|Home | About | Journals | Submit | Contact Us | Français|
Myocilin is a secreted glycoprotein that is expressed in ocular and non-ocular tissues. Mutations in the MYOCILIN gene may lead to juvenile- and adult-onset primary open-angle glaucoma. Here we report that myocilin is expressed in bone marrow-derived mesenchymal stem cells (MSCs) and plays a role in their differentiation into osteoblasts in vitro and in osteogenesis in vivo. Expression of myocilin was detected in MSCs derived from mouse, rat, and human bone marrow, with human MSCs exhibiting the highest level of myocilin expression. Expression of myocilin rose during the course of human MSC differentiation into osteoblasts but not into adipocytes, and treatment with exogenous myocilin further enhanced osteogenesis. MSCs derived from Myoc-null mice had a reduced ability to differentiate into the osteoblastic lineage, which was partially rescued by exogenous extracellular myocilin treatment. Myocilin also stimulated osteogenic differentiation of wild-type MSCs, which was associated with activation of the p38, Erk1/2, and JNK MAP kinase signaling pathways as well as up-regulated expression of the osteogenic transcription factors Runx2 and Dlx5. Finally, cortical bone thickness and trabecular volume, as well as the expression level of osteopontin, a known factor of bone remodeling and osteoblast differentiation, were reduced dramatically in the femurs of Myoc-null mice compared with wild-type mice. These data suggest that myocilin should be considered as a target for improving the bone regenerative potential of MSCs and may identify a new role for myocilin in bone formation and/or maintenance in vivo.
Bone marrow-derived mesenchymal stem cells, also known as mesenchymal stromal cells or MSCs,3 have attracted significant interest in recent years because of their potential clinical application to tissue regeneration and repair in multiple organ systems (see Refs. 1–3 for recent reviews). The relative ease of MSC isolation from bone marrow aspirates and their potential for autologous transplantation have made this cell type especially attractive for this purpose. Cultured MSCs are readily expandable and capable of differentiating along multiple lineages, including those of osteoblasts, chondrocytes, adipocytes, and myoblasts, under appropriate conditions (4, 5). Several growth factors and signaling pathways have been implicated in controlling MSC differentiation along specific lineages (6). For example, it has been suggested that differentiation of MSCs into osteoblasts may involve the integration of the bone morphogenic protein, Notch, Wnt, Hedgehog, and fibroblast growth factor signaling pathways. These pathways lead to activation of a set of transcription factors, including Runx2 and NFAT2, that regulate the expression of genes essential for osteoblast differentiation, such as collagen type 1, osteopontin, sparc, and osteocalcin (see Ref. 7 for a recent review). The translation of MSC therapy into clinical application, however, will require a more complete understanding of how signaling pathways integrate to control cell maintenance, proliferation, commitment, and differentiation.
Human myocilin is a secreted glycoprotein with a length of 504 amino acids. Mutations in the MYOCILIN (MYOC) gene are found in more than 10% of juvenile open-angle glaucoma cases and in 3–4% of patients with adult-onset primary open-angle glaucoma (8–12). MYOC is expressed in ocular (trabecular meshwork, iris, ciliary body, sclera, and retinal pigmented epithelial cells) and non-ocular (sciatic nerve, skeletal muscle, mammary gland, thymus, and testis) tissues. Myocilin is not known to play a role in human disease outside of the eye. However, we have demonstrated recently that overexpression of wild-type myocilin in transgenic mouse skeletal muscle leads to increased muscle size and suggested that myocilin may regulate muscle hypertrophy (13). We have also shown that myocilin may serve as a modulator of Wnt signaling by interacting with several Wnt receptors, frizzled receptors, and antagonists of Wnt, including Wif-1 and sFRP (secreted Frizzled-related) proteins (14). In the experimental models we used, myocilin stimulated noncanonical but not canonical Wnt signaling pathways. Treatment of different cell types with extracellular myocilin induces formation of stress fibers and intracellular redistribution of β-catenin with accumulation on the cellular membrane but not in cell nuclei (14).
Here we demonstrate that myocilin is expressed in MSCs and plays a role in MSC differentiation along the osteoblastic lineage. MSCs exhibit an increase in myocilin expression when cultured in osteogenic conditions, and addition of exogenous myocilin increases the efficiency of osteogenic differentiation in wild-type MSC cultures. MSCs isolated from the bone marrow of Myoc-null mice have an impaired ability to differentiate along the osteoblastic lineage, but addition of exogenous myocilin improves osteogenic differentiation in MSC cultures from these mice. Stimulation of MSC differentiation into osteoblasts by extracellular myocilin is associated with activation of the p38, Erk1/2, and JNK MAPK signaling pathways, and blocking these pathways attenuates osteogenic differentiation by MSCs. We also speculate that myocilin may play a role in osteogenesis more generally because Myoc-null mice exhibit reduced cortical bone thickness and trabecular volume compared with wild-type mice.
Mice (B6/129 mixed background) and Sprague-Dawley rats were maintained in accordance with the guidelines set forth by the National Eye Institute Committee on the Use and Care of Animals. Myoc-null mice have been described previously (15).
Human MSCs (hMSCs) were purchased from the ATCC and grown according to the instructions of the manufacturer. Rat MSCs were isolated from femoral bone marrow aspirates as described previously (16). Mouse MSCs (mMSCs) were isolated from femoral bone marrow aspirates of wild-type and Myoc-null mice as described previously (17). In brief, 6-week-old rats or mice were sacrificed by increasing exposure to CO2 followed by cervical dislocation. The body of the animal was rinsed in 70% ethanol, and the cutaneous tissues were removed. Lower limbs were disarticulated and stored briefly on ice in DMEM supplemented with penicillin (100 units/ml) and streptomycin (100 μg/ml). Bone marrow was aspirated from femurs by flushing them with 1–2 ml of DMEM. Mononucleated cells from the bone marrow were plated at 2 × 107 cells/100 mm2 on plastic culture dishes in DMEM supplemented with 10% FBS and penicillin (100 units/ml) and streptomycin (100 μg/ml) at 37 °C in 5% CO2. Non-adherent cells were removed after 48–72 h, and adherent cells (MSCs) were replenished with fresh medium every 2–3 days until confluent. MSCs between passages 7 and 9 were used in all experiments.
MSCs were seeded into culture vessels at 5 × 103 cells/cm2. For osteogenic differentiation, the cultures were incubated in osteoblast differentiation medium (StemPro osteogenesis differentiation kit, Invitrogen) for up to 3 weeks. The medium was changed two times per week. The cells were fixed with 10% formalin for 20 min at room temperature and stained with Alizarin Red S (pH 4.1) (Sigma) for 20 min at room temperature. For adipogenic differentiation, the cultures were incubated in adipocyte differentiation medium (StemPro adipogenesis differentiation kit, Invitrogen) for 3 weeks. The medium was changed twice weekly. The cells were then fixed with 10% formalin for 20 min at room temperature and stained with 0.5% Oil Red O (Sigma) in methanol (Sigma) for 20 min at room