Granulin epithelin precursor: a bone morphogenic protein 2-inducible growth factor that activates Erk1/2 signaling and JunB transcription factor in chondrogenesis.
Journal: 2010/July - FASEB Journal
ISSN: 1530-6860
Abstract:
Granulin epithelin precursor (GEP) has been implicated in development, tissue regeneration, tumorigenesis, and inflammation. Herein we report that GEP stimulates chondrocyte differentiation from mesenchymal stem cells in vitro and endochondral ossification ex vivo, and GEP-knockdown mice display skeleton defects. Similar to bone morphogenic protein (BMP) 2, application of the recombinant GEP accelerates rabbit cartilage repair in vivo. GEP is a key downstream molecule of BMP2, and it is required for BMP2-mediated chondrocyte differentiation. We also show that GEP activates chondrocyte differentiation through Erk1/2 signaling and that JunB transcription factor is one of key downstream molecules of GEP in chondrocyte differentiation. Collectively, these findings reveal a novel critical role of GEP growth factor in chondrocyte differentiation and the molecular events both in vivo and in vitro.
Relations:
Content
Citations
(51)
References
(76)
Chemicals
(6)
Organisms
(6)
Processes
(3)
Anatomy
(5)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
FASEB J 24(6): 1879-1892

Granulin epithelin precursor: a bone morphogenic protein 2-inducible growth factor that activates Erk1/2 signaling and JunB transcription factor in chondrogenesis

MATERIALS AND METHODS

Plasmid constructs

An ∼1.6-kb GEP promoter fragment (corresponding to bases −1573 to +62) was subcloned into the HindIII and SacI sites of the pGL3 basic luciferase reporter vector (Promega, Madison, WI, USA) to generate −1573GEPluc. The fragments of the human GEP gene between −1393 and +62, −1175 to +62, −887 to +62, −570 to +62, −275 to +62, and −51 to +62 were amplified by PCR using −1573GEPluc as a template with the 3′ oligonucleotide (HindIII restriction site is underscored) of 5′-GTCAAGCTTCTGTTGTCTCCGGCTGAGACT-3′) and the following 6 different 5′ oligonucleotides (SacI restriction site is underscored): −1393 and +62, 5′-TACGAGCTCCACTGGCATTGAACATGGCA-3′; −1175 to +62, 5′-TACGAGCTCGCCGATCTGGAGACTAGGAA-3′; −887 to +62, 5′-TACGAGCTCCTGGCGCACAACCTTGTATC-3′; −570 to +62, 5′-TACGAGCTCCAGGCCGCAGTGAGCTATGA3′; −275 to +62, 5′-TACGAGCTCAGATCGTGCCACTGCACTCC3′; and −51 to +62, 5′TACGAGCTCCCGACGTCACATGATTCTCC3′. The amplified fragments were subcloned into the HindIII and SacI sites of the pGL3 basic for generation of −1393GEPluc, −1175GEPluc, −887GEPluc, −570GEPluc, −275GEPluc, and −51GEPluc.

The cDNA encoding the full-length human GEP was amplified with the sequence specific primers 5-tatAAGCTTGGCAGACCATGTGGACCCTGGTGAGC-3′, containing a HindIII restriction site (underscored), and 5′-GAATCTAGATCACAGCAGCTGTCTCAAGG-3′, containing an XbaII restriction site (underscored). After the double digestion with HindIII and XbaII, the cDNA was cloned in-frame into the HindIII/XbaII sites of pcDNA3.1/myc-His (A) vector (Invitrogen, Carlsbad, CA, USA). The construct was verified by nucleic acid sequencing; subsequent analysis was performed using CuraTools (Curagen, New Haven, CT, USA) and BLAST software (available at http://www.ncbi.nlm.nih.gov/BLAST/).

Generation of GEP stable line and purification of recombinant GEP protein

A total of 293 EBNA cells were cultured in tissue culture dishes in DMEM supplemented with 10% heat-inactivated FCS (Invitrogen) and antibiotics. Cells were cultured for 1 d before transfection using Lipofectamine reagent (Invitrogen), following the manufacturer’s instructions, at a density of 1.5 × 10 cells/30-mm plate. The plasmid pHis/Myc-GEP was transfected into 293 cells. Two days after transfection, cells were divided into 100-mm dishes at a density of 10 cells/dish in 10 ml of DMEM containing 1,000 μg/ml G418. After 14 d in selective medium (medium changed every 3 d), cells were expanded in DMEM containing 500 μg/ml G418.

To prepare recombinant GEP protein, we collected medium from the GEP stable line. For each liter of medium, 30 ml of nickel-nitrilotriacetic acid (Ni-NTA)-agarose was added, and the suspension was incubated overnight with agitation at 4°C. After sedimentation, the Ni-NTA-agarose was packed into a 3- × 30-cm column. With extensive washing, the bound protein was eluted, and the eluted protein was adjusted to 10 ng/μl (0.2% BSA was added to stabilize GEP for storage). Then 5 μl of purified proteins was analyzed by 8% SDS-PAGE and visualized by Coomassie Blue staining and Western blotting with polyclonal anti-GEP antibodies.

RNA preparation and real-time PCR

Total RNA was extracted from micromass cultures of C3H10T1/2 cells treated with 300 ng/ml recombinant BMP2 for various time points using an RNeasy kit (Qiagen, Valencia, CA, USA). One microgram of total RNA per sample was reverse-transcribed using the ImProm-II Reverse Transcription system (Promega). The following sequence-specific primers were synthesized: 5′-TGGTGGAGCAGCAAGAGCAA-3′ and 5′-CAGTGGACAGTAGACGGAGGAAA-3′ for mouse GEP; 5′-TGGTGGAGCAGCAAGAGCAA-3′ and 5′-CAGTGGACAGTAGACGGAGGAAA-3′ for mouse collagen II; 5′-CTGCTGCTAATGTTCTTGAC-3′ and 5′-ACTGGAATCCCTTTACTCTTT-3′ for mouse collagen X; 5′-TGATGACACTGCCACCTGTG-3′ and 5′-ACTCTGGCTTTGGGAAGAGC-3′ for mouse core-binding factor α1 (Cbfa1); 5′-CGCTCGCAATACGACTACGC-3′ and 5′-TAGAGCCCTGAGCCCTGTCC-3′ for mouse Sox9; 5′-CAGTGGAGTGTCCTGGTATT-3′ and 5′-GATCTCCGCGATCAGATGGT-3′ for mouse parathyroid hormone-related peptide (PTHrP); and 5′-GCTCGTGCCTCTTGCCTACA-3′ and 5′-CGTGTTCTCCTCGTCCTTGA-3′ for mouse Indian hedgehog (IHH). The following pair of oligonucleotides was used as internal controls: 5′-AGGTCGGTGTGAACGGATTTG-3′ and 5′-TGTAGACCATGTAGTTGAGGTCA-3′ for mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Reactions were performed in a 50-μl SYBR Green PCR volume in a 96-well optical reaction plate formatted in the 7300 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using the following PCR conditions: 40 cycles, 95°C for 15 s, and 60°C for 1 min. The transcript of GAPDH mRNA was used as an internal control for RNA quality. For each gene, three independent PCRs from the same reverse transcription sample were performed. The presence of a single specific PCR product was verified by melting curve analysis and confirmed on an agarose gel and further sequenced by the Applied Biosystems sequencing system.

To examine the effects of cytokines on GEP expression, human chondrocytes adapted to serum-free medium conditions for 24 h were treated with the following factors separately: TGF-β (5 ng/ml), BMP2 (100 ng/ml), IL-1β (5 ng/ml), TNF-α (5 ng/ml), or noggin (an antagonist of BMPs, 100 ng/ml). Twenty-four hours later, the cells were harvested for RNA isolation and quantitation of GEP by real-time PCR.

To examine the effect of GEP on BMP2, TGF-β, and TNF-α, human chondrocytes were cultured in the absence [control (CTR)] or presence of BMP2 (50 or 100 ng/ml) or GEP (50 or 100 ng/ml) for 2 d. Total RNA was then isolated for RT-PCR.

To determine whether blockage of Erk1/2 affects GEP-mediated chondrogenesis, micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis. Then 0.01% DMSO (v/v, GEP) or 1 μM U0126 (GEP+U0126) was added to the GEP-pretreated groups and incubated for additional 7 d, and real-time PCR was performed as described above.

To determine whether JunB is important for GEP-mediated chondrogenesis, micromass cultures of C3H10T1/2 cells stably transfected with either pSuper vector, pSuper-JunB encoding a small interfering RNA (siRNA) against JunB, or pCMV-JunB expression plasmid as well as various combinations were cultured in the absence or presence of 100 ng/ml GEP for 7 d for induction of chondrogenesis. Then real-time PCR was performed as described above.

In situ hybridization

The relevant digoxigenin-labeled mRNA antisense probes were prepared from cDNA templates for GEP, alkaline phosphatase, collagen type II, and collagen type X. Sections were dewaxed, treated with proteinase K, and incubated in hybridization buffer containing the riboprobe. Probe was added at an approximate concentration of 0.25 μg/ml. Stringency washes of saline sodium citrate solution were done at 52°C and a further wash was done in maleic acid buffer with 1% Tween 20. Slides were treated with antidigoxigenin antibody (Roche Diagnostics, Indianapolis, IN, USA). For color detection, slides were incubated in nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate solution (Vector Red Substrate Kit; Vector Laboratories, Burlingame, CA, USA) and counterstained with methyl green.

Immunohistochemistry

All tissue samples were fixed in 4% paraformaldehyde/PBS after dissection and dehydrated with gradually increasing concentrations of ethanol. Tissue samples from postnatal pups or mice were decalcified in 2% EDTA for 2 wk before being embedded in paraffin. The 5-μm-thick sections were immunostained for GEP. The sections were pretreated with chondroitinase (Sigma-Aldrich Corp., St. Louis, MO, USA) for 30 min at 37°C followed by protein block (Serum-Free Protein Block; DakoCytomation A/S, Copenhagen, Denmark) for 10 min at room temperature to reduce nonspecific staining. Affinity-purified polyclonal anti-GEP (46) was diluted at 1:100 and incubated overnight at 4°C. For detection, anti-rabbit link and streptavidin label from the Super Sensitive Multilink-Alk Phos Detection Kit (BioGenex Laboratories Inc., San Ramon, CA, USA) was used. 3,3′-Diaminobenzidine (0.5 mg/ml) in 50 mM Tris-Cl substrate (Sigma-Aldrich) was used for visualization, and sections were then counterstained with methyl green.

Western blotting for examining the induction of JunB by GEP

Micromass cultures of C3H10T1/2 progenitor cells were incubated in the presence of 100 ng/ml recombinant GEP for induction of chondrocyte differentiation. Total cell extracts were mixed with 5× sample buffer (312.5 mM Tris-HCl, pH 6.8; 5% β-mercaptoethanol; 10% SDS; 0.5% bromphenol blue; and 50% glycerol). Proteins were resolved on a 10% SDS-polyacrylamide gel and electroblotted onto a nitrocellulose membrane. After blocking in 5% nonfat dry milk in Tris buffer-saline-Tween 20 (10 mM Tris-HCl, pH 8.0; 150 mM NaCl; and 0.5% Tween 20), blots were incubated with rabbit polyclonal anti-JunB antibody (diluted 1:1000) for 1 h. After washing, the secondary antibody (horseradish peroxidase-conjugated anti-rabbit immunoglobulin; 1:2000 dilution) was added, and bound antibody was visualized using an enhanced chemiluminescence system (Amersham Life Science, Arlington Heights, IL, USA).

Assay for GEP-mediated chondrogenesis of murine bone marrow stem cells (BMSCs)

Chondrogenic differentiation medium consists of high-glucose (4.5 g/L) DMEM supplemented with 6.25 μg/ml insulin, 6.25 μg/ml transferrin, 6.25 μg/ml selenous acid, 5.33 μg/ml linoleic acid, and 1.25 mg/ml BSA (ITS+; Collaborative Research, Cambridge, MA), 2 mM pyruvate, and antibiotics, 0.1 μM dexamethasone, 50 μg/ml ascorbate 2-phosphate, 100 ng/ml BMP2, or 100 ng/ml GEP. BMP2 should be aliquoted for single use and stored at −80°C to avoid cycles of freeze/thawing that can cause degradation.

Assay for GEP-mediated chondrogenesis of human mesenchymal stem cells (hMSCs)

Chondrogenic differentiation was induced by placing 2.5 × 10 hMSCs into the defined chondrogenic medium and subjecting the cells to gentle centrifugation (800 g for 5 min) in a 15-ml conical polypropylene tube, after which the cap was loosened, and the tube was placed in the incubator, where the cells adhered to one another and consolidated into a cell pellet within 24 h. Three to 4 d later, the hMSCs had formed a 1-mm ball in the bottom of the tube. The chondrogenic medium was made with fresh BMP2 or GEP every 3–4 d, and the medium was changed by careful aspiration, because the cell pellets were free floating. If the cell pellets were found attached to the tube wall, they were gently dislodged with a pipette tip. Three weeks later, for histological analysis, the pellets were fixed in 4% formaldehyde, paraffin-embedded, sectioned, and analyzed by immunostaining for collagen II or collagen X expression. Sections can also be stained with 0.1% Safranin O for detection of proteoglycans.

Culture of fetal mouse bone explants

Fetal mouse metatarsal bones were dissected from 15-d-old pregnant fetal FVB/N mice and cultured in aMEM (Gibco; Invitrogen) containing 1% heat-inactivated FCS and 100 U/ml penicillin/streptomycin in the absence or presence of 100 ng/ml recombinant GEP for 5 d in sextuplicate. For Alizarin red/Alcian blue staining (staining for bone and cartilage), the explants were placed in 4% paraformaldehyde in PBS for overnight fixation. Subsequently, explants were placed in staining solution (0.05% Alizarin red, 0.015% Alcian blue, and 5% acetic acid in 70% ethanol) for 45–60 min. Digital images of stained bones were analyzed. For Safranin O/Fast Green staining, explants were fixed in 96% alcohol and processed for paraffin embedding. Sections were stained with 0.1% Safranin O (orange stain) to evaluate cartilage matrices, and 0.03% Fast Green to evaluate morphological features, as described previously (54).

Generation and characterization of silencing GEP (siGEP)-knockdown transgenic mice

A 19-bp 5′-GCCTATCCAAGAACTACAC-3′ oligo (siGEP) and its antisense oligo with a loop were cloned into a pBS/U6-ploxPneo vector [a gift provided by Chuxia Deng, U.S. National Institutes of Health (NIH), Bethesda, MD, USA]. The transgene was digested by ApaI and EcoRI restriction enzymes and purified from agarose gel after electrophoretic migration using the QIAquick Gel Extraction Kit (Qiagen). Seven transgenic founders with a C57B6 background were obtained by pronuclear injection according to standard techniques. The transgenic lines were genotyped using PCR with a pair of primers (5′-CGAAGTTATCTAGAGTCGAC-3′ and 5′-AAACAAGGCTTTTCTCCAAGG-3′) that amplifies ∼100 bp from the U6 promoter and the connecting neo gene. Two of 7 independent founders were used for crossing to Sox2 Cre mice (stock number 004783 from the Jax mice database). All animal studies were performed in accord with the guidelines of and approved by the Institutional Animal Care and Use Committee of Baylor College of Dentistry. Antibodies used for BrdU and alkaline phosphatase were purchased from Sigma-Aldrich.

Cartilage repair assay

Approval was obtained from the Institutional Animal Care and Use Committee of Baylor College of Dentistry before this arm of the study was performed. After induction of anesthesia with an intramuscular injection of ketamine and xylazine, full-thickness articular cartilage defects were created in the distal femora of skeletally mature (9-mo-old) male New Zealand White rabbits following our published protocol (55). A midline longitudinal incision and medial arthrotomy with lateral subluxation of the patella was followed by the creation of a full-thickness (2-mm-deep), 3-mm-diameter femoral trochlear osteochondral defect using a Dremel power tool (Robert Bosch, Mount Prospect, IL, USA) under steady irrigation. Each defect was then grafted, the patella was reduced, and the wound was closed. Rabbits were allowed unrestricted cage movement. Defects were treated with one of 3 implants: collagen sponge only (n=6); sponge containing 6 μg of recombinant GEP protein (n=6); or sponge containing 6 μg of recombinant human BMP2 protein (n=6; R&D Systems, Minneapolis, MN, USA; dosage based on our previous studies, ref. 11). At 6 wk postoperatively, animals were sacrificed, and specimens were processed for routine histological analysis with hematoxylin and eosin stain. Additional sections were stained with toluidine blue to highlight glycosaminoglycan distribution in the repair. Quality of repair was evaluated blindly using a modified O’Driscoll histological grading system for cartilage repair (55).

Chromatin immunoprecipitation (ChIP)

In vivo binding of Smad4 to the GEP promoter was investigated using a ChIP assay kit (Upstate Biotechnology, Lake Placid, NY, USA). RCS cells transfected with GEP-specific reporter construct −1575GEPluc were cultured in the absence or presence of 100 ng/ml BMP2 for 24 h and then were treated with formaldehyde (final 1%) to cross-link Smad4 to the DNA. Cells were washed with cold PBS and lysed with SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.1). The lysate was sonicated to shear DNA to a length between 200 and 1000 bp. The sonicated supernatant was diluted 10-fold with ChIP dilution buffer (0.01% SDS; 1% Triton X-100; 2 mM Tris-HCl, pH 8.1; and 150 mM NaCl) and incubated with either antibody against Smad4 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) or control IgG overnight at 4°C with rotation. To collect DNA-Smad4-antibody complex, salmon sperm DNA/protein A-agarose slurry was added to the mixture and incubated for 1 h at 4°C with rotation, and the DNA/protein A-agarose complex was pelleted by centrifugation. After extensive washing of the pellet in a series of washing buffers, the pellet was dissolved with 250 μl of elution buffer and centrifuged to remove agarose. The supernatant was treated with 20 μl of 5 M NaCl and heated to 65°C for 4 h to reverse the Smad4-DNA cross-link. After treatment with EDTA and proteinase K, the supernatant was extracted with phenol/chloroform and precipitated with ethanol to recover the DNA. For PCR of the GEP minimal promoter region using the chromatin-immunoprecipitated DNA, one-tenth of the DNA was PCR-amplified using forward primer 5′-AGATCGTGCCACTGCACTCC-3′ and reverse primer 5′-CTGTTGTCTCCGGCTGAGACT-3′. Thirty-five cycles of PCR at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s were performed. PCR products were analyzed by 1% agarose gel.

Reporter gene assay

RCS cells grown to ∼50% confluence in 35-mm culture dishes were transfected with 1 μg of the reporter construct (either −1573GEPPluc, −1393GEPluc, −1175GEPluc, −887GEPluc, −570GEPluc, −275GEPluc, −51GEPluc, or mpGEPluc that contains only the monomial promoter of GEP, −275 to −51) along with 1 μg of pSVGal plasmid (internal control) in the absence or presence of 100 ng/ml BMP2 or transfected with Smad1, Smad4, or Smad5 expression plasmid or various combinations. At 48 h after transfection, the cultures were harvested and lysed. Luciferase assays were performed using 20 μl of cell extract and 100 μl of luciferin substrate (Promega). β-Galactosidase assays were performed using a β-galactosidase assay kit (Tropix, Bedford, MA) according to the manufacturer’s protocol. β-Galactosidase and luciferase activities were measured using a Mini-Lum liminometer (BioScan, Washington, DC, USA).

Assay for GEP-mediated signaling and target genes in chondrocytes

Human C28I2 chondrocytes (provided by Dr. Mary B. Goldring, Hospital for Special Surgery, New York, NY, USA) were starved for 24 h and treated with 50 ng/ml GEP for various time points, and cell lysates were analyzed using the PathScan Multiplex Western Cocktail I (Cell Signaling Technology Inc., Danvers, MA, USA) following the manufacturer’s instructions.

To identify GEP target molecules, a genome-wide DNA chip analysis was performed. In brief, total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP for various time points and analyzed by microarray analysis (Affymetrix, Santa Clara, CA, USA) at the New York University Genomic Core Facility.

Statistical tests

One-way ANOVA was performed in R software to determine the significant differences (F>3.35, α<0.05) of the activity among different doses. In addition, Tukey’s test was also used in conjunction with ANOVA to find significant differences (P<0.05; P<0.001) of the levels of genes of interest.

Plasmid constructs

An ∼1.6-kb GEP promoter fragment (corresponding to bases −1573 to +62) was subcloned into the HindIII and SacI sites of the pGL3 basic luciferase reporter vector (Promega, Madison, WI, USA) to generate −1573GEPluc. The fragments of the human GEP gene between −1393 and +62, −1175 to +62, −887 to +62, −570 to +62, −275 to +62, and −51 to +62 were amplified by PCR using −1573GEPluc as a template with the 3′ oligonucleotide (HindIII restriction site is underscored) of 5′-GTCAAGCTTCTGTTGTCTCCGGCTGAGACT-3′) and the following 6 different 5′ oligonucleotides (SacI restriction site is underscored): −1393 and +62, 5′-TACGAGCTCCACTGGCATTGAACATGGCA-3′; −1175 to +62, 5′-TACGAGCTCGCCGATCTGGAGACTAGGAA-3′; −887 to +62, 5′-TACGAGCTCCTGGCGCACAACCTTGTATC-3′; −570 to +62, 5′-TACGAGCTCCAGGCCGCAGTGAGCTATGA3′; −275 to +62, 5′-TACGAGCTCAGATCGTGCCACTGCACTCC3′; and −51 to +62, 5′TACGAGCTCCCGACGTCACATGATTCTCC3′. The amplified fragments were subcloned into the HindIII and SacI sites of the pGL3 basic for generation of −1393GEPluc, −1175GEPluc, −887GEPluc, −570GEPluc, −275GEPluc, and −51GEPluc.

The cDNA encoding the full-length human GEP was amplified with the sequence specific primers 5-tatAAGCTTGGCAGACCATGTGGACCCTGGTGAGC-3′, containing a HindIII restriction site (underscored), and 5′-GAATCTAGATCACAGCAGCTGTCTCAAGG-3′, containing an XbaII restriction site (underscored). After the double digestion with HindIII and XbaII, the cDNA was cloned in-frame into the HindIII/XbaII sites of pcDNA3.1/myc-His (A) vector (Invitrogen, Carlsbad, CA, USA). The construct was verified by nucleic acid sequencing; subsequent analysis was performed using CuraTools (Curagen, New Haven, CT, USA) and BLAST software (available at http://www.ncbi.nlm.nih.gov/BLAST/).

Generation of GEP stable line and purification of recombinant GEP protein

A total of 293 EBNA cells were cultured in tissue culture dishes in DMEM supplemented with 10% heat-inactivated FCS (Invitrogen) and antibiotics. Cells were cultured for 1 d before transfection using Lipofectamine reagent (Invitrogen), following the manufacturer’s instructions, at a density of 1.5 × 10 cells/30-mm plate. The plasmid pHis/Myc-GEP was transfected into 293 cells. Two days after transfection, cells were divided into 100-mm dishes at a density of 10 cells/dish in 10 ml of DMEM containing 1,000 μg/ml G418. After 14 d in selective medium (medium changed every 3 d), cells were expanded in DMEM containing 500 μg/ml G418.

To prepare recombinant GEP protein, we collected medium from the GEP stable line. For each liter of medium, 30 ml of nickel-nitrilotriacetic acid (Ni-NTA)-agarose was added, and the suspension was incubated overnight with agitation at 4°C. After sedimentation, the Ni-NTA-agarose was packed into a 3- × 30-cm column. With extensive washing, the bound protein was eluted, and the eluted protein was adjusted to 10 ng/μl (0.2% BSA was added to stabilize GEP for storage). Then 5 μl of purified proteins was analyzed by 8% SDS-PAGE and visualized by Coomassie Blue staining and Western blotting with polyclonal anti-GEP antibodies.

RNA preparation and real-time PCR

Total RNA was extracted from micromass cultures of C3H10T1/2 cells treated with 300 ng/ml recombinant BMP2 for various time points using an RNeasy kit (Qiagen, Valencia, CA, USA). One microgram of total RNA per sample was reverse-transcribed using the ImProm-II Reverse Transcription system (Promega). The following sequence-specific primers were synthesized: 5′-TGGTGGAGCAGCAAGAGCAA-3′ and 5′-CAGTGGACAGTAGACGGAGGAAA-3′ for mouse GEP; 5′-TGGTGGAGCAGCAAGAGCAA-3′ and 5′-CAGTGGACAGTAGACGGAGGAAA-3′ for mouse collagen II; 5′-CTGCTGCTAATGTTCTTGAC-3′ and 5′-ACTGGAATCCCTTTACTCTTT-3′ for mouse collagen X; 5′-TGATGACACTGCCACCTGTG-3′ and 5′-ACTCTGGCTTTGGGAAGAGC-3′ for mouse core-binding factor α1 (Cbfa1); 5′-CGCTCGCAATACGACTACGC-3′ and 5′-TAGAGCCCTGAGCCCTGTCC-3′ for mouse Sox9; 5′-CAGTGGAGTGTCCTGGTATT-3′ and 5′-GATCTCCGCGATCAGATGGT-3′ for mouse parathyroid hormone-related peptide (PTHrP); and 5′-GCTCGTGCCTCTTGCCTACA-3′ and 5′-CGTGTTCTCCTCGTCCTTGA-3′ for mouse Indian hedgehog (IHH). The following pair of oligonucleotides was used as internal controls: 5′-AGGTCGGTGTGAACGGATTTG-3′ and 5′-TGTAGACCATGTAGTTGAGGTCA-3′ for mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Reactions were performed in a 50-μl SYBR Green PCR volume in a 96-well optical reaction plate formatted in the 7300 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using the following PCR conditions: 40 cycles, 95°C for 15 s, and 60°C for 1 min. The transcript of GAPDH mRNA was used as an internal control for RNA quality. For each gene, three independent PCRs from the same reverse transcription sample were performed. The presence of a single specific PCR product was verified by melting curve analysis and confirmed on an agarose gel and further sequenced by the Applied Biosystems sequencing system.

To examine the effects of cytokines on GEP expression, human chondrocytes adapted to serum-free medium conditions for 24 h were treated with the following factors separately: TGF-β (5 ng/ml), BMP2 (100 ng/ml), IL-1β (5 ng/ml), TNF-α (5 ng/ml), or noggin (an antagonist of BMPs, 100 ng/ml). Twenty-four hours later, the cells were harvested for RNA isolation and quantitation of GEP by real-time PCR.

To examine the effect of GEP on BMP2, TGF-β, and TNF-α, human chondrocytes were cultured in the absence [control (CTR)] or presence of BMP2 (50 or 100 ng/ml) or GEP (50 or 100 ng/ml) for 2 d. Total RNA was then isolated for RT-PCR.

To determine whether blockage of Erk1/2 affects GEP-mediated chondrogenesis, micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis. Then 0.01% DMSO (v/v, GEP) or 1 μM U0126 (GEP+U0126) was added to the GEP-pretreated groups and incubated for additional 7 d, and real-time PCR was performed as described above.

To determine whether JunB is important for GEP-mediated chondrogenesis, micromass cultures of C3H10T1/2 cells stably transfected with either pSuper vector, pSuper-JunB encoding a small interfering RNA (siRNA) against JunB, or pCMV-JunB expression plasmid as well as various combinations were cultured in the absence or presence of 100 ng/ml GEP for 7 d for induction of chondrogenesis. Then real-time PCR was performed as described above.

In situ hybridization

The relevant digoxigenin-labeled mRNA antisense probes were prepared from cDNA templates for GEP, alkaline phosphatase, collagen type II, and collagen type X. Sections were dewaxed, treated with proteinase K, and incubated in hybridization buffer containing the riboprobe. Probe was added at an approximate concentration of 0.25 μg/ml. Stringency washes of saline sodium citrate solution were done at 52°C and a further wash was done in maleic acid buffer with 1% Tween 20. Slides were treated with antidigoxigenin antibody (Roche Diagnostics, Indianapolis, IN, USA). For color detection, slides were incubated in nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate solution (Vector Red Substrate Kit; Vector Laboratories, Burlingame, CA, USA) and counterstained with methyl green.

Immunohistochemistry

All tissue samples were fixed in 4% paraformaldehyde/PBS after dissection and dehydrated with gradually increasing concentrations of ethanol. Tissue samples from postnatal pups or mice were decalcified in 2% EDTA for 2 wk before being embedded in paraffin. The 5-μm-thick sections were immunostained for GEP. The sections were pretreated with chondroitinase (Sigma-Aldrich Corp., St. Louis, MO, USA) for 30 min at 37°C followed by protein block (Serum-Free Protein Block; DakoCytomation A/S, Copenhagen, Denmark) for 10 min at room temperature to reduce nonspecific staining. Affinity-purified polyclonal anti-GEP (46) was diluted at 1:100 and incubated overnight at 4°C. For detection, anti-rabbit link and streptavidin label from the Super Sensitive Multilink-Alk Phos Detection Kit (BioGenex Laboratories Inc., San Ramon, CA, USA) was used. 3,3′-Diaminobenzidine (0.5 mg/ml) in 50 mM Tris-Cl substrate (Sigma-Aldrich) was used for visualization, and sections were then counterstained with methyl green.

Western blotting for examining the induction of JunB by GEP

Micromass cultures of C3H10T1/2 progenitor cells were incubated in the presence of 100 ng/ml recombinant GEP for induction of chondrocyte differentiation. Total cell extracts were mixed with 5× sample buffer (312.5 mM Tris-HCl, pH 6.8; 5% β-mercaptoethanol; 10% SDS; 0.5% bromphenol blue; and 50% glycerol). Proteins were resolved on a 10% SDS-polyacrylamide gel and electroblotted onto a nitrocellulose membrane. After blocking in 5% nonfat dry milk in Tris buffer-saline-Tween 20 (10 mM Tris-HCl, pH 8.0; 150 mM NaCl; and 0.5% Tween 20), blots were incubated with rabbit polyclonal anti-JunB antibody (diluted 1:1000) for 1 h. After washing, the secondary antibody (horseradish peroxidase-conjugated anti-rabbit immunoglobulin; 1:2000 dilution) was added, and bound antibody was visualized using an enhanced chemiluminescence system (Amersham Life Science, Arlington Heights, IL, USA).

Assay for GEP-mediated chondrogenesis of murine bone marrow stem cells (BMSCs)

Chondrogenic differentiation medium consists of high-glucose (4.5 g/L) DMEM supplemented with 6.25 μg/ml insulin, 6.25 μg/ml transferrin, 6.25 μg/ml selenous acid, 5.33 μg/ml linoleic acid, and 1.25 mg/ml BSA (ITS+; Collaborative Research, Cambridge, MA), 2 mM pyruvate, and antibiotics, 0.1 μM dexamethasone, 50 μg/ml ascorbate 2-phosphate, 100 ng/ml BMP2, or 100 ng/ml GEP. BMP2 should be aliquoted for single use and stored at −80°C to avoid cycles of freeze/thawing that can cause degradation.

Assay for GEP-mediated chondrogenesis of human mesenchymal stem cells (hMSCs)

Chondrogenic differentiation was induced by placing 2.5 × 10 hMSCs into the defined chondrogenic medium and subjecting the cells to gentle centrifugation (800 g for 5 min) in a 15-ml conical polypropylene tube, after which the cap was loosened, and the tube was placed in the incubator, where the cells adhered to one another and consolidated into a cell pellet within 24 h. Three to 4 d later, the hMSCs had formed a 1-mm ball in the bottom of the tube. The chondrogenic medium was made with fresh BMP2 or GEP every 3–4 d, and the medium was changed by careful aspiration, because the cell pellets were free floating. If the cell pellets were found attached to the tube wall, they were gently dislodged with a pipette tip. Three weeks later, for histological analysis, the pellets were fixed in 4% formaldehyde, paraffin-embedded, sectioned, and analyzed by immunostaining for collagen II or collagen X expression. Sections can also be stained with 0.1% Safranin O for detection of proteoglycans.

Culture of fetal mouse bone explants

Fetal mouse metatarsal bones were dissected from 15-d-old pregnant fetal FVB/N mice and cultured in aMEM (Gibco; Invitrogen) containing 1% heat-inactivated FCS and 100 U/ml penicillin/streptomycin in the absence or presence of 100 ng/ml recombinant GEP for 5 d in sextuplicate. For Alizarin red/Alcian blue staining (staining for bone and cartilage), the explants were placed in 4% paraformaldehyde in PBS for overnight fixation. Subsequently, explants were placed in staining solution (0.05% Alizarin red, 0.015% Alcian blue, and 5% acetic acid in 70% ethanol) for 45–60 min. Digital images of stained bones were analyzed. For Safranin O/Fast Green staining, explants were fixed in 96% alcohol and processed for paraffin embedding. Sections were stained with 0.1% Safranin O (orange stain) to evaluate cartilage matrices, and 0.03% Fast Green to evaluate morphological features, as described previously (54).

Generation and characterization of silencing GEP (siGEP)-knockdown transgenic mice

A 19-bp 5′-GCCTATCCAAGAACTACAC-3′ oligo (siGEP) and its antisense oligo with a loop were cloned into a pBS/U6-ploxPneo vector [a gift provided by Chuxia Deng, U.S. National Institutes of Health (NIH), Bethesda, MD, USA]. The transgene was digested by ApaI and EcoRI restriction enzymes and purified from agarose gel after electrophoretic migration using the QIAquick Gel Extraction Kit (Qiagen). Seven transgenic founders with a C57B6 background were obtained by pronuclear injection according to standard techniques. The transgenic lines were genotyped using PCR with a pair of primers (5′-CGAAGTTATCTAGAGTCGAC-3′ and 5′-AAACAAGGCTTTTCTCCAAGG-3′) that amplifies ∼100 bp from the U6 promoter and the connecting neo gene. Two of 7 independent founders were used for crossing to Sox2 Cre mice (stock number 004783 from the Jax mice database). All animal studies were performed in accord with the guidelines of and approved by the Institutional Animal Care and Use Committee of Baylor College of Dentistry. Antibodies used for BrdU and alkaline phosphatase were purchased from Sigma-Aldrich.

Cartilage repair assay

Approval was obtained from the Institutional Animal Care and Use Committee of Baylor College of Dentistry before this arm of the study was performed. After induction of anesthesia with an intramuscular injection of ketamine and xylazine, full-thickness articular cartilage defects were created in the distal femora of skeletally mature (9-mo-old) male New Zealand White rabbits following our published protocol (55). A midline longitudinal incision and medial arthrotomy with lateral subluxation of the patella was followed by the creation of a full-thickness (2-mm-deep), 3-mm-diameter femoral trochlear osteochondral defect using a Dremel power tool (Robert Bosch, Mount Prospect, IL, USA) under steady irrigation. Each defect was then grafted, the patella was reduced, and the wound was closed. Rabbits were allowed unrestricted cage movement. Defects were treated with one of 3 implants: collagen sponge only (n=6); sponge containing 6 μg of recombinant GEP protein (n=6); or sponge containing 6 μg of recombinant human BMP2 protein (n=6; R&amp;D Systems, Minneapolis, MN, USA; dosage based on our previous studies, ref. 11). At 6 wk postoperatively, animals were sacrificed, and specimens were processed for routine histological analysis with hematoxylin and eosin stain. Additional sections were stained with toluidine blue to highlight glycosaminoglycan distribution in the repair. Quality of repair was evaluated blindly using a modified O’Driscoll histological grading system for cartilage repair (55).

Chromatin immunoprecipitation (ChIP)

In vivo binding of Smad4 to the GEP promoter was investigated using a ChIP assay kit (Upstate Biotechnology, Lake Placid, NY, USA). RCS cells transfected with GEP-specific reporter construct −1575GEPluc were cultured in the absence or presence of 100 ng/ml BMP2 for 24 h and then were treated with formaldehyde (final 1%) to cross-link Smad4 to the DNA. Cells were washed with cold PBS and lysed with SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.1). The lysate was sonicated to shear DNA to a length between 200 and 1000 bp. The sonicated supernatant was diluted 10-fold with ChIP dilution buffer (0.01% SDS; 1% Triton X-100; 2 mM Tris-HCl, pH 8.1; and 150 mM NaCl) and incubated with either antibody against Smad4 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) or control IgG overnight at 4°C with rotation. To collect DNA-Smad4-antibody complex, salmon sperm DNA/protein A-agarose slurry was added to the mixture and incubated for 1 h at 4°C with rotation, and the DNA/protein A-agarose complex was pelleted by centrifugation. After extensive washing of the pellet in a series of washing buffers, the pellet was dissolved with 250 μl of elution buffer and centrifuged to remove agarose. The supernatant was treated with 20 μl of 5 M NaCl and heated to 65°C for 4 h to reverse the Smad4-DNA cross-link. After treatment with EDTA and proteinase K, the supernatant was extracted with phenol/chloroform and precipitated with ethanol to recover the DNA. For PCR of the GEP minimal promoter region using the chromatin-immunoprecipitated DNA, one-tenth of the DNA was PCR-amplified using forward primer 5′-AGATCGTGCCACTGCACTCC-3′ and reverse primer 5′-CTGTTGTCTCCGGCTGAGACT-3′. Thirty-five cycles of PCR at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s were performed. PCR products were analyzed by 1% agarose gel.

Reporter gene assay

RCS cells grown to ∼50% confluence in 35-mm culture dishes were transfected with 1 μg of the reporter construct (either −1573GEPPluc, −1393GEPluc, −1175GEPluc, −887GEPluc, −570GEPluc, −275GEPluc, −51GEPluc, or mpGEPluc that contains only the monomial promoter of GEP, −275 to −51) along with 1 μg of pSVGal plasmid (internal control) in the absence or presence of 100 ng/ml BMP2 or transfected with Smad1, Smad4, or Smad5 expression plasmid or various combinations. At 48 h after transfection, the cultures were harvested and lysed. Luciferase assays were performed using 20 μl of cell extract and 100 μl of luciferin substrate (Promega). β-Galactosidase assays were performed using a β-galactosidase assay kit (Tropix, Bedford, MA) according to the manufacturer’s protocol. β-Galactosidase and luciferase activities were measured using a Mini-Lum liminometer (BioScan, Washington, DC, USA).

Assay for GEP-mediated signaling and target genes in chondrocytes

Human C28I2 chondrocytes (provided by Dr. Mary B. Goldring, Hospital for Special Surgery, New York, NY, USA) were starved for 24 h and treated with 50 ng/ml GEP for various time points, and cell lysates were analyzed using the PathScan Multiplex Western Cocktail I (Cell Signaling Technology Inc., Danvers, MA, USA) following the manufacturer’s instructions.

To identify GEP target molecules, a genome-wide DNA chip analysis was performed. In brief, total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP for various time points and analyzed by microarray analysis (Affymetrix, Santa Clara, CA, USA) at the New York University Genomic Core Facility.

Statistical tests

One-way ANOVA was performed in R software to determine the significant differences (F>3.35, α<0.05) of the activity among different doses. In addition, Tukey’s test was also used in conjunction with ANOVA to find significant differences (P<0.05; P<0.001) of the levels of genes of interest.

RESULTS

Differential expression of GEP in the chondrogenesis of a micromass culture of C3H10T1/2 cells

To investigate GEP function in chondrogenesis, we first studied GEP expression profiles during chondrocyte differentiation using the C3H10T1/2 cell line, a pluripotent murine stem cell line that is widely used for in vitro chondrogenic studies (56,57,58). Micromass cultures of these cells were incubated in the presence of 100 ng/ml recombinant BMP2 for induction of chondrocyte differentiation. Cells were harvested at various time points followed by real-time PCR for measurements of GEP (Fig. 1A) and collagen X (Fig. 1B). As shown in Fig. 1, the level of GEP was relatively low until d 7, when it is doubled and thereafter remained at high levels during the differential stage. Note that the peak level of GEP was ∼2 d earlier than that of collagen X, a marker for hypertrophic chondrocytes, suggesting that GEP may regulate collagen X expression.

An external file that holds a picture, illustration, etc.
Object name is z380061077830001.jpg

Expression of GEP during chondrogenesis both in vitro and in vivo. A, B) Expressions of GEP and collagen X were examined in the course of chondrogenesis of a micromass culture of C3H10T1/2 cells. Micromass cultures of C3H10T1/2 cells were stimulated by BMP2 protein at various time points, as indicated, and the mRNA levels of GEP (A) and collagen type X (Col X; B) were assayed using real-time PCR. Units are arbitrary; the normalized values were calibrated against the d 0 time point, here given the value of 1. C) GEP immunostaining assay shows that GEP (arrows) is first detected in the late condensation stage (E12.5) and then at the differentiation stage (E13.5) and continues during entire chondrogenic developmental stages in both proliferating and hypertrophic zones (from E17.5 to 1 mo old). Scale bar = 100 μm.

GEP expression patterns in chondrocytes during both embryonic and postnatal development stages

Next we characterized the temporal and spatial expression pattern of GEP during skeletal development using an immunostaining assay at multiple time points, including embryonic day (E) 10.5 (onset of chondrogenesis that begins with the proliferation and subsequent condensation of mesenchymal cells), E12.5 (condensation), E13.5 (right after cartilage formation but before endochondral bone formation), and E17.5 (onset of skeletal growth), as well as postnatal developmental stages (newborn, 10-d, and 30-d). As revealed in Fig. 1C, GEP is detected at E12.5, and its level is increased in the center of the condensation and around it at E13.5. It demonstrates prominent expression in prehypertrophic chondrocytes at E15.5 and E17.5 and in newborn mice. A high level of GEP throughout the whole growth plate is observed at the age of 1 month, suggesting that the expression profile of GEP is closely linked to the entire chondrogenic period.

GEP stimulates chondrogenesis in vitro and endochondral bone formation ex vivo

Prominent expression of GEP in chondrocytes and its stimulation of chondrocyte proliferation (46) prompted us to determine whether GEP was able to induce chondrocyte differentiation. We first generated a stable line expressing His-tagged GEP in 293 EBNA cells, and purified recombinant GEP using commercial beads. Purified proteins (5 μl) were analyzed by 8% SDS-PAGE and visualized by Coomassie Blue staining with verification by Western blotting with anti-GEP antibody (Fig. 2A). Next we compared the roles of GEP (100 ng/ml) and BMP2 (100 ng/ml), a well-documented growth factor, during chondrogenesis using BMSCs, which are capable of differentiation into various lineages, including chondrocytes (59, 60). In brief, the high-density culture system was incubated in the absence (CTR) or presence of 100 ng/ml GEP or 100 ng/ml BMP2 (serving as a positive control) for 7 d. Chondrogenesis was monitored by analyzing the expressions of marker genes specific for chondrocytes (Fig. 2B). As for BMP2, GEP markedly induced the expression of aggrecan, collagen II, and collagen X. Note that GEP is twice as potent as BMP2 in inducing collagen X.

An external file that holds a picture, illustration, etc.
Object name is z380061077830002.jpg

GEP stimulates chondrogenesis in vitro. A) Characterization of recombinant GEP protein. A purified recombinant human GEP (10 ng/μl in 0.2% BSA) was either stained with Coomassie Brilliant Blue R-250 (left) or detected by Western blotting with a polyclonal anti-GEP antibody (right). B) Comparisons of GEP and BMP2 in stimulations of chondrogenesis of murine BMSCs. BMSC aggregate cultures were incubated in the absence (CTR) or presence of either 100 ng/ml BMP2 or 100 ng/ml GEP for 2 wk, followed by quantitative measurements of aggrecan, collagen II (Col II), and collagen X (Col X) using real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1. C) Comparable potency of GEP and BMP2 in stimulation of chondrogenesis and activation of collagen II and collagen X expressions. hMSC aggregate cultures were incubated in the absence (CTR) or presence of 100 ng/ml BMP2 or GEP for 3 wk followed by Safranin O staining (left, red) or immunostains of collagen II (middle, green) or collagen X (right, green).

Using a long-term culture system (3 wk) of high-density hMSC pellets, we demonstrated that GEP, like BMP2, induced chondrogenesis as reflected by positive stains with Safranin O (pink color, left panel, Fig. 2C) and immunostains for collagen II and collagen X (right panels, Fig. 2C).

The effect of GEP on endochondral bone formation was then studied in an ex vivo model of 14.5-d-old fetal mouse metatarsal bones. At the time of explantation, these explants consisted of undifferentiated cartilage. In a 5-d culture period, these explants underwent all sequential stages of endochondral bone formation. As shown in Fig. 3, GEP significantly stimulated chondrocyte hypertrophy, mineralization, and bone length.

An external file that holds a picture, illustration, etc.
Object name is z380061077830003.jpg

GEP stimulates chondrocyte hypertrophy, mineralization, and endochondral bone growth. A, B) Safranin O/Fast Green staining of metatarsal bones. Metatarsals from 14.5-d-old mouse embryos were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d and stained with Safranin O/Fast Green, shown in low-power (A) and high-power (B) microphotographs. C) Alizarin red/Alcian blue staining of metatarsals. Explants were fixed and processed for staining. A representative photograph of an explanted metatarsal is presented. D) Percentage changes in total (T) and mineralization (M) length of metatarsal bones. Percentage changes in bone length were calculated as (length at d 5 − length at d 0)/length at d 0. *P < 0.05 vs. control.

In vivo knockdown of GEP leads to striking defects in skeleton development

To define the in vivo physiological role of GEP in chondrogenesis, we created conditional GEP knockdown mice using an siRNA technique and the Cre/loxP system. This in vivo knockdown system was originally developed by C. X. Deng at the National Institute of Diabetes and Digestive and Kidney Diseases, NIH and has been successfully used to knock down FGF receptor 2 in cartilage (61). In U6-ploxPneo-GEP transgenic mice, siGEP (5′-AGAACTCCGGCTCCTACTA-3′) is driven by a U6 promoter that is disturbed by a loxP-flanked neomycin cassette (61). To generate conditional-knockdown GEP mice, the U6-ploxPneo-GEP transgenic line was crossed to Sox2-Cre transgenic mice, in which Cre enzyme is activated in early mesenchymal cells. Immunostains of GEP showed that GEP is largely undetectable in the GEP-knockdown growth plate compared with the same litters (wild-type control) (Fig. 4A); Safranin O stain, staining cartilage red, showed that the knockdown growth plate is only half the size of that in wild-type mice (Fig. 4B). The bromodeoxyuridine (BrdU) incorporation assay, measuring cell proliferation, showed a sharp reduction of chondrocyte proliferation in the GEP-knockdown proliferation zone (Fig. 4C). Furthermore, we tested whether knockdown of GEP would change marker genes critical for chondrogenesis by using an in situ hybridization assay. As shown in Fig. 4, alkaline phosphatase in all growth plate zones (Fig. 4D), collagen II in the proliferating zone (Fig. 4E), and collagen X in the hypertrophic zone (Fig. 4F) are dramatically reduced in GEP-knockdown mice. Collectively, the above in vivo data, consistent with the in vitro data, support the novel concept that GEP plays an important role in cartilage development.

An external file that holds a picture, illustration, etc.
Object name is z380061077830004.jpg

GEP knockdown (KD) leads to a sharp reduction of chondrogenic markers. A) Immunostaining of GEP in the growth plate of the femur from 3-wk-old wild-type (WT) (top) and conditional KD (bottom) mice crossing between siGEP transgenic mice and Sox2 Cre mice. B) Reduction in thickness of the growth plate stained with Safranin O stain in GEP KD (top) compared with control (bottom). C) Reduction of BrdU (reflection of cell proliferation) in the GEP KD proliferation zone (top) compared with control (bottom). D) Reduction of alkaline phosphatase mRNA in the GEP KD growth plate (top) compared with control (bottom). E) Sharp reduction of collagen II mRNA in the GEP KD growth plate (top) compared with control (bottom). F) Collagen X is largely undetectable in the GEP KD hypertrophic zone (top) compared with control (bottom).

GEP induces cartilage repair

Because GEP is able to induce chondrogenesis of mesenchymal stem cells and plays an important role in cartilage development (Figs. 223334),4), and GEP was shown previously to stimulate wound healing (42), we next determined whether GEP could be used for cartilage repair. Articular cartilage defects were created in the femoral trochlea of rabbits and were treated with one of 3 implants for 6 wk: a collagen sponge only (n=6), a collagen sponge containing 6 μg of GEP (n=6), or a collagen sponge containing 6 μg of BMP2 (n=6) (the dose selected was based on our previous studies; ref. 55). As shown in Fig. 5, implants with collagen alone had little effect on cartilage repair, as the defect region was filled with a smooth-surfaced mixture of fibrous tissue and bone, and the adjacent tissue showed loss of matrix staining and reduction of cellularity (Fig. 5A). The implants containing BMP2 (Fig. 5B) or GEP (Fig. 5C) show enhanced cartilage repair, as reflected by good restoration of anatomic features and positive Safranin O staining of cartilage under the repairing region. Although adherence to adjacent host cartilage was best with BMP2 treatment, the mean ± sd histological scores for GEP (17.5±3.11) and BMP2 (17.25±2.63) did not differ significantly, and both were significantly better than that for the collagen treatment alone (7±1.00). The results support a positive correlation between implantation of GEP and cartilage repair.

An external file that holds a picture, illustration, etc.
Object name is z380061077830005.jpg

GEP accelerates cartilage repair. AC) Articular cartilage and subchondral bone sections stained with Safranin O/Fast Green. There is a lack of cartilage repair in the trauma region of the control group that was implanted with a collagen sponge only (A; arrow). Implantations of the collagen sponge containing a recombinant BMP2 protein accelerates high-quality hyaline-appearing repair as reflected by well-integrated and organized new tissues stained positive with Safranin O (B; pink, arrow). Implantations of the collagen sponge containing the equivalent GEP display a similar hyaline-like repairing response (C; pink, arrow). D) Quantitation of tissue repair. Collagen sponge with BMP2 or GEP accelerates cartilage repair.

GEP is a BMP2-inducible growth factor and required for BMP-mediated chondrocyte differentiation

In searching for upstream molecules of GEP, we tested a few growth factors known to be important for chondrogenesis in the primary human chondrocytes. Our results showed that GEP mRNA is up-regulated 3-fold by BMP2 and 1.5-fold by TGF-β. IL-1β, and noggin had no apparent effects on GEP expression (Fig. 6A).

An external file that holds a picture, illustration, etc.
Object name is z380061077830006.jpg

GEP is a downstream molecule of BMP2 and is required for BMP2 stimulation of chondrogenesis. A) Effects of growth factors and cytokines on GEP mRNA in chondrocytes by real-time PCR. Expression of GEP mRNA was normalized against 18S rRNA. ***P < 0.001. B) siGEP, a GEP siRNA construct, reduces 80% of endogenous GEP mRNA. C3H10T1/2 cells were transfected with either pSuper (CTR) or siGEP, and level of GEP mRNA was measured by real-time PCR. C) Suppression of GEP by siRNAs inhibits BMP2-induced chondrogenesis. Micromass cultures of C3H10T1/2 cells transfected with either pSuper (CTR) or siGEP were used to test whether BMP2 (100 ng/ml)-induced chondrogenesis is GEP dependent. Expressions of marker genes, as indicated, were determined by real-time PCR.

Next, we tested whether GEP was required for BMP2-mediated chondrogenesis using the siRNA approach. As shown in Fig. 6B, siGEP markedly reduced 80% of the endogenous GEP in C3H10T1/2 cells. We then performed micromass cultures of C3H10T1/2 cells with either a pSuper vector (CTR) or a siGEP construct transfected first and followed by additions of BMP2 (100 ng/ml). Our real-time PCR assay showed that reductions of the endogenous GEP by siGEP sharply decrease chondrogenic responses induced by BMP2: 65% down of Sox9, 60% down of collagen II, 80% down of collagen X, and 70% down of aggrecan compared with the control group responses (Fig. 6C). These results support the concept that GEP is a key downstream molecule of BMP2 during chondrogenesis.

BMP2 and Smads activate GEP-specific reporter genes

To elucidate the molecular mechanism by which BMP2 activates GEP expression, we cloned regulatory elements in the human GEP 5′-flanking region from −1573 to +62 (62) with a series of 5′ deletion promoter-luciferase constructs (Supplemental Fig. 1) and tested their transcriptional activity in RCS cells known to express GEP (46). Deletions of the region from −1393 to −1175 and the region from −570 to −275 result in ∼2- and 2.4-fold increases in promoter activity, respectively, suggesting that these two regions function as silencers in controlling GEP gene transactivation. On the other hand, deletion of the region from −270 to −51 leads to the complete loss of the reporter activity, indicating that this region is probably the basic promoter of the GEP gene. Applications of BMP2 were able to activate all GEP promoter constructs containing the region between −275 and −51 (Fig. 7A) Furthermore, this basic promoter region directly responded to BMP2 (Fig. 7B).

An external file that holds a picture, illustration, etc.
Object name is z380061077830007.jpg

BMP2 and its mediators, Smads, activate the GEP-specific reporter genes. A) BMP2 activates GEP-specific reporter genes in RCS cells. Indicated segments from the 5′-flanking region of the GEP gene were linked to an simian virus 40 promoter and a DNA segment encoding luciferase. Numbers indicate distances in nucleotides from first nucleotide of intron 1. Indicated reporter gene and a pSVgal internal control plasmid were transfected into RCS cells in the presence or absence of 100 ng/ml BMP2 for 48 h, and β-galactosidase and luciferase activities were determined. Luciferase activities were normalized to β-galactosidase activities. B) BMP2 activates the minimal promoter of GEP gene (−275 to −51). RCS cells were transfected with either an internal deletion or a negative control reporter construct, as indicated above, and luciferase and β-galactosidase assays were performed. C) BMP2-activated SMAD4 binds to the GEP promoter (ChIP). RCS cells treated with or without 100 ng/ml BMP2 for 12 h were cross-linked by formaldehyde treatment and lysed. Cell lysates were subjected to immunoprecipitation with control IgG or anti-Smad4. Purified DNA from the cell lysate (input DNA serves as a positive control) and DNA recovered from immunoprecipitation were amplified by PCR using specific primers for GEP minimal promoter. D) BMP2 downstream transcription factors (Smads) activate the GEP-specific reporter genes. GEP-specific reporter construct −1575GEPluc was transfected into RCS cells together with the indicated Smad expression plasmids (i.e., Smad1, Smad4, and Smad5), as well as a pSVgal internal control plasmid. At 48 h after transfection, cultures were harvested and processed as described in A.

Because BMP2 activates the cellular signaling through Smads, we then tested interaction of Smad4, a coregulatory Smad that binds to Smad1 or Smad5 for transducing BMP2 signaling, with GEP promoter regions (in particular, the region of −275 and −51) in vivo using the ChIP assay. As shown in Fig. 7C, we observed a clear PCR product using DNA isolated from immunoprecipitated complexes with anti-Smad4 antibodies from BMP2-treated cells but not from BMP2-untreated cells, suggesting that the Smad4 is recruited into this GEP promoter region after exposure to BMP2. Next, we determined whether Smad transcription factors could directly activate the GEP at the transcription level. Cotransfection of the GEP-luciferase plasmid (−1575GEPluc) with an expression plasmid encoding either Smad1, Smad4, Smad5 (cDNA constructs kindly provided by Drs. Riko Nishimura and Regis O’Keefe, University of Rochester, Rochester, NY, USA), or a combination of either Smad1/Smad4 or Smad5/Smad4 markedly increased the expression of the GEP reporter gene, although the combinations of Smad1 and Smad4 gave the highest value (Fig. 7D). The above data support the notion that BMP2 controls GEP expression through Smad signaling.

GEP activates chondrogenesis through Erk1/2 signaling pathway

In searching for GEP-activated signaling in chondrocytes, we tested its effects on the MAPK signal using the PathScan Multiplex Western Cocktail I that allows us to simultaneously detect levels of phospho-p90RSK, phospho-Akt, phospho-p44/42 MAPK (Erk1/2), and phospho-S6 ribosomal protein on a single membrane. Human C28I2 chondrocytes were starved for 24 h and treated with 50 ng/ml GEP at various time points, and cell lysates were analyzed using Western blots. As shown in Fig. 8A, robust activation of p44/p42 (Erk1/2) was observed after a 10-min treatment with GEP, and activation of the MAPK signal reached a peak at 30 min and continued for 2 h. Intriguingly, moderate activation of Akt signaling was also observed, suggesting that GEP may also activate the PI3K-Akt pathway. To further test the specificity of this activation we determined whether U0126, a specific inhibitor of MAPK, could block the GEP-induced chondrogenesis. The micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis Then 0.01% DMSO (v/v, GEP) or 1 μM U0126 (GEP+U0126) was added to the GEP-pretreated groups and incubated for an additional 7 d, and real-time PCR was performed (Fig. 8B). As expected, GEP strongly induced both collagen II (5.5-fold) and collagen X (10-fold), whereas these inductions were largely abolished by treatment with U0126, indicating that Erk1/2 signaling is required for GEP-mediated chondrocyte differentiation.

An external file that holds a picture, illustration, etc.
Object name is z380061077830008.jpg

GEP activates chondrogenesis through the Erk1/2 signal. A) GEP activates Akt and Erk1/2 pathways in chondrocytes. Human C28I2 chondrocytes were starved for 24 h and treated with 50 ng/ml GEP at various time points, as indicated, and cell lysates were analyzed using the PathScan Multiplex Western Cocktail I. Left panel: long-time exposure (5 min). Right panel: short-time exposure (30 s). B) U0126 abolishes GEP-induced type II and type X collagen expression during chondrogenesis. Micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis, then 0.01% DMSO (GEP) or 1 μM U0126 in 0.01% DMSO (GEP+U0126) was added to the GEP-pretreated groups and incubated for additional 7 d, and expression levels of type II (Col II) and type X (Col X) collagen were analyzed by real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1.

JunB is one of the early response genes of GEP and is involved in GEP-mediated chondrogenesis

To identify GEP downstream molecules, we performed genome-wide DNA chip analysis. Total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP at various time points and analyzed by microarray analysis (Affymetrix,). Approximately 50 genes were up-regulated (>2-fold) by GEP, as determined by hierarchical clustering (63), and some of the GEP-inducible genes (Fig. 9A), including Gadd45β, JunB, Dlx2, KLF2, Smad7, Sox4, and Tcf8, are also known to be activated by the TGF-β subfamily (64,65,66). However, we focused on JunB, because JunB was also one of the downstream molecules of BMP2 signaling and is required for BMP2-mediated differentiation (67,68,69). Using micromass cultures of C3H10T1/2 progenitor cells treated with 100 ng/ml recombinant GEP, we showed that JunB mRNA was immediately up-regulated 1 h after exposure to GEP, reached its peak at the 2 h time point and then quickly returned to the control level. Note that JunB protein level was markedly elevated at the 5-h time point and thereafter remained at high levels (Fig. 9B, C).

An external file that holds a picture, illustration, etc.
Object name is z380061077830009.jpg

JunB is an early responsive gene of GEP and required in GEP-induced chondrogenesis. A) Genome-wide DNA chip analysis for isolating GEP-responsive genes. Total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP for various time points, as indicated, and analyzed by microarray analysis. Several up-regulated genes after GEP treatment were determined by hierarchical clustering. B, C) Induction of JunB by GEP was examined in a micromass culture of C3H10T1/2 cells. Micromass cultures of C3H10T1/2 progenitor cells were incubated in the presence of 100 ng/ml recombinant GEP for various time points, as indicated, and cells were harvested followed by real-time PCR and Western blotting for measurements of JunB mRNA (B) and protein (C), respectively. D) GEP-mediated chondrogenesis is largely dependent on JunB. Micromass cultures of C3H10T1/2 cells transfected with pSuper-JunB encoding a siRNA against JunB, JunB expression plasmid, or combinations, as indicated, were incubated with 100 ng/ml GEP for 7 d, and expression of type II (Col II) and type X (Col X) collagen was analyzed by real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1. E) Proposed model to explain the role and regulation of GEP in chondrogenesis.

We next determined whether JunB, as a downstream molecule of GEP, was involved in the GEP-stimulated chondrogenesis by knocking down the JunB level using the siRNA approach. In brief, the micromass cultures of C3H10T1/2 cells were first stably transfected with pSuper vector, pSuper-JunB encoding a siRNA against JunB, or pCMV-JunB expression plasmid, as well as various combinations, as indicated. Next, these JunB knocked down stable cell lines or the JunB restored cell lines were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 7 d for induction of chondrogenesis followed by the real-time PCR analysis (Fig. 9D). The inductions of collagen II and collagen X by GEP were largely undetectable in the JunB knocked down cell lines, but chondrogenetic responses to GEP were restored when JunB was reexpressed in these cell lines. Taken together, these data indicated that JunB is a critical mediator of GEP activity in the course of chondrogenesis.

Differential expression of GEP in the chondrogenesis of a micromass culture of C3H10T1/2 cells

To investigate GEP function in chondrogenesis, we first studied GEP expression profiles during chondrocyte differentiation using the C3H10T1/2 cell line, a pluripotent murine stem cell line that is widely used for in vitro chondrogenic studies (56,57,58). Micromass cultures of these cells were incubated in the presence of 100 ng/ml recombinant BMP2 for induction of chondrocyte differentiation. Cells were harvested at various time points followed by real-time PCR for measurements of GEP (Fig. 1A) and collagen X (Fig. 1B). As shown in Fig. 1, the level of GEP was relatively low until d 7, when it is doubled and thereafter remained at high levels during the differential stage. Note that the peak level of GEP was ∼2 d earlier than that of collagen X, a marker for hypertrophic chondrocytes, suggesting that GEP may regulate collagen X expression.

An external file that holds a picture, illustration, etc.
Object name is z380061077830001.jpg

Expression of GEP during chondrogenesis both in vitro and in vivo. A, B) Expressions of GEP and collagen X were examined in the course of chondrogenesis of a micromass culture of C3H10T1/2 cells. Micromass cultures of C3H10T1/2 cells were stimulated by BMP2 protein at various time points, as indicated, and the mRNA levels of GEP (A) and collagen type X (Col X; B) were assayed using real-time PCR. Units are arbitrary; the normalized values were calibrated against the d 0 time point, here given the value of 1. C) GEP immunostaining assay shows that GEP (arrows) is first detected in the late condensation stage (E12.5) and then at the differentiation stage (E13.5) and continues during entire chondrogenic developmental stages in both proliferating and hypertrophic zones (from E17.5 to 1 mo old). Scale bar = 100 μm.

GEP expression patterns in chondrocytes during both embryonic and postnatal development stages

Next we characterized the temporal and spatial expression pattern of GEP during skeletal development using an immunostaining assay at multiple time points, including embryonic day (E) 10.5 (onset of chondrogenesis that begins with the proliferation and subsequent condensation of mesenchymal cells), E12.5 (condensation), E13.5 (right after cartilage formation but before endochondral bone formation), and E17.5 (onset of skeletal growth), as well as postnatal developmental stages (newborn, 10-d, and 30-d). As revealed in Fig. 1C, GEP is detected at E12.5, and its level is increased in the center of the condensation and around it at E13.5. It demonstrates prominent expression in prehypertrophic chondrocytes at E15.5 and E17.5 and in newborn mice. A high level of GEP throughout the whole growth plate is observed at the age of 1 month, suggesting that the expression profile of GEP is closely linked to the entire chondrogenic period.

GEP stimulates chondrogenesis in vitro and endochondral bone formation ex vivo

Prominent expression of GEP in chondrocytes and its stimulation of chondrocyte proliferation (46) prompted us to determine whether GEP was able to induce chondrocyte differentiation. We first generated a stable line expressing His-tagged GEP in 293 EBNA cells, and purified recombinant GEP using commercial beads. Purified proteins (5 μl) were analyzed by 8% SDS-PAGE and visualized by Coomassie Blue staining with verification by Western blotting with anti-GEP antibody (Fig. 2A). Next we compared the roles of GEP (100 ng/ml) and BMP2 (100 ng/ml), a well-documented growth factor, during chondrogenesis using BMSCs, which are capable of differentiation into various lineages, including chondrocytes (59, 60). In brief, the high-density culture system was incubated in the absence (CTR) or presence of 100 ng/ml GEP or 100 ng/ml BMP2 (serving as a positive control) for 7 d. Chondrogenesis was monitored by analyzing the expressions of marker genes specific for chondrocytes (Fig. 2B). As for BMP2, GEP markedly induced the expression of aggrecan, collagen II, and collagen X. Note that GEP is twice as potent as BMP2 in inducing collagen X.

An external file that holds a picture, illustration, etc.
Object name is z380061077830002.jpg

GEP stimulates chondrogenesis in vitro. A) Characterization of recombinant GEP protein. A purified recombinant human GEP (10 ng/μl in 0.2% BSA) was either stained with Coomassie Brilliant Blue R-250 (left) or detected by Western blotting with a polyclonal anti-GEP antibody (right). B) Comparisons of GEP and BMP2 in stimulations of chondrogenesis of murine BMSCs. BMSC aggregate cultures were incubated in the absence (CTR) or presence of either 100 ng/ml BMP2 or 100 ng/ml GEP for 2 wk, followed by quantitative measurements of aggrecan, collagen II (Col II), and collagen X (Col X) using real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1. C) Comparable potency of GEP and BMP2 in stimulation of chondrogenesis and activation of collagen II and collagen X expressions. hMSC aggregate cultures were incubated in the absence (CTR) or presence of 100 ng/ml BMP2 or GEP for 3 wk followed by Safranin O staining (left, red) or immunostains of collagen II (middle, green) or collagen X (right, green).

Using a long-term culture system (3 wk) of high-density hMSC pellets, we demonstrated that GEP, like BMP2, induced chondrogenesis as reflected by positive stains with Safranin O (pink color, left panel, Fig. 2C) and immunostains for collagen II and collagen X (right panels, Fig. 2C).

The effect of GEP on endochondral bone formation was then studied in an ex vivo model of 14.5-d-old fetal mouse metatarsal bones. At the time of explantation, these explants consisted of undifferentiated cartilage. In a 5-d culture period, these explants underwent all sequential stages of endochondral bone formation. As shown in Fig. 3, GEP significantly stimulated chondrocyte hypertrophy, mineralization, and bone length.

An external file that holds a picture, illustration, etc.
Object name is z380061077830003.jpg

GEP stimulates chondrocyte hypertrophy, mineralization, and endochondral bone growth. A, B) Safranin O/Fast Green staining of metatarsal bones. Metatarsals from 14.5-d-old mouse embryos were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d and stained with Safranin O/Fast Green, shown in low-power (A) and high-power (B) microphotographs. C) Alizarin red/Alcian blue staining of metatarsals. Explants were fixed and processed for staining. A representative photograph of an explanted metatarsal is presented. D) Percentage changes in total (T) and mineralization (M) length of metatarsal bones. Percentage changes in bone length were calculated as (length at d 5 − length at d 0)/length at d 0. *P < 0.05 vs. control.

In vivo knockdown of GEP leads to striking defects in skeleton development

To define the in vivo physiological role of GEP in chondrogenesis, we created conditional GEP knockdown mice using an siRNA technique and the Cre/loxP system. This in vivo knockdown system was originally developed by C. X. Deng at the National Institute of Diabetes and Digestive and Kidney Diseases, NIH and has been successfully used to knock down FGF receptor 2 in cartilage (61). In U6-ploxPneo-GEP transgenic mice, siGEP (5′-AGAACTCCGGCTCCTACTA-3′) is driven by a U6 promoter that is disturbed by a loxP-flanked neomycin cassette (61). To generate conditional-knockdown GEP mice, the U6-ploxPneo-GEP transgenic line was crossed to Sox2-Cre transgenic mice, in which Cre enzyme is activated in early mesenchymal cells. Immunostains of GEP showed that GEP is largely undetectable in the GEP-knockdown growth plate compared with the same litters (wild-type control) (Fig. 4A); Safranin O stain, staining cartilage red, showed that the knockdown growth plate is only half the size of that in wild-type mice (Fig. 4B). The bromodeoxyuridine (BrdU) incorporation assay, measuring cell proliferation, showed a sharp reduction of chondrocyte proliferation in the GEP-knockdown proliferation zone (Fig. 4C). Furthermore, we tested whether knockdown of GEP would change marker genes critical for chondrogenesis by using an in situ hybridization assay. As shown in Fig. 4, alkaline phosphatase in all growth plate zones (Fig. 4D), collagen II in the proliferating zone (Fig. 4E), and collagen X in the hypertrophic zone (Fig. 4F) are dramatically reduced in GEP-knockdown mice. Collectively, the above in vivo data, consistent with the in vitro data, support the novel concept that GEP plays an important role in cartilage development.

An external file that holds a picture, illustration, etc.
Object name is z380061077830004.jpg

GEP knockdown (KD) leads to a sharp reduction of chondrogenic markers. A) Immunostaining of GEP in the growth plate of the femur from 3-wk-old wild-type (WT) (top) and conditional KD (bottom) mice crossing between siGEP transgenic mice and Sox2 Cre mice. B) Reduction in thickness of the growth plate stained with Safranin O stain in GEP KD (top) compared with control (bottom). C) Reduction of BrdU (reflection of cell proliferation) in the GEP KD proliferation zone (top) compared with control (bottom). D) Reduction of alkaline phosphatase mRNA in the GEP KD growth plate (top) compared with control (bottom). E) Sharp reduction of collagen II mRNA in the GEP KD growth plate (top) compared with control (bottom). F) Collagen X is largely undetectable in the GEP KD hypertrophic zone (top) compared with control (bottom).

GEP induces cartilage repair

Because GEP is able to induce chondrogenesis of mesenchymal stem cells and plays an important role in cartilage development (Figs. 223334),4), and GEP was shown previously to stimulate wound healing (42), we next determined whether GEP could be used for cartilage repair. Articular cartilage defects were created in the femoral trochlea of rabbits and were treated with one of 3 implants for 6 wk: a collagen sponge only (n=6), a collagen sponge containing 6 μg of GEP (n=6), or a collagen sponge containing 6 μg of BMP2 (n=6) (the dose selected was based on our previous studies; ref. 55). As shown in Fig. 5, implants with collagen alone had little effect on cartilage repair, as the defect region was filled with a smooth-surfaced mixture of fibrous tissue and bone, and the adjacent tissue showed loss of matrix staining and reduction of cellularity (Fig. 5A). The implants containing BMP2 (Fig. 5B) or GEP (Fig. 5C) show enhanced cartilage repair, as reflected by good restoration of anatomic features and positive Safranin O staining of cartilage under the repairing region. Although adherence to adjacent host cartilage was best with BMP2 treatment, the mean ± sd histological scores for GEP (17.5±3.11) and BMP2 (17.25±2.63) did not differ significantly, and both were significantly better than that for the collagen treatment alone (7±1.00). The results support a positive correlation between implantation of GEP and cartilage repair.

An external file that holds a picture, illustration, etc.
Object name is z380061077830005.jpg

GEP accelerates cartilage repair. AC) Articular cartilage and subchondral bone sections stained with Safranin O/Fast Green. There is a lack of cartilage repair in the trauma region of the control group that was implanted with a collagen sponge only (A; arrow). Implantations of the collagen sponge containing a recombinant BMP2 protein accelerates high-quality hyaline-appearing repair as reflected by well-integrated and organized new tissues stained positive with Safranin O (B; pink, arrow). Implantations of the collagen sponge containing the equivalent GEP display a similar hyaline-like repairing response (C; pink, arrow). D) Quantitation of tissue repair. Collagen sponge with BMP2 or GEP accelerates cartilage repair.

GEP is a BMP2-inducible growth factor and required for BMP-mediated chondrocyte differentiation

In searching for upstream molecules of GEP, we tested a few growth factors known to be important for chondrogenesis in the primary human chondrocytes. Our results showed that GEP mRNA is up-regulated 3-fold by BMP2 and 1.5-fold by TGF-β. IL-1β, and noggin had no apparent effects on GEP expression (Fig. 6A).

An external file that holds a picture, illustration, etc.
Object name is z380061077830006.jpg

GEP is a downstream molecule of BMP2 and is required for BMP2 stimulation of chondrogenesis. A) Effects of growth factors and cytokines on GEP mRNA in chondrocytes by real-time PCR. Expression of GEP mRNA was normalized against 18S rRNA. ***P < 0.001. B) siGEP, a GEP siRNA construct, reduces 80% of endogenous GEP mRNA. C3H10T1/2 cells were transfected with either pSuper (CTR) or siGEP, and level of GEP mRNA was measured by real-time PCR. C) Suppression of GEP by siRNAs inhibits BMP2-induced chondrogenesis. Micromass cultures of C3H10T1/2 cells transfected with either pSuper (CTR) or siGEP were used to test whether BMP2 (100 ng/ml)-induced chondrogenesis is GEP dependent. Expressions of marker genes, as indicated, were determined by real-time PCR.

Next, we tested whether GEP was required for BMP2-mediated chondrogenesis using the siRNA approach. As shown in Fig. 6B, siGEP markedly reduced 80% of the endogenous GEP in C3H10T1/2 cells. We then performed micromass cultures of C3H10T1/2 cells with either a pSuper vector (CTR) or a siGEP construct transfected first and followed by additions of BMP2 (100 ng/ml). Our real-time PCR assay showed that reductions of the endogenous GEP by siGEP sharply decrease chondrogenic responses induced by BMP2: 65% down of Sox9, 60% down of collagen II, 80% down of collagen X, and 70% down of aggrecan compared with the control group responses (Fig. 6C). These results support the concept that GEP is a key downstream molecule of BMP2 during chondrogenesis.

BMP2 and Smads activate GEP-specific reporter genes

To elucidate the molecular mechanism by which BMP2 activates GEP expression, we cloned regulatory elements in the human GEP 5′-flanking region from −1573 to +62 (62) with a series of 5′ deletion promoter-luciferase constructs (Supplemental Fig. 1) and tested their transcriptional activity in RCS cells known to express GEP (46). Deletions of the region from −1393 to −1175 and the region from −570 to −275 result in ∼2- and 2.4-fold increases in promoter activity, respectively, suggesting that these two regions function as silencers in controlling GEP gene transactivation. On the other hand, deletion of the region from −270 to −51 leads to the complete loss of the reporter activity, indicating that this region is probably the basic promoter of the GEP gene. Applications of BMP2 were able to activate all GEP promoter constructs containing the region between −275 and −51 (Fig. 7A) Furthermore, this basic promoter region directly responded to BMP2 (Fig. 7B).

An external file that holds a picture, illustration, etc.
Object name is z380061077830007.jpg

BMP2 and its mediators, Smads, activate the GEP-specific reporter genes. A) BMP2 activates GEP-specific reporter genes in RCS cells. Indicated segments from the 5′-flanking region of the GEP gene were linked to an simian virus 40 promoter and a DNA segment encoding luciferase. Numbers indicate distances in nucleotides from first nucleotide of intron 1. Indicated reporter gene and a pSVgal internal control plasmid were transfected into RCS cells in the presence or absence of 100 ng/ml BMP2 for 48 h, and β-galactosidase and luciferase activities were determined. Luciferase activities were normalized to β-galactosidase activities. B) BMP2 activates the minimal promoter of GEP gene (−275 to −51). RCS cells were transfected with either an internal deletion or a negative control reporter construct, as indicated above, and luciferase and β-galactosidase assays were performed. C) BMP2-activated SMAD4 binds to the GEP promoter (ChIP). RCS cells treated with or without 100 ng/ml BMP2 for 12 h were cross-linked by formaldehyde treatment and lysed. Cell lysates were subjected to immunoprecipitation with control IgG or anti-Smad4. Purified DNA from the cell lysate (input DNA serves as a positive control) and DNA recovered from immunoprecipitation were amplified by PCR using specific primers for GEP minimal promoter. D) BMP2 downstream transcription factors (Smads) activate the GEP-specific reporter genes. GEP-specific reporter construct −1575GEPluc was transfected into RCS cells together with the indicated Smad expression plasmids (i.e., Smad1, Smad4, and Smad5), as well as a pSVgal internal control plasmid. At 48 h after transfection, cultures were harvested and processed as described in A.

Because BMP2 activates the cellular signaling through Smads, we then tested interaction of Smad4, a coregulatory Smad that binds to Smad1 or Smad5 for transducing BMP2 signaling, with GEP promoter regions (in particular, the region of −275 and −51) in vivo using the ChIP assay. As shown in Fig. 7C, we observed a clear PCR product using DNA isolated from immunoprecipitated complexes with anti-Smad4 antibodies from BMP2-treated cells but not from BMP2-untreated cells, suggesting that the Smad4 is recruited into this GEP promoter region after exposure to BMP2. Next, we determined whether Smad transcription factors could directly activate the GEP at the transcription level. Cotransfection of the GEP-luciferase plasmid (−1575GEPluc) with an expression plasmid encoding either Smad1, Smad4, Smad5 (cDNA constructs kindly provided by Drs. Riko Nishimura and Regis O’Keefe, University of Rochester, Rochester, NY, USA), or a combination of either Smad1/Smad4 or Smad5/Smad4 markedly increased the expression of the GEP reporter gene, although the combinations of Smad1 and Smad4 gave the highest value (Fig. 7D). The above data support the notion that BMP2 controls GEP expression through Smad signaling.

GEP activates chondrogenesis through Erk1/2 signaling pathway

In searching for GEP-activated signaling in chondrocytes, we tested its effects on the MAPK signal using the PathScan Multiplex Western Cocktail I that allows us to simultaneously detect levels of phospho-p90RSK, phospho-Akt, phospho-p44/42 MAPK (Erk1/2), and phospho-S6 ribosomal protein on a single membrane. Human C28I2 chondrocytes were starved for 24 h and treated with 50 ng/ml GEP at various time points, and cell lysates were analyzed using Western blots. As shown in Fig. 8A, robust activation of p44/p42 (Erk1/2) was observed after a 10-min treatment with GEP, and activation of the MAPK signal reached a peak at 30 min and continued for 2 h. Intriguingly, moderate activation of Akt signaling was also observed, suggesting that GEP may also activate the PI3K-Akt pathway. To further test the specificity of this activation we determined whether U0126, a specific inhibitor of MAPK, could block the GEP-induced chondrogenesis. The micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis Then 0.01% DMSO (v/v, GEP) or 1 μM U0126 (GEP+U0126) was added to the GEP-pretreated groups and incubated for an additional 7 d, and real-time PCR was performed (Fig. 8B). As expected, GEP strongly induced both collagen II (5.5-fold) and collagen X (10-fold), whereas these inductions were largely abolished by treatment with U0126, indicating that Erk1/2 signaling is required for GEP-mediated chondrocyte differentiation.

An external file that holds a picture, illustration, etc.
Object name is z380061077830008.jpg

GEP activates chondrogenesis through the Erk1/2 signal. A) GEP activates Akt and Erk1/2 pathways in chondrocytes. Human C28I2 chondrocytes were starved for 24 h and treated with 50 ng/ml GEP at various time points, as indicated, and cell lysates were analyzed using the PathScan Multiplex Western Cocktail I. Left panel: long-time exposure (5 min). Right panel: short-time exposure (30 s). B) U0126 abolishes GEP-induced type II and type X collagen expression during chondrogenesis. Micromass cultures of C3H10T1/2 cells were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 5 d for initial induction of chondrogenesis, then 0.01% DMSO (GEP) or 1 μM U0126 in 0.01% DMSO (GEP+U0126) was added to the GEP-pretreated groups and incubated for additional 7 d, and expression levels of type II (Col II) and type X (Col X) collagen were analyzed by real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1.

JunB is one of the early response genes of GEP and is involved in GEP-mediated chondrogenesis

To identify GEP downstream molecules, we performed genome-wide DNA chip analysis. Total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP at various time points and analyzed by microarray analysis (Affymetrix,). Approximately 50 genes were up-regulated (>2-fold) by GEP, as determined by hierarchical clustering (63), and some of the GEP-inducible genes (Fig. 9A), including Gadd45β, JunB, Dlx2, KLF2, Smad7, Sox4, and Tcf8, are also known to be activated by the TGF-β subfamily (64,65,66). However, we focused on JunB, because JunB was also one of the downstream molecules of BMP2 signaling and is required for BMP2-mediated differentiation (67,68,69). Using micromass cultures of C3H10T1/2 progenitor cells treated with 100 ng/ml recombinant GEP, we showed that JunB mRNA was immediately up-regulated 1 h after exposure to GEP, reached its peak at the 2 h time point and then quickly returned to the control level. Note that JunB protein level was markedly elevated at the 5-h time point and thereafter remained at high levels (Fig. 9B, C).

An external file that holds a picture, illustration, etc.
Object name is z380061077830009.jpg

JunB is an early responsive gene of GEP and required in GEP-induced chondrogenesis. A) Genome-wide DNA chip analysis for isolating GEP-responsive genes. Total RNA was isolated from human C28I2 chondrocytes treated with 50 ng/ml GEP for various time points, as indicated, and analyzed by microarray analysis. Several up-regulated genes after GEP treatment were determined by hierarchical clustering. B, C) Induction of JunB by GEP was examined in a micromass culture of C3H10T1/2 cells. Micromass cultures of C3H10T1/2 progenitor cells were incubated in the presence of 100 ng/ml recombinant GEP for various time points, as indicated, and cells were harvested followed by real-time PCR and Western blotting for measurements of JunB mRNA (B) and protein (C), respectively. D) GEP-mediated chondrogenesis is largely dependent on JunB. Micromass cultures of C3H10T1/2 cells transfected with pSuper-JunB encoding a siRNA against JunB, JunB expression plasmid, or combinations, as indicated, were incubated with 100 ng/ml GEP for 7 d, and expression of type II (Col II) and type X (Col X) collagen was analyzed by real-time PCR. Units are arbitrary; normalized values were calibrated against controls, here given the value of 1. E) Proposed model to explain the role and regulation of GEP in chondrogenesis.

We next determined whether JunB, as a downstream molecule of GEP, was involved in the GEP-stimulated chondrogenesis by knocking down the JunB level using the siRNA approach. In brief, the micromass cultures of C3H10T1/2 cells were first stably transfected with pSuper vector, pSuper-JunB encoding a siRNA against JunB, or pCMV-JunB expression plasmid, as well as various combinations, as indicated. Next, these JunB knocked down stable cell lines or the JunB restored cell lines were cultured in the absence (CTR) or presence of 100 ng/ml GEP for 7 d for induction of chondrogenesis followed by the real-time PCR analysis (Fig. 9D). The inductions of collagen II and collagen X by GEP were largely undetectable in the JunB knocked down cell lines, but chondrogenetic responses to GEP were restored when JunB was reexpressed in these cell lines. Taken together, these data indicated that JunB is a critical mediator of GEP activity in the course of chondrogenesis.

DISCUSSION

GEP, as a growth factor, has been linked to development, tissue regeneration, tumorigenesis, and inflammation (22, 23, 41, 42). Our recent studies showed that GEP was directly associated with COMP and chondrocyte proliferation (46). The current study focused on the role of GEP in chondrogenesis and cartilage repair, its regulation by BMP2, and the signaling pathway. Our data show that GEP is highly induced in the course of chondrogenesis in vitro and is active during the entire cartilage development (Figs. 1 and and2).2). The in vitro, ex vivo, and in vivo studies support the concept that GEP, a growth factor previously unrecognized in cartilage, is a potent stimulator of chondrocyte differentiation and endochondral bone growth (Figs. 223334)4) and seems to stimulate early cartilage repair (Fig. 5).

The most powerful tool in elucidating the role of genes in development is the gene-knockout approach (70). Currently there are 3 ways to produce the GEP knockout/knockdown mouse model. The first approach is to generate the targeted deletion to the locus at which endogenous GEP is removed by using the classic homologous recombination technique. The second option is to use the Cre-LoxP system combining conditional deletion of the endogenous GEP in a tissue- and temporal-specific pattern. The third one is to use Cre-LoxP RNA interference, a newly developed approach that has been successfully used in knockdown of FGF receptor 2 with in cartilage (61). We selected the third option based on the following justifications. 1) The first option, the targeted deletion of GEP in the locus where endogenous GEP is, has much higher risk, partly because of difficulties in construction of sophisticated gene-targeting vectors, in homolog recombination while embryonic stem cells are manipulated, and in getting the targeted cell into the germline. 2) GEP is expressed in both cartilage tissue and many soft tissues such as brain during developmental processes (12,13,14, 18), and this mouse model would be useful for testing multiple functions of GEP in different tissues at different developmental stages without risk of embryonic lethality. 3) Our in vitro GEP-RNA interference construct has been validated in inhibition of markers critical for endogenous chondrogenesis or chondrogenesis markers induced by BMP2 (Fig. 6). 4) *Cre-LoxP RNA interference, which acts dominantly and only needs one allele of the transgene to suppress the endogenous gene, has been shown to be simple and fast (only 3 mo instead of ≥1 yr for options 1 or 2, which require both alleles to be deleted). This advantage becomes more obvious when the genetic interaction of ≥2 genes is studied (71, 72). 5) Finally, it is cost-effective, with high efficiency (95% in reduction of the endogenous gene; ref. 61). In this study, we successfully knocked down the expression of GEP in growth plate chondrocytes of GEP-knockdown mice. More important, knockdown of GEP led to dwarfism and striking defects in the skeletal system, including delayed endochondral bone growth, reduced bone length and volume, diminished growth plates, and reduced chondrocyte proliferation and differentiation (Fig. 4), suggesting that GEP is an essential previously unknown growth factor for cartilage development.

Our work also supports the concept that GEP is a key downstream molecule of BMP2 in chondrogenesis and cartilage repair based on the following evidence. 1) Both BMP2 and GEP are potent in inducing in vitro chondrogenesis and induction of chondrogenetic markers such as collagen II, collagen X, and aggrecan (Fig. 2) (similar to the effects of BMP2, the effects of GEP on chondrogenesis are probably stage specific, unlike many factors that have opposite effects on collagen II and collagen X). 2) Knockout of BMP2 (40) or knockdown of GEP led to abnormalities in skeletal development (Fig. 4). 3) Both BMP2 (55) and GEP enhance early cartilage repair (Fig. 3). 4) BMP2 induced GEP in chondrocytes (Fig. 6), but GEP had no effect on BMP2 expression (data not shown). 5) BMP2 activates GEP-specific reporter genes through Smad transcription factors (Fig. 7). 6) Finally, and notably, knockdown of GEP strongly inhibited BMP2-mediated chondrogenesis (Fig. 6).

GEP was reported to activate Erk1/2 signaling in SW-13 adrenal carcinomas (73). Our assay with the PathScan Multiplex Western Cocktail I that allowed us to simultaneously detect several signal pathways on a single membrane revealed that GEP also strongly activated Erk1/2 signaling in chondrocytes (Fig. 8A). Furthermore, GEP-stimulated chondrocyte differentiation depends on activation of the Erk1/2 pathway, because blocking Erk1/2 activation via its inhibitor abolished more than 60% of chondrogenetic responses induced by GEP, as reflected by changes of collagen II and collagen X, two critical marker genes for chondrogenesis (Fig. 8B). Furthermore, we showed that GEP also activated Akt signaling (Fig. 8A) and that blocking Akt signaling by 0.2 μM wortmannin had a moderate effects on chondrogenetic responses induced by GEP (data not shown).

In further screening of GEP downstream molecules that might be critical in chondrogenesis, we identified ∼50 GEP-responsive genes using a whole-genome DNA microarray. Among these GEP-inducible genes, JunB, Gadd45β, Dlx2, and Sox4 are also known to be up-regulated by BMP2 (74, 75), which further supports our finding that GEP is a downstream target of BMP2. Our work supports an idea that JunB is probably a direct early response gene of GEP in chondrogenesis based on the following evidence: JunB mRNA was up-regulated 1 h after exposure to GEP and reached its peak at the 2-h time point; JunB can induce chondrogenic responses; and knocking down JunB using the siRNA approach abolishes >50% of collagen II and collagen X induced by GEP (Fig. 9).

Recently we reported that a disintegrin and metalloproteinase with thrombospondin type 1 motif (ADAMTS)-7 metalloproteinase associated with GEP and converted GEP into its processed fragments; in addition, ADAMTS-7 inhibited chondrocyte differentiation and endochondral bone formation, probably via inactivating the chondrogenic activity of GEP (56). These findings, together with our previous reports that ADAMTS-7 binds to and degrades COMP (76) and that COMP interacts with GEP and potentiates GEP-stimulated chondrocyte functions (46), indicate that ADAMTS-7, GEP, and COMP form an interaction and interplay network in regulating chondrocyte functions (46, 76). It remains to be determined how the interaction network among ADAMTS-7 metalloproteinase, GEP growth factor, and COMP extracellular matrix molecule acts in concert in regulating chondrocyte differentiation and endochondral ossification.

In summary, this study provides novel insights into the role of GEP, a key downstream molecule of BMP2, in regulating chondrocyte differentiation and endochondral bone formation, and cartilage repair. Our work supports a hypothesis that GEP regulates chondrogenesis through Erk1/2 signaling and its target gene, including JunB transcription factor (Fig. 9E). In addition, this study also provides us with potential molecule targets, especially recombinant GEP protein, for treatment of cartilage disorders and arthritic conditions.

Supplementary Material

Supplemental Data:
Department of Biomedical Sciences, Baylor College of Dentistry, Texas A&amp;M University System Health Science Center, Dallas, Texas, USA; and
Department of Orthopaedic Surgery and
Department of Cell Biology, New York University School of Medicine, New York, New York, USA
These authors contributed equally to this work.
Current address: Department of Cell Biology and Genetics, Chongqing Medical University, Chongqing 400016, China.
Correspondence: Department of Orthopaedic Surgery, New York University School of Medicine, 301 E. 17th St., New York, NY 10003, USA. E-mail: ude.uyn.dem@uil.ujnauhc
Received 2009 Sep 2; Accepted 2010 Jan 7.

Abstract

Granulin epithelin precursor (GEP) has been implicated in development, tissue regeneration, tumorigenesis, and inflammation. Herein we report that GEP stimulates chondrocyte differentiation from mesenchymal stem cells in vitro and endochondral ossification ex vivo, and GEP-knockdown mice display skeleton defects. Similar to bone morphogenic protein (BMP) 2, application of the recombinant GEP accelerates rabbit cartilage repair in vivo. GEP is a key downstream molecule of BMP2, and it is required for BMP2-mediated chondrocyte differentiation. We also show that GEP activates chondrocyte differentiation through Erk1/2 signaling and that JunB transcription factor is one of key downstream molecules of GEP in chondrocyte differentiation. Collectively, these findings reveal a novel critical role of GEP growth factor in chondrocyte differentiation and the molecular events both in vivo and in vitro.—Feng, J. Q., Guo, F.-J., Jiang, B.-C., Zhang, Y., Frenkel, S., Wang, D.-W., Tang, W., Xie, Y., Liu, C.-J. Granulin epithelin precursor: a bone morphogenic protein 2-inducible growth factor that activates Erk1/2 signaling and JunB transcription factor in chondrogenesis.

Keywords: GEP, BMP-2, chondrocyte, differentiation
Abstract

Chondrogenesis plays a fundamental role in skeletal patterning, bone formation, and joint development. Well-orchestrated chondrogenesis is controlled exquisitely by cellular interactions with the growth factors, surrounding matrix proteins, and other environmental factors that mediate cellular signaling pathways and transcription of specific genes in a temporal-spatial manner (1,2,3). Production of and response to different growth factors are observed at all times, and autocrine and paracrine cell stimulations are key elements of the process (4, 5). Particularly relevant is the role of the TGF-β superfamily and more specifically of the bone morphogenic protein (BMP) subfamily. Other factors include IGF, fibroblast growth factor (FGF), growth hormone (GH), and retinoids (6,7,8). The growing evidence demonstrates that complicated cellular signaling language and informational content of chondrogenesis lie not in an individual growth factor but in the entire set of growth factors and other signals to which a cell is exposed (4, 5, 9). The ways in which growth factors exert their combinatorial effects are becoming clearer as the molecular mechanisms of growth factor actions are being investigated.

GEP, also known as PC cell-derived growth factor (PCDGF), progranulin, acrogranin, or GP80, was first purified as a growth factor from conditioned tissue culture media (10, 11). It has been identified from different sources by several independent laboratories (12,13,14,15). GEP is a 593-aa secreted glycoprotein with an apparent molecular mass of 80 kDa (12, 16), which acts as an autocrine growth factor. GEP contains 7.5 repeats of a cysteine-rich motif (CX5–6CX5CCX8CCX6CCXDX2HCCPX4CX5–6C) in the order P-G-F-B-A-C-D-E, in which A–G are full repeats and P is the half-motif. The C-terminal region of the consensus sequence contains the conserved sequence CCXDX2HCCP and is suggested to have a metal binding site and to be involved in regulatory function (17).

GEP is abundantly expressed in rapidly cycling epithelial cells, in cells of the immune system, and in neurons (12,13,14, 18). High levels of GEP expression are also found in several human cancers and contribute to tumorigenesis in diverse cancers, including breast cancer, clear-cell renal carcinoma, invasive ovarian carcinoma, glioblastoma, adipocytic teratoma, and multiple myeloma (19,20,21,22,23,24,25,26). Although GEP functions mainly as a secreted growth factor, it was also found to be localized inside cells and to directly modulate intracellular activities (14, 27,28,29). The role of GEP in the stimulation of cellular proliferation has been well characterized using embryo fibroblasts derived from mice with a targeted deletion of the IGF-I receptor (IGF-IR) gene (R cells). These cells are unable to proliferate in response to IGF-I or to other growth factors (epidermal growth factor and platelet-derived growth factor) necessary for progression through the cell cycle (30). In contrast, GEP is the only known growth factor able to bypass the requirement for IGF-IR, thus promoting growth of R cells (15, 31). Evidence implicating GEP in the regulation of differentiation, development, and pathological processes has also been increasing. GEP has been isolated as a differentially expressed gene from mesothelial differentiation (32), sexual differentiation of the brain (33), macrophage development (34), and synovium in rheumatoid arthritis and osteoarthritis (35). Mutations in GEP cause tau-negative frontotemporal dementia linked to chromosome 17 (36,37,38,39). In addition, GEP was shown to be a crucial mediator of wound response and tissue repair (23, 40, 41).

Several GEP-associated proteins have been reported to affect GEP action in various processes. One example is the secretory leukocyte protease inhibitor (SLPI). Elastase digests GEP exclusively in the intergranulin linkers, resulting in the generation of granulin peptides, suggesting that this protease may be an important GEP convertase. SLPI blocks this proteolysis either by directly binding to elastase or by sequestering granulin peptides from the enzyme (42). GEP was also found to interact with perlecan, a heparan sulfate proteoglycan; perlecan-null mice exhibit severe skeletal defects (21, 43,44,45). Our global screens also led to the isolation of GEP as a binding partner of cartilage oligomeric matrix protein (COMP) (46), a noncollagenous matrix protein whose mutations lead to pseudoachondroplasia, multiple epiphyseal dysplasia, and short-limb dwarfism (47,48,49,50,51,52,53).

In this study, we attempt to determine whether GEP is essential for skeletal development using both in vitro and in vivo approaches. Second, we tested the ex vivo roles of the recombinant GEP in a rabbit cartilage repair model. Third, we studied its upstream and downstream molecules during chondrogenesis, as well as its molecular mechanisms by which GEP regulate chondrogenesis. Our results support a novel role of GEP, a key downstream molecule of BMP2 in controls of chondrogenesis via activating Erk1/2 signaling and JunB transcription factor.

Click here to view.

Acknowledgments

This work was aided by U.S. NIH research grants AR053210, {"type":"entrez-nucleotide","attrs":{"text":"AR050620","term_id":"5973345","term_text":"AR050620"}}AR050620, and {"type":"entrez-nucleotide","attrs":{"text":"AG029388","term_id":"16556261","term_text":"AG029388"}}AG029388 (to C.J.L.) and {"type":"entrez-nucleotide","attrs":{"text":"DE018486","term_id":"62261594","term_text":"DE018486"}}DE018486 and AR51587 (to J.Q.F).

Acknowledgments
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.