Lithium Chloride Enhances Cathepsin H Expression and BMP-4 Degradation in C3H10T1/2 Cells

The effect of canonical Wnt/β-catenin signaling on chondrogenic differentiation induced by transfection of BMP4 expressing plasmid was analyzed. Lithium chloride (LiCl) which mimics canonical Wnt/β-catenin signaling was added to cells transfected with BMP4 expressing plasmid. Although BMP4 mRNA expression was not affected by LiCl, LiCl decreased BMP4 protein accumulation. Gene expression analysis exhibited upregulation of cathepsin H by LiCl treatment. Gene silencing of cathepsin H enhanced BMP4 protein accumulation from BMP4 expressing cells. These results suggested that cathepsin H is regulated by Wnt/β-catenin signaling and plays an important role in the regulation of BMP4 biological activity.


Introduction
Bone morphogenetic proteins (BMPs) play critical roles in chondrogenic and osteogenic differentiation of mesenchymal multipotent cells [1]. The use of BMPs is a promising approach for tissue engineering of bone and cartilage. Actions of BMPs were regulated by various molecules and signaling pathways both in extracellular matrices and intracellular signal transduction. Wnt signaling is one of such pathways which regulates the cell fate in chondrogenesis and osteogenesis [2]. So far, three distinct intracellular cascades are known: canonical Wnt/ -catenin pathway, JNK pathway, and Wnt/Ca2+ pathway. From observations of mutant mice, deletion of canonical Wnt/ -catenin pathway leads to the defect of osteoblastogenesis both in endochondral and membranous ossification [3,4]. These studies implicated that canonical Wn/catenin pathway prevents osteoblast from differentiation into chondrocyte and is required for osteogenesis. Interestingly, stabilized -catenin negatively affected both osteogenesis and chondrogenesis in gain-of-function mutant study [4]. Despite these findings using mutant mice, the direct relationship between BMPs and canonical Wnt signaling is still unclear.
Gene transfer of BMP expressing plasmid drives recipient cells to produce BMP continuously [5]. High density micromass culture after transfection of BMP4 expressing plasmid differentiates mesenchymal C3H10T1/2 cells to chondrogenic lineage [6]. In the current study, lithium chloride (LiCl) was added on the micromass culture of BMP4 gene transfected C3H10T1/2 cells to examine the effects of Wnt/ -catenin signaling on the BMP4 mediated chondrogenic differentiation. LiCl inhibits glycogen synthase kinase-3 (GSK-3 ) and leads to accumulation of -catenin. Therefore, LiCl is known to mimic the activity of Wnt/ -catenin signaling [7].
From our findings in this study, we focused on cathepsin H. Cathepsin H is a member of cathepsins, which are proteases ubiquitously expressed in animal tissues. Cathepsin family includes the serine proteases cathepsins A and G, the aspartic proteases cathepsins D and E, and the lysosomal cysteine cathepsins B, C, F, H, K, L, O, S, V, X, and W [8]. Cathepsins have numerous functions in normal and pathological processes. It is widely known that cathepsin K is highly expressed in osteoclasts and plays an essential role in bone resorption [9]. Cathepsins including cathepsin H are found activated in colorectal [10], pancreatic [11], and prostate cancer [12], glioma [13], and melanoma [14]. So, cathepsins could be a therapeutic target in such malignant tumors. Cathepsin H is known to be expressed in the adult lung and process pulmonary surfactant protein [15]. From the investigation of the development of lung, Lü et al. [16] reported that cathepsin H was involved in the degradation of BMP4. We hypothesized that cathepsin H and degradation of BMP4 are related also in C3H10T1/2 cells. In the current study, we examined the gene expression and gene-silencing effects of cathepsin H.

Plasmid Construction and Small
Interfering RNA. A 1.6 kb mouse BMP-4 cDNA is a gift from Hogan et al. [17]. It was inserted into blunted EcoRI site of the pCAGGS expression vector which was kindly provided by Miyazaki [18]. A green fluorescent protein expressing plasmid pCAGGS-EGFP was also constructed. TOPFLASH reporter plasmid containing three copies of the TCF/LEF-1-binding site and FOPFLASH containing three copies of the mutated TCF/LEF-binding site were purchased from Upstate Biotechnology (Lake Placid, NY). A sea pansy luciferase expressing plasmid with SV40, pRL-SV40, was purchased from TOYO ink. The siRNAs for cathepsin H (siCTSH: Stealth Select RNAi) and negative control siRNA (siControl: 743519) were purchased from Invitrogen. The most effective siCTSH was selected from three candidates in our preliminary experiments.

Cell
Culture. C3H10T1/2 clone 8 cells were obtained from RIKEN cell bank (Tsukuba, Japan) and maintained in Dulbecco's modified Eagle's medium (DMEM: Invitrogen, Carlsbad, CA) supplemented with 10% fatal bovine serum (Invitrogen), 50 U/mL penicillin, and 50 mg/mL streptomycin (Penstrep: Invitrogen), that is, growth medium. The micromass culture technique was modified from Ahrens et al. [19]. Trypsinized cells were resuspended in growth medium at a concentration of 10 7 cells/mL. A 10 L drop of this cell suspension was placed in the center of a well in 12-well tissue culture dish (Greiner Bio-one, Kremsmuenster, Austria). The cells were allowed to adhere for 60 min and, then, the wells were filled with 1 mL of DMEM/F12 (1 : 1) supplemented with 1% fetal bovine serum (Invitrogen), 0.2% BSA (Sigma, St. Louis, MO), 50 g/mL ascorbic acid 2 phosphate (Sigma), Penstrep, and ITS+ (BD bioscience, Franklin Lakes, NJ), that is, differentiation medium. This medium was changed every 3 days.

In Vitro Electroporation.
Before electroporation, cells were lifted with 0.25% Trypsin EDTA (Invitrogen) and dissolved in electroporation buffer [20] (75% cytosalts; 120 mM KCl, 0.15 mM CaCl 2 , 10 mM K 2 HPO 4 pH7.6, 5 mM MgCl 2 ; 25% Opti-MEM1) at a concentration of 2.5 × 10 6 cells/mL. Plasmid solution was added into 600 L of cell suspension (1.5 × 10 6 cells). Electric pulses were generated by CUY-21 in vitro (BEX, Tokyo, Japan) and applied through 4 mm gap cuvette. Pulse settings were 480 V, 2.5 ms, 2 pulses at the interval of 1 second. After electroporation, the cells were recovered in growth medium overnight. On the next day of electroporation, dead cells were removed by washing with phosphate buffered saline (PBS: Invitrogen) and micromass culture was started as mentioned above. The timeline of experiments was counted from the start of micromass culture in this study.

Quantitative Reverse Transcription Polymerase Chain
Reaction. For quantitative reverse transcription polymerase chain reaction (qRT-PCR), total RNA was isolated using RNeasy Mini kit and RNase-free DNase set (QIAGEN, West Sussex, UK) according to the manufacturer's protocol. Oncolumn DNA digestion was routinely done. Single stranded cDNA was synthesized using High-Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA). qRT-PCR analyses were carried out with StepOnePlus and Power SYBR green master mix (Applied Biosystems). The fractional cycle number at which the fluorescence passes the threshold (Ct values) was used for quantification by using a comparative Ct method. Values of gene-of-interest (GOI) were normalized to the threshold value for GAPDH: ΔCt = Ct (GOI) − Ct (GAPDH). The Ct value of control was used as a reference. ΔΔCt = ΔCt (experiment) − ΔCt (control). The fold change in mRNA expression was calculated by the following formula: 2 −ΔΔCt . Primers for RT-PCR were listed on Table 1.

Western
Blotting. Culture medium was subjected to SDS-PAGE with 10% acrylamide gel. Samples were blotted to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA). The membranes were blocked with 2% nonfat dry milk for 1 hour at room temperature and incubated with anti-BMP4 primary antibody (1 : 400 Santa Cruz) at 4 ∘ C overnight. Then HRP-conjugated secondary anti-mouse IgG (1 : 1000; Santa Cruz) was applied for 2 hours at room temperature. Signals were visualized by chemiluminescence using ECL plus western blotting detection reagents (GE Healthcare, UK) with a digital luminescent image analyzer LAS-1000 (Fujifilm, Tokyo, Japan). Band intensities were analyzed by ImageJ 1.37v software program (National Institutes of Health, MD).

2.7.
Statistics. Data were presented as means ± standard deviation of independent experiments. One-way analysis of variance (ANOVA) was performed using GraphPad Prism version 5.0 program (Graphpad software, San Diego, CA). Statistical analyses of differences between experimental groups and control were performed using Dunnett's post hoc

Electroporatic Transfer of BMP4 Expressing Plasmid
Induced Chondrogenesis. To confirm the chondrogenesis, pCAGGS-BMP4 transfected C3H10T1/2 cells were cultured in micromass. From day 3, the cells metamorphosed into round shape. Alcian blue staining at day 6 exhibited positive stain. The intensity of Alcian blue staining showed dosedependent increase depending on the amount of plasmid used at electroporation (Figure 1(a)). Gene expression profiles were analyzed by qRT-PCR at day 3. BMP4 mRNA expression increased in the dose-dependent manner of the plasmid amount. Sox9 mRNA expression was not affected by the dose of plasmid. Aggrecan and Col2a1 mRNA showed marked dose-dependent increase (Figure 1(b)).

The Effect of LiCl Treatment on Gene Transfected Cells.
LiCl was added to the micromass culture of pCAGGS-BMP4 transfected C3H10T1/2 cells. To correct the osmolality of medium, NaCl was added up to 25 mM together with LiCl (NaCl;LiCl = 25;0, 20;5, 15;10, 25;0 mM). Alcian blue staining of micromass at day 6 was inhibited by the addition of LiCl in dose-dependent manner. The supplementation of wnt3a protein also inhibited Alcian blue staining of micromass at the concentration of 50 mM significantly (Figures 2(a) and 2(b)). To test the effect of LiCl on the expression levels of other plasmid, pCAGGS-EGFP or pRL-SV40 was transfected into C3H10T1/2 cells, and cells were cultured in micromass under the same concentrations of LiCl. The expression of EGFP was similar to that of the control in 5 and 10 mM of LiCl treatment and higher in the 25 mM LiCl treatment (Figure 2(c)). The luciferase assay showed similar tendencies to the EGFP expression (Figure 2(d)).

LiCl Treatment Activated -Catenin/TCF Pathway.
To confirm the activation of canonical Wnt signaling in the LiCl treated micromass, TOPFLASH or FOPFLASH (20 g/ cuvette), pRL-SV40 (20 g) and pCAGGS-BMP4 (20 g), was cotransfected by electroporation. After a 3-day culture in micromass with or without LiCl, luciferase activity of cell lysate was assessed. The promoter activity of TOPFLASH was enhanced by addition of LiCl (Figure 2(e)).

LiCl Treatment Suppressed Chondrogenic Differentiation
of C3H10T1/2. The gene expression profiles of C3H10T1/2 cells transfected with pCAGGS-BMP4 (30 g/cuvette) were analyzed at day 3 (Figure 3(a)). The levels of BMP4 mRNA did not show significant difference by LiCl treatment. Sox9 mRNA expression was significantly suppressed at 25 mM LiCl treatment. Aggrecan and Col2a1 gene were dosedependently inhibited by LiCl. The expression levels of antagonists against BMPs were also investigated. Noggin mRNA was dose-dependently suppressed. Gremlin 1 expression were not affected by LiCl. The levels of follistatin were increased in dose-dependent manner. The expression levels of cathepsin H was increased in a dose-dependent manner of LiCl.

LiCl Treatment Remarkably Suppressed BMP4 Protein
Accumulation. The conditioned medium used in the micromass culture of BMP4 transfected cells was subjected to western blot analysis (Figure 3(b)). LiCl treatment dose dependently reduced the accumulation of BMP4 protein in the culture medium. Localization of BMP4 protein was assessed by immunocytochemical staining of micromass (Figure 3(c)). BMP4 protein in the cells and extracellular matrix was also reduced by LiCl treatment in dose dependent manner.  (Figure 4(a)). Western blotting for BMP4 showed that cathepsin H gene silencing remarkably increased BMP4 protein expression by 2.9 ± 1.4-folds (Figure 4(b)). Alcian blue staining of micromass at day 6 showed significant increase in siCTSH transfected cells (Figure 3(c)).

