Tyrosine Phosphorylation Is Required for IκB Kinase-β (IKKβ) Activation and Function in Osteoclastogenesis*

The transcription factor NF-κB is crucial for numerous cellular functions such as survival, differentiation, immunity, and inflammation. A key function of this family of transcription factors is regulation of osteoclast differentiation and function, which in turn controls skeletal homeostasis. The IκB kinase (IKK) complex, which contains IKKα, IKKβ, and IKKγ, is required for activation of NF-κB, and deletion of either IKKα or IKKβ resulted with defective osteoclast differentiation and survival. We have recently investigated the details of the mechanisms governing the role of IKKβ in osteoclastogenesis and found that constitutively active IKKβ in which serine residues 177/181 were mutated into negatively charged glutamic acids instigates spontaneous bona fide receptor activator of NF-κB ligand (RANKL)-independent osteoclastogenesis. To better understand and define the functional role of IKKβ domains capable of regulating the osteoclastogenic activity of IKK, we investigated key motifs in the activation T loop of IKKβ, which are potentially capable of modulating its osteoclastogenic activity. We discovered that dual serine (traditional serine residues 177/181) and tyrosine (188/199) phosphorylation events are crucial for IKKβ activation. Mutation of the latter tyrosine residues blunted the NF-κB activity of wild type and constitutively active IKKβ, and tyrosine 188/199-deficient IKKβ inhibited osteoclastogenesis. Thus, tyrosines 188/199 are a novel target for regulating IKKβ activity, at least in osteoclasts.

Nuclear factor -B (NF-B) 3 comprises a family of transcription factors crucial for numerous cellular functions including survival, differentiation, and apoptosis (1,2). This family of transcription factors is ubiquitously expressed in various cell types of the body and regulates expression of numerous gene products including its own (3)(4)(5)(6). Advances made in the last decade have implicated NF-B as an effector of a wide range of inflammatory diseases, including rheumatoid arthritis and inflammatory bone erosion (5,7,8). NF-B is regarded as the centerpiece of the inflammatory process and induces expression and production of pro-inflammatory and osteoclastogenic mediators (5,9,10). Furthermore, it has been established that persistence of NF-B activation in inflammatory settings, which is accompanied with secretion of a wide array of inflammatory mediators, leads to dysregulation of skeletal homeostasis primarily through accelerated pathologic bone loss because of exacerbated osteoclast recruitment and activity (7)(8)(9). On the other hand, using gene deletion studies, the role of NF-B in osteoclastogenesis was established and involves most members of this family (11).
The transcription factor NF-B family is composed of several members including p50, p52, RelA/p65, RelB, c-Rel, the precursors NF-B1/p105 and NF-B2/p100 (which undergo processing into p50 and p52, respectively), and the inhibitory subunits IB␣, IB␤, and IB⑀. Activation of NF-B dimers is regulated by the inhibitory proteins, of which IB␣ has been widely investigated (4). Under unstimulated conditions, most of the NF-B is bound to IB␣ and retained in the cytosol in its inactive form. Various stimuli, such as TNF, IL-1, UV, endotoxins, lymphotoxins, viruses, the osteoclastogenic factor RANK ligand (RANKL), and other factors, prompt activation of a proximal kinase complex, leading to phosphorylation and dissociation of IB␣ from the NF-B complex, allowing nuclear translocation of the transcription factor (4,12). Phosphorylation of IB␣ occurs on N-terminal serine residues and was found to be induced by a large IB␣ serine kinase (IKK) complex. The predominant IKK complex found in most cells contains two catalytic subunits, IKK␣ (also known as IKK1) and IKK␤ (IKK2), and a regulatory subunit, IKK␥ (also known as NEMO) (13)(14)(15)(16). Although the catalytic serine kinases IKK␣ and IKK␤ were found to target serines 32 and 36 of the IB␣ (and p100 processing by IKK␣), NEMO was found to act as a scaffold protein. NEMO contains several protein interaction motifs with no apparent catalytic domains, yet it is essential for staging the assembly of the IKK signalosome (17,18).
IKK␤ is the dominant activator of the classical NF-B, mediates the effects of pro-inflammatory and pro-osteoclastic factors, and as such, plays a pivotal role in inflammatory responses. Recent studies have shown that IKK␤ is crucial for osteoclastogenesis and mediates osteolysis (19,20). In this regard, targeted deletion of IKK␤ in the myeloid lineage abrogates osteoclastogenesis in vitro and in vivo (19,20). Consistent with this role, inhibiting IKK␤ activation abolished osteoclastogenesis, inflammatory osteolysis, and inflammatory arthritis in various models of inflammation (21)(22)(23). Furthermore, dele-tion of IKK␤ renders immune cells, macrophages, and osteoclast progenitors susceptible for TNF-induced apoptosis through a gain of function in the c-Jun N-terminal kinase (JNK) pathway. This is supported by the findings that inhibition of TNF or JNK restores the osteoclastogenic and inflammatory potential of these cells (19,20).
IKK␤ is activated by proximal kinases, the identity of which has not been fully clarified. Several in vitro studies using mouse embryonic fibroblast cells have implicated TAK1 as a potential activator of IKK␤ (24,25). However, the proximal events of NF-B activation are rather complex and involve a series of polyubiquitination events culminating with activation of IKK␤ (24,25). The vast majority of published studies suggest that IKK␤ undergoes serine phosphorylation on residues Ser-177/ 181 located in the amino-terminal activation T loop. In support of this mechanism, substitution of these two serine residues to alanines hinders the activation of IKK␤ (1). Conversely, substitution of the same serine residues to negatively charged glutamic acids renders IKK␤ constitutively active. This phosphomimetic form is biologically superactive in various cellular targets (26,27).
In a recent study, we have discovered that this phospho-mimetic form of IKK␤ is capable of inducing the classical NF-B pathway in osteoclast progenitors independent of the entire proximal signaling of RANK, TNF receptor, and IL-1 receptor. Importantly, this signaling was also found to be independent of other NF-B partners including IKK␣, NEMO, and Rel-B (27). To provide better understanding of the mechanism(s) underlying IKK␤ activation, we examined the amino-terminal activation T loop of this kinase with further scrutiny. We observed that tyrosines 188 and 199, which are juxtaposed to serines 177/181, were found to be conserved among different kinases, including IKK␣, AKT, cAMP-dependent protein kinase, and all protein kinase C isoforms (28,29). Interestingly, using osteoclast progenitors, we show that IKK␤ undergoes dual tyrosine and serine phosphorylation in response to RANKL and TNF. This observation is supported by an earlier study in which, using A549 cells, in vitro transfections showed that IKK␤ undergoes tyrosine phosphorylation on residues 188/199 (30,31). To better understand the role of tyrosines 188/199 in IKK␤ activation mechanism and cellular function, we engineered various forms of IKK␤ in which tyrosines 188/199 and serines 177/ 181 were substituted at various combinations. Our findings suggest that activation of IKK␤ requires dual phosphorylation of tyrosines 188/199 followed by serines 177/181. Furthermore, interruption of tyrosines in the activation loop hinders the activity and function of all forms of IKK␤ including the constitutively active IKK␤ SSEE form. Most importantly, tyrosine-mutated IKK␤ inhibits expression of inflammatory mediators and arrests osteoclastogenesis.

EXPERIMENTAL PROCEDURES
Animals and Cells-All mice were housed in a controlled barrier facility at Washington University (St. Louis, MO). Floxed IKK␤ (32) mice were from Dr. Pasparakis (University of Cologne, Germany). For in vitro experiments, wild-type C57BL/6 mice at 5-6 weeks of age were used.
Generation of Monocyte/Macrophages (Osteoclast Progenitors/Precursors)-Marrow was flushed from long bones into ␣-minimum essential medium (MEM). Cell pellets were resuspended in whole medium (␣-MEM with 1ϫ penicillin/ streptomycin, 10% heat-inactivated fetal bovine serum (FBS)). Monocytes/macrophages were produced by growing cell suspensions in the presence of 10 ng/ml M-CSF. Monocytes/macrophages were allowed to proliferate for 3 days at 37°C in 5% CO 2 , at which point they were infected with retrovirus (50% virus supernatant, 50% ␣-MEM containing 10% FBS, 10 ng/ml M-CSF, penicillin/streptomycin, and 4 g/ml hexadimethrine bromide). Twenty-four h after infection, cells were selected in ␣-MEM containing 10% FBS, 10 ng/ml M-CSF, penicillin/streptomycin, and 2 g/ml puromycin for 72 h, at which point selection medium was removed, and cells were washed and grown for 24 additional h without puromycin. At this point, cells were lifted, counted, and plated for downstream experiments.
Generation of Retrovirus-The use of Plat-E retrovirus packaging cells stably expressing retroviral structural proteins gagpol and env for transient production of high titer retrovirus was described previously (35). Briefly, 8 g of pMx vectors expressing our gene of interest were transfected into 5 million plat-E cells (grown in Dulbecco's modified Eagle's medium supplemented with 10% FBS, 10 ng/ml M-CSF, and penicillin/streptomycin) using FuGENE 6 (Roche Applied Science) according to the manufacturer's instructions. Twenty-four h after transfection, medium was exchanged to remove transfection reagent. Twenty-four and 48 h after medium exchange, supernatant was collected and pooled for infection of monocytes (see above).
In Vitro Osteoclastogenesis-For osteoclastogenesis assays, 3 ϫ 10 4 monocytes were plated in 200 l of ␣-MEM with 10% FBS. IKK␤ SSEE -expressing cells were cultured in 10 ng/ml M-CSF, whereas GFP-and IKK␤ WT -expressing cells were cultured in 10 ng/ml M-CSF plus 100 ng/ml RANKL for 4 days. At this point, cells were fixed and TRAP-stained using the leukocyte acid phosphatase kit (Sigma). TRAP-positive cells with three or more nuclei were scored as osteoclasts.
RNA Isolation and cDNA Production-RNA was isolated from macrophage or osteoclast cultures using the total RNA isolation mini kit (Agilent Technologies, Santa Clara, CA) according to the manufacturer's instructions. Reverse transcription was described previously (19).
Quantitative Real-time PCR-Quantitative real-time PCR procedure was described in detail previously (19).
Western Blotting-The Western blot procedure was described previously (19). One million cells were used for protein extraction and demonstration of protein expression.
Microscopy-Cells were imaged under white or ultraviolet light on an inverted microscope (Olympus IX-51). Digital images were captured using a CCD camera (Olympus DP70, 12 megapixel resolution).
Statistics-A Student's two-tailed t test for comparison between means was used for all analyses.

IKK␤ Undergoes Dual Serine and Tyrosine Phosphorylation-
Serine phosphorylation of IKK␤ at residues Ser-177/181 is crucial for its activation. In fact, substitution of these two serines into alanines renders this protein irresponsive to stimuli, nonphosphorylatable, and catalytically inactive. Conversely, substituting serines 177/181 into negatively charged amino acid such as glutamic or aspartic acid render IKK␤ constitutively active. This form of active IKK␤ phosphorylates target proteins and activates the downstream canonical NF-B subunits in complete absence of external stimuli. Surprisingly, we have recently discovered that this constitutive activity of IKK␤ empowers spontaneous osteoclastogenesis, which is independent of all known proximal signals including RANK stimulation (27). To better understand the mechanism underlying the phosphorylation events occurring in the activation loop of IKK␤ that are crucial for its osteoclastogenic activity, we tested the potential contribution of other potentially important neighboring motifs. In this regard, phosphorylation of IKK␤ on tyrosine residues located in the activation T loop has been suggested (31); however, neither functional significance nor meaningful mechanistic insights have been provided.
In this study, we contemplated that phosphorylation of tyrosines 188/199 located in close proximity to serines 177/ 181 in the activation loop of IKK␤ may regulate activation of this kinase. To provide the foundation for this assumption, we first provide evidence that IKK␤ undergoes tyrosine phosphorylation in osteoclast progenitors treated with RANKL or TNF (Fig. 1). Tyrosine phosphorylation of IKK␤ was enhanced severalfold when compared with untreated (lane 1, denoted C) condition. This tyrosine phosphorylation event was specific, as supported by the presence of tyrosine phosphorylation of the phosphatidylinositol 3-kinase subunit p85 (positive control) under TNF-treated conditions and absence of such phosphorylation with IgG (negative control). To determine the relevance and specificity of this event, we examined tyrosine and serine phosphorylation of IKK␤ in the presence of the tyrosine kinase inhibitor PP2. Using antibodies specific for the T-loop serine and tyrosine residues, we demonstrate that RANKL-induced serine and tyrosine phosphorylation, as evident after 15 and 30 min of stimulation respectively, was entirely obliterated in the presence of the tyrosine kinase inhibitor PP2 ( Fig. 2A), whereas protein levels of IKK␤ remained unchanged. This result suggests that serine and tyrosine phos- Osteoclast progenitors were cultured for 3 days, after which time they were starved for 2 h in serum-free ␣MEM and treated with RANKL (RL) (50 ng/ml) or TNF (10 ng/ml) for 30 min. Cells were then lysed and subjected for immunoprecipitation with anti IKK␤ antibody, p85 (phosphatidylinositol 3-kinase) antibody, or IgG followed by immunoblots with phospho-tyrosine (PY), IKK␤, and p85 antibodies. A fraction of the whole cell lysate (prior to immunoprecipitation) was used to determine ␤-actin levels as a loading control. C denotes control. phorylation steps leading to activation of IKK␤ are coupled events. To provide further support for this mechanism, we employed a molecular approach. We transfected IKK␤-null mouse embryonic fibroblast cells with GFP, wild-type IKK␤ (IKK␤ WT ), tyrosine-mutated IKK␤ (IKK␤ YYFF ), IKK␤ in which serines were mutated into alanines (IKK␤ SSAA ), and constitutively active IKK␤ in which serines 177/181 were substituted to glutamic acids (IKK␤ SSEE ). Cell lysates were then processed, and both protein expression as well as serine phosphorylation of the various IKK␤ forms were examined and normalized by ␤-actin expression. The data show that although IKK␤ WT underwent serine phosphorylation, IKK␤ YYFF did not. As expected, IKK␤ SSEE exhibited hyperphosphorylation (positive control), and IKK␤ SSAA failed to do so in response to TNF (Fig.  2B). The detection of baseline phosphorylation in the latter condition is within the margin of antibody specificity evident by lack of up-regulation in the stimulated when compared with the unstimulated conditions. Altogether, the results of this experiment demonstrate that integrity and phosphorylation of tyrosine residues proceed and are essential for serine phosphorylation of IKK␤.

Mutation of Tyrosines 188/199 in the Activation T Loop Abolishes Function of IKK␤-
To provide further mechanistic support for this finding, we surmised that mutation of tyrosine residues 188/199 is expected to nullify activity of constitutively active IKK␤ (IKK␤ SSEE ). To this end, tyrosines 188/199 were singly or doubly mutated to phenylalanines in wild-type and constitutively active IKK␤ backbones. These IKK␤ forms (IKK␤ Y188F , IKK␤ Y199F , WT-IKK␤ YYFF , IKK␤ SSEE/YYFF (referred to as IKK␤ SEYF )), alongside other forms, which were used as positive and negative controls, were transfected in IKK␤-null mouse embryonic fibroblast cells in the absence or presence of TNF or lipopolysaccharide stimulation, and activation of NF-B was measured using an NF-B luciferase reporter assay. As expected, NF-B transcriptional activity was elevated in cells expressing IKK␤ WT and stimulated with TNF or lipopolysaccharide. Furthermore, baseline NF-B transcriptional activity was vastly increased in cells expressing IKK␤ SSEE (Fig. 3). In contrast, IKK␤ KM failed to mount any meaningful activity, and baseline NF-B transcriptional activity elicited by IKK␤ SSAA was not inducible, as expected. Surprisingly and consistent with our previous observations, NF-B transcriptional activity was dramatically reduced when single or double tyrosine mutations were introduced. More importantly, tyrosine mutation of constitutively active IKK␤ (IKK␤ SEYF ) rendered this kinase entirely inactive (Fig. 3), supporting the notion that integrity of the T-loop tyrosines is obligatory for IKK␤ activity. Interestingly, NF-B activity under conditions of the absence of amino-terminal tyrosine residues was below basal activity values, suggesting that tyrosine-mutated IKK␤ forms may act as dominant negative proteins. Taken together, this finding suggests that serine phosphorylation and tyrosine phosphorylation of IKK␤ are coupled events.
Tyrosines 188/199 Are Essential for the Inflammatory and Osteoclastogenic Activity of IKK␤-Having established that mutation of tyrosines 188/199 abolishes NF-B transcriptional activity, we asked whether this inhibition will impact the signaling of IKK␤ as an inflammatory signal inducer. Indeed, using quantitative PCR, we demonstrate that although IKK␤ SSEE induces mRNA levels of TNF and IL-1␤, tyrosine-mutated wild-type (WT) or SSEE forms of IKK␤ fail to do so (Fig. 4, A  and B).  We have shown that IKK␤ is crucial for osteoclastogenesis and that constitutively active IKK␤ stimulates spontaneous osteoclastogenesis. Based on these findings, we surmised that if our hypothesis is correct, we expect that mutating tyrosines 188/199 in IKK␤ WT and IKK␤ SSEE may alter their function in osteoclastogenesis. To test this proposition, we employed the various IKK␤ modified forms and tested their osteoclastogenic activity using the osteoclast formation assay. For this experiment, all relevant IKK␤ forms were expressed retrovirally to facilitate efficient delivery to osteoclast progenitors. Wild-type osteoclast progenitors expressing IKK␤ WT generate osteoclasts in response to RANKL treatment. In contrast, when WT osteoclast progenitors were infected with tyrosine mutant IKK␤, their osteoclastogenic response was significantly impaired. Furthermore, we have shown that IKK␤ SSEE potently induces osteoclastogenesis regardless of RANKL (27). Even this powerful response was dramatically impaired when tyrosine mutations were introduced in the IKK␤ SSEE backbone (Fig. 5, A and B).
As a proof of principle, we also tested the potential of tyrosine mutated IKK␤ forms to rescue osteoclastogenesis by IKK␤-null cells (without RANKL). To this end, osteoclast precursors derived from IKK␤-null mice were infected with various forms of IKK␤ in which serines 177/181 or tyrosines 188/199 were mutated at different combinations. In addition, a constitutively active form of IKK␣ (right bottom panel) was used as a negative control. As expected, only IKK␤ SSEE was capable of supporting osteoclastogenesis by IKKnull osteoclast progenitors. However, when Tyr-188/199 were mutated (S177E/S181E/Y188F/Y199F), the osteoclastogenic activity of this constitutively active IKK form was nullified (Fig.  5, C and D). Interestingly, similar to the tyrosine mutated form, IKK␤ in which serines 177/181 were mutated into alanines failed to rescue osteoclastogenesis.
Having demonstrated that PP2 inhibits tyrosine and serine phosphorylation of IKK␤ ( Fig. 2A), we asked whether this approach carries physiological significance in osteoclastogenesis. To address this prospect, we conducted an osteoclast formation assay in the presence of increasing doses of PP2 using both RANKL-stimulated and IKK␤ SSEE -induced osteoclastogenesis (Fig. 6, A and C). Although the effect of the carrier DMSO on osteoclasts was indistinguishable from control, PP2 dose dependently inhibited osteoclast formation (Fig. 6, A and  B) without affecting cell survival (Fig. 6C). The effect was significant at submicromolar concentrations and maximal at 1 M. At this optimal concentration, PP2 inhibited RANKL-stimulated and IKK␤ SSEE -induced osteoclastogenesis entirely (Fig.  6C). We also tested the effect of administering a tyrosine phosphatase directly to osteoclast precursors. Specifically, cells were infected with a retroviral SHP1 or its dominant negative form SHP1(C435S) and subjected to an osteoclast formation assay.
The results indicate that although SHP1 significantly inhibited osteoclast formation, its dominant negative form, which has been reported to inhibit endogenous SHP1 (36), appears to enhance osteoclastogenesis (Fig. 6C, bottom panels). These latter data suggest, but do not provide definitive proof, that the tyrosine phosphatase SHP1 may regulate IKK␤ activity. Further studies are required to investigate the details of this mechanism. Taken together, these observations suggest that tyrosine phosphorylation and serine phosphorylation of IKK␤ are pivotal for its osteoclastogenic activity. In parallel experiments, we show that although IKK␤ SSEE (similar to RANKL) induces the osteoclast differentiation markers TRAP and NFATc1, mutation of Tyr-188/199 abolishes this expression (Fig. 7, A  and B). These observations were further supported biochemically where we observed that mutating Tyr-188/199 abolished expression of the osteoclast markers NFATc1, RelB, c-Fos, and ␤3-integrin, which was induced by the IKK␤ SSEE form and by RANKL-generated osteoclasts (compare lane 7 with lanes 5 and 8, respectively) (Fig. 7C). Thus, tyrosines 188/199 are crucial for IKK␤ function as osteoclastogenic mediators.

DISCUSSION
The activation loop of IKK␤ contains unique elements crucial for the function of this kinase. The most studied elements are serines 177/181 and lysine 44. It has been established that serine kinases, such as TAK-1, phosphorylate IKK␤ at residues 177/181. This event is essential for ATP binding by Lys-44, which is a key step in the activation of IKK␤. Mutational studies have confirmed that substitution of serines 177/181 to alanines (S177A/S181A) or lysine 44 to methionine (K44M) abrogate the activity of this kinase. Conversely, substitution of serines 177/181 to negatively charged glutamic acids (S177E/S181E) results in the constitutively active form of IKK␤ because of the phospho-mimetic nature of glutamic acids (37,38).
Several studies have demonstrated that constitutively active IKK␤ is capable of executing biological functions in a ligandfree fashion at levels and durations exceeding stimulus-driven activation (26,39). In addition, constant and continuous hyperactivity of IKK␤, typical for IKK␤ SSEE , often leads to dysregulated functions and may pave the way for various pathologies. In this regard, we have shown recently that when introduced into osteoclast progenitors, this active form of IKK␤ leads to differentiation of bona fide osteoclasts in the complete absence of RANK signaling. This is consistent with a recent study by Sasaki et al. (26), which reported that activation of the canonical NF-B pathway by constitutively active IKK␤ renders B cell survival independent of B cell activation factor signaling cascade. Furthermore, constitutive IKK␤ signaling (canonical NF-B signaling), which is considered a hallmark of various B cell lymphomas, leads to the accumulation of resting B cells and supports their proliferation and survival upon activation. Therefore, regulation of this stimulus-independent activity of IKK␤ is crucial to regulate cellular functions, specificity of which appears to be dictated by the cellular context.
In our study, we examined key elements in the activation loop of IKK␤, which has been overlooked or underappreciated, and investigated their potential contribution to IKK␤ activation and regulation. In this regard, tyrosine residues 188/199 are within the activation T loop of IKK␤ and are in close proximity to serines 177/181. The fact that the tyrosine kinase c-Src binds to IKK␤ in endothelial cells suggests that tyrosine phosphorylation of IKK␤ is potentially important for regulation of its activity. Indeed, we find that mutating these two key tyrosine residues in the activation loop renders IKK␤ inactive. Even more strikingly is the finding that this double tyrosine residue mutation nullifies the activity of constitutively active IKK␤. This observation led us to posit that tyrosine phosphorylation on residues 188/199 is a crucial step that facilitate serine phosphorylation of IKK␤. Moreover, our data suggest that intact serines 177/181, or when substituted into glutamic acid (SS 3 EE), are not sufficient to maintain activation of the classical NF-B pathway when tyrosines 188/199 are substituted with phenylalanines. These observations suggest that a number of scenarios are possible. First, phosphorylation of tyrosines 188/199 may serve as a first event in a multistep process leading to serine phosphorylation and ATP binding culminating with activation of the kinase. Second, phosphorylation of tyrosines 188/199 may be essential for maintaining a structural change that facilitates activation of IKK␤. Third, it is possible that phospho-tyrosines 188/199 act as a docking site for the SH2containing protein that consequently contributes to IKK␤ activation. This possibility remains to be verified experimentally. By way of analogy, phosphorylation of conserved tyrosine residues was found to be crucial for activation of a large number of kinases including protein kinase B/Akt, PDK1, cAMP-dependent protein kinase, and over a dozen isoforms of protein kinase C (28,29,40).
We have shown previously that the tyrosine kinase c-Src phosphorylates the IKK substrate IB␣ on the amino-terminal residue Tyr-42 (41). Subsequent studies in various systems have confirmed tyrosine phosphorylation of IB␣ and implicated c-Src family members in such process (42)(43)(44)(45). We have confirmed the notion that c-Src or a closely related tyrosine kinase associates with the IKK complex and that IKK␤ undergoes substantial tyrosine phosphorylation on residues 188/199. The fact that both IKK␤ and its substrate IB␣ associate with and undergo tyrosine phosphorylation, events that appear in our studies as important for osteoclastogenesis, underscores the significance of this mechanism for regulating osteolysis. However, future studies will determine whether tyrosine phosphorylation of both proteins is dependent.
We demonstrated the functional relevance of mutating tyrosines 188/199 by examining the inflammatory and osteoclastogenic activities of IKK␤. TNF and IL-1␤ are hallmarks of the inflammatory response, which are induced by canonical NF-B activation. Our data demonstrate that levels of these two cytokines is blunted in osteoclast progenitors expressing IKK␤ YYFF or IKK␤ SEYF , bolstering the notion that intact tyrosines 188/199 are essential for this activity by IKK␤. More importantly, our data confirm that IKK␤ YYFF acts as a dominant negative protein that hinders osteoclastogenesis induced by RANKL or by IKK␤ SSEE , as shown by osteoclast differentiation assay and by regulating mRNA and protein expression of key osteoclast differentiation markers. Consistent with these findings, administration of the tyrosine kinase inhibitor PP2 or expression of the tyrosine phosphatase SHP1, both FIGURE 6. The tyrosine kinase inhibitor PP2 and tyrosine phosphatase SHP1 inhibit osteoclastogenesis. Osteoclast precursors were treated with RANKL or infected with IKK␤ SSEE and cultured at the presence of PP2 (at the concentrations shown) or the carrier DMSO. A, cultures were fixed on the 4th day of culture and stained for TRAP expression. B, the osteoclasts in panel A were counted under a light microscope. Each bar represents the average of four wells. The experiment was repeated three times with consistent results. * denotes p Ͻ 0.05, and ** denotes p Ͻ 0.001. Error bars indicate S.E. C, upper panels represent osteoclasts from RANKL-treated cells in the absence or presence of PP2. Middle panel represents cells infected with retroviral IKK␤ SSEE and cultured in the absence or presence of PP2 as indicated. Lower panels represent cells infected with retroviral SHP1 or mutant SHP1 (m-SHP1) and subjected to osteoclastogenic conditions in the presence of RANKL. Arrows mark multinucleated osteoclasts.
of which target phospho-tyrosine residues, dramatically inhibited osteoclastogenesis. This is further supported by our finding that IKK␤ YYFF inhibits NF-B transactivation as demonstrated by luciferase reporter assay. These observations are consistent with previous findings by our group and others demonstrating that IKK␤ is crucial for osteoclast differentiation and function and plays a key role in the regulation of skeletal integrity.
Taken together, our data propose a potentially novel mechanism for IKK␤ activation. This mechanism requires intact tyrosine residues at positions 188/199 in the activation loop, phosphorylation of which facilitates and/or acts in concert with phosphorylation of serines 177/181 for proper activation of IKK␤. Furthermore, tyrosines 188/199 appear to be crucial for this process because their mutation blunts the activity of the serine phospho-mimetic form of IKK␤ (IKK␤ SEYF ). Finally, this mechanism provides another level at which activation of IKK␤ can be regulated and presents a potential target for therapeutic intervention.