Oncogene (2010) 29, 3619–3629 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc ORIGINAL ARTICLE TRE17/USP6 oncogene translocated in induces matrix metalloproteinase production via activation of NF-jB

YYe1, LM Pringle1, AW Lau1, DN Riquelme1, H Wang2, T Jiang3, D Lev4, A Welman5, GA Blobel2, AM Oliveira6 and MM Chou1

1Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA, USA; 2Division of Pediatrics, Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA, USA; 3Abramson Cancer Center, University of Pennsylvania School of Medicine, Philadelphia, PA, USA; 4Department of Cancer Biology, University of Texas MD Anderson Cancer Center, Houston, TX, USA; 5Edinburgh Cancer Research Center, University of Edinburgh, Edinburgh, UK and 6Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA

Aneurysmal bone cyst (ABC) is an aggressive, pediatric Introduction bone tumor characterized by extensive destruction of the surrounding bone. Although first described over 60 years -specific proteases (USPs) constitute a large ago, its molecular etiology remains poorly understood. subfamily of de-ubiquitinating enzymes that function in Recent work revealed that ABCs harbor translocation diverse cellular processes, including the DNA-damage of TRE17/USP6, leading to its transcriptional upregula- response, inflammation and oncogenesis (Amerik tion. TRE17 encodes a ubiquitin-specific protease (USP), and Hochstrasser, 2004; Singhal et al., 2008; Komander and a TBC domain that mediates binding to the Arf6 et al., 2009). Several USPs have been implicated in GTPase. However, the mechanisms by which TRE17 human tumorigenesis. HAUSP/USP7 is a key regulator overexpression contributes to tumor pathogenesis, and the of the p53 tumor suppressor (Li et al., 2002). HAUSP role of its USP and TBC domains, are unknown. ABCs are de-ubiquitinates p53 to rescue it from proteasomal characterized by osteolysis, inflammatory recruitment and degradation. CYLD encodes a tumor suppressor that extensive vascularization, the processes in which matrix is mutated in familial cylindromatosis, an auto- proteases have a prominent role. This led us to explore somal dominant disease characterized by benign tumors whether TRE17 regulates the production of matrix arising from cells of skin appendages. CYLD is a nega- metalloproteinases (MMPs). In this study we show that tive regulator of the transcription factor nuclear factor-kB TRE17 is sufficient to induce expression of MMP-9 and (NF-kB), and functions by de-ubiquitinating the ubiquitin MMP-10, in a manner requiring its USP activity, but not ligases, TRAF-2 (tumor necrosis factor receptor-associated its ability to bind Arf6. TRE17 induces transcription of factor 2) and TRAF-6 (Brummelkamp et al., 2003; MMP-9 through activation of nuclear factor-jB(NF-jB), Kovalenko et al., 2003; Trompouki et al., 2003). mediated in part by the GTPase RhoA and its effector Another USP strongly implicated in human neoplasia kinase, ROCK. Furthermore, xenograft studies show that is TRE17/USP6, which is translocated at high frequency TRE17 induces formation of tumors that reproduce multiple in aneurysmal bone cyst (ABC), a pediatric osseous features of ABC, including a high degree of vascularization, neoplasm (Cottalorda and Bourelle, 2007; Saccomanni, with an essential role for the USP domain. In sum, these 2008). Although benign, ABCs cause extensive bone studies reveal that TRE17 is sufficient to initiate tumor- degradation and impinge on surrounding tissue because igenesis, identify MMPs as novel TRE17 effectors that of their expansile nature, causing pain, swelling, likely contribute to ABC pathogenesis and define the deformity, neurological symptoms and pathologic frac- underlying signaling mechanism of their induction. ture (Cottalorda and Bourelle, 2007; Saccomanni, 2008). Oncogene (2010) 29, 3619–3629; doi:10.1038/onc.2010.116; Although first described over 60 years ago, the etiology published online 26 April 2010 of ABC has remained poorly understood. They were long thought to be reactive lesions, arising from local Keywords: TRE17/USP6; aneurysmal bone cyst; MMP-9; vascular disturbances. However, recent work showed Arf6; NF-kB; RhoA that translocation of TRE17 occurs in over 60% of ABC cases, revealing their neoplastic origin (Panoutsakopou- lous et al., 1999; Althof et al., 2004; Oliveira et al., 2004a, b, 2005; Sukov et al., 2008). Five different fusion Correspondence: Dr MM Chou, Department of Pathology and partners of TRE17 have been identified, and in every Laboratory Medicine, Children’s Hospital of Philadelphia, University case translocation resulted in promoter swapping, in of Pennsylvania School of Medicine, 3615 Civic Center Boulevard, which the complete coding sequence of TRE17 was Philadelphia, PA 19104, USA. E-mail: [email protected] placed downstream of the highly active promoter of its Received 1 October 2009; revised 2 March 2010; accepted 9 March 2010; fusion partner, leading to its transcriptional upregula- published online 26 April 2010 tion (Oliveira et al., 2004a, b, 2005). Notably, TRE17 is TRE17/USP6 induces MMP through NF-jB YYeet al 3620 absent or expressed at very low levels in most adult showed that the fibrous stromal regions of ABC stain human tissues except for testes (Paulding et al., 2003). strongly for MMP-9 (Kumta et al., 2003). This led us to These studies strongly suggest that TRE17 over- explore whether TRE17 contributes to ABC pathogen- expression has a key role in the etiology of ABC. esis by regulating production of MMPs. Using an Histologically, ABCs are complex lesions consisting of unbiased antibody-based array to screen the conditioned blood-filled cysts, separated by fibrous stromal areas medium (CM) of transfected cells, we found that TRE17 containing spindle cells, inflammatory cells and numer- induced the accumulation of MMP-9 and MMP-10 ous capillaries. Strikingly, TRE17 translocation occurs (Supplementary Figure 1). These array results were exclusively in spindle cells within the fibrous stroma confirmed by direct immunoblotting, performing studies that are believed to represent an early phase of in osteoblastic cells, the lineage that harbors transloca- osteoblast differentiation (Oliveira et al., 2004b). This tion/overexpression of TRE17 in ABC. As shown in is supported by the fact that the promoters of TRE17’s Figure 1a, transient expression of TRE17(long) in fusion partners are highly active during osteoblast hFOB1.19, a human pre-osteoblastic cell line (Subra- differentiation and bone formation (Oliveira et al., maniam et al., 2002), induced production of MMP-9 2004a, b, 2005). Indeed, one fusion partner, OMD,is and MMP-10, as determined by immunoblotting of CM expressed selectively in osteoblasts. Furthermore, from serum-starved cells. osteoid (the organic component of bone matrix pro- We next examined the roles of the TBC and USP duced by osteoblasts) is commonly present in these domains of TRE17 in MMP induction. To determine lesions. Together, these observations suggest that cells the requirement for the USP activity of TRE17, we overexpressing TRE17 in ABCs are immature cells of tested the naturally occurring TRE17(short) isoform, as the osteoblast lineage. However, the mechanism by well as a catalytically inactive point mutant (TRE17 which overexpression of TRE17 in osteoblasts contri- (long)/USP–) (Shen et al., 2005; Figure 1b). As seen in butes to the formation of these complex, cellularly Figure 1a, both of these alleles failed to induce MMP heterogeneous lesions is completely unknown. production in hFOB1.19 cells. We also assessed the role Two naturally occurring splice variants of TRE17 exist of the TBC domain, which we previously reported (Nakamura et al.,1992;Pauldinget al., 2003), giving rise to mediates binding to Arf6 (Martinu et al., 2004). We isoforms termed TRE17(long) and TRE17(short) (pre- generated a triple-point mutant within the TBC domain viously termed TRE17(onco)). They are identical through that ablates binding to Arf6 (denoted TRE17 (A6–); the first 773 amino acids, but then diverge; TRE17(long) Figures 1b and c). This mutant was competent to induce encodes a USP, which comprises cysteine and histidine MMP production as well as TRE17(long) (Figure 1a). subdomains, both of which are required for catalytic Together, these data indicate that the USP activity of activity (Papa and Hochstrasser, 1993). TRE17(short) is TRE17, but not its ability to bind Arf6, is required for truncated shortly after the cysteine subdomain. Both the induction of MMP-9 and MMP-10. isoforms contain an N-terminal TBC (Tre2/Bub2/Cdc16) We wished to confirm these results in an inde- domain. TBC domains are generally predicted to encode pendent osteoblastic system, under conditions of stable GTPase-activating proteins for Rab family GTPases expression. For this purpose, MC3T3-E1 murine (Bernards, 2003). However, we and others have shown pre-osteoblasts (Wang et al., 1999), referred to here- that the TBC domain of TRE17 lacks catalytic activity, after as MC3T3, were engineered to stably express and that it functions instead to mediate interaction with the TRE17 alleles in a doxycyclin (dox)-inducible manner. small GTPase, Arf6 (Martinu et al., 2004; Bizimungu and Figure 1d shows uniform expression of the various Vandenbol, 2005; Bizimungu et al., 2007). We further mutants, strictly in the presence of dox. MC3T3 cell reported that TRE17 promotes activation of Arf6 in vivo lines expressing TRE17(short) were also generated; (Martinu et al., 2004). However, the mechanism by which however, its expression levels were significantly reduced TRE17 contributes to ABC pathogenesis, and the roles of relative to the other alleles (data not shown). There- its TBC and USP domains, remain largely unknown. In fore, cells expressing TRE17 (long)/USP– were used this study, we identify matrix metalloproteinases (MMPs) in subsequent experiments to examine the role of the as novel effectors of TRE17, and dissect the signaling USP domain. Dox-dependent induction of MMP-9 was pathways leading to their induction. We further show that observed in the MC3T3 lines expressing TRE17(long) TRE17 is sufficient to induce the formation of tumors and TRE17 (A6–), as determined by both immunoblot- that recapitulate multiple features of ABC, in a manner ting and zymography, which measures gelatinolytic that requires its USP domain. activity (Figure 1d). In contrast, TRE17(long)/USP– was significantly impaired in stimulating MMP-9 production (Figure 1d). Effects on MMP-10 could not Results be confirmed in these cells by immunoblotting, as antibodies that recognize the murine protein are TRE17 induces production of MMP-9 and MMP-10 not available. However, alternative detection methods in a USP-dependent manner revealed that it was also induced (see next section). ABCs are characterized by extensive bone degradation, We next explored whether MMP induction by TRE17 inflammatory recruitment and angiogenesis, the was specific to osteoblasts, or could be elicited in other cell processes in which matrix proteases have a key role. types. To test this, TRE17 alleles were transfected into Furthermore, a previous immunohistochemical study HeLa cells, a cervical carcinoma. As seen in Figure 1e,

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3621

vec Short Long USP- A6-

MMP-9 TBC C H TRE17(long)

MMP-10 TBC C TRE17(short)

TRE17(long) TBC CX H TRE17(long)/USP- ns * TBCX C H TRE17/A6- TRE17 TRE17(short)

actin

hFOB1.19

vec Long Short USP- vec Long USP- A6- MMP-9 - + - + - + - + : Dox MMP-10 Long vec A6- TRE17 Arf6, I.P. MMP-9 Arf6, WCL TRE17 zymography TRE17, I.P. MC3T3 actin actin MC3T3 HeLa Figure 1 TRE17 induces secretion of MMP-9 and MMP-10 in a USP-dependent manner. (a) hFOB1.19 human fetal osteoblasts were transfected with the indicated hemagglutinin (HA)-tagged TRE17 constructs and then serum starved. MMP-9 and MMP-10 in the CM were concentrated as described in the Materials and methods , and detected by immunoblotting. Whole cell lysates were blotted for TRE17 (arrowheads indicate migration of TRE17(long) and TRE17(short) isoforms; asterisk denotes nonspecific (NS) band recognized by antibody) and actin as a loading control. (b) Domain structure of TRE17 alleles. TBC, TBC domain; C and H, cysteine and histidine subdomains of the USP domain. TRE17(long)/USP– and TRE17/A6– harbor point mutations in the indicated domains, as detailed in the Materials and methods. (c) MC3T3 osteoblasts stably expressing hemagglutinin (HA)-tagged TRE17(long) or TRE17 (A6–) in a doxycycline (dox)-inducible manner were treated with dox (2 mg/ml) for 24 h. Cell extracts were immunoprecipitated with anti-HA, and then blotted with anti-Arf6 (top panel) or anti-TRE17 (bottom panel). IP, anti-HA immunoprecipitate; WCL, whole cell lysate. (d)StableMC3T3celllines expressing the indicated TRE17 alleles were grown with or without dox for 24 h, and then starved for 24 h in the continued absence or presence of dox. MMP-9 was purified from the CM and then subjected to blotting or zymography to detect gelatinolytic activity. (e)HeLa cells were transfected as indicated, and accumulation of MMP-9 and MMP-10 in the CM was monitored as in a.

TRE17(long) was capable of inducing robust expression levels of MMP-9 and MMP-10 mRNAs were elevated in of MMP-9 and MMP-10 in these cells, again in a cells expressing TRE17(long), but not TRE17(short) or USP-dependent manner. Thus, the ability of TRE17 to TRE17(long)/USP– (Figures 2a–c). stimulate matrix protease production is not restricted to We next tested whether increased MMP expression osteoblasts. As mentioned above, previous work showed resulted from transcriptional induction. We focused on that ABC tumors stain strongly for MMP-9 (Kumta et al., MMP-9asitsexpressioninABChadbeenpreviously 2003), but the cells responsible for its production had not shown, and there exists a wealth of information on its been identified. Our data show that TRE17 is sufficient, in regulation as well as reagents for its analysis. Reporter the absence of any serum-derived factors, to induce assays were performed using a luciferase construct production of MMPs, and suggest that it may be directly driven by base pairs À674 to þ 3 of the human MMP-9 responsible for their production in ABCs. promoter (Liu et al., 2006). In both hFOB1.19 and Hela, TRE17(long) activated the MMP-9 promoter by approxi- mately fivefold over control vector-transfected samples TRE17 stimulates MMP expression on a transcriptional (Figure 2d). TRE17(short) and TRE17(long)/USP– were level significantly compromised in activation of the reporter. We next sought to determine the mechanism by which Transient reporter assays could not be performed in the TRE17 induces production of MMPs. Because trans- MC3T3 cell lines as they were engineered to stably and criptional induction is the most common mode of MMP inducibly express luciferase. Nevertheless, these data in regulation (Sternlicht and Werb, 2001; Mott and Werb, sum reveal that TRE17 induces transcription of MMP-9 in 2004), we examined whether TRE17 increased the a USP-dependent manner. abundance of MMP mRNA. RNA was isolated from transfected hFOB1.19 or HeLa, or the MC3T3 stable lines. Real-time quantitative reverse transcription PCR NF-kB is required for induction of MMP-9 by TRE17 was performed using two independent primer pairs for Two key regulators of the MMP-9 promoter are NF-kB each MMP, and normalized against glyceraldehyde (Liu et al., 2006) and activator protein 1 (Cho et al., 3-phosphate dehydrogenase. In all three systems, the 2000; Genersch et al., 2000; Kim and Lee, 2005; Han

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3622 abhFOB 1.19 Hela 10 50 50 50 ** * MMP9-1 MMP10-1 MMP9-1 MMP10-1 MMP9-2 MMP10-2 MMP9-2 MMP10-2 8 40 40 40 ** ** * * * * * * 6 30 30 ** 30 * 4 20 20 20 Relative mRNA level Relative Relative mRNA level Relative 2 10 10 10

0 0 0 0 Vec Short Long USP- A6- Vec Short Long USP- A6- Vec Long Vec Long

MC3T3 c 10 50 MMP9-1 MMP10-1 MMP9-2 MMP10-2 8 40

6 30

4 20

Relative mRNA level Relative 2 10

0 0 Dox: - + - + - + Dox: - + - + - + Vec Long USP- Vec Long USP-

d hFOB1.19 Hela 7 7 * 6 ** 6

5 5

4 4

3 3

2 2

1 1

0 MMP-9 promoter activity Relative 0 Relative MMP-9 promoter activity Relative VecLong Short VecLong Short USP- Figure 2 TRE17 induces MMP expression on a transcriptional level. (a) hFOB1.19 or (b) HeLa cells were transfected, and then serum starved for 24–48 h. MC3T3-derived cell lines (c) were treated with or without dox for 24 h, and then starved for an additional 24 h. For all three cell systems, RNA was isolated, and real-time quantitative reverse transcription PCR (RT–qPCR) was performed to monitor MMP-9 and MMP-10 mRNA, using two independent primer pairs for each (denoted MMP9-1 and -2, and MMP10-1 and -2), normalizing against glyceraldehyde 3-phosphate dehydrogenase (GAPDH). For hFOB1.19 and HeLa cells, data represent the mean±s.d. of three experiments. Samples with statistically significant differences from vector control cells are indicated (*Po0.05; **Po0.01). (d) hFOB1.19 or HeLa cells were co-transfected with the indicated TRE17 constructs and a luciferase reporter driven by a fragment of the human MMP-9 promoter encoding base pairs À674 to þ 3 relative to the transcription start site. Luciferase assays were performed; data represent the mean±s.d. of four experiments. Samples with statistically significant differences from vector control cells are indicated (*Po0.05; **Po0.01).

et al., 2006). To ascertain their roles in MMP-9 phosphorylation by extracellular regulated kinase, had induction, we first analyzed a reporter construct mut- no effect on MMP-9 secretion (Figure 3b). To confirm ated in the NF-kB response element (Liu et al., 2006). the requirement for NF-kB, we tested the effects of Loss of the NF-kB-binding site abrogated activation of three additional structurally unrelated NF-kB pathway the MMP-9 promoter by TRE17(long) in hFOB1.19 and inhibitors: wedelolactone, SC-514 and IKK inhibitor XI. HeLa cells (Figure 3a), indicating that NF-kB is the key As shown in Figure 3c, all three drugs significantly mediator of TRE17-induced MMP-9 transcription. inhibited MMP-9 production in both HeLa and MC3T3 To confirm the dominant role of NF-kB in induction cells. Together, these results confirm that induction of of endogenous MMP-9 by TRE17, we examined the MMP-9 by TRE17(long) is predominantly mediated effects of inhibitors of the NF-kB and activator protein through NF-kB. 1 pathways on MMP-9 secretion. NF-kB activation was blocked using Bay11-7082 (Pierce et al., 1997). Bay11-7082 caused a dose-dependent reduction in levels TRE17 activates NF-kB in a USP-dependent manner of MMP-9 induced by TRE17(long) in hFOB1.19 The requirement for NF-kB in MMP-9 production led and MC3T3 cells (Figure 3b). However, U0126, which us to explore whether TRE17 activates NF-kB. As inhibits activator protein 1 activation by preventing its shown in Figure 4a, TRE17 (long) potently activated an

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3623 hFOB1.19 HeLa vec MMP9 WT vec MMP9 WT MMP9/NFκB mut MMP9/NFκB mut TRE17(short) TRE17(short)

TRE17(long) ] * TRE17(long) ] **

TRE17(long)/USP- TRE17(long)/USP-

071 2 3 4 5 6 071 2 3 4 5 6 Relative MMP9 promoter activity Relative MMP9 promoter activity

vecTRE17(long) vec TRE17(long) vec Long vec Long vec Long Bay: 0 0 2.5 5 10 20 0 0 2.5 5 10 - - U U B B : Inhib

MMP9

TRE17

Actin

hFOB1.19 MC3T3

vec TRE17(long) vec TRE17(long) + - + + + + :Dox +-- W SC XI - - W SC XI : Inhib

MMP9

TRE17

Actin

HeLa MC3T3 Figure 3 NF-kB is required for induction of MMP-9 by TRE17. (a) hFOB1.19 or HeLa cells were transfected with the indicated TRE17 constructs and either a luciferase reporter driven by the WT MMP-9 promoter or a mutant in which the NF-kB response element was ablated. Luciferase assays were performed; data represent the mean±s.d. of four experiments (*Po0.05; **Po0.01). (b, left) hFOB1.19 were transfected as indicated and then starved for 24 h. (b, middle) Control or TRE17 (long)/MC3T3 cells were treated with dox and then starved in the presence of dox. Bay11-7082 was added at the indicated micromolar concentration during starvation. (b, right) Control or TRE17 (long)/MC3T3 cells were treated with Bay11-7082 (B, 10 mM) or U0126 (U, 50 mM) where indicated. For all three, MMP-9 in the CM was detected; whole cell lysates were blotted with TRE17 and actin antibodies. (c) MMP-9 was assayed as in b, except the IKK inhibitors wedelolactone (W, 50 mM), SC514 (SC, 100 mM) or IKK inhibitor XI (XI, 20 mM) were used where indicated.

NF-kB-responsive luciferase reporter in both hFOB1.19 activated by a wide variety of signaling molecules, and HeLa cells, whereas TRE17(short) and TRE17 including Rho family GTPases. During the course of (long)/USP– were attenuated by comparison. Activation our studies, we noted that cells expressing TRE17 showed of NF-kB by TRE17(long) was confirmed by electro- alterations in morphology that were consistent with phoretic mobility shift assays (Figure 4b), and by activation of RhoA (Nobes and Hall, 1995a, b). Specifi- monitoring its subcellular localization (Figure 4c). cally, they had a greatly increased number of focal NF-kB, which comprises p65 and p50 subunits, is adhesions distributed throughout the cell (data not maintained in an inactive state through its cytoplasmic shown). To determine whether RhoA was activated by retention by IkB; activation is reflected by nuclear TRE17, we performed pulldown assays using the translocation of p65/p50. In control hFOB1.19, p65 Rho-binding domain of Rhotekin (which binds to RhoA showed diffuse cytoplasmic staining and was largely in a GTP-dependent manner) as an affinity reagent. excluded from the nucleus (Figure 4c). In contrast, in Expression of TRE17 (long) in hFOB1.19 or HeLa cells expressing TRE17(long), but not TRE17(short) or induced robust activation of RhoA (Figure 5a). Activa- TRE17 (long)/USP–, p65 accumulated in the nucleus tion of RhoA was also observed in TRE17(long)/MC3T3, (Figure 4c). Similar results were observed in Hela cells in a dox- and USP-dependent manner (Figure 5a). and in the MC3T3 cell lines (data not shown). Together, Previous work has shown that RhoA potently these data confirm that TRE17 activates NF-kBina activates NF-kB, functioning through its effector kinase, USP-dependent manner. ROCK (Benitah et al., 2003; Anwar et al., 2004; Nakakuki et al., 2005; Cui et al., 2006). This led us to investigate whether RhoA and ROCK mediate TRE17- RhoA/ROCK mediate NF-kB activation and MMP-9 induced activation of NF-kB. Coexpression of domi- production by TRE17 nant-negative RhoA partially attenuated activation of We next explored the signaling mechanism by which the NF-kB-luc reporter by TRE17(long) in both TRE17 induces activation of NF-kB. NF-kBcanbe hFOB1.19 and HeLa cells (Figure 5b). Furthermore,

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3624 12 hFOB1.19 * 12 HeLa ** 10 10

8 8

6 6 *

4 4 Fold activation Fold Fold activation Fold 2 2

0 0 vecShort Long USP- vecShort Long

-TNFα +TNFα T17(long) -+ - + - + :cold oligo

HeLa

+TNFα T17(long) USP- + + - + + + : Dox - + - - + - : cold oligo

MC3T3

Figure 4 TRE17(long) activates NF-kB in a USP-dependent manner. (a) hFOB1.19 or Hela cells were co-transfected with the indicated TRE17 construct and a luciferase reporter driven by tandem NF-kB response elements. Cells were starved and luciferase assays were performed. Data represent the mean±s.d. of four independent experiments (*Po0.05; **Po0.01). (b) Electrophoretic mobility shift assays (EMSAs) were performed in transiently transfected HeLa (top) or MC3T3 lines expressing vector, TRE17(long) or TRE17(long)/USP– (bottom). As a positive control, vector control cells were treated with tumor necrosis factor-a (TNFa; 10 ng/ml, 30 min) where indicated. (c) hFOB1.19 cells were transfected with the indicated TRE17 construct, starved for 24 h, and then stained with anti-HA, anti-p65 and 4,6-diamidino-2-phenylindole (DAPI). Arrows highlight transfected cells. Scale bar, 20 mm.

treatment of cells with the ROCK inhibitor Y27632 also early as 5–7 days after injection in all nine mice injected partially inhibited NF-kB activation by TRE17(long), in with TRE17(long)/MC3T3, exclusively in the presence a dose-dependent manner (Figure 5b). of dox. Tumors reached maximal size (from 3 to 10 mm As we showed that NF-kB is essential for induction of in diameter) within 2 weeks. No palpable tumors were MMP-9, we predicted that Y27632 should also block observed in any mice injected with MC3T3 expressing MMP-9 production. Accordingly, we found that Y27632 TRE17(long)/USP–, TRE17(short) or luciferase alone. caused a dose-dependent reduction in MMP-9 levels To assess tumorigenesis quantitatively, in vivo biolumi- induced by TRE17(long)/MC3T3, as determined by nescence imaging was performed. Animals were injected immunoblotting and zymography (Figure 5c). Collec- intraperitoneally with luciferin, and then imaged using the tively, our data define a pathway in which TRE17, in a Xenogen in vivo Imaging System (Caliper Life Sciences, USP-dependent manner, activates RhoA/ROCK, leading Hopkinton, MA, USA) to detect the luciferase-expressing to NF-kB activation and resultant production of MMP-9. xenografted cells. Although all animals injected with TRE17(long)/MC3T3 cells emitted a strong biolumines- cence signal, mice injected with cells expressing TRE17 TRE17 induces formation of tumors that recapitulate (long)/USP–, TRE17(short) or luciferase alone gave sig- multiple features of ABC in a USP-dependent manner nals approximately 10-fold weaker (Figure 6a and data not In sum, our in vitro analysis shows that TRE17 is shown). These results reveal that the USP activity of sufficient to induce expression of MMPs, which we TRE17 is essential for tumorigenesis. propose contributes to ABC pathogenesis. MMPs Dissection of animals injected with TRE17(long)/ (MMP-9 in particular) have a central role in angiogen- MC3T3 cells revealed the presence of solid tumors esis and inflammation, the processes that typify ABC with a striking hemorrhagic appearance (Figure 6b). lesions. This led us to examine whether xenografts of Hematoxylin and eosin staining revealed that this TRE17(long)-expressing MC3T3 recapitulate molecular arose from extensive vascularization as well as blood features of this tumor. We elected to introduce cells into extravasation (Figure 6b), both of which are prominent mice subcutaneously rather than intraosseously to features of ABC. Lesions were composed of a spindle facilitate both delivery and monitoring of the xenograft. cell population associated with a fibrous or myxoid Subcutaneous placement of the xenograft is physiologi- matrix (Figure 6b), also highly reminiscent of ABC cally relevant as ABCs can arise in soft tissues, and are (Oliveira et al., 2004b; Shinde et al., 2006). No tumors or histologically identical to osseous cases (Nielsen et al., vascularization were observed upon dissection of mice 2002; Sukov et al., 2008). injected with control or TRE17 (long)/USP– expressing MC3T3-derived lines were injected subcutaneously MC3T3 cells (Figure 6b). Immunohistochemical ana- into flanks of nude mice, and expression of TRE17(long) lysis of serial tumor sections revealed strong MMP-9 and luciferase was induced in vivo by addition of dox to staining of the TRE17-positive cells (Figure 6c). These the drinking water. Palpable growths were visible as results not only provide in vivo confirmation of our

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3625 vec Long USP- its pathogenesis. The discovery in 2004 that transloca- vec Long vec Long - + - + - + : Dox tion of TRE17/USP6 occurs in the majority of ABC cases represented a significant advance. However, many Active RhoA key questions have remained unanswered regarding the role of TRE17 in the etiology of this tumor, including Total RhoA whether TRE17 overexpression is sufficient to initiate tumor formation, the molecular pathways it engages TRE17 and the role of its USP domain. Our study provides insights into all of these Actin key issues. We show that overexpression of TRE17 in pre-osteoblastic MC3T3 cells is sufficient to drive hFOB1.19 HeLa MC3T3 the formation of tumors that reproduce molecular and histological features of ABC. Interestingly, TRE17 hFOB1.19 HeLa 10 10 (long) tumors remained as solid masses, and did not Y27632 Y27632 form the multicystic lesions typically observed in ABC. 0 μM 0 μM 8 Thus, although TRE17 is sufficient to initiate tumor 8 10 μM μ 10 M formation, additional events may be required to produce 50 μM 50 μM 6 6 the full spectrum of characteristics of these complex lesions. Alternatively, the MC3T3 system might not 4 4 precisely mimic the cellular context of the TRE17- Fold activation Fold translocated cells in ABC. Nevertheless, our results 2 2 refute the prevailing model for the pathogenesis of ABC, 0 0 which posits that it originates from a local circulatory vec Long Long vec Long Long disturbance, with resultant tissue damage. Our data +RhoA N19 +RhoA N19 instead suggest a direct role for TRE17 in establishing a degradative and vascularized microenvironment. vec TRE17(long) This work further identifies MMPs as novel effectors 0 2 10 50 0 2 10 50 : Y27632 of TRE17. Previous immunohistochemistry showed that MMP9 MMP-9 is expressed in the fibrous stroma of ABC, but the cells responsible for its production had not been zymography determined (Kumta et al., 2003). Furthermore, it was TRE17 unclear whether MMP-9 had a causal role in ABC pathogenesis, or was produced secondarily as a result of Actin tissue damage and inflammation. Our finding that TRE17 directly induces its production supports the Figure 5 RhoA and ROCK partially mediate activation of NF-kB former scenario, and suggests that MMP-9 is a relevant by TRE17. (a) RhoA activity was measured in transiently in vivo transfected hFOB1.19 (left), HeLa (middle) and MC3T3 cell lines effector of TRE17 . (right). Cell extracts were subjected to pulldowns using a There are multiple means by which MMP-9 and glutathione-S-transferase (GST) fusion of the RhoAGTP-binding MMP-10 could contribute to ABC pathogenesis. Both domain (RBD) of Rhotekin, followed by blotting with anti-RhoA. of these MMPs have previously been implicated in Active RhoA, RBD pulldowns; total RhoA, whole cell lysates. remodeling/degradation of bone matrix, under physio- (b) hFOB1.19 or HeLa were transfected with NF-kB luciferase and TRE17(long) in the absence or presence of dominant-negative logical and pathological conditions (Bord et al., 1998; RhoA (RhoAN19) or Y27632, and subjected to luciferase assays. Giambernardi et al., 1998; Engsig et al., 2000; Sternlicht (c) Control or TRE17 (long)/MC3T3 cells were treated with dox and Werb, 2001; Onodera et al., 2002; Tolboom et al., for 24 h, and then starved with dox and the indicated concentration 2002; Kumta et al., 2003; Haeusler et al., 2005; Barksby of Y27632 (mM) for 24 h. MMP-9 in the CM was detected by et al. et al. immunoblotting and zymography. Whole cell extracts were blotted , 2006; Kanangat , 2006). MMP-9 and to confirm TRE17 and actin levels. MMP-10 have multiple substrates including other MMPs, which induce their zymogenic activation (Stern- in vitro results, but also recapitulate the MMP-9 licht and Werb, 2001; Van den Steen et al., 2002; enrichment observed in the stroma of ABC tumors Krampert et al., 2004), potentially further contributing (Kumta et al., 2003). Thus, our data show that this to the degradative microenvironment in ABC. They osteoblastic model reproduces multiple features of ABC, can also proteolyze components of the ECM, which and strongly validate its use for dissecting the molecular serves not only to degrade the ECM but also to functions of TRE17 in ABC pathogenesis. liberate matrix-bound growth factors that further promote inflammation and angiogenesis (Sternlicht and Werb, 2001). For example, numerous studies indicate that MMP-9 has an important role in the Discussion release of vascular endothelial growth factor from bone and cartilage matrix to promote angiogenesis Although ABC was first described over 60 years ago, and vascular permeability (Vu et al., 1998; John and remarkably little progress has been made in elucidating Tuszynski, 2001; Rundhaug, 2005; Ghajar et al., 2008).

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3626

220 600 3000 200 500 180 2500 160 2000 400 140 1500 300 120 100 1000 200 80 500 100

vector TRE17(long) TRE17(long)/USP-

vector TRE17

TRE17(long) MMP-9 Figure 6 TRE17(long) induces formation of tumors that recapitulate multiple features of ABC in a USP-dependent manner. (a) In vivo bioluminescence imaging of mice injected with MC3T3 cell lines expressing TRE17(long), TRE17(long)/USP– or luciferase alone. (b, left) Macroscopic images of tumor and injection site of mice injected with vector or TRE17 (long)-expressing MC3T3. (b, right) Hematoxylin and eosin (H&E) stain of TRE17(long) tumor section. Arrowheads highlight vascular elements. (c) Immunohistochemistry was performed on serial sections of TRE17(long)-induced tumor using antibodies against TRE17 and MMP-9. Nuclear staining for TRE17 is nonspecific; only central cells show specific cytoplasmic staining of TRE17.

MMP-9 has also been proposed to contribute to of the NF-kB regulatory pathway. NFkB activation is angiogenesis by degrading the endothelial and inter- exquisitely controlled by a complex array of ubiquitina- stitial matrix to facilitate endothelial cell migration and tion and phosphorylation events (Hacker and Karin, branching. Thus, the known functions of MMP-9 are 2006; Sebban et al., 2006). Notably, three other USPs, fully consistent with the histological features of ABC, CYLD, A20 and USP11, have previously been linked which contain numerous capillaries and extravasated with NF-kB regulation (Brummelkamp et al., 2003; blood. Future studies aimed at blocking MMP-9 Kovalenko et al., 2003; Trompouki et al., 2003; Wertz expression/activity will determine whether it is indeed et al., 2004; Yamaguchi et al., 2007). However, in required for the extensive vascularization observed in contrast to TRE17, they function as negative regulators our xenograft model. of NF-kB, suggesting that TRE17 targets a distinct Our studies further dissect the mechanism by which substrate from these other USPs. TRE17 induces MMP expression. Induction of MMP-9 In conclusion, this work defines a signaling pathway, and MMP-10 is dependent on the USP activity of TRE17 4 RhoA/ROCK 4 NF-kB 4 MMP-9, which TRE17, but not its ability to bind Arf6. Furthermore, we propose contributes to ABC pathogenesis. Our NF-kB is required for transcriptional induction of results show that the USP activity of TRE17 and MMP-9. Activation of NF-kB is also dependent on NF-kB are required for MMP-9 induction, raising the the USP activity of TRE17, and is partially mediated possibility that development of USP-specific inhibitors through RhoA/ROCK. The substrate(s) of the USP or NF-kB antagonists might be effective novel strategies domain of TRE17 and the mechanism by which it leads for the treatment of these tumors. to activation of RhoA remain to be determined. Previous work showed that RhoA can be ubiquitinated to trigger its proteasomal degradation (Wang et al., 2003, 2006; Bryan et al., 2005). However, it is unlikely Materials and methods that TRE17 functions by de-ubiquitinating and stabiliz- ing RhoA, as total RhoA levels are unaffected by Cell culture TRE17. A more likely possibility is that the USP HeLa cells were grown in Dulbecco’s modified Eagle’s medium domain of TRE17 targets a guanine nucleotide exchange containing 10% fetal bovine serum, penicillin and streptomycin, factor for RhoA to rescue it from proteasomal sodium pyruvate and GlutaMax (Invitrogen, Carlsbad, CA, USA). The murine pre-osteoblast cell line MC-3T3-E1 (Clone 14) degradation. (referred to hereafter as MC3T3) was obtained from American As inhibition of RhoA/ROCK only partially attenu- Type Culture Collection (Manassas, VA, USA), and grown in ated activation of NF-kB by TRE17, this suggests that a-modified Eagle’s media with tetracyclin-free fetal bovine serum, TRE17 functions through additional parallel path- penicillin and streptomycin, sodium pyruvate, GlutaMax and ways, in a manner that requires its USP activity. One G418 (400 mg/ml). Both lines were maintained at 37 1Cin5% possibility is that it might de-ubiquitinate a component CO2. Human fetal osteoblast hFOB1.19 cells were grown in 1:1

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3627 Dulbecco’s modified Eagle’s medium without phenol red:Ham’s luciferase reporter was used for normalization. Hemagglutinin- F12, with the same supplements and G418 (300 mg/ml active), tagged TRE17 constructs were described previously (Masuda- andkeptat341Cin5%CO2. Robens et al., 2003; Martinu et al., 2004; Shen et al., 2005). Cells were seeded in 35 mm wells, transfected with Lipofecta- Antibodies and reagents mine 2000 (Invitrogen) or FuGENE6 (Roche, Indianapolis, Antibodies for hemagglutinin, actin, p65 and RhoA were from IN, USA), starved and then treated with drugs as indicated. Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA), Anti- Dual luciferase assays were performed according to the human MMP-9 (AF911), anti-mouse MMP-9 (MAB9091) manufacturer’s instructions (Promega, Madison, WI, USA). and anti-human MMP-10 (MAB9101) were from R&D Evaluations of statistical significance were conducted using Systems (Minneapolis, MN, USA). Gelatin-agarose was from Student’s t-test. A two-tailed P-value of o0.05 was considered Sigma (St Louis, MO, USA). TRE17 was detected using statistically significant. Electrophoretic mobility shift assay anti-hemagglutinin, or an antibody generated against the was performed as previously described (Solt et al., 2007). N-terminus of TRE17. Y27632, Bay11-7082, wedelolactone, SC514 and IKK inhibitor XI (PS1145) were from EMD Immunofluorescence microscopy Biosciences (Gibbstown, NJ, USA). Cells were seeded on coverslips, transfected using FuGENE6 and then treated as described previously (Masuda-Robens Generation of MC3T3 cell lines et al., 2003). Samples were viewed on a Zeiss confocal TRE17 complementary DNAs (TRE17(long), TRE17(short), microscope (Thornwood, NJ, USA) with LSM510 soft- TRE17(long)/USP– and TRE17(A6–)) were subcloned into ware, using excitation wavelengths of 488 nm (fluorescein pBI-L (Clontech, Mountain View, CA, USA), which contains isothiocyanate), 546 nm (Cy3) or 358 nm (4,6-diamidino-2- a bi-directional tetracyclin/dox-inducible promoter that drives phenylindole). simultaneous expression of TRE17 and firefly luciferase. Resulting plasmids were AseI digested and then ligated to RhoA activity and TRE17-Arf6 binding assays AseI-digested pN1b-actin-rtTA2S-IRES-EGFP (Welman For RhoA activity assays, cells were lysed in Rho-binding et al., 2006). As negative control, pBI-L encoding luciferase domain buffer (20 mM HEPES pH 7.5, 1% Nonidet P-40, 10% alone was ligated to pN1b-actin-rtTA2S-M2-IRES-GFP to glycerol, 1 mM EDTA, 150 mM NaCl, 10 mM MgCl2,1mM generate pSMV (super module vector) (Welman et al., 2006). Na3VO4,1mM phenylmethylsulfonyl fluoride, pepstatin, leu- MC3T3 cells were transfected with the pSMV-derived con- peptin and aprotinin). Extracts were precipitated with 20 mgof structs, and then selected in medium containing tetracyclin- glutathione-S-transferase fused to the Rho-binding domain of free fetal bovine serum and G418 (400 mg/ml). Green Rhotekin conjugated to agarose for 2 h, washed three times in fluorescent protein-positive cells were isolated from the Rho-binding domain buffer and then immunoblotted with G418-resistant pool by flow cytometry. RhoA antibody. Association of TRE17 and Arf6 was monitored as described (Martinu et al., 2004). Detection of MMPs from conditioned medium Cells were starved under serum-free conditions for 24–48 h Xenografting, bioluminescence imaging and histology before harvesting. CM was subjected to low-speed centrifuga- All mouse procedures were in accordance with the institutional tion to remove cell debris. To monitor MMP-9, CM was animal care and use committee of the University of Pennsyl- supplemented with protease inhibitors, incubated with gelatin- vania. Nude mice (Jackson Laboratories, Bar Harbor, ME, agarose beads (Sigma) at 4 1C for 6 h, washed three times in USA) were fed water with dox (1 mg/ml containing 5% phosphate-buffered saline and then subjected to immunoblot- sucrose) for 1 to 2 weeks before xenografting. MC3T3 cell lines ting. For zymography assays, gelatin-agarose-purified MMP-9 were pre-treated with dox 1 day before subcutaneous injection was fractionated on 8% gels containing 30 mg/ml gelatin (2.5E6 cells per site). For dox-treated animals, nine mice were (Sigma). Gels were washed with 2.5% Triton X-100, incubated injected per cell line; dox was maintained in water for the in substrate buffer (50mM Tris-HCl pH 8.0, 5 mM CaCl2)at duration of the experiment. For animals not treated with dox, 37 1C for 36 h and then stained with Coommassie Brilliant three mice were injected per cell line. Blue. To detect MMP-10, CM was precipitated with 10% For bioluminescence imaging, mice were anesthetized with trichloroacetic acid. Pellets were washed twice in cold acetone, Avertin. Animals were injected intraperitoneally with D- air dried, and then fractionated by sodium dodecyl sulfate– luciferin (Biotium, Hayward, CA, USA; 120 ml at 25 mg/ml), polyacrylamide gel electrophoresis. and imaged within 8–20 min using the Xenogen in vivo Imaging System. Tumors were paraffin embedded and histological analysis was performed by the Abramson Cancer Center Reverse transcription–quantitative PCR Pathology Core of the University of Pennsylvania. Sections RNA was extracted with Trizol (Invitrogen). Reverse trans- were subjected to immunohistochemistry using antibodies cription reactions were performed on 1–2 mgoftotalRNAusing against TRE17 or MMP-9 (AF909; R&D Systems). a Superscript First-Strand Synthesis System kit (Invitrogen). Real-time PCR was performed using TaqMan probes on an ABI Prism 7900 system (Foster City, CA, USA). Relative Conflict of interest expression levels were determined from a standard curve of serial dilutions of complementary DNA samples and normal- ized to glyceraldehyde 3-phosphate dehydrogenase levels. The authors declare no conflict of interest. Primer sequences are available upon request. Acknowledgements Luciferase assays and electrophoretic mobility shift assays NF-kB-dependent pBIIx (Zhong et al., 1997) and human This work was supported by the NIH Grants CA081415 MMP-9 promoter (Liu et al., 2006) reporter constructs have (MMC), DK58044 (GAB) and P30 AR050950 (Penn Center been previously described. b-actin promoter-driven Renilla for Musculoskeletal Disorders). Additional support was from

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3628 the Mayo Foundation CR20 Program (AMO), the Pennsyl- cally disclaims responsibility for any analysis, interpretations vania Muscle Institute Pilot Project Grant (MMC) and the or conclusions. We thank Yvette Liu for invaluable assistance Pennsylvania Department of Health (MMC), which specifi- with bioluminescence imaging.

References

Althof PA, Ohmori K, Zhou M, Bailey JM, Bridge RS, Nelson M Haeusler G, Walter I, Helmreich M, Egerbacher M. (2005). Localiza- et al. (2004). Cytogenetic and molecular cytogenetic findings in 43 tion of matrix metalloproteinases, (MMPs) their tissue inhibitors, aneurysmal bone cysts: aberrations of 17p mapped to 17p13.2 by and vascular endothelial growth factor (VEGF) in growth plates of fluorescence in situ hybridization. Mod Pathol 17: 518–525. children and adolescents indicates a role for MMPs in human Amerik AY, Hochstrasser M. (2004). Mechanism and function of postnatal growth and skeletal maturation. Calcif Tissue Int 76: deubiquitinating enzymes. Biochim Biophys Acta 1695: 189–207. 326–335. Anwar KN, Fazal F, Malik AB, Rahman A. (2004). RhoA/Rho- Han S, Ritzenthaler JD, Sitaraman SV, Roman J. (2006). Fibronectin associated kinase pathway selectively regulates thrombin-induced increases matrix metalloproteinase 9 expression through activation intercellular adhesion molecule-1 expression in endothelial cells of c-Fos via extracellular-regulated kinase and phosphatidylinositol via activation of I kappa B kinase beta and phosphorylation of 3-kinase pathways in human lung carcinoma cells. J Biol Chem 281: RelA/p65. J Immunol 173: 6965–6972. 29614–29624. Barksby HE, Milner JM, Patterson AM, Peake NJ, Hui W, Robson T John A, Tuszynski G. (2001). The role of matrix metalloproteinases et al. (2006). Matrix metalloproteinase 10 promotion of collageno- in tumor angiogenesis and tumor metastasis. Pathol Oncol Res 7: lysis via procollagenase activation: implications for cartilage 14–23. degradation in arthritis. Arthritis Rheum 54: 3244–3253. Kanangat S, Postlethwaite A, Hasty K, Kang A, Smeltzer M, Appling Benitah SA, Valeron PF, Lacal JC. (2003). ROCK and nuclear factor- W et al. (2006). Induction of multiple matrix metalloproteinases in kappaB-dependent activation of cyclooxygenase-2 by Rho human dermal and synovial fibroblasts by Staphylococcus aureus: GTPases: effects on tumor growth and therapeutic consequences. implications in the pathogenesis of septic arthritis and other soft Mol Biol Cell 14: 3041–3054. tissue infections. Arthritis Res Ther 8: R176. Bernards A. (2003). GAPs galore! A survey of putative Ras Kim KC, Lee CH. (2005). MAP kinase activation is required for superfamily GTPase activating proteins in man and Drosophila. the MMP-9 induction by TNF-stimulation. Arch Pharm Res 28: Biochim Biophys Acta 1603: 47–82. 1257–1262. Bizimungu C, Thomas A, Brasseur R, Vandenbol M. (2007). Komander D, Clague MJ, Urbe S. (2009). Breaking the chains: Mutational analysis of the TRE2 oncogene encoding an inactive structure and function of the deubiquitinases. Nat Rev Mol Cell Biol RabGAP. Biotechnol Lett 29: 1927–1937. 10: 550–563. Bizimungu C, Vandenbol M. (2005). At least two regions of the Kovalenko A, Chable-Bessia C, Cantarella G, Israel A, Wallach D, oncoprotein Tre2 are involved in its lack of GAP activity. Biochem Courtois G. (2003). The tumour suppressor CYLD negatively Biophys Res Commun 335: 883–890. regulates NF-kappaB signalling by deubiquitination. Nature 424: Bord S, Horner A, Hembry RM, Compston JE. (1998). Stromelysin-1 801–805. (MMP-3) and stromelysin-2 (MMP-10) expression in developing Krampert M, Bloch W, Sasaki T, Bugnon P, Rulicke T, Wolf E et al. human bone: potential roles in skeletal development. Bone 23: 7–12. (2004). Activities of the matrix metalloproteinase stromelysin-2 Brummelkamp TR, Nijman SM, Dirac AM, Bernards R. (2003). Loss (MMP-10) in matrix degradation and keratinocyte organization in of the cylindromatosis tumour suppressor inhibits apoptosis by wounded skin. Mol Biol Cell 15: 5242–5254. activating NF-kappaB. Nature 424: 797–801. Kumta SM, Huang L, Cheng YY, Chow LTC, Lee KM, Zheng MH. Bryan B, Cai Y, Wrighton K, Wu G, Feng XH, Liu M. (2005). (2003). Expression of VEGF and MMP-9 in giant cell tumor Ubiquitination of RhoA by Smurf1 promotes neurite outgrowth. of bone and other osteolytic lesions. Life Sci 73: 1427–1436. FEBS Lett 579: 1015–1019. Li M, Chen D, Shiloh A, Luo J, Nikolaev AY, Qin J et al. (2002). Cho A, Graves J, Reidy MA. (2000). Mitogen-activated protein Deubiquitination of p53 by HAUSP is an important pathway for kinases mediate matrix metalloproteinase-9 expression in vascular p53 stabilization. Nature 416: 648–653. smooth muscle cells. Arterioscler Thromb Vasc Biol 20: 2527–2532. Liu J, Zhan M, Hannay JA, Das P, Bolshakov SV, Kotilingam D et al. Cottalorda J, Bourelle S. (2007). Modern concepts of primary (2006). Wild-type p53 inhibits nuclear factor-kappaB-induced aneurysmal bone cyst. Arch Orthop Trauma Surg 127: 105–114. matrix metalloproteinase-9 promoter activation: implications Cui R, Tieu B, Recinos A, Tilton RG, Brasier AR. (2006). RhoA for soft tissue sarcoma growth and metastasis. Mol Cancer Res 4: mediates angiotensin II-induced phospho-Ser536 nuclear factor 803–810. kappaB/RelA subunit exchange on the interleukin-6 promoter in Martinu L, Masuda-Robens JM, Robertson SE, Santy LC, Casanova VSMCs. Circ Res 99: 723–730. JE, Chou MM. (2004). The TBC (Tre-2/Bub2/Cdc16) domain Engsig MT, Chen QJ, Vu TH, Pedersen AC, Therkidsen B, Lund LR protein TRE17 regulates plasma membrane-endosomal trafficking et al. (2000). Matrix metalloproteinase 9 and vascular endothelial through activation of Arf6. Mol Cell Biol 24: 9752–9762. growth factor are essential for osteoclast recruitment into develop- Masuda-Robens JM, Kutney SN, Qi H, Chou MM. (2003). The ing long bones. J Cell Biol 151: 879–889. TRE17 oncogene encodes a component of a novel effector pathway Genersch E, Hayess K, Neuenfeld Y, Haller H. (2000). Sustained ERK for Rho GTPases Cdc42 and Rac 1 and stimulates actin remodeling. phosphorylation is necessary but not sufficient for MMP-9 Mol Cell Biol 23: 2151–2161. regulation in endothelial cells: involvement of Ras-dependent and Mott JD, Werb Z. (2004). Regulation of matrix biology by matrix -independent pathways. J Cell Sci 113(Pt 23): 4319–4330. metalloproteinases. Curr Opin Cell Biol 16: 558–564. Ghajar CM, George SC, Putnam AJ. (2008). Matrix metalloproteinase Nakakuki T, Ito M, Iwasaki H, Kureishi Y, Okamoto R, Moriki N control of capillary morphogenesis. Crit Rev Eukaryot Gene Expr et al. (2005). Rho/Rho-kinase pathway contributes to C-reactive 18: 251–278. protein-induced plasminogen activator inhibitor-1 expression in Giambernardi TA, Grant GM, Taylor GP, Hay RJ, Maher VM, endothelial cells. Arterioscler Thromb Vasc Biol 25: 2088–2093. McCormick JJ et al. (1998). Overview of matrix metalloproteinase Nakamura T, Hillova J, Mariage-Samson R, Onno M, Huebner K, expression in cultured human cells. Matrix Biol 16: 483–496. Cannizzaro LA et al. (1992). A novel transcriptional unit of the tre Hacker H, Karin M. (2006). Regulation and function of IKK and oncogene widely expressed in human cancer cells. Oncogene 7: IKK-related kinases. Sci STKE 2006: re13. 733–741.

Oncogene TRE17/USP6 induces MMP through NF-jB YYeet al 3629 Nielsen GP, Fletcher CD, Smith MA, Rybak L, Rosenberg AE. Sternlicht MD, Werb Z. (2001). How matrix metalloproteinases (2002). Soft tissue aneurysmal bone cyst: a clinicopathologic study regulate cell behavior. Annu Rev Cell Dev Biol 17: 463–516. of five cases. Am J Surg Pathol 26: 64–69. Subramaniam M, Jalal SM, Rickard DJ, Harris SA, Bolander ME, Nobes CD, Hall A. (1995a). Rho, rac and cdc42 GTPases: regulators Spelsberg TC. (2002). Further characterization of human fetal of actin structures, cell adhesion and motility. Biochem Soc Trans osteoblastic hFOB 1.19 and hFOB/ER alpha cells: bone formation 23: 456–459. in vivo and karyotype analysis using multicolor fluorescent in situ Nobes CD, Hall A. (1995b). Rho, rac, and cdc42 GTPases regulate the hybridization. J Cell Biochem 87: 9–15. assembly of multimolecular focal complexes associated with actin Sukov WR, Franco MF, Erickson-Johnson M, Chou MM, Unni KK, stress fibers, lamellipodia, and filopodia. Cell 81: 53–62. Wenger DE et al. (2008). Frequency of USP6 rearrangements in Oliveira AM, Hsi BL, Weremowicz S, Rosenberg AE, Dal Cin P, myositis ossificans, brown tumor, and cherubism: molecular Joseph N et al. (2004a). USP6 (Tre2) fusion oncogenes in cytogenetic evidence that a subset of ‘myositis ossificans-like lesions’ aneurysmal bone cyst. Cancer Res 64: 1920–1923. are the early phases in the formation of soft-tissue aneurysmal bone Oliveira AM, Perez-Atayde AR, Dal Cin P, Gebhardt MC, Chen CJ, cyst. Skeletal Radiol 37: 321–327. Neff JR et al. (2005). Aneurysmal bone cyst variant translocations Tolboom TC, Pieterman E, van der Laan WH, Toes RE, upregulate USP6 transcription by promoter swapping with Huidekoper AL, Nelissen RG et al. (2002). Invasive properties of the ZNF9, COL1A1, TRAP150, and OMD . Oncogene 24: fibroblast-like synoviocytes: correlation with growth characteristics 3419–3426. and expression of MMP-1, MMP-3, and MMP-10. Ann Rheum Dis Oliveira AM, Perez-Atayde AR, Inwards CY, Medeiros F, Derr V, 61: 975–980. Hsi BL et al. (2004b). USP6 and CDH11 oncogenes identify Trompouki E, Hatzivassiliou E, Tsichritzis T, Farmer H, Ashworth A, the neoplastic cell in primary aneurysmal bone cysts and are absent Mosialos G. (2003). CYLD is a that in so-called secondary aneurysmal bone cysts. Am J Pathol 165: negatively regulates NF-kappaB activation by TNFR family 1773–1780. members. Nature 424: 793–796. Onodera S, Nishihira J, Iwabuchi K, Koyama Y, Yoshida K, Tanaka S Van den Steen PE, Dubois B, Nelissen I, Rudd PM, Dwek RA, et al. (2002). Macrophage migration inhibitory factor up-regulates Opdenakker G. (2002). Biochemistry and molecular biology of matrix metalloproteinase-9 and -13 in rat osteoblasts. Relevance to gelatinase B or matrix metalloproteinase-9 (MMP-9). Crit Rev intracellular signaling pathways. JBiolChem277: 7865–7874. Biochem Mol Biol 37: 375–536. Panoutsakopoulous G, Pandis N, Kyriazoglou I, Gustafson P, Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA, Hanahan D Mertens F, Mandahl N. (1999). Recurrent t{16:17}{q22;p13} in et al. (1998). MMP-9/gelatinase B is a key regulator of growth plate aneurysmal bone cysts. Genes Cancer 26: 265–266. angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93: Papa FR, Hochstrasser M. (1993). The yeast DOA4 gene encodes a 411–422. deubiquitinating enzyme related to a product of the human tre-2 Wang D, Christensen K, Chawla K, Xiao G, Krebsbach PH, oncogene. Nature 366: 313–319. Franceschi RT. (1999). Isolation and characterization of MC3T3- Paulding CA, Ruvolo M, Haber DA. (2003). The Tre2 (USP6) E1 preosteoblast subclones with distinct in vitro and in vivo oncogene is a hominoid-specific gene. Proc Natl Acad Sci USA 100: differentiation/mineralization potential. J Bone Miner Res 14: 2507–2511. 893–903. Pierce JW, Schoenleber R, Jesmok G, Best J, Moore SA, Collins T Wang HR, Ogunjimi AA, Zhang Y, Ozdamar B, Bose R, Wrana JL. et al. (1997). Novel inhibitors of cytokine-induced IkappaBalpha (2006). Degradation of RhoA by Smurf1 . Meth phosphorylation and endothelial cell adhesion molecule Enzymol 406: 437–447. expression show anti-inflammatory effects in vivo. J Biol Chem Wang HR, Zhang Y, Ozdamar B, Ogunjimi AA, Alexandrova E, 272: 21096–21103. Thomsen GH et al. (2003). Regulation of cell polarity and Rundhaug JE. (2005). Matrix metalloproteinases and angiogenesis. protrusion formation by targeting RhoA for degradation. Science J Cell Mol Med 9: 267–285. 302: 1775–1779. Saccomanni B. (2008). Aneurysmal bone cyst of spine: a review of Welman A, Cawthorne C, Ponce-Perez L, Barraclough J, Danson S, literature. Arch Orthop Trauma Surg 128: 1145–1147. Murray S et al. (2006). Increases in c-Src expression level and Sebban H, Yamaoka S, Courtois G. (2006). Posttranslational activity do not promote the growth of human colorectal carcinoma modifications of NEMO and its partners in NF-kappaB signaling. cells in vitro and in vivo. Neoplasia 8: 905–916. Trends Cell Biol 16: 569–577. Wertz IE, O’Rourke KM, Zhou H, Eby M, Aravind L, Seshagiri S Shen C, Ye Y, Robertson SE, Lau AW, Mak DO, Chou MM. (2005). et al. (2004). De-ubiquitination and ubiquitin ligase domains of A20 Calcium/calmodulin regulates ubiquitination of the ubiquitin- downregulate NF-kappaB signalling. Nature 430: 694–699. specific protease TRE17/USP6. J Biol Chem 280: 35967–35973. Yamaguchi T, Kimura J, Miki Y, Yoshida K. (2007). The Shinde A, Mehlman CT, Collins MH. (2006). Aneurysmal bone cysts deubiquitinating enzyme USP11 controls an IkappaB kinase alpha express vascular markers. Pediatr Dev Pathol 9: 38–43. (IKKalpha)-p53 signaling pathway in response to tumor necrosis Singhal S, Taylor MC, Baker RT. (2008). Deubiquitylating enzymes factor alpha (TNFalpha). J Biol Chem 282: 33943–33948. and disease. BMC Biochem 9(Suppl 1): S3. Zhong H, SuYang H, Erdjument-Bromage H, Tempst P, Ghosh S. Solt LA, Madge LA, Orange JS, May MJ. (2007). Interleukin- (1997). The transcriptional activity of NF-kappaB is regulated by 1-induced NF-kappaB activation is NEMO-dependent but does not the IkappaB-associated PKAc subunit through a cyclic AMP- require IKKbeta. J Biol Chem 282: 8724–8733. independent mechanism. Cell 89: 413–424.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene