The IκB Kinase (IKK) Inhibitor, NEMO-binding Domain Peptide, Blocks Osteoclastogenesis and Bone Erosion in Inflammatory Arthritis*

Activation of NF-κB leads to expression of ample genes that regulate inflammatory and osteoclastogenic responses. The process is facilitated by induction of IκB kinase (IKK) complex that phosphorylates IκB and leads to its dissociation from the NF-κB complex, thus permitting activation of NF-κB. The IKK complex contains primarily IKKα, IKKβ, and the regulatory kinase IKKγ, also known as NEMO. NEMO regulates the IKK complex activity through its binding to carboxyl-terminal region of IKKα and IKKβ, termed NEMO-binding domain (NBD). In this regard, a cell-permeable NBD peptide has been shown to block association of NEMO with the IKK complex and inhibit activation of NF-κB. Given the pivotal role of cytokine-induced NF-κB in osteoclastogenesis and inflammatory bone loss, we deduced that cell-permeable TAT-NBD peptide may hinder osteoclastogenesis and bone erosion in inflammatory arthritis. Using NBD peptides, we show that wild type, but not mutant, NBD blocks IKK activation and reduces cytokine-induced promoter and DNA binding activities of NF-κB and inhibits cytokine-induced osteoclast formation by osteoclast precursors. Consistent with the key role of NF-κB in osteoinflammatory responses in vivo, wild type TAT-NBD administered into mice prior to induction of inflammatory arthritis efficiently block in vivo osteoclastogenesis, inhibits focal bone erosion, and ameliorates inflammatory responses in the joints of arthritic mice. The mutant NBD peptide fails to exert these functions. These results provide strong evidence that IKKs are potent regulators of cytokine-induced osteoclastogenesis and inflammatory arthritis. More importantly, blockade of NEMO assembly with the IKK complex is a viable strategy to avert inflammatory osteolysis.

The transcription factor NF-B is ubiquitously expressed in various cell types in the body and regulates expression of numerous gene products including its own (1)(2)(3)(4). Advances made in the last decade have implicated NF-B as an effector of a wide range of inflammatory diseases, including rheumatoid arthritis and associated bone erosion (4 -6). However, the role of NF-B in osteoclastogenesis was not clear until the double cross of two single NF-B knockouts, namely p50 and p52 NF-Bs, which resulted in osteopetrosis due to osteoclast deficiency (7).
Activation of NF-B dimers is regulated by the inhibitory proteins, IB␣, IB␤, and IB⑀, of which IB␣ was widely investigated (1). Under non-stimulated conditions most of the NF-B is bound to IB and retained in the cytosol in its inactive form. Various stimuli prompt phosphorylation and dissociation of IB from the NF-B complex, allowing nuclear translocation of the transcription factor (1,8). The phosphorylation sites of IB␣ have also been identified, as well as the kinases involved in the process, namely the IB kinase (IKK). 1 The IKK complex is comprised primarily from three IKKs, ␣, ␤, and ␥ (9 -12). While the serine kinases IKK␣ and IKK␤ were found to target serines 32 and 36 of the IB␣, the role of IKK␥ (also known as NEMO) was deemed regulatory of the other IKKs (13,14).
Genetic studies have demonstrated that IKK␣ and IKK␤ play distinct roles with respect to NF-B activation. In this regard, IKK␤ is responsible for pro-inflammatory cytokineinduced IB␣ phosphorylation and subsequent activation of classical NF-B complexes containing the p50 and p65 subunits, whereas IKK␣ plays a significant albeit poorly defined role in keratinocyte differentiation independent of its catalytic activity (15,16). Interestingly, however, it was shown recently that IKK␣ catalytic activity is required for RANKL-induced NF-B activation in mammary epithelial cells (17) and B-lymphocytes (18). Intriguingly, these mechanisms were associated with the non-classical (non-canonical) NF-B activation pathway, namely targeting the NF-B2/p100 to process p52NF-B (19). These observations strongly suggest that IKK␣ and IKK␤ are prone to execute distinct signaling pathways based on the nature of their own activation. This activation of the IKKs is tightly regulated by the non-catalytic IKK␥/NEMO (14,20). NEMO is critical for pro-inflammatory activation of the IKK complex (15). Although the precise mechanism of NEMO action is poorly understood, it was speculated that it recruits the IKK complex to ligated cytokine receptors and facilitates transphosphorylation events (14,21). More intriguingly, NEMO may facilitate the recruitment of upstream IKK activators such as kinases that specifically target the activation loops within the catalytic domains of the IKK subunits (22).
Recent advances have identified that an NH 2 -terminal ␣-helical region of NEMO associates with a hexapeptide sequence within the extreme carboxyl terminus of IKK␤ and IKK␣, termed NEMO-binding domain (NBD) (23). Importantly, short cell-permeable peptide spanning the IKK␤ NBD disrupted the association of NEMO with IKK␤, blocked NF-B activation, and ameliorated responses in animal models of inflammation  (23,24). These observations position the IKK␤ NBD as a viable target for the design of anti-inflammatory drugs. Furthermore, given the central role of NF-B (the downstream target of the IKK complex) in osteoclastogenesis and inflammatory bone loss diseases (4,6,25,26), NBD is indeed an attractive target to be tested in this subset of diseases. Having shown that direct inhibition of NF-B by dominant-negative forms of IB blocks osteoclastogenesis, inhibits bone resorption in vivo, and ameliorates arthritic bone erosion, we deduced that inhibiting the assembly of IKK complex through administration of NBD peptide will result in NF-B inhibition and consequently to inhibition of inflammatory osteolysis. Our results indicate that TAT-NBD peptide inhibits IKK and NF-B activation and arrests osteoclastogenesis in vitro. More intriguingly, NBD is sufficient to block in vivo activation of NF-B, dampens osteoclasts formation and bone erosion, and ameliorates inflammatory arthritis.

MATERIALS AND METHODS
Reagents-All cytokines were from R&D Industries. All antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). All other chemicals were from Sigma unless otherwise indicated.
Cell Isolation and Purification-Osteoclast precursors (OCPs) in the form of marrow macrophages were isolated from whole bone marrow of 4 -6-week-old mice and incubated in tissue culture plates, at 37°C in 5% CO 2 , in the presence of 10 ng/ml M-CSF (27). After 24 h in culture, the non-adherent cells were collected and layered on a Ficoll-Hypaque gradient. Cells at the gradient interface were further purified with CD11b (Pharmingen) antibody and Dynabeads (Dynal) following the manufacturer's instructions, collected, and plated in ␣-minimal essential medium, supplemented with 10% heat-inactivated fetal bovine serum, at 37°C in 5% CO 2 in the presence of 10 ng/ml M-CSF, and plated according to each experimental conditions.
Mice-KRN-TCR (K/B) transgenic mice on a B6 background were kindly provided by D. Mathis and C. Benoist (Harvard Medical School, Boston, MA). K/BϫN mice were generated by breeding KRN-TCR with non-obese diabetic mice (Taconic Farms, Germantown, NY). All mice were housed under a controlled barrier facility at Washington Univer-sity (St. Louis, MO). Balb/c mice (4 -6 weeks old) were used to obtain primary OCPs and for in vivo experiments with NBD peptides.
Arthritogenic Serum-Serum is obtained from K/BϫN mice (6 -12 weeks old), pooled, and stored in aliquots in Ϫ70°C until used. A single dose of 150 l of serum per mouse was selected as optimal to induce arthritis in 100% of the mice. Three-to four-week-old Balb/c mice will be injected intra-peritoneally with phosphate-buffered saline, serum (150 l) with or without wild type TAT-NBD peptide, or mutant NBD peptide (ϳ3 mg/kg body weight was established as optimal dose for NBD). NBD peptides were injected on first and third day of experiment. Control mice developed swelling and redness of paws starting as early as 48-h post-serum injection. All mice were sacrificed on day 7 for examination.
Tissue Collection-Knee and ankle joints were excised and skin and soft tissues removed. These joints were snap-frozen immediately in liquid nitrogen for further analysis. For cellular and nuclear isolation, extracts were prepared from mouse paws following the procedure described by Han et al. (28). In brief, whole mouse joints and paws were flash-frozen in liquid nitrogen, crushed using a stainless steel motar and pestel, and then pulverized using CertiPrep Freezer Mill 6570 (Spex CertiPrep Inc., Metuchen, NJ) under liquid nitrogen to achieve a fine powder of ϳ2 ϫ 1 mm. Equal amounts were then subjected to lysis or nuclear isolation as described previously (25). For histology, intact limbs were skinned and processed for histology as described elsewhere (25).
Histology-For histology intact limbs were preserved in 10% buffered formalin (24 h), skinned, and subjected to a decalcification process using 10% EDTA, pH 7.0, for 7 days with gentle rocking and daily replacement of solution. Decalcified bones were then dehydrated in graded alcohol, cleared through xylene, and embedded in paraffin. Paraffin blocks were sectioned longitudinally. Five-micron sections were then stained with either hematoxylin and eosin or histochemically for tartrate-resistant acid phosphatase (TRAP) to determine osteoclasts.

FIG. 1. NBD peptide inhibits IKK and NF-B activation in
OCPs. Murine OCPs were treated with Me 2 SO or TAT-NBD (50 or 200 M) for 1 h followed by stimulation with RANKL (40 ng/ml/30 min). A, cell lysates were subjected to immunoprecipitation with anti-IKK␤ antibody followed by IKK kinase assay using GST-IB as a substrate. Equal amounts of crude cell lysates were immunobloted with anti-IKK antibody to document protein expression levels. B, nuclear extracts from the same treatment groups were subjected to EMSA as described under "Materials and Methods" using NF-B consensus sequence as a probe. TAT-NBD Peptide-For TAT-NBD peptides, we synthesized and high performance liquid chromatography-purified (using Washington University facility) two peptides: 1) functional wild type TAT-NBD (YGRKKRRQRRR-G-TTLDWSWLQME) and 2) negative control mutant TAT-NBD (YGRKKRRQRRR-G-TTLDASALQME).
Electrophoretic Mobility Shift Assay (EMSA)-Nuclear fractions were prepared as described previously (29,30). In brief, cells retrieved from joint tissue were washed twice with ice-cold phosphate-buffered saline. Cells were then resuspended in hypotonic lysis buffer A (10 mM HEPES, pH 7.8, 10 mM KCl, 1.5 mM MgCl, 0.5 mM dithiothreitol 0.5 mM 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), 5 g/ml leupeptin) and incubated on ice for 15 min, and Nonidet P-40 was added to a final concentration of 0.64%. Nuclei were pelleted, and the cytosolic fraction was carefully removed. The nuclei were then resuspended in nuclear extraction buffer B (20 mM HEPES, pH 7.8, 420 mM NaCl, 1.2 mM MgCl, 0.2 mM EDTA, 25% glycerol, 0.5 mM dithiothreitol, 0.5 mM AEBSF, 5 g/ml pepstatin A, 5 g/ml leupeptin), vortexed for 30 s, and rotated for 30 min in 4°C. The samples were then centrifuged, and the nuclear proteins in the supernatant were transferred to fresh tubes and protein content was measured using standard BCA kit (Pierce). Nuclear extracts (10 g) were incubated with an end-labeled double-stranded oligonucleotide probe containing the sequence 5Ј-AAA CAG GGG GCT TTC CCT CCT C-3Ј (31) derived from the B3 site of the TNF promoter. The reaction was performed in a total of 20 l of binding buffer (20 mM HEPES, pH 7.8, 100 mM NaCl, 0.5 mM dithiothreitol, 1 g poly(dI-dC), and 10% glycerol) for 30 min at room temperature. Samples were then fractionated on a 4% polyacrylamide gel and visualized by exposing dried gel to film.

Wild Type NBD Peptide Inhibits IKK and NF-B Activation
in OCPs-IKK induction is essential for phosphorylation of IB and subsequent activation of NF-B. Given that activation of IKKs requires recruitment of NEMO, we utilized an excess TAT-NBD peptide designed in its wild type form to sequester NEMO and prevent binding to IKK␤. Our results show that wild type NBD (50 and 200 M) is sufficient to block RANKL (Fig. 1A) and TNF-activation (not shown) of IKK, whereas mutated NBD fails to do so (Fig. 1A). Furthermore, wild type NBD, but not mutant NBD, blocks NF-B DNA binding (Fig.  1B) and promoter activity by OCPs (data not shown).

Inhibition of NF-B Activation by NBD Peptide Arrests Osteoclastogenesis-We and others (32) have shown that NF-B is essential for osteoclast differentiation and survival and that inhibition of NF-B activation by dominant-negative IBs attenuates osteoclastogenesis and induce osteoclast apoptosis.
Given that NBD appears to be a potent inhibitor of this pathway in OCPs, we examined whether the peptide impacts differentiation of OCPs into mature osteoclasts. The data presented in Fig. 2a demonstrate that NBD strongly inhibits basal (RANKL) and TNF-stimulated osteoclast differentiation in vitro as evident by significant reduction of multinucleated TRAP-expressing cells. Significant inhibition of osteoclasts was achieved with 50 M, whereas a dose of 200 M appears maximal (Fig. 2b). These results were complemented by lack of such inhibition when mutant NBD peptide is used. Thus, wild type, not mutant, NBD peptide inhibits in vitro osteoclastogenesis.