temperature. Alkaline phosphatase (AP) activity assays were carried out by washing cells in PBS and lysing them with three cycles of freezing and thawing in 0.05% Triton X-100. An aliquot of the lysate was incubated with naphthol AS-MX phosphate alkaline solution with fast blue RR salt at 37 °C for 30 min. The reactions were stopped by adding 5 μl of 2 m NaOH to 50 μl of extracts, and the absorbance was measured at 405 nm using a multifunctional reader (Bio-Rad). AP activity was normalized to total protein concentration, which was determined using Bradford reagent (Bio-Rad). For chondrocyte differentiation, the cultures were incubated in chondrocyte differentiation medium (StemPro chondrogenesis differentiation kit, Invitrogen) for 2 weeks. The medium was changed every 2–3 days. MAP kinase inhibitors SB203580 and SP600125 (Sigma-Aldrich), BIRB 796 (Selleck Chemicals LLC, TX) and BI 78D3 (R&D Systems, MN) were prepared and used as recommended by the manufacturers. Si-JNK#1 (5′-GGAGCUCAAGGAAUAGUAU-3′), si-p38#1 (5′-GCCCAUAAGGCCAGAAACU), and control scrambled si-S1 siRNA oligonucleotides (Cell Signaling Technology, Inc., MA) were used at a final concentration of 10 and 100 nm. Si-p38#2 (GCCCUGAGGUUCUGGCAA); si-p38#3 (CGACGAGCACGUUCAAUUC); si-p38#4 (CCAUAGACCUCCUUGGAAG); si-JNK#2 (GCCCAGUAAUAUAGUAGUA); si-JNK#3 (GGCAUGGGCUACAAGGAAA); si-JNK#4 (GAAUAGUAUGCGCAGCUUA); and the non-targeting pool of control si-S2 (5′-UGGUUUACAUGUCGACUAA-3′, 5′-UGGUUUACAUGUUGUGUGA-3′, 5′-UGGUUUACAUGUUUUCUGA-3′, and 5′-UGGUUUACAUGUUUUCCUA-3′) siRNAs ( Dharmacon, CO) were used at final concentration of 100 nm. hMSCs were transfected using siLentFect (Bio-Rad) according to the instructions of the manufacturer 4 days after the beginning of osteoblastic differentiation.
Total RNA was extracted from cell cultures using the RNeasy mini kit (Qiagen, Valencia, CA). Human muscle RNA was purchased from ORIGENE (Rockville, MD). 20 ng of RNA were used for each qRT-PCR reaction. qRT-PCR was performed using the SuperScript® III one-step qRT-PCR system with Platinum® Taq kit according to the instructions of the manufacturer (Invitrogen). Primers for qRT-PCR were as follows: human MYOC, 5′-ATTTGAAGGAGAGCCCATC-3′ and 5′-GCTACCCCTTCTAAGGTTCAC-3′; human GAPDH, 5′-ACCCATCACCATCTTCCAGGAGCG-3′ and 5′-CGGGAAGCTCACTGGCATGGCCT-3′; mouse Dlx5, 5′-CCAACCAGCCAGAGAAAGAA-3′ and 5′-GCAAGGCGAGGTACTGAGTC-3′; mouse Runx2, 5′-AACCCACGAATGCACTATCCA-3′ and 5′-CGGACATACCGAG GGACATG-3′; mouse osteocalcin, 5′-AGCAAAGGTGCAGCCTTTGT-3′ and 5′-GCGCCTGGGTCTCTTCACT-3′; and mouse Gapdh, 5′-CCCATCACCATCTTCCAGGAGCG-3′ and 5′-CGGGAAGCTCACTGGCATGGCCT-3′. GAPDH was used for normalization. qRT-PCR was carried out on a 7900HT real-time thermocycler (Applied Biosystems, CA). Control experiments showed that the efficiency of amplification was very similar for different primer sets (less than 5% difference). To quantifying the relative changes in gene expression, we used the 2−ΔΔCT method. The average CT was calculated for the target genes and internal control (GAPDH), and the ΔCT (CT,target-CT,GAPDH) values were determined. All reactions were performed in triplicate using three independent samples.
Cells were washed once with PBS and then lysed in lysis buffer containing 1% Nonidet P-40, 5% sodium deoxycholate, 1 mm phenylmethylsulfonyl fluoride, 100 mm sodium orthovanadate, and 1:100 protease inhibitor mixtures (Sigma). 5- to 10-μg aliquots of protein extracts were separated by 4–12% SDS-PAGE and transferred to polyvinylidene difluoride membranes (Invitrogen). The membranes were blocked with 5% nonfat milk overnight at 4 °C and then probed with primary antibodies. The immunocomplexes were visualized with horseradish peroxidase-coupled goat anti-rabbit or anti-mouse IgG (GE Healthcare, NJ) using SuperSignal reagents (Pierce). Primary antibodies used were as follows: anti-phosphorylated JNK and anti-total JNK monoclonal antibodies (Sigma); anti-HSC70 monoclonal antibodies (Santa Cruz Biotechnology, Inc., CA); anti-Runx2 (Millipore, MA); anti-Dlx5 (Chemicon, CA); anti-phospho-p38 polyclonal and anti-p38 (Cell Signaling Technology, Inc.); and anti-phospho-Erk1/2, anti-Erk1/2, and anti-PPARγ (Cell Signaling Technology, Inc.). Antibodies against myocilin have been described previously (18). For digital quantification, membranes were scanned using a Typhoon 9410 variable-mode imager (Amersham Biosciences) and analyzed using Image Scion Alpha 22.214.171.124 (Scion Cor, Frederick, MD). In some cases, the chemiluminescent signals were captured using an 8-megapixel scientific-grade CCD camera (Flurochem M., Santa Clara, CA), and the signal intensities were quantified and analyzed using AlphaView software (Proteinsimple, Santa Clara, CA). All experiments were repeated at least twice.
MSCs were seeded on 2-well glass chamber slides (Nalge Nunc, NY) and cultured for 24 h in complete DMEM. The medium was removed, and cells were washed 2–3 times with PBS. The MSCs were fixed with fresh 3.7% formaldehyde for 10 min at room temperature and permeabilized with 0.1% Triton X-100 in PBS for 5 min. After blocking with 5% BSA in PBS at room temperature for 1 h, cells were incubated with anti-myocilin (1:500 dilution) polyclonal and anti-golgin-97 monoclonal antibody (1:200) (Invitrogen) or anti-protein-disulfide isomerase monoclonal antibody (1:200 dilution) (Abcam, Cambridge, MA) at room temperature for 1 h. Cells were washed with PBS-Tween 20 and then incubated with Alexa Fluor 488- or 594-conjugated secondary antibodies for 30 min. An Axioplan 2 fluorescent microscope and Zeiss 700 confocal microscope (Carl Zeiss MicroImaging, Inc. NY) were used to detect fluorescence. The images were processed with Adobe Photoshop Elements 2.0 (Adobe Inc., CA). Identical processing parameters were used for all images within any single experiment. p-p38 detection was performed with polyclonal anti-phosphorylated p38 antibody (1:100) (Cell Signaling Technology, Inc.) and revealed with secondary Alexa Fluor 569 anti-mouse antibody (1:500) (Invitrogen). DAPI (Sigma) was used for counterstaining. Anti-osteopontin (1:200) and anti-CD106 (1:200) antibodies (R&D Systems) were used as primary antibodies for immunostaining of frozen sections of the femurs.
Mouse femurs were removed from alcohol storage and dried superficially on paper tissue before being wrapped in parafilm to prevent drying during scanning. Each parafilm-wrapped femur (n = 3 for each group) was placed in a plastic/polystyrene foam tube that was mounted vertically in the scanner. Micro-CT of mouse anatomy was performed with a SkyScan 1172 Micro x-ray CT scanner (Micro Photonics, Inc., Allentown PA) with the x-ray source (focal spot size, 4 μm, energy range 20–100 kV) biased at 50 kV/200 microamps and with a 0.5 A mm filter to reduce beam hardening. The images were acquired with a pixel size of 6.87 μm with the camera-to-source distance of 221.553 mm and an object-to-source distance of 65.935 mm. 799 projections were acquired with an angular resolution of 0.4 degrees through 180 degrees of rotation. Four frames were averaged for each projection radiograph with an exposure time of 295 ms/frame. The scan duration was ~30 min/femur. Tomographic images were reconstructed using vendor-supplied software on the basis of the Feldkamp cone beam algorithm.
Quantitative data are expressed as mean ± S.E. Two-tailed Student's t tests were used where comparisons between two groups were made. One-way analyses of variance with Bonferroni multiple comparisons post hoc tests (only computed if overall p < 0.05) were used when comparisons of three or more treatment groups were made (IBM SPSS Statistics 17).
Available data suggest that the mouse Myoc gene is expressed highly in joint tissues. Because cultured MSCs are capable of differentiating into chondrocytes under appropriate conditions (19, 20), we examined whether myocilin is also present in MSCs. Myocilin protein was detected in mouse, rat, and human MSCs, with hMSCs demonstrating the highest levels of myocilin antibody reactivity (Fig. 1, A and B). The level of myocilin in mMSCs was about 4-fold lower than in mouse skeletal muscles, a tissue known to express relatively high levels of myocilin (13, 21). Similarly, the level of MYOC mRNA in hMSCs was about 2-fold lower than in human skeletal muscles as judged by qRT-PCR analysis (Fig. 1C). Immunostaining of hMSCs demonstrated that myocilin was located preferentially in the endoplasmic reticulum and Golgi apparatus (Fig. 1, D and E), similar to its localization in trabecular meshwork cells (18, 22).
Myocilin was discovered as a protein that is induced by glucocorticoids in the ocular trabecular meshwork (23). Likewise, we found that myocilin could be induced by glucocorticoids in MSCs with kinetics typical for an indirectly responsive glucocorticoid-inducible gene (23, 24). The level of myocilin induction was more than 2-fold at 4 days and more than 4-fold at 8 days following addition of 100 nm dexamethasone (Supplemental Fig. S1, A and B). To test whether MSC differentiation is associated with changes in myocilin expression, we compared myocilin levels in the lysates of undifferentiated hMSCs and hMSC cultured in chondrogenic, adipogenic, or osteogenic conditions. Myocilin levels were increased by more than 2-fold in hMSCs differentiating into osteoblasts as compared with undifferentiated hMSCs or hMSCs differentiating into chondrocytes or adipocytes (Fig. 2, A–D). The levels of MYOC mRNA were also increased about 2-fold in hMSCs differentiating into osteoblasts as compared with undifferentiated hMSCs or hMSCs differentiating into adipocytes (Fig. 2E). In summary, these results demonstrate that myocilin is expressed in MSCs and that its level increases when these cells differentiate into osteoblasts.
Myocilin is a secreted glycoprotein that may interact with frizzled receptors on the cell surface (14) and with extracellular proteins (25, 26). Because the expression of myocilin was increased in MSCs differentiating into osteoblasts, we tested whether treatment with extracellular myocilin might affect MSC osteogenesis. Addition of myocilin (3 μg/ml) to hMSCs did not affect their differentiation into adipocytes as judged by Oil Red O staining and PPARγ expression (Fig. 3, A–D), but stimulated their differentiation into osteoblasts as evaluated by Alizarin Red S staining (Fig. 3, E and F). Moreover, the effect of exogenously applied myocilin on MSC osteogenesis was dose-dependent through a range of 1–3 μg/ml, as judged by Alizarin Red S staining and measurements of AP activity (Fig. 4A). Higher tested concentrations of exogenous myocilin (≥ 6 μg/ml) did not lead to further elevation of AP activity compared with 3 μg/ml of myocilin (Fig. 4A), so we used 3 μg/ml of myocilin in most subsequent experiments. Addition of antiserum against myocilin (1:1000 dilution) simultaneously with myocilin eliminated the pro-osteogenic effect of myocilin (Fig. 4A).
To determine whether the presence of endogenous myocilin contributes to osteoblastic differentiation by MSCs, we compared differentiation of mMSCs from wild-type and Myoc-null mice. As judged by AP activity measurements, mMSCs from Myoc-null mice demonstrated less osteogenesis when cultured in osteoblast differentiation medium (Fig. 4B). Addition of myocilin (1–3 μg/ml) to differentiating mMSCs from Myoc-null mice led to elevation of AP activity in a dose-dependent manner but with higher concentrations of myocilin (6 μg/ml) showing reduced stimulatory effects compared with the peak effect at 3 μg/ml. The maximal level of myocilin-mediated stimulation of relative AP activity in Myoc-null mMSCs (~4-fold) was lower than that for wild-type mMSCs (~8-fold, Fig. 4, A–C).
Several signaling pathways have been implicated in the regulation of MSC osteoblastic differentiation (7). Although myocilin may serve as a modulator of Wnt signaling, in vitro experiments studying effects of canonical and noncanonical Wnt signaling on osteogenesis have been inconclusive (7). Therefore, we sought to determine how myocilin affects other signaling pathways with known involvement in osteogenesis by osteoblasts and/or MSCs, such as the JNK and p38 MAPK pathways (27, 28). Addition of myocilin (3 μg/ml) simultaneously with osteoblast differentiation medium to hMSCs increased the levels of activated p38 (p-p38), activated JNK (p-JNK), and activated Erk1/2 (p-Erk1/2, a downstream mediator of MAPK signaling) measured 1 h after initiation of differentiation (Fig. 5, A–F). Addition of the inhibitors of p38 (SB203580) (29, 30) or JNK (SP600125) (31) eliminated the stimulating effects of myocilin (Fig. 5, A–D). Although SB203580 and SP600125 have been utilized throughout the published literature as specific inhibitors of p38 and JNK, respectively, it has been reported that they may have additional targets (32). To address the issue of inhibitor specificity, we employed another set of inhibitors, BIRB 796 (32, 33) and BI 78D3 (34), that act by mechanisms that are different from SB203580 and SP600125. Similar to the results with SB203580 and SP600125, treatment with alternative inhibitors of p38 (BIRB 796) or JNK (BI 78D3) significantly reduced the stimulating effects of myocilin, as measured 1 h after initiation of differentiation (Fig. 6, A–C).
Prolonged exposure (6 days) of hMSCs to myocilin in the course of their differentiation into osteoblasts produced similar effects on MAPK activation (data not shown). The levels of p-p38, p-JNK, and p-Erk1/2 were increased about 4-, 2-, and 2.5-fold, respectively, compared with hMSCs that were differentiated in the absence of myocilin, whereas addition of antiserum against myocilin together with myocilin completely eliminated a stimulating effect of myocilin, confirming the specificity of myocilin action (Figs. 7 and and8).8). Myocilin treatment led to migration of activated p38 to the nucleus, whereas addition of antiserum against myocilin prevented accumulation of p-p38 in the nucleus (Fig. 7C). Addition of SB203580 or SP600125 completely eliminated or significantly reduced the osteogenesis-stimulating effects of myocilin, as judged by Alizarin Red S staining and AP assays (Figs. 7, D and E, and and88E). As an alternative method of suppressing p38 and JNK we used siRNAs. Transfection of hMSCs with 100 nm of different p38 siRNAs or JNK siRNAs 4 days after initiation of differentiation significantly reduced the levels of p38 or JNK, respectively, 48 h after transfection (supplemental Fig. S2, A–D). We selected two different p38 siRNAs and two different JNK siRNAs for subsequent experiments. Transfection of hMSCs with p38 or JNK siRNAs, but not with scrambled siRNAs, reduced the stimulatory effect of myocilin (3 μg/ml) on osteogenic differentiation, as judged by AP activity measurements 6 days after initiation of differentiation (supplemental Fig. S2E). The inhibitory effects of siRNAs were less pronounced than the ones observed with specific inhibitors of p38 and JNK. This may be due to the fact that siRNAs were present only during last 48 h of differentiation of hMSCs into osteoblasts.
MSC osteogenesis involves many steps controlled by a hierarchy of transcription factors. Molecular and genetic studies have established a critical role for the Runx2 transcription factor in osteogenesis (35). Another important transcription factor that is involved in osteoblastic differentiation and may operate up-stream of Runx2 is Dlx5 (36). We measured mRNA levels for these transcription factors as well as for a late and specific marker of bone formation, osteocalcin, in mMSCs isolated from wild-type and Myoc-null mice that were differentiating into osteoblasts in the presence or absence of myocilin for 4 days. The level of Dlx5 mRNA was higher, whereas the levels of Runx2 and osteocalcin mRNAs did not differ significantly in MSCs from wild-type mice compared with MSCs from Myoc-null mice differentiating in the absence of Myocilin treatment (Fig. 9A). The levels of Dlx5 and Runx2 mRNAs were elevated in mMSCs from both wild-type and Myoc-null mice differentiating into osteoblasts in the presence of myocilin compared with without myocilin treatment, although the rise in expression of these factors was blunted in Myoc-null mMSCs compared with WT mMSCs. The levels of osteocalcin mRNA did not change significantly as a result of myocilin treatment after 4 days. Osteocalcin is a late osteogenic differentiation marker as compared with Dlx5 and Runx2, and longer myocilin treatment is probably necessary to produce elevated levels of osteocalcin. Addition of antibodies against myocilin eliminated the stimulating effect of myocilin on Runx2 and Dlx5 expression. The levels of Runx2 and Dlx5 proteins were also increased in hMSCs differentiating into osteoblasts after treatment with myocilin (Fig. 9B). Thus, it appears that myocilin stimulates the osteogenic differentiation of MSCs by activating osteogenic transcription factors that potentially lie downstream of the JNK and p38 MAP kinase pathways.
The impaired ability of mMSCs from Myoc-null mice to differentiate into osteoblasts in culture suggested that this may correlate with defects in bone formation in these mice compared with wild-type littermates. As the first step to test this hypothesis, we immunostained frozen sections of the femur bone marrow from wild-type and Myoc-null mice with antibodies against osteopontin and CD106, a positive marker for mouse MSCs. A significant decrease in the staining intensity was observed for both markers (Fig. 10, A and B). Subsequently, we employed microcomputed tomography (micro-CT) to characterize the three-dimensional trabecular and cortical bone structures of wild-type and Myoc-null mouse femurs. The results showed that cortical bone thickness (Fig. 10, E and F) as well as trabecular volume (C and D) were reduced in the Myoc-null mice compared with wild-type mice. These data suggest that myocilin plays a role in bone formation and/or maintenance in vivo.
Although several major signaling pathways have been implicated in the differentiation of MSCs into osteoblasts (7), the mechanisms controlling osteogenic differentiation are still not well understood. In this work we demonstrate that myocilin is expressed in MSCs and identify myocilin as a novel molecular participant in the differentiation of MSCs into osteoblasts.
Myocilin was originally identified as a protein that was highly induced by glucocorticoids in human trabecular meshwork cells in vitro. There are only a few cell types that express myocilin in culture, and myocilin is not induced by glucocorticoids in cell lines that do not express myocilin. Myocilin expression in MSCs and its inducibility by glucocorticoids suggest that these cells would serve as a reasonable model to study regulation of the MYOC gene. The trabecular meshwork is a part of the aqueous humor outflow pathway located at the angle of the anterior chamber of the eye. In humans, the trabecular meshwork is made up of collagen beams covered by endothelial-like cells. The space between the beams is filled with extracellular material/matrix. Microarray analysis of gene expression has demonstrated that some genes essential for cartilage and bone formation, such as matrix Gla protein and osteopontin, are also abundantly expressed in the trabecular meshwork (37, 38). These observations led to a suggestion that calcification of the trabecular meshwork may occur in certain conditions (see (39) for review). Mutations in the MYOC gene may lead to glaucoma. Although the exact mechanisms are still not clear, it has been suggested that accumulation of mutated myocilin induces the unfolded protein response, makes cells more sensitive to oxidative stress, and, consequently, leads to deterioration of trabecular meshwork function and/or trabecular meshwork cell death (40–42). At the same time, no other phenotypes have been reported in humans carrying MYOC mutations or in mice expressing mutated myocilin (43–45). Myoc-null mice appeared to be normal and do not develop glaucoma, but careful analysis of non-ocular phenotypes have not been performed. Our recent data demonstrated that the absence of myocilin led to a moderate reduction in the amount of syntrophin associated with dystrophin and reduced levels of phospho-Akt in the skeletal muscle of Myoc-null mice compared with wild-type littermates (13). Data presented in this study show that Myoc-null mice exhibit reduced cortical bone thickness and trabecular volume as well as reduced levels of femoral osteopontin. Osteopontin is an abundant non-collagenous sialoprotein in the bone matrix produced by osteoblasts that plays a role their differentiation as well as in bone remodeling (46), osteoclast attachment, and resorption (47).