Discussion
In the current study, BMP4 expressing cells were treated with LiCl which stabilize -catenin and activate Wnt/ -catenin signaling. LiCl treatment did not affect the mRNA expression from transfected BMP4 expressing plasmid. However, LiCl dramatically reduced the accumulation of BMP4 protein both in conditioned medium and micromass. This discrepancy between gene expression and protein accumulation could be attributed to the degradation of BMP4 protein. Cathepsin H which is known to be involved in the degradation of BMP4 protein in the developmental stage of lung [16] showed increase in the dose-dependent manner of LiCl. Gene silencing of cathepsin H enhanced the protein accumulation of BMP4 from expressing plasmid. These results confirmed that cathepsin H is involved in the degradation of BMP4 protein in the C3H10T1/2 and its expression was regulated by Wnt/catenin signaling. The current study suggested that LiCl treatment which mimics Wnt/ -catenin signaling controls the biological activity of BMP4 through the degradation process by cathepsin H, for the first time.
Cathepsin H is a protease which has both endopeptidase and aminopeptidase activities [21]. From the spatiotemporal BioMed Research International    exhibit BMP4 degradation activity. Lysosomal cathepsins require slightly acidic environment, such as that found in the lysosomes [8]. So, it is likely that cathepsin H degrades BMP4 in the lysosomes or endosomes. Ishihara et al. [22] analyzed global gene expression using microarray at the fracture site adenovirally infected by BMP genes in an equine model. The expression of cathepsin H was greater in BMP-2 and -6 treated fracture callus than that control. This may imply cathepsin H plays a certain role in the activity of BMPs at fracture site. Only a few researches have shed light on the degradation of BMPs. There are many issues awaiting investigation, such as degradation of other members of TGFsuperfamily, BMP degradation ability of other members of cathepsin family, and the details of degradation process. BMP4 protein is synthesized as an inactive precursor and undergoes proteolytic cleavage by subtilisin-like proprotein convertase at the specific site to be activated [23,24]. So far, no evidence has been shown that cathepsins are involved in the cleavage of BMP precursors. According to suppliers information, the antibody for BMP4 used in the current study reacts with both BMP4 precursor and mature BMP4. In our immunocytostaining for BMP4, LiCl treatment also suppressed the BMP4 accumulation in the cells. Cathepsin H is a lysosomal cysteine protease. Therefore, cathepsin H seemed to degrade both BMP4 precursors and mature BMP4 in cytoplasm before its extracellular secretion.
Noggin is a well-known antagonist of BMPs. Noggin binds to BMPs with high affinity and blocks interaction between BMPs and their receptors [25]. BMPs including BMP-4 induce the expression of noggin [26]. This phenomenon suggests that there exists a negative feedback loop to limit the activities of BMPs. The suppression of noggin expression along with the reduction of BMP4 accumulation in our observation is consistent with previous findings. Gremlin expression was not influenced by LiCl in our data. Sun et al. suggested that gremlin is also expressed in developing lung with BMP4 [27]. Interestingly, gremlin interacts with BMP4 precursor intracellularly and efficiently inhibits the activity of BMP4. Cathepsin H is also expressed in developing lung and inhibits BMP4. There may exist an interplay between gremlin and cathepsin H. Follistatin also binds BMP4 and inhibits its activity extracellularly. On the contrary to noggin, it is known that follistatin expression is downregulated by BMP4 in chondrocytes [28]. The decrease of BMP4 accumulation by LiCl seemed to stimulate the expression of follistatin by the reduction of inhibition in our experiments.
In the current study, LiCl was used to mimic the action of canonical Wnt/ -catenin signaling. The presence of LiCl inhibits glycogen synthase kinase-3 (GSK-3 ). Intracellular -catenin undergoes digestion via ubiquitin proteasome system when the receptor frizzled and coreceptor LRP were not bound with Wnt ligand. The inhibition of GSK-3 blocks this digestion and increases the accumulation of intracellular -catenin [7]. LiCl inhibited the chondrogenic differentiation of C3H10T1/2 micromass culture treated with BMP2 protein supplementation [29]. This data suggested that Wnt/ -catenin signaling directly inhibits chondrogenesis. In our experiments, Wnt/ -catenin signaling synergistically inhibits chondrogenesis of C3H10T1/2 cells through both the degradation of BMP4 and direct inhibition by Wnt/catenin signaling.
Gene silencing of cathepsin H in BMP4 expressing cells exhibited greater accumulation of BMP4 protein in the absence of LiCl. This result without mimicking Wnt/ -catenin signaling may suggest that catalytic processes through cathepsin H are active at a certain level in C3H10T1/2. Although this activity could vary according to types of cells and conditions, the inhibition of cathepsin H has a potential to enhance the effects of BMPs in the gene-based tissue engineering of bone and cartilage.

Conclusion
In conclusion, upregulation of canonical wnt/ -catenin signaling mimicked by LiCl caused discrepancy between mRNA and protein expression of BMP4. Cathepsin H, a cysteine protease which degrades BMP4 mRNA expression, was also upregulated by LiCl treatment. Gene silencing of cathepsin H enhanced BMP4 accumulation from BMP4 expressing cells. Wnt/ -catenin signaling may regulate the biological activity of BMP4 through the regulation of cathepsin H.