NBD Peptide Inhibits IKK and NF-B Activation in Inflammatory Arthritis-We have demonstrated that NF-B plays a pivotal role in erosive arthritis in the serum-transfer model of inflammatory arthritis (25). Thus, we decided to examine FIG. 3. NBD peptide inhibits IKK and NF-B activation in inflammatory arthritis. a-c, mice (ϳ35 g) were injected (intraperitoneally) with Me 2 SO (DMSO) or NBD peptides (3 and 6 mg/kg body weight). At the same time mice were injected with 0.2 ml of arthritogenic serum to engender inflammatory arthritis. Me 2 SO and NBD peptides were injected again on day 3. On day 7, cells were retrieved form joint synovium as described under "Materials and Methods" and processed for IKK kinase assays (a) and NF-B-DNA binding assays (EMSA) (c). Intact joints were processed for histology and immunostaining to detect active IKK (b, arrows). Panels represent histological sections from control (A) or arthritic (B) mice injected with mutant (C) or wild type NBD (D) as indicated. Digital images were taken with 10ϫ light microscope objective. whether NBD inhibits this pathway in this inflammatory model. Wild type NBD peptide blocks IKK activation in cellular extracts (Fig. 3a) and histological sections (Fig. 3b) obtained form the joints of arthritic mice as demonstrated by kinase assays and immunoblots for active IKKs. These observations were consistent with dramatic inhibition of NF-B activation observed in cells retrieved from the same joints and processed for EMSA experiments (Fig. 3c). In contrast, activation of IKK and NF-B in arthritic mice was unaffected in the presence of excess mutant NBD peptide at the optimal doses used.
NBD Peptide Inhibits Osteoclast Formation and Bone Destruction in the Joints of Mice Bearing Inflammatory Arthritis-NF-B activation modulates in vivo differentiation of osteoclasts (25). Thus, we examined the effect of NBD peptide on in vivo osteoclastogenesis and on bone erosion associated with inflammatory arthritis. We document that the peptide strongly inhibits osteoclast formation in the joints of arthritic mice (Fig.  4, arrows). Moreover, examination of ankle and knee joints from these mice indicates that NBD peptide significantly reduces focal bone erosion and plummets cellularity (Fig. 4, asterisks), whereas mutant NBD peptide has no such effect. Altogether, these observations point out that IKK complex assembly and activation leading to activation of NF-B are central to erosive arthritis. More importantly, the data provide evidence that targeted inhibition of this pathway by NBD peptide is sufficient to attenuate this erosive process. DISCUSSION In our previous studies, we have shown that inhibition of NF-B activation through the use of dominant-negative forms of IB, inhibits osteoclast differentiation in vitro (33), induces apoptosis of osteoclasts and their precursors (32), and, most importantly, attenuates in vivo osteoclastogenesis and bone erosion and alleviates inflammatory erosive arthritis (25). The principal of this inhibition entails introduction of excess dominant-negative IB that displaces the in vivo moiety and that does not undergo phosphorylation. The resultant DN-IB/ NF-B complex is then presumed stable for a longer time leading to significant delays and attenuation of NF-B-dependent gene activation functions. In the current study, and along similar principles, we have utilized a NEMO-binding domain that binds NEMO and compromises formation of active IKK complex. This sequestration disables IKK functions, primarily the immediate phosphorylation of IB and activation of NF-B. However, it should be noted that levels of IB under these circumstances are likely to be physiologically relevant unlike supraphysiological levels of the dominant-negative forms required to achieve blockade of the NF-B pathway, thus avoiding likely complications of cytotoxicity associated with previous models.
The current study provides clear indications that IKKs modulate osteoclastogenesis and inflammatory responses leading to bone erosion. These findings were not obvious in IKK knock-out animals due to their lethality, notwithstanding ample studies with NF-B-compromised systems, clearly pointed to the central role of this pathway in inflammatory osteolysis. Despite the fact that NF-B activates a wide range of genes, inflammatory responses and osteoclastogenesis are increasingly perceived to be central to the activation of this transcription factor. This is supported by this and a previous report (25) showing little or no adverse effects resulting from blockade of NF-B pathway for therapeutic purposes. This seemingly high selectivity could be interpreted by targeting specific heterodimeric compositions of NF-B. Although not clear at the moment, this speculation is in agreement with the diverse classical and non-classical activation pathways of NF-B and their differential preference to IKK subtypes.
Our data also point to a high selectivity of the NBD peptide toward in vivo blockade of osteoclastogenesis and inflammatory bone erosion. The basis of such selectivity mandates further investigation in light of previous findings, including our own, that implicate the NF-B pathway as a prosurvival mechanism. In this regard, we have shown that optimal apoptosis of macrophages and osteoclasts under conditions of NF-B inhibition requires stimulation of the cells with pro-inflammatory cytokines such as TNF (32). This is consistent with previous observations where we documented that inhibition of NF-B dampens the levels of pro-osteoclastogenic and pro-inflammatory cytokines (e.g. RANKL and TNF) in vivo (25). Taken together, inhibition of NF-B appears to dampen expression of pro-inflammatory cytokines and as a result thwarts undesired apoptotic and other responses. Until the precise mechanism by which NF-B modulates inflammatory osteolysis is uncovered, the use of the NBD peptide could be considered as a promising tool for therapeutic intervention to alleviate inflammatory focal bone erosion and ameliorate inflammatory arthritis.