Modulation of connexin43 alters expression of osteoblastic differentiation markers
Abstract
Gap junctional channels between cells provide a pathway for exchange of regulatory ions and small molecules. We previously demonstrated that expression of connexins and cell-to-cell communication parallel osteoblastic differentiation and that nonspecific pharmacological inhibitors of gap junctional communication inhibit alkaline phosphatase activity. In this study, we stably transfected connexin (Cx)43 antisense cDNA into the immortalized human fetal osteoblastic cell line hFOB 1.19 (hFOB/Cx43−). hFOB/Cx43− cells express lower levels of Cx43 protein and mRNA and display a 50% decrease in gap junctional intercellular communication relative to control [hFOB/plasmid vector control (pvc)]. This suggests that other connexins, such as Cx45, which is expressed to a similar degree in hFOB/Cx43− cells and hFOB/pvc cells, contribute to cell-to-cell communication in hFOB 1.19 cells. We observed almost total inhibition of alkaline phosphatase activity in hFOB/Cx43− cells despite only a 50% decrease in cell-to-cell communication. This suggests the intriguing possibility that Cx43 expression per se, independent of cell-to-cell communication, influences alkaline phosphatase activity and perhaps bone cell differentiation. Quantitative real-time RT-PCR revealed that mRNA levels for osteocalcin and core binding factor α1 (Cbfa1) increased as a function of time in hFOB/pvc but were inhibited in hFOB/Cx43−. Osteopontin mRNA levels were increased in hFOB/Cx43− relative to hFOB/pvc and decreased as a function of time in both hFOB/Cx43− and hFOB/pvc. Transfection with Cx43 antisense did not affect expression of type I collagen in hFOB 1.19 cells. These results suggest that gap junctional intercellular communