF-�Bp65 subunit DNA binding activity induced by TNF� stimula-tion. Nuclear extracts of NEMO/IKK�(�/�) and p65/p50(�/�)MEFs were employed as negative controls. As shown in Fig. 3,comparable levels of NF-�B p65 DNA binding activity wereinduced upon TNF� stimulation of wild type, IKK�(�/�) � EV,IKK�(�/�) � Wt. IKK�, and IKK�(�/�) � IKK�(K44M)
FIG. 1. Physiological levels of expression of Wt. IKK� andIKK�(K44M) in rescued populations of IKK�(�/�) MEFs. Popu-lations of IKK�(�/�) MEFs were retrovirally transduced to expressstably either murine Wt. IKK�-HA or an IKK�(K44M)-HA mutantprotein. SDS-PAGE was performed to determine the levels of IKK�expression obtained in the infected populations (top) (see “ExperimentalProcedures”). The membrane was stripped and reprobed with a NEMO-specific antibody as a reference control (bottom), which showed compa-rable expression in each cell background.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14059
MEFs. To validate the specificity of the DNA binding reactions,NF-�B p65-dependent DNA binding was abolished in nuclearextracts of TNF�-induced IKK�(�/�) � Wt. IKK� andIKK�(�/�) � IKK�(K44M) cells by an excess of a wild typeNF-�B binding oligonucleotide but not by a mutant NF-�Bbinding sequence (Fig. 3). Importantly, this experiment dem-onstrates that when expressed at near physiological levels, theIKK�(K44M) mutant protein does not function in a generaldominant-negative manner with respect to the nuclear local-ization and activation of NF-�B DNA binding activity.
Physiological Rescue of IKK�(�/�) MEFs with Wt. IKK� IsSufficient to Restore the Global Expression of NF-�B-dependentGenes—To identify the cohort of TNF�-responsive genes inIKK�(�/�) MEFs, which are not expressed because of theirIKK� deficiency, we compared DNA microarray screens ofIKK�(�/�) MEFs rescued with Wt. IKK� with IKK�(�/�) cellsexpressing an empty retroviral vector (EV) as a matched neg-
ative control. IKK�(�/�) � Wt. IKK� screens were performedin duplicate to rule out any gene-specific variations and werecompared with TNF�-stimulated IKK�(�/�)-EV cells. 118genes were rescued based on the stringent criteria of doubleincrease calls (Affymetrix MAS 5.1) and average -fold changevalues of 1.5 or greater. Hierarchical clustering of the signalvalues of these 118 genes shows that two independent TNF�
stimulations of the Wt. IKK�-rescued IKK�(�/�) cell popula-tion have very similar expression profiles (see first and secondcolumns of Fig. 4). Some variations in the degrees of expressionof specific genes were observed, but more importantly all ofthese genes are expressed at significantly higher levels in thetwo Wt. IKK�-rescued samples compared with TNF-stimulatedIKK�(�/�) cells harboring the empty retroviral vector or pa-rental IKK�(�/�) cells.
To determine whether these IKK�-dependent genes werealso dependent on NF-�B, we employed hierarchical clusteringanalysis to cross-compare these screens with others performedwith two varieties of NF-�B-compromised MEFs: 1) Wt. MEFsstably expressing an I�B�(S32A,S36A) super-repressor (26),and 2) NF-�B p50(�/�) MEFs (43). Both cell lines were stim-ulated with TNF� for 2 h and compared with their wild typecounterparts. Some variation in the degrees of NF-�B depend-ence of subsets of genes in the p50(�/�) and Wt. � I�B�SRscreens were observed (see Fig. 4 Treeview image). This lattereffect was most likely the result of a combination of the pen-etrance of the I�B� super-represssor as well as the differentthresholds of specific NF-�B subunits required for the expres-sion of their specific downstream target genes. Known genesand expressed sequence tags with significant decreases in ex-pression in either the I�B�SR-expressing MEFs or in p50(�/�)MEFs compared with Wt. MEFs were judged to be dependenton NF-�B for their activity.
The -fold change values of a representative set of 40 genesrescued by restoring Wt. IKK� expression in IKK�-null MEFsare shown in Table I. The relative NF-�B dependences of thesegenes are presented as positive -fold change values in one orboth of the two microarray screens (Wt. MEF versus Wt. MEF� I�B�SR or Wt. MEF versus p50(�/�)). Importantly, many ofthese genes were also rescued to levels of expression observedin Wt. MEFs. Comparisons of the relative mRNA expressionlevels of genes in duplicate TNF�-stimulated and unstimulatedsamples revealed that the genes rescued by Wt. IKK� proteinin IKK�-null cells fell into two distinct stimulus-dependent and-independent classes (Table I, Fig. 5, and data not shown). Thesignal values (equivalent to mRNA quantities) of three repre-sentative examples of the stimulus-dependent (Fas ligand,C/EBP-�, and CXCL10) and independent (Clast1, NOV, andC1r) classes of these Wt. IKK�-rescued genes are presented inbar graph format in Fig. 5. Collectively, these results revealthat the expression of a large number of NF-�B-dependentgenes was restored in IKK�-null MEFs by retroviral transduc-tion of a Wt. IKK� protein in a TNF�-responsive manner,including, IL-6, GADD45�, RANTES, ScyB5/LIX, A20, I�B�,IFITR-1, C/EBP-�, ATF3, Fas ligand, caspase-11, M/CSF-1,serum amyloid A3, MIP2�, VCAM, JunB, ScyD1, ISG15, Gro1,MMP3, MMP13, and Met 1 (see Table I and selected signalvalue comparisons in Fig. 5A). Among these 40 representativegenes rescued by the Wt. IKK� protein, examples of signal-independent rescues by Wt. IKK� include ClastI/LR8, C1r,IFP35, Plf2, Plf3, Itm2b, NOV/CNN3, Decorin, Snx10, andIFITR2 (see genes highlighted in gray in Table I and alsoselected signal value comparisons in Fig. 5B). In agreementwith these results, the expression of this same class of geneswithout extracellular stimulation was also found to be simi-larly reduced in Wt. MEFs harboring a constitutively expressed
FIG. 2. Immunostaining of Wt. IKK�-HA and IKK�(K44M)-HAproteins in retrovirally transduced populations of IKK�(�/�)cells. Wt. IKK�-HA and IKK�(K44M)-HA proteins, stably expressed bypopulations of retrovirally transduced IKK�(�/�) MEFs, were visual-ized by in situ immunostaining. Cells were plated on coverslips coatedwith poly-L-lysine, fixed, and immunostained using 12CA5 anti-HAmonoclonal antibody or no primary antibody as a negative control.Stained cells were viewed by a phase-contrast Nikon Diaphot micro-scope and photographed using a Nikon D1X digital camera, as de-scribed under “Experimental Procedures.” In the absence of primaryantibody (top panels) or in IKK�(�/�) parental cells, no immuno-staining is seen, whereas populations of IKK�(�/�) cells stably trans-duced by either Wt. IKK�-HA- or IKK�(K44M)-HA-expressing retro-viruses (bottom panels) show uniform cytoplasmic staining.
FIG. 3. Physiological expression of Wt. IKK� or IKK�(K44M) inIKK�(�/�) MEFs does not interfere with stimulus-dependentNF-�B DNA binding. RelA/p65 DNA binding was assayed using theTransAM NF-�B p65 Transcription factor assay kit (Active Motif),following the manufacturer’s instructions for the preparation of nuclearextracts. All samples are presented as (�) unstimulated or (�) stimu-lated for 2 h with 20 ng/ml TNF� prior to lysis and nuclear extractpreparation. Data shown represent each data point done in quadrupli-cate, with standard deviations presented as error bars. Nuclear p65 wasmeasured as the absorbance at 450 nm, with a reference wavelength of650 nm using a fluorescent plate reader. Specificity of p65-DNA bindingwithin the Wt. IKK�- and IKK�(K44M)-infected populations was de-termined by competition for binding using an excess of either Wt. (W) ormutant (M) NF-�B synthetic oligonucleotide.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14060
I�B� super-repressor or in p50-null MEFs compared with theirwild type counterparts (data not shown). Thus, these observa-tions show that the Wt. IKK�-containing signalsome is re-quired for both the stimulus-dependent and basal levels ofexpression of NF-�B-dependent genes. Because this class ofNF-�B-dependent genes required IKK� without TNF� stimu-lation, they define a novel class of IKK�-dependent genes thatare downstream of basally activated NF-�B.
In addition, the vast majority of the 118 genes rescued byphysiological restoration of IKK� in IKK�-null MEFs wereco-dependent on NF-�B, based on their reduced or severelycompromised expression in either Wt. MEFs expressing anI�B� super-repressor or NF-�B p50(�/�) MEFs (Fig. 4 andTable I). Thus, our global expression results also show thatIKK� is not likely required for the transcription of a largenumber of genes outside of the NF-�B pathway. It also directlyfollows that despite the reported ability of IKK� to facilitatepotentially more general aspects of chromatin activation byeither phosphorylating serine 10 of histone H3 or the SMRTtranscriptional corepressor (27–29), IKK� must somehow stillremain preferentially targeted to NF-�B-dependent genes.
A Kinase-inactive IKK� Mutant Rescues a Portion of theGenes Dependent on Wt. IKK� in IKK�(�/�) MEFs—Becausethe IKK� mechanism of action to activate NF-�B-dependenttranscription in the canonical pathway remains poorly under-stood, we next investigated whether the kinase activity of IKK�was essential for the expression of the 118 genes rescued by the
Wt. IKK� protein in IKK�(�/�) MEFs. Lysine 44 in the IKK�kinase domain is essential for its binding of ATP, and itsmutation to methionine prevents ATP binding, thereby com-pletely destroying IKK� kinase activity (32, 44). To determinethe contribution of IKK� kinase function for the rescue ofNF-�B-dependent genes, we used duplicate microarray analy-sis to determine the ability of a kinase-dead IKK�(K44M) mu-tant to rescue IKK�/NF-�B-dependent targets when expressedat near physiological levels in the IKK�(�/�) cells (see hierar-chical Treeview comparisons of Wt. IKK� and IKK�(K44M)expressing IKK�(�/�) cells in the first and second columns andthird and fourth columns, respectively, in Fig. 6). These resultsdemonstrate that �72% of the genes rescued by Wt. IKK� inIKK�(�/�) cells were unaffected by the presence of comparablelevels of an IKK�(K44M) mutant protein. Surprisingly, �28%(33 of the 118 IKK�-dependent genes) were reproducibly res-cued by the IKK�(K44M) mutant. Most of these 33IKK�(K44M)-rescued genes can be visualized as two coclus-tered groups in the upper and lower portions of the hierarchicalTreeview image presented in Fig. 6.
The -fold change values of 20 representative genes rescuedby the IKK�(K44M) mutant protein are presented in Table I.Comparable degrees of rescue of each of these genes wereachieved with Wt. IKK� and the kinase-inactive IKK� mutantproteins. However, in contrast to the expression of genes re-stored by Wt. IKK�, pairs of TNF�-stimulated and -unstimu-lated microarray screens of IKK�(K44M)-transduced IKK�-
FIG. 4. Hierarchical cluster imageof the NF-�B dependences of genesrescued by Wt. IKK� in IKK�(�/�)MEFs. Signal values of 118 IKK�-depend-ent genes derived from two independentmicroarray screens of IKK�(�/�) cells ex-pressing Wt. IKK� (first two columns) weresubjected to hierarchical clustering com-pared with the following samples: third col-umn, IKK�(�/�) MEFs � EV 2T, fourthcolumn, IKK�(�/�) MEFs 2T; fifth col-umn, p50(�/�) 2T; and sixth column, Wt.MEFs�I�B�(S32A,S36A)2T.The118IKK�-dependent genes employed in these com-parisons were selected on the basis of av-erage -fold change values of 1.5 or greater(minimum of 1.3-fold each) and differencecalls of increase or marginal increase intwo independent samples of 2-h TNF�(2T)-stimulated IKK�(�/�) MEF �Wt.IKK� 2T versus IKK�(�/�) MEF �EV 2T MEFs. The locations of a number ofgenes are indicated. Signal values wereall derived from MAS 5.1 calculation andnormalized as described under “Experi-mental Procedures.” Gene expression val-ues are shown in color according to theindicated expression scale bar.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14061
TABLE IA Wt. kinase-competent IKK� protein rescues the expression of two classes of NF-�B-dependent genes in IKK�-null cells
40 representative IKK�-dependent genes, which met the stringent selection criteria outlined-under “Experimental Procedures,” are shown. Thefirst data column shows the -fold change values of each gene obtained in two independent microarray screens of IKK�(�/�) � Wt. IKK� 2T versusIKK�(�/�) � EV 2T. The second column displays their relative dependences on IKK� in the context of wt. MEFs (i.e. Wt. MEF 2T versusIKK�(�/�) 2T as described previously) (26) (and data not shown). The third and fourth columns show the NF-�B dependences of each gene bycomparing their induced expressions in Wt. MEF 2T compared with either p50-null MEF 2T (third column) or Wt. MEF � I�B�SR(super-repressor)2T (26) (fourth column). The criteria for assigning the TNF� responsiveness of each rescued gene was determined on the basis of duplicate S (2T)versus US microarray screens of Wt. IKK�-rescued IKK�(�/�) MEFs (see description of criteria under “Experimental Procedures”). Examples ofsignal values of three genes are shown in Fig. 5, and TaqMan real time PCR analyses of selected genes are shown in Fig. 9. Genes induced by TNF�are assigned a (�) sign, and genes whose expressions were not significantly stimulated by TNF� were given a (�) sign. The expressions of 9 of thisrepresentative group of 40 genes were rescued independently of TNF� stimulation and are highlighted in gray.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14062
null cells revealed that only a small fraction of the genesrescued by the IKK�(K44M) mutant responded to TNF� stim-ulation (see examples of A20, M/CSF-1 and VL30 highlightedin gray in Table II). Bar graphs of signal values (comparablewith amounts of mRNAs) of three representative examples ofthe stimulus-dependent (A20, mVL30, and B94) and independ-ent (Coagulation factor III, Sgk, and NDPP1) classes ofIKK�(K44M)-rescued genes are shown in Fig. 7. In addition tothese two distinct classes of genes, a third class ofIKK�(K44M)-rescued genes only responded to TNF� stimula-tion after rescue by Wt. IKK� and not if rescued by theIKK�(K44M) kinase-inactive mutant (see examples GADD45�,ATF3, and JunB in Table II). Interestingly, compared with theNF-�B-dependent genes solely rescued by the Wt. IKK� pro-tein, a larger fraction of the IKK�/NF-�B-dependent genesrescued by the kinase-inactive IKK�(K44M) mutant preferen-tially encoded proteins associated with NF-�B autoregulation,growth arrest, apoptosis, proliferation, and survival (see rep-resentative genes in Table II and under “Discussion”). In addi-tion, most of these NF-�B-dependent genes, which are rescuedby wild type and kinase-inactive IKK�, depend on IKK� fortheir basal levels of expression in the absence of an extracel-lular stimulus. Thus, our global expression profiling analysisreveals the surprising result that different NF-�B-dependentgenes differentially require IKK� kinase activity for their basaland TNF�-dependent expression, with the majority of genesencoding proteins associated with pro-inflammatory stress-likeresponses requiring IKK� kinase activity.
TaqMan Real Time PCR Validations of Wt. IKK� andIKK�(K44M)-rescued Genes—As an additional validation of theduplicate microarray screens, TaqMan real time PCR experi-ments were performed on eight NF-�B target genes (IL-6,ISG15, RANTES, SAA3, VCAM1, GADD45B, A20, and ATF3).Importantly, TaqMan validations were performed at least induplicate on a third independent set of unstimulated and 2 h
(2T) TNF�-stimulated samples, which were not employed inthe duplicate microarray screens. Fig. 8 shows the absoluteexpression levels in the context of TNF� stimulation of each ofthese eight genes as mRNA copy numbers in IKK�(�/�) MEFsexpressing Wt. IKK�, IKK�(K44M), or an EV compared withWt. MEFs. Each of these genes is expressed in Wt. MEFs andWt. IKK�-rescued IKK�(�/�) MEFs above their low to negli-gible levels in IKK�(�/�)�EV cells. Some variations in thedegrees of the IKK�-dependent rescues compared with wildtype control cells are noted with some genes being expressed athigher levels in the wild type control and others expressed athigher levels in the Wt. IKK�-rescued cells. Fig. 9 illustratesthe TNF� dependences of the same eight genes. In agreementwith the duplicate microarray screens, A20, ATF3, andMyD118/GADD45�, were rescued by the IKK�(K44M) mutantcompared with IKK�(�/�)�EV MEFs, whereas IL-6, ISG15,RANTES, SAA3, and VCAM1 failed to exhibit expression abovebackground in IKK�(K44M)-expressing IKK�(�/�) cells. Ofthese three IKK�(K44M)-rescued genes, only A20 respondedsignificantly to TNF� stimulation. In agreement with our du-plicate microarray screens, after their rescue by Wt. IKK� eachof these eight genes was confirmed to be TNF�-responsive.
DISCUSSION
Studies of IKK�(�/�) and IKK�(�/�) MEFs have defini-tively shown that IKK� is the in vivo I�B� kinase and thatIKK� is not needed for I�B degradation, NF-�B nuclear local-ization, nor for inducing NF-�B DNA binding activity in re-sponse to proinflammatory NF-�B stimuli such as TNF�. How-ever, IKK� functions in the canonical NF-�B pathway toensure or modulate the transcriptional competence of DNA-bound NF-�B. In support of this view, we have shown hereinthat restoration of IKK�(�/�) MEFs with near physiologicallevels of a Wt. IKK� kinase globally and specifically activatedNF-�B-dependent genes in response to TNF� stimulation. In
FIG. 5. Relative mRNA expression levels of examples of stimulus-dependent and -independent classes of genes only rescued by aWt. IKK� protein. A, comparisons of signal values of three NF-�B-dependent genes, which were rescued only by a kinase-competent Wt. IKK�protein, in a TNF�-responsive manner. B, signal value comparisons of three additional genes, whose basal levels of NF-�B-dependent expressionwere rescued only by Wt. IKK� and were unaffected by TNF� stimulation.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14063
addition, these experiments also revealed a hitherto unknownrequirement for IKK� to maintain the basal levels of expres-sion of specific NF-�B-dependent genes in the absence of anextracellular stress-like stimulus. Furthermore, the ability of akinase-inactive IKK� mutant to rescue a portion of theseNF-�B target genes reveals that even though the IKK� proteinis globally required for the expression of NF-�B-dependentgenes, its role as a functional kinase is also target gene-specific.In summary, our findings show that genes dependent on IKK�and NF-�B can be formally divided into five distinct classes ofresponsive genes: 1) genes that require a functional Wt. IKK�kinase for their stimulus-dependent, NF-�B-dependent expres-sion; 2) genes that require a functional Wt. IKK� kinase fortheir stimulus-independent basal NF-�B-dependent expres-sion; 3) genes that require a functional Wt. IKK� kinase fortheir signal-dependent rescue but only require an IKK� proteinfor their basal, stimulus-independent expression; 4) genes thatrequire the IKK� protein regardless of its kinase activity fortheir stimulus-dependent, NF-�B-dependent expression; and 5)genes that require an IKK� protein regardless of its kinaseactivity for their stimulus-independent, basal NF-�B-depend-ent expression.
Modifications of NF-�B subunits and other post-transla-tional nuclear processes are also necessary for the induction ofNF-�B target genes, and a number of reports have implicatedIKK� in these events. Phosphorylation of serines 276 (in theRel Homology domain/RHD) and serines 529 and 536 in thetranscriptional activation domain (TAD) of the NF-�B p65/
RelA subunit have been suggested to play activating roles, anda number of kinases have been directly implicated in this stepincluding the IKKs (45–52). Transactivation by p65/RelA inresponse to TNF� was localized to serine 529 within the p65/RelA TAD and found to mediate its transcriptional activationindependent of the ability of NF-�B to bind DNA (89). Phos-phorylation of serine 276 was also found to be involved in theactivation of p65/RelA at least in part by controlling its asso-ciation with the p300/CBP coactivator or the histone deacety-lase-1 (53). IKK� and IKK� were both implicated as down-stream effectors of Akt-dependent signaling targeted to serines529 and 536 in the p65/RelA TAD, which in part appeared toinvolve the engagement of the CBP/p300 transcriptional coac-tivator (90, 91). PTEN, a negative upstream effector of Akt, wasalso reported to inhibit TNF�-induced NF-�B activation (92,93), which was subsequently shown to occur solely at the levelof p65 transactivation (94). Additionally, IL-1- and Akt-medi-ated NF-�B activation was found to involve p65 TAD phospho-rylations with a codependence on IKK� and IKK� (25). TNF�-induced phosphorylation of p65/RelA serine 536 was recentlyshown to be dependent on both IKK� and IKK� and mediatedby TRAF2, TRAF5, and TAK1 signaling (54), and a require-ment for IKK� in p65 serine 529 phosphorylation and NF-�B-dependent transcriptional activation in response to LT� stim-ulation has also recently been reported (55). In addition, theactivation of NF-�B by the HTLV-Tax1 protein involves thespecific phosphorylation of p65 serines 529 and 536, requiringIKK�, but not IKK� (56). The IKK� mechanism of action in the
FIG. 6. Hierarchical cluster imageof genes rescued by Wt. IKK� com-pared with an IKK�(K44M) mutant.Signal values of the 118 genes rescued byWt. IKK� in duplicate screens (first twocolumns) were evaluated by hierarchicalclustering (as described in Fig. 3) com-pared with their signal values in dupli-cate screens of 2-h TNF� (2T)-stimulatedIKK�(�/�) MEFs expressing a kinase-in-active IKK�(K44M) mutant (third andfourth columns). As in Fig. 4, the IKK�specificities of the rescues can be visual-ized in the fifth and sixth columns, whichdisplay the signal values of the 118 genesin IKK�(�/�)-null MEFs or the samecells expressing EV.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14064
canonical NF-�B pathway has also been proposed to be purelynuclear in nature. In this context, IKK� has been shown tomigrate into the nucleus (57) and associate with the promotersof NF-�B-dependent genes upon TNF� stimulation (27, 28).More recently mitogenic activation of the c-fos gene by epider-mal growth factor-dependent signaling was found to requirethe constitutive and induced recruitment of p65/RelA andIKK�, respectively, to the c-fos promoter (58). Because IKK�was also found to phosphorylate serine 10 of histone H3 invitro, and the phosphorylation status of histone H3 inIKK�(�/�) MEFs was enhanced by introduction of Wt. IKK�(27, 28), these results suggested that IKK� might be function-ing by enhancing the transcriptional accessibility of the chro-matin of NF-�B target genes and potentially a broader range ofgenes that also respond to TNF� stimulation. More recently,IKK� was also shown to be critical for the signal-dependent
expression of the cIAP-2 and IL-8 NF-�B-dependent genes inseveral cellular contexts by virtue of its required ability tophosphorylate the SMRT corepressor, thereby also inducingchromatin activation by facilitating the exchange of transcrip-tional corepressors for coactivators (29).
Wt. IKK� Kinase Restores the Expression of NF-�B-depend-ent Genes in IKK�-null MEFs—Our comparative genomic anal-ysis of the abilities of Wt. IKK� and a kinase-deadIKK�(K44M) ATP binding domain mutant indicate that IKK�kinase activity is required for the activation of most but not allgenes whose expressions are dependent on NF-�B. Indeed, wefind that IKK� kinase activity is required for the induction ofnumerous TNF�-responsive NF-�B target genes (includingRANTES, I�B�, IL-6, ISG-15, IFITR-1, mGBP-2, mGBP-3,Gro1, Fas ligand, Jun-B, caspase-11, ScyD1, serum amyloid A3,MMP3, MMP13, ScyB5/LIX, Ptx3, Schaflen 2, Met1, and
TABLE IINF-�B-dependent genes rescued by IKK� independently of its kinase activity
20 representative genes meeting the criteria of being rescued by both the Wt. IKK� and IKK�(K44M) proteins are shown. -Fold change valuesof these 20 genes obtained from duplicate microarray comparisons using independent samples of IKK�(�/�) � IKK�(K44M) 2T versus IKK�(�/�)� EV 2T are indicated in the first column. The second column shows the results obtained for the same genes in duplicate microarray comparisonsof IKK�(�/�) � Wt. IKK� 2T versus IKK�(�/�) � EV 2T. The third column displays their relative dependences on IKK� in the context of Wt.MEFs (i.e. Wt. MEF 2T versus IKK�(�/�) 2T as described previously (26 and data not shown). The fourth and fifth columns show their NF-�Bdependences by comparing their expressions in Wt. MEF 2T with either p50-null MEF 2T (fourth column) or Wt. MEF � I�B�SR(super-repressor)2T (26) (fifth column). The TNF� dependences for the IKK�-rescued expression of each gene are shown in the context of their independent rescuesby physiological levels of either IKK�(K44M) (sixth column 6) or Wt. IKK� (seventh column). The criteria for assigning the TNF� responsivenessof each rescued gene was determined on the basis of duplicate S (2T) versus US microarray screens (for Wt. IKK� or IKK�(K44M)-rescuedIKK�(�/�) MEFs as indicated) (see description of criteria under “Experimental Procedures”). Examples of signal values of three genes are shownin Fig. 7, and TaqMan real time PCR analyses of selected genes are in Fig. 9. Genes induced by TNF� are assigned a (�) sign, and genes whoseexpression was not significantly stimulated by TNF� were given a (�). The IKK�(K44M) mutant in a TNF�-responsive manner comparable to thatachieved by Wt. IKK� rescued three genes highlighted in gray.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14065
C/EBP�), many of which we identified previously as targets ofproinflammatory cytokine-mediated NF-�B activation inWt. MEFs but not in IKK�-null MEFs (26).
The expression of the majority of IKK�-rescued NF-�B targetgenes in IKK�-null MEFs was rescued by Wt. IKK� to within2-fold of their levels in Wt. MEFs (see -fold change values ofrepresentative set of 40 genes in Table I) with a portion ofNF-�B targets being expressed at higher levels in Wt. MEFs.This latter observation is not that surprising given that longterm inactive genes are known to differentially acquire the
attributes of transcriptionally inaccessible chromatin states,which is unlikely to be fully overcome by IKK� reexpression. Inaddition, some NF-�B-dependent genes were also expressed athigher levels in IKK�-rescued IKK�(�/�) cells compared withtheir wild type counterpart (see -fold change values forIFITR-1, Gro1, and caspase-11 in Wt. IKK�-rescued IKK�-nullcells compared with their levels of expression in Wt. MEFs inTable I), which could reflect intrinsic differences in the abso-lute levels of NF-�B target gene expression in different cellularbackgrounds. A fraction of the TNF�-responsive genes rescued
FIG. 7. Relative mRNA expression levels of examples of stimulus-dependent and -independent classes of genes rescued by theIKK� protein regardless of its kinase activity. A, comparisons of signal values of three NF-�B-dependent genes, which were comparablyrescued by both Wt. IKK� and the IKK�(K44M) mutant, in a TNF�-responsive manner. B, signal value comparisons of three additional NF-�Btarget genes, whose basal levels of NF-�B-dependent expression were rescued by either a wild type or a kinase-inactive IKK� mutant but wereunresponsive to TNF� stimulation.
FIG. 8. TaqMan real time PCR analysis of selected genes differentially rescued by Wt. IKK� and IKK�(K44M). Independent samplesof Wt. MEFs and IKK�(�/�) MEFs expressing Wt. IKK�, EV, or IKK�(K44M) were stimulated with 20 ng/ml TNF� for 2 h., and RNAs wereprepared and subjected to TaqMan real time PCR analysis. The levels of expression of eight representative genes (IL-6, ISG15, SAA3, RANTES,VCAM1, A20, ATF3, and GADD45�) were quantitatively compared. Each bar represents data obtained at least in duplicate with the indicatedstandard deviations. All samples were normalized to a glyceraldehyde-3-phosphate dehydrogenase probe set, and mRNA copy numbers weredetermined compared with a genomic DNA standard for each probe set.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14066
by Wt. IKK� in IKK�-null fibroblasts did not appear to bedependent on NF-�B for their expression employing the criteriaof being unaffected by the inhibitory effects of the I�B� super-repressor in Wt. MEFs nor by the loss of the NF-�B p50 sub-unit. However, because genes such as MMP13, an NF-�B tar-get gene in other cellular contexts (59–61), were among thislatter gene subset (see Figs. 4 and 6 and data not shown), thissupports our conclusion that the large majority of TNF�-re-sponsive, IKK�-dependent genes in these cells are alsoNF-�B-dependent.
Surprisingly, a number of NF-�B-dependent genes were de-pendent on IKK� for their basal, stimulus-independent expres-sion in IKK�-compromised MEFs (21 of the 53 representativegenes rescued by IKK� in Tables I and II). Furthermore, thesesame genes were also expressed at significantly lower levels inNF-�B-compromised MEFs compared with Wt. MEFs (data notshown). Thus our findings show that the IKK�-containing sig-nalsome is also required to maintain the basal, intrinsic expres-sion levels of specific NF-�B-dependent genes. Interestingly, 14of 21 of these stimulus-independent/Wt. IKK�-rescued geneswere also preferentially rescued by the IKK�(K44M) mutant inthe absence of stimulation (see Table II), also demonstratingthat IKK� kinase activity is not required to restore their basallevels of NF-�B-dependent gene expression in most but not allcases.
IKK� Does Not Always Function as a Kinase to Ensure theExpression of NF-�B-dependent Genes—We also employed du-plicate microarray screens to compare the abilities of Wt. IKK�or a kinase-dead IKK�(K44M) mutant protein to rescue NF-�B-dependent gene expression in IKK�-null fibroblasts on agenomic scale. These findings, in combination with quantita-tive TaqMan real time PCR assays on selected genes, revealthat even though IKK� kinase activity is required to activatethe majority of NF-�B-dependent genes, it is not functioningdirectly as a kinase for the expression of �28% of the NF-�B-dependent genes in this cell background. Interestingly, most ofthe latter subset of IKK�(K44M)-rescued genes were unrespon-sive to TNF�, with the exception of several genes (includingM/CSF-1, A20, mVL30, and B94) (see Table II, Figs. 7 and 9).In contrast to their TNF�-independent rescues by theIKK�(K44M) mutant, GADD45�/MyD118, ATF3, and JunBwere rescued by Wt. IKK� in a TNF�-dependent manner (Ta-ble II and Fig. 9). It is also noteworthy in this context that our
quantitative TaqMan analysis revealed that Wt. IKK� com-pared with IKK�(K44M) also rescued higher levels ofGADD45�, ATF3, and A20 expression (Figs. 8 and 9). Takentogether these observations show that the nature of the Wt.IKK�- versus IKK�(K44M)-mediated rescues of GADD45�,ATF3, JunB, and perhaps even A20 are intrinsically differentfrom each other, revealing that their dependences on the IKK�protein occurs at more than one regulatory level.
Of particular interest, a larger proportion of the IKK�-res-cued genes, which were also rescued by the kinase-inactiveIKK�(K44M) mutant, encode proteins with functional proper-ties in NF-�B autoregulation, cellular proliferation, growtharrest, apoptosis, or cellular survival (M/CSF-1, PLF2, PLF3,GADD45�/MyD118, A20, Sequestosome/p62, NDPP1/CARD 8,JunB, and ATF3). GADD45�/MyD118 was first described tofunction as an effector of myeloid cell differentiation and as amember of a class of cell cycle checkpoint protein arrestingcellular growth in response to DNA damage or in associationwith terminal cellular differentiation (62, 63). In contrast andmore recently, GADD45� has also been shown to be an NF-�B-dependent antiapoptotic effector in response to TNF�, where itappears to act by directly blocking MKK7/JNKK2 activitythereby leading to the suppression of the c-Jun NH2-terminalkinase (JNK) cascade in response to TNF� stimulation (64,65)and also by suppressing Fas/CD95/APO-1-induced caspaseactivation in response to CD40 triggering in B lymphocytes(66). Our results show that the induction of GADD45� by TNF�requires IKK� kinase activity for its NF-�B-dependent expres-sion. However, because we also find that a kinase-inactiveIKK�(K44M) mutant restored unstimulated levels ofGADD45� expression in IKK�-null MEFs, our results alsoindicate that other intrinsic properties of the IKK� proteincontribute to the expression of GADD45� under different phys-iological circumstances, perhaps leading to other functionalproperties ascribed to GADD45�/MyD118. Sequestosome/p62,an atypical protein kinase C-interacting protein, interacts withRIP (receptor-interacting protein), linking it to TNF�-mediatedNF-�B induction with its inhibition or down-modulation alsointerfering with IL-1 and TRAF 6-dependent NF-�B activation(67). A20 is a TNF�- and IL-1-induced zinc finger protein,which has been reported to act as a negative effector of NF-�B(68–70), mediated by RIP and TRAF2 signaling (70). A20 wasalso recently shown to attenuate TNF�-mediated NF-�B acti-
FIG. 9. TaqMan real time PCR analysis of the TNF� dependences of selected genes rescued by Wt. IKK� or IKK�(K44M). Therelative levels of expression in TNF�-stimulated and unstimulated cells of the eight selected genes in Fig. 8 (IL-6, ISG15, SAA3, RANTES, VCAM1,A20, ATF3, and GADD45�) are shown. TaqMan PCRs were performed and quantitated as described under “Experimental Procedures” and in thelegend of Fig. 8.
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14067
vation via the cooperative action of its two intrinsic ubiquitin-editing domains resulting in the polyubiquitination of RIP, anessential mediator of the TNF receptor 1 signaling complex,thereby targeting RIP proteasomal destruction (71). Akin toGADD45�, A20 has also been shown to block TNF�-dependentapoptosis at least in part by preventing TRADD and RIP re-cruitment to TNFRI (72, 73). NDPP1 a novel member of thecaspase-associated recruitment domain (CARD) family of pro-teins, has also been reported either to promote or suppressapoptotic responses in specific cell types and to block TNF�-induced NF-�B activation (74, 75). ATF3, a member of theATF/CREB family of leucine zipper transcription factors (alsoknown as LRF-1/TI-241) (76, 77) and a stress-induced tran-scriptional repressor (78–80), has recently been shown to bedependent on both the NF-�B and JNK signaling pathways forits expression in response to TNF� and nitric oxide (81). ATF3has also been shown to play roles in either protection against orinduction of apoptotic responses, dependent on the nature ofthe signal and cellular context (79, 82, 83). Finally, Proliferin 2and Proliferin 3/mitogen-regulated protein 3, which were alsorescued by both Wt. IKK� and IKK�(K44M) in a stimulus-independent manner, are members of the prolactin growthhormone family of proliferins whose functional properties havebeen associated with cellular proliferation and migration,wound healing, and angiogenesis (84–88). Because anIKK�(K44M) kinase-inactive mutant was found to be incapableof phosphorylating serine 10 of histone H3 (28), our observa-tions also indicate that not all genes whose expressions arecodependent on IKK� and NF-�B require IKK� kinase activityfor histone H3 phosphorylation.
In summary, our global expression profiling analysis showsthat IKK� functions in both kinase-dependent and -independ-ent modes, thereby revealing the existence of five distinctclasses of genes codependent on IKK� and NF-�B in mouseembryonic fibroblasts. In addition to the importance of aWt. IKK�-containing signalsome for the activities of manystimulus-dependent target genes of NF-�B, the Wt. IKK� pro-tein also plays a role in the regulation of a distinct class ofNF-�B-dependent genes requiring only basal levels of activeNF-�B for their regulated expression. We envision that theIKK� kinase-independent mode of action to ensure the expres-sion of a subset of NF-�B-dependent genes may be attributableto a novel regulatory/docking-like property of the IKK� protein,which facilitates the recruitment of other regulatory factorsrequired for the expression of specific downstream targets ofthe NF-�B pathway.
Acknowledgments—We gratefully acknowledge Dr. Michael Karinfor providing IKK�(�/�), IKK�(�/�), and IKK�/NEMO(�/�) MEFsand Dr. Amer Beg for p65/p50(�/�) and p50(�/�) MEFs.
REFERENCES
1. Karin, M., and Ben-Neriah, Y. (2000) Annu. Rev. Immunol. 18, 621–6632. Ghosh, S., and Karin, M. (2002) Cell 109, (suppl.) S81–S963. Karin, M., and Lin, A. (2002) Nat. Immunol. 3, 221–2274. Li, Q., and Verma, I. M. (2002) Nat. Rev. Immunol. 2, 725–7345. Kucharczak, J., Simmons, M. J., Fan, Y., and Gelinas, C. (2003) Oncogene 22,
8961–89826. Weih, F., and Caamano, J. (2003) Immunol. Rev. 195, 91–1057. Karin, M., Yamamoto, Y., and Wang, Q. M. (2004) Nat. Rev. Drug Discov. 3,
17–268. Bonizzi, G., and Karin, M. (2004) Trends Immunol. 25, 280–2889. Patke, A., Mecklenbrauker, I., and Tarakhovsky, A. (2004) Curr. Opin. Immu-
nol. 16, 251–25510. Weil, R., and Israel, A. (2004) Curr. Opin. Immunol. 16, 374–38111. Hayden, M. S., and Ghosh, S. (2004) Genes Dev. 18, 2195–222412. Baldwin, A., Jr. (1996) Annu. Rev. Immunol. 14, 649–68313. Ghosh, S., May, M. J., and Kopp, E. B. (1998) Annu. Rev. Immunol. 16,
225–26014. Pahl, H. L. (1999) Oncogene 18, 6853–686615. Yamamoto, Y., and Gaynor, R. B. (2004) Trends Biochem. Sci. 29, 72–7916. Pomerantz, J. L., and Baltimore, D. (2002) Mol. Cell 10, 693–69417. Kato, T., Delhase, M., Hoffmann, A., and Karin, M. (2003) Mol. Cell 12,
829–839
18. Tergaonkar, V., Bottero, V., Ikawa, M., Li, Q., and Verma, I. M. (2003) Mol.Cell. Biol. 23, 8070–8083
19. Cao, Y., Bonizzi, G., Seagroves, T. N., Greten, F. R., Johnson, R., Schmidt,E. V., and Karin, M. (2001) Cell 107, 763–775
20. Hu, Y., Baud, V., Oga, T., Kim, K. I., Yoshida, K., and Karin, M. (2001) Nature410, 710–714
21. Sil, A. K., Maeda, S., Sano, Y., Roop, D. R., and Karin, M. (2004) Nature 428,660–664
22. Ohazama, A., Hu, Y., Schmidt-Ullrich, R., Cao, Y., Scheidereit, C., Karin, M.,and Sharpe, P. T. (2004) Dev. Cell 6, 219–227
23. Dejardin, E., Droin, N. M., Delhase, M., Haas, E., Ca, Y., Makris, C., Karin, M.,Ware, C. F., and Green, D. R. (2002) Immunity 17, 525–535
24. Amir, R. E., Haecker, H., Karin, M., and Ciechanover, A. (2004) Oncogene 23,2540–2547
25. Sizemore, N., Lerner, N., Dombrowski, N., Sakurai, H., and Stark, G. R. (2002)J. Biol. Chem. 277, 3863–3869
26. Li, X., Massa, P. E., Hanidu, A., Peet, G. W., Aro, P., Savitt, A., Mische, S., Li,J., and Marcu, K. B. (2002) J. Biol. Chem. 277, 45129–45140
27. Anest, V., Hanson, J. L., Cogswell, P. C., Steinbrecher, K. A., Strahl, B. D., andBaldwin, A. S. (2003) Nature 423, 659–663
28. Yamamoto, Y., Verma, U. N., Prajapati, S., Kwak, Y. T., and Gaynor, R. B.(2003) Nature 423, 655–659
29. Hoberg, J. E., Yeung, F., and Mayo, M. W. (2004) Mol. Cell 16, 245–25530. Connelly, M. A., and Marcu, K. B. (1995) Cell. Mol. Biol. Res. 41, 537–54931. Geleziunas, R., Ferrell, S., Lin, X., Mu, Y., Cunningham, E. J., Grant, M.,
Connelly, M. A., Hambor, J. E., Marcu, K. B., and Greene, W. C. (1998) Mol.Cell. Biol. 18, 5157–5165
32. McKenzie, F. R., Connelly, M. A., Balzarano, D., Muller, J. R., Geleziunas, R.,and Marcu, K. B. (2000) Mol. Cell. Biol. 20, 2635–2649
33. Morgenstern, J. P., and Land, H. (1990) Nucleic Acids Res. 18, 3587–359634. Mahadevappa, M., and Warrington, J. A. (1999) Nat. Biotechnol. 17,
1134–113635. Lockhart, D. J., Dong, H., Byrne, M. C., Follettie, M. T., Gallo, M. V., Chee,
M. S., Mittmann, M., Wang, C., Kobayashi, M., Horton, H., and Brown, E. L.(1996) Nat. Biotechnol. 14, 1675–1680
36. Eisen, M. B., Spellman, P. T., Brown, P. O., and Botstein, D. (1998) Proc. Natl.Acad. Sci. U. S. A. 95, 14863–14868
37. Livak, K. J., Flood, S. J., Marmaro, J., Giusti, W., and Deetz, K. (1995) PCRMethods Appl. 4, 357–362
38. Field, J., Nikawa, J., Broek, D., MacDonald, B., Rodgers, L., Wilson, I. A.,Lerner, R. A., and Wigler, M. (1988) Mol. Cell. Biol. 8, 2159–2165
39. Renard, P., Ernest, I., Houbion, A., Art, M., Le Calvez, H., Raes, M., andRemacle, J. (2001) Nucleic Acids Res. 29, E21
40. Li, Q., Lu, Q., Hwang, J. Y., Buscher, D., Lee, K. F., Izpisua-Belmonte, J. C.,and Verma, I. M. (1999) Genes Dev. 13, 1322–1328
41. Hu, Y., Baud, V., Delhase, M., Zhang, P., Deerinck, T., Ellisman, M., Johnson,R., and Karin, M. (1999) Science 284, 316–320
42. Takeda, K., Takeuchi, O., Tsujimura, T., Itami, S., Adachi, O., Kawai, T.,Sanjo, H., Yoshikawa, K., Terada, N., and Akira, S. (1999) Science 284,313–316
43. Sha, W. C., Liou, H. C., Tuomanen, E. I., and Baltimore, D. (1995) Cell 80,321–330
44. Mercurio, F., Zhu, H., Murray, B. W., Shevchenko, A., Bennett, B. L., Li, J.,Young, D. B., Barbosa, M., Mann, M., Manning, A., and Rao, A. (1997)Science 278, 860–866
45. Gerristen, M. E., Williams, A. J., Neish, A. S., Moore, S., Shi, Y., and Collins,T. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 2927–2932
46. Perkins, N. D., Felzien, L. K., Betts, J. C., Leung, K., Beach, D. H., and Nabel,G. J. (1997) Science 275, 523–527
48. Zhong, H., Voll, R. E., and Ghosh, S. (1998) Mol. Cell 1, 661–67149. Bergmann, M., Hart, L., Lindsay, M., Barnes, P. J., and Newton, R. (1998)
J. Biol. Chem. 273, 6607–661050. Sizemore, N., Leung, S., and Stark, G. R. (1999) Mol. Cell. Biol. 19, 4798–480551. Wang, D., Westerheide, S. D., Hanson, J. L., and Baldwin, A. S., Jr. (2000)
J. Biol. Chem. 275, 32592–3259752. Vermeulen, L., De Wilde, G., Damme, P. V., Vanden Berghe, W., and Haege-
man, G. (2003) EMBO J. 22, 1313–132453. Zhong, H., May, M. J., Jimi, E., and Ghosh, S. (2002) Mol. Cell 9, 625–63654. Sakurai, H., Suzuki, S., Kawasaki, N., Nakano, H., Okazaki, T., Chino, A., Doi,
T., and Saiki, I. (2003) J. Biol. Chem. 278, 36916–3692355. Jiang, X., Takahashi, N., Matsui, N., Tetsuka, T., and Okamoto, T. (2003)
J. Biol. Chem. 278, 919–92656. O’Mahony, A. M., Montano, M., Van Beneden, K., Chen, L. F., and Greene,
W. C. (2004) J. Biol. Chem. 279, 18137–1814557. Birbach, A., Gold, P., Binder, B. R., Hofer, E., de Martin, R., and Schmid, J. A.
(2002) J. Biol. Chem. 277, 10842–1085158. Anest, V., Cogswell, P. C., and Baldwin, A. S. (2004) J. Biol. Chem. 279,
31183–3118959. Mengshol, J. A., Vincenti, M. P., Coon, C. I., Barchowsky, A., and Brinckerhoff,
C. E. (2000) Arthritis Rheum. 43, 801–81160. Vincenti, M. P., and Brinckerhoff, C. E. (2002) Arthritis Res. 4, 157–16461. Liacini, A., Sylvester, J., Li, W. Q., Huang, W., Dehnade, F., Ahmad, M., and
Zafarullah, M. (2003) Exp. Cell Res. 288, 208–21762. Abdollahi, A., Lord, K. A., Hoffman-Liebermann, B., and Liebermann, D. A.
(1991) Oncogene 6, 165–16763. Liebermann, D. A., and Hoffman, B. (1998) Oncogene 17, 3319–332964. De Smaele, E., Zazzeroni, F., Papa, S., Nguyen, D. U., Jin, R., Jones, J., Cong,
R., and Franzoso, G. (2001) Nature 414, 308–31365. Papa, S., Zazzeroni, F., Bubici, C., Jayawardena, S., Alvarez, K., Matsuda, S.,
Nguyen, D. U., Pham, C. G., Nelsbach, A. H., Melis, T., Smaele, E. D., Tang,W. J., D’Adamio, L., and Franzoso, G. (2004) Nat. Cell Biol. 6, 146–153
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes14068
66. Zazzeroni, F., Papa, S., Algeciras-Schimnich, A., Alvarez, K., Melis, T., Bubici,C., Majewski, N., Hay, N., De Smaele, E., Peter, M. E., and Franzoso, G.(2003) Blood 102, 3270–3279
67. Sanz, L., Diaz-Meco, M. T., Nakano, H., and Moscat, J. (2000) EMBO J. 19,1576–1586
68. Ferran, C., Stroka, D. M., Badrichani, A. Z., Cooper, J. T., Wrighton, C. J.,Soares, M., Grey, S. T., and Bach, F. H. (1998) Blood 91, 2249–2258
69. Grey, S. T., Arvelo, M. B., Hasenkamp, W., Bach, F. H., and Ferran, C. (1999)J. Exp. Med. 190, 1135–1146
70. Heyninck, K., De Valck, D., Vanden Berghe, W., Van Criekinge, W., Contreras,R., Fiers, W., Haegeman, G., and Beyaert, R. (1999) J. Cell Biol. 145,1471–1482
71. Wertz, I. E., O’Rourke, K. M., Zhou, H., Eby, M., Aravind, L., Seshagiri, S., Wu,P., Wiesmann, C., Baker, R., Boone, D. L., Ma, A., Koonin, E. V., and Dixit,V. M. (2004) Nature 430, 694–699
72. He, K. L., and Ting, A. T. (2002) Mol. Cell. Biol. 22, 6034–604573. Malewicz, M., Zeller, N., Yilmaz, Z. B., and Weih, F. (2003) J. Biol. Chem. 278,
32825–3283374. Razmara, M., Srinivasula, S. M., Wang, L., Poyet, J. L., Geddes, B. J., DiSte-
fano, P. S., Bertin, J., and Alnemri, E. S. (2002) J. Biol. Chem. 277,13952–13958
75. Zhang, H., and Fu, W. (2002) Int. J. Oncol. 20, 1035–104076. Hsu, J. C., Laz, T., Mohn, K. L., and Taub, R. (1991) Proc. Natl. Acad. Sci.
U. S. A. 88, 3511–351577. Nawa, T., Nawa, M. T., Cai, Y., Zhang, C., Uchimura, I., Narumi, S., Numano,
F., and Kitajima, S. (2000) Biochem. Biophys. Res. Commun. 275, 406–41178. Hai, T., Wolfgang, C. D., Marsee, D. K., Allen, A. E., and Sivaprasad, U. (1999)
Gene Expr. 7, 321–33579. Mashima, T., Udagawa, S., and Tsuruo, T. (2001) J. Cell. Physiol. 188,
352–35880. Nawa, T., Nawa, M. T., Adachi, M. T., Uchimura, I., Shimokawa, R., Fujisawa,
K., Tanaka, A., Numano, F., and Kitajima, S. (2002) Atherosclerosis 161,281–291
81. Hartman, M. G., Lu, D., Kim, M. L., Kociba, G. J., Shukri, T., Buteau, J.,Wang, X., Frankel, W. L., Guttridge, D., Prentki, M., Grey, S. T., Ron, D.,and Hai, T. (2004) Mol. Cell. Biol. 24, 5721–5732
82. Nobori, K., Ito, H., Tamamori-Adachi, M., Adachi, S., Ono, Y., Kawauchi, J.,Kitajima, S., Marumo, F., and Isobe, M. (2002) J. Mol. Cell. Cardiol. 34,1387–1397
83. Nakagomi, S., Suzuki, Y., Namikawa, K., Kiryu-Seo, S., and Kiyama, H. (2003)J. Neurosci. 23, 5187–5196
84. Connor, A. M., Waterhouse, P., Khokha, R., and Denhardt, D. T. (1989)Biochim. Biophys. Acta 1009, 75–82
85. Groskopf, J. C., Syu, L. J., Saltiel, A. R., and Linzer, D. I. (1997) Endocrinology138, 2835–2840
86. Fassett, J. T., and Nilsen-Hamilton, M. (2001) Endocrinology 142, 2129–213787. Toft, D. J., Rosenberg, S. B., Bergers, G., Volpert, O., and Linzer, D. I. (2001)
Proc. Natl. Acad. Sci. U. S. A. 98, 13055–1305988. Corbacho, A. M., Martinez De La Escalera, G., and Clapp, C. (2002) J. Endo-
crinol. 173, 219–23889. Wang, D., and Baldwin, A. S., Jr. (1998) J. Biol. Chem. 273, 29411–2941690. Madrid, L. V., Wang, C. Y., Guttridge, D. C., Schottelius, A. J., Baldwin, A. S.,
Jr., and Mayo, M. W. (2000) Mol. Cell. Biol. 20, 1626–163891. Madrid, L. V., Mayo, M. W., Reuther, J. Y., Baldwin, A. S., Jr. (2001) J. Biol.
Chem. 276, 18934–1894092. Koul, D., Yao, Y., Abbruzzese, J. L., Yung, W. K., and Reddy, S. A. (2001) J.
Biol. Chem. 276, 11402–1140893. Gustin, J. A., Maehama, T., Dixon, J. E., and Donner, D. B. (2001) J. Biol.
Chem. 276, 27740–2774494. Mayo, M. W., Madrid, L. V., Westerheide, S. D., Jones, D. R., Yuan, X. J.,
Baldwin, A. S., Jr., and Whang, Y. E. (2002) J. Biol. Chem. 277,11116–11125
Wt. and Mutant IKK� Control Specific NF-�B-dependent Genes 14069