There is a growing number of cases showing that mutations in interacting proteins may produce similar phenotypes (48). It has been shown that myocilin interacts with sparc (25), a secreted protein that is present in high levels in MSCs and bones (49). MSCs from sparc-null mice contain fewer osteoblastic precursors than control mice, and sparc-null osteoblasts are less committed to osteoblastic differentiation than wild-type osteoblasts (50). Sparc-null mice show decreased bone mineral density and bone mineral content and increased mechanical fragility of bone (49). Future work should determine whether the biomechanical properties of bones from Myoc-null mice are affected similarly.
We have shown that myocilin is expressed in MSCs from different species and that its expression level increases when these cells differentiate along the osteoblastic lineage. Although mMSCs express lower levels of myocilin compared with rat MSCs or hMSCs, even these levels of myocilin appear to play a physiological role in vitro and in vivo. mMSCs from Myoc-null mice have a reduced ability to differentiate into the osteoblastic lineage. Addition of extracellular myocilin stimulates differentiation of MSCs along the osteoblastic lineage in vitro, and this stimulation includes the activation of the p38, Erk1/2, and, to a lesser extent, JNK signaling pathways. It has been shown that the Erk-MAPK pathway plays an important role in Runx2 regulation, osteoblast differentiation, and fetal bone development in vivo (51). Activation of p38, but not Erk1/2 or JNK MAPK, and the induction of MSC differentiation into osteoblasts by gold nanoparticles have been reported previously reported (52). At the same time, integrin α5 has been shown to induce osteogenic differentiation of MSCs by activating Erk1/2 but not p38 MAPK (53). After addition of myocilin, similar as in other systems, phosphorylated p38 is translocated into the nucleus where it may phosphorylate Dlx5 and Runx2, resulting in increased transcriptional activity of those factors (54, 55).
It is now well recognized that MSCs may facilitate bone repair and regeneration (56). In particular, MSCs from bone marrow and other sources contribute to bone repair in rodent models (56). MSCs can be introduced therapeutically to sites of regeneration by systemic infusion or via scaffolds loaded with MSCs. Several reports have demonstrated that genetic modification of MSCs (53) or treatment of MSCs with certain growth factors, such as epidermal growth factor (57), before application may accelerate bone repair and regeneration. We have demonstrated that treatment of MSCs with myocilin activates MAPK pathways and accelerates differentiation of these cells onto osteoblasts. Potentially, myocilin may be considered as a potential target that could be employed to improve the bone regenerative potential of MSCs. Moreover, the role that myocilin may play in natural osteogenesis or bone remodeling in vivo should be explored further.
We thank Danielle Donahue (National Institutes of Health Mouse Imaging Facility) for help with micro-CT scanning and data analysis.
*This work was supported, in whole or in part, by the Intramural Research Program of the National Institutes of Health, National Eye Institute.
This article contains supplemental Figs. S1 and S2.
3The abbreviations used are: