Ubiquitin C-terminal hydrolase-L3 regulates Smad1 ubiquitination and osteoblast differentiation

Abstract

UCHL3 physically interacts with SMAD1 by anti tag coimmunoprecipitation (View interaction) The process of ubiquitination/deubiquitination is a pivotal mechanism of controlling protein action in various biological processes and diseases [1-5]. The loss of function in ubiquitin–proteasome system-related proteins results in the development of various diseases including neurodegenerative disease, cancers, and osteoporosis [1, 2, 6-8]. Deubiquitination, the removal of covalently attached ubiquitin from the target protein, is catalyzed by deubiquitinating enzymes to regulate the abundance and/or functional activity of target proteins. These enzymes consist of four groups, ubiquitin C-terminal hydrolases (UCHs), ubiquitin-specific proteases (USPs), otubain domain-containing proteins (OTUs), and Machado-Joseph domain (Josephin domain)-containing proteins (MJD) [9]. The UCHs family includes Uch-L1, Uch-L3, Uch-L5, and BAP1. Uch-L3 is expressed in various tissues [10]. The Uch-L3 gene is located on mouse chromosome 14 (human chromosome 13). Bone morphogenetic proteins (BMPs) play important roles in skeletal development and bone formation. In vivo bone formation is induced by recombinant human BMP2 and BMP7 [11]. BMP2 and BMP7 initiate signal transduction by binding to BMP receptor type I and II, and cause the phosphorylation and activation of Smad 1/5/8. The activated Smads heterodimerize with Smad4 and translocate from the cytosol to the nucleus leading to expression of the down-stream target genes. In TGFβ signaling, two known deubiquitinating (DUB) enzymes, Usp9X (FAM) and Uch-L5 (Uch37) are involved in the stability and activation of target proteins [12, 13]. In Wnt/β-catenin signaling, Uch-L1 positively regulates the process by the increased stability of β-catenin [14]. The regulation of ubiquitination and/or deubiquitination is also critical in the regulation of cellular metabolism [15-17]. The loss of Uch-L3 expression showed retinal degeneration, skeletal muscle degeneration, and weakness of testis following cryptorchid injury [18-20]. Two recent reports suggest a function for Uch-L3 in obesity and insulin signaling [21, 22]. Furthermore, the induction of the stress response was observed in the mouse embryonic fibroblast (MEF) cells from Uch-L3 deficient mice; the MEF cells showed up-regulation of cleaved ATF6, Grp78, and some HSPs and the accumulation of poly-ubiquitinated proteins [23]. It suggests the defective role of Uch-L3's deubiquitinating activity might cause the increase of poly-ubiquitination on target proteins. Uch-L3 was also shown to be involved in oncogenesis [24-26]. In our recent proteomic study, Uch-L3 was found to be a regulator for identifying the specific targets in bone morphogenetic protein (BMP) 2-induced osteoblast differentiation [27]. There is limited information on the function of Uch-L3 in bone metabolism. Therefore, the current study focuses on the function of Uch-L3 in BMP2-induced osteoblast differentiation. The mouse premyoblast C2C12 cells and the human embryonic kidney HEK293 cells were maintained as previously described [28, 29]. The primary osteoblast cells were isolated from mouse long bone according to the previous reports [30, 31]. The construct of pCMV2-Flag-tagged Smad1 and pCI-neo-Flag-Uch-L3 were gifts from Dr. T. Katagiri (Department of Biochemistry, School of Dentistry, Showa University, Tokyo, Japan) and Dr. K. Wada (Department of Degenerative Neurological Diseases, National Institute of Neuroscience, Tokyo, Japan), respectively. The retroviral expression vector for Uch-L3 plasmid 22564 and pCMV5-HA-Smad1 plasmid 14956 were purchased from Addgene [32, 33]. Anti-Uch-L3, anti-Samd1, anti-phospho-Smad 1/5/8 antibodies were from Cell Signaling (MA, USA) and anti-HA, anti-Flag, anti-β-actin antibodies, and alkaline phosphatase staining kit were from Sigma–Aldrich (MO, USA). The total RNA was extracted by TRIZOL reagent (Invitrogen, CA, USA) according to the manufacturer's instructions. The real-time PCR was performed with the DNA Engine OPTICON 2 system (MJ research, MA, USA) and Go Taq Flexi DNA polymerase (Promega, WI, USA) used for PCR reactions as described previously [33, 34]. The sequences of primers used in this study are 5′-TGAAGGTCAGACTGAGGCACC-3′ and 5′-AATTGGAAATGGTTTCCGTCC-3′ for Uch-L3, 5′-GATAACGAGATGCCACCAGAGG-3′ and 5′-TCCACGTCGGTTCTGTTCTTC-3′ for Alp, 5′-TGCACCCAGATCCTATAGCC-3′ and 5′-CTCCATCGTCATCATCATCG-3′ for Opn, 5′-GAGTCTGACAAAGCCTTCA-3′ and 5′-AGCCATACTGGTCTGATAG-3′ for Oc, 5′-ACAGCGGGCGAGGTGGTACTTG-3′ and 5′-TCCGGTGGTCCCGACTTCAGA-3′ for ID1, 5′-GCCCAGCACCCAGC GACAA-3′ and 5′-CGGCACCTCCAGCCGGTC-3′ for Noggin, and 5′-GGCATTGCTCTCAATGACAA-3′ and 5′-TGTGAGGGAGATGCTCAGTG-3′ for Gapdh. Uch-L3 was overexpressed in C2C12 cells by retroviral transduction using previously described method [35]. Briefly, the HEK 293 cells were co-transfected with the retroviral Uch-L3 vector, pGag, and pVSV, and then incubated for 48 h. The viral supernatant were collected and incubated in C2C12 cells. The cells were selected 24 h after the transduction by puromycin (4 μg/ml) for 5 days. The mouse Uch-L3 siRNA and scrambled siRNA (negative control) were purchased from Bioneer Co. (Daejeon, Korea). The duplex was transiently transfected into C2C12 cells with HiPerfect transfection reagent (Qiagen, CA, USA) according to the manufacture's protocol. Briefly, C2C12 cells were seeded on six-well plates. Each siRNA (200 nM) was transfected and incubated for 24 h, and then the media was changed before the treatment of BMP2 (100 ng/ml, jPC, Incheon, Korea). The cells were collected after 24 h for RNA isolation. The cells were lysed with a buffer containing 10 mM Tris, 200 mM NaCl, 5 mM MgCl2, 10 mM EDTA, 10% glycerol, and a protease inhibitor cocktail (Roche, Mannheim, Germany) [36]. The lysates were immunoprecipitated with either anti-Flag antibody or anti-HA antibody, and the eluents were loaded on a 12% SDS–PAGE gel. The blots were incubated overnight at 4 °C with either anti-HA antibody or anti-Flag antibody. After washing in TBS containing 0.05% Tween 20, the blots were incubated for 1 h at room temperature with anti-mouse IgG horseradish peroxidase-linked secondary antibody. The protein bands were visualized by ECL (Amersharm Pharmacia Biotech, NJ, USA). The HEK 293 cells were transiently transfected with the plasmids, pcDNA-HA/Flag-Smad1, pcDNA-Flag/HA-Ub, and/or pcDNA-Flag-Uch-L3. After 16 h, the transfected cells were treated with the proteasome inhibitor MG-132 (10 μM) for an additional 8 h. The cell lysates were immunoprecipitated with anti-HA antibody overnight at 4 °C. After 3 times washing, the eluents were immunoblotted against anti-Flag or Smad1 antibodies. ALP staining was performed as previously reported [37]. Briefly, osteogenic transdifferentiation was induced in C2C12 cells for 24 h with BMP2 stimulation. The cells were then washed with 1× PBS, fixed with 2% PFA, and stained for alkaline phosphatase. The C2C12 cells were seeded on cover glass in six-well plate. The mouse Uch-L3 siRNA and scrambled siRNA (negative control) were transfected. After 24 h of transfection, the cells were treated with BMP2 (100 ng/ml) for 3 h and fixed in 4% paraformaldehyde solution for 1 h. After the fixation, the cells were incubated in the freezer in the presence of ice-cold 100% EtOH. After the blocking with 3% BSA, the anti-phospho Smad1 antibody (1:500) was incubated overnight at 4 °C. Then, the cells were incubated in the anti-rabbit alexa-Fluor-555-conjugated secondary antibody (1:500) in the dark. The counterstaining was performed with DAPI for visualization of nuclei (Vector Lab., Burlingame, CA). The 1× PBS was used in the washing steps (3 times) between steps. The photography was taken by a fluorescent microscope (Zeiss, Exton, PA). Ubiquitin C-terminal hydrolase L3 (Uch-L3) was found in our previous proteomic study as a possible modulator sufficient to identify the specific targets in bone morphogenetic protein (BMP) 2-induced osteoblast differentiation [27]. To investigate the interaction between Uch-L3 and Smad1, HA-Smad1 and Flag-Uch-L3 plasmids were transfected into HEK 293 cells. Co-immunoprecipitation (IP) was performed against HA-Smad1 using anti-HA antibody. Uch-L3 protein (Flag-Uch-L3) was detected in the eluent of the immuno-complex by western blot using anti-Flag antibody (Fig. 1 A). Reciprocally, the co-IP against Flag-Uch-L3 demonstrated that the eluent contained HA-Smad1 (Fig. 1B). Furthermore, it wondered whether the endogenous Smad1 interacts with Uch-L3. We determined the endogenous Uch-L3 levels in primary osteoblast (Ob), MC3T3-E1 (Mc), and C2C12 (C2) cells. In those cells, similar levels of Uch-L3 proteins (Fig. 1C, upper panels) and mRNA levels (Fig. 1C, lower panels) were observed. Among them, C2C12 cells were selected to make the Uch-L3 stable cells. As shown in Fig. 1D, the endogenous Smad1 was also detected in the immunoprecipitated complex of Uch-L3 (Flag-Uch-L3). Furthermore, the endogenous Uch-L3 was also interacted with the endogenous Smad1 in C2C12 cells as confirmed by co-IP against anti-Uch-L3 antibody and WB using anti-Smad1 antibody (Fig. 1E). All of these results showed that Uch-L3 interacts with Smad1. After confirming the interaction between Smad1 and Uch-L3, we tested whether the deubiquitinating activity of Uch-L3 is able to affect the ubiquitination pattern of Smad1. After the transfection of HA-Smad1 and Flag-Ub constructs with increasing amount of Uch-L3 or without Uch-L3 plasmid into HEK 293 cells, the degree of Smad1 ubiquitination was analyzed. The addition of Uch-L3 construct showed the significant deubiquitination of Smad1 by a dose-dependent manner (Fig. 2 A, left panel). In the presence of Uch-L3, the amount of poly-ubiquitinated Smad1 was dramatically diminished compared to the control (in the absence of Uch-L3). Reciprocally, poly-ubiquitinated Smad1 was also detected in the HA-labeled ubiquitin/ubiquitinated protein complex immunoprecipitated with HA-Ub by anti-HA antibody. In addition, the amount of poly-ubiquitinated Smad1 was also decreased in the presence of Uch-L3 (Fig. 2B, top panel). These results suggest that Uch-L3 can deubiquitinate the poly-ubiquitinated Smad1. We then tested whether Uch-L3, a possible DUB enzyme of Smad1, is able to accelerate the differentiation of osteoblast. The efficient stable expression of Uch-L3 was produced by the retroviral transduction onto C2C12 cells. The Uch-L3-expressing stable cells were selected by puromycin for 5 days. After the selection, we confirmed the expression of exogenous Flag-HA-Uch-L3 from the stable cell by Western blot (Fig. 3 A). The activity of alkaline phosphatase (ALP), a downstream molecule in Smad1 pathway, was clearly enhanced as detected by ALP staining in Uch-L3 over-expressing cells compared to that in mock control cells (Fig. 3B, upper panels). The enhanced ALP activity may be caused by the deubiquitinating activity of Uch-L3 stabilizing Smad1. Moreover, BMP2-induced stimulation further increased ALP activity in Uch-L3 over-expressing cells compared to control cells (Fig. 3B, lower panels). It is known that BMP2 activates Smad1 and induces the expression of the downstream genes. We checked the expression of some downstream target genes: alkaline phosphatase (Alp), osteopontin (Opn), osteocalcin (Oc), ID1, and Noggin. The induction of these genes was confirmed by real-time RT-PCR with BMP2 treatment (100 ng/ml). The expressions of Opn, Alp, Oc, ID1, and Noggin were higher in Uch-L3 stably-expressing cells compared to control cells (Fig. 3C). These results of increased expression of differentiation markers and down-stream target genes suggest that Uch-L3 may enhance osteoblast differentiation by elevating the stability of Smad1. To determine the effect of Uch-L3 knock-down in osteoblast differentiation, Uch-L3 specific siRNA was transfected into C2C12 cells. The expressions of Uch-L3 mRNA and protein after transfection were evaluated by real-time RT-PCR and immunoblotting (Fig. 4 A). We then checked the osteoblast differentiation markers in the cells transfected with the Uch-L3 specific siRNA and the scrambled siRNA (negative control). The ALP staining was greatly reduced in Uch-L3 knock-down cells compared to the scrambled siRNA control cells (Fig. 4B). The mRNA expression levels of the markers, Opn, Alp, and Oc were also reduced in the Uch-L3 siRNA cells compared to the control cells (Fig. 4C). Furthermore, in the Uch-L3 knock-down cells transfected with L3-siRNA, the level of BMP2-induced phosphorylated Smad1 (pSmad1) was dramatically reduced 3 h after BMP2 treatment (Fig. 4D). In the immunocytochemical staining, the Uch-L3 knock-down cells showed the reduced expression of pSmad1 in the nucleus 3 h after BMP2 treatment, while the cells transfected with scrambled siRNA (Ctl) showed strong pSmad1 expression in the nucleus (Fig. 4E). These results suggest that the reduction of Uch-L3 decreases the active pSmad1 signal and thus reduces the osteoblast differentiation. This study showed a novel function for Uch-L3 in osteoblast differentiation induced by the regulation of BMP2 signaling. The physical interaction between Uch-L3 and Smad1 was revealed. The activity of deubiquitinating enzyme Uch-L3 dramatically reduced the amount of ubiquitinated Smad1. In Uch-L3 stably-expressing cells, the expression of osteoblast differentiation markers, specifically osteocalcin, osteopontin, and alkaline phosphatase, was increased. In the Uch-l3 knock-down experiment, loss of Uch-L3 resulted in reduced ALP staining and the reduced expression of the markers. These results suggest that Uch-L3 plays a role as a modulator in the BMP2-mediated osteoblastogenesis and thus enhancing the differentiation of osteoblasts. In our previous study, we reported elevated phosphorylation of Uch-L3 in BMP2-induced osteoblast differentiation [27]. The loss of Uch-L3 affects the accumulation of ubiquitinated proteins and the induction of stress [23]. A defect of Uch-L3 also resulted in retinal degeneration and neurodegeneration [6, 18]. The up-regulation of Uch-L3 expression was observed in breast invasive cancer and cervical carcinoma [25, 26]. Recent in vivo studies showed that the Uch-L3 deficiency reduced insulin signaling and adipogenesis [22]. However, there is no report on the possibility of Uch-L3 regulating osteoblast differentiation. It has been reported that two proteins, Nedd8 and caseinolytic peptidase B (CLPB), can interact with Uch-L3 [32, 38]. The C-terminal of Nedd8, an ubiquitin-like protein, is cleaved by Uch-L3. The CLPB, a protein in disaggregating machines, was revealed as a partner protein of Uch-L3 from a global proteomic approach. Inhibiting Uch-L3 activity caused increased ubiquitination in epithelial sodium channel (ENaC), which is regulated by the ubiquitin E3 ligase Nedd4-2 [39]. Although these studies highlight many important functions of Uch-L3, there is little known about the role of Uch-L3 in BMP signaling. In TGFβ signaling, two known deubiquitinating enzymes, Usp9X (FAM) or Uch-L5 (Uch37) are known to act on the stability of the target proteins, Smad4 or Smad7 and type I TGFβ receptor, respectively [12, 13]. Recently, it was reported that Uch-L1 positively regulates the stability of β-Catenin in Wnt signaling [14]. As those reports, UCH family members Uch-L1 and Uch-L5 which have 31–55% identity, function as deubiquitinating enzyme. However, there is no information on the involvement of deubiquitinating enzymes in BMP signaling and/or the differentiation of osteoblasts. In this study, it was found Smad1 is physically interacted with Uch-L3, which deubiquitinates Smad1. That is, Smad1 poly-ubiquitination degrees were decreased by the forced expression of Uch-L3. Furthermore, the forced expression of Uch-L3 promoted the osteoblast differentiation, and Uch-L3 knock-down showed reduced differentiation. This study shows a novel finding that osteoblast differentiation could be regulated by Uch-L3, a deubiquitinating enzyme, through the BMP2-induced Smad1 stabilization. Future studies using the Uch-L3 knock-out mouse model will be necessary to understand the in vivo functions of Uch-L3. This research was supported by a grant from the (NRF) funded by the Ministry of Education, Science and Technology (No. 2010-0020553 & No. 2010-0011231) and partly by a grant from the Korea Health 21 R&D Project, Ministry of Health & Welfare (No. A090610 & No. A010252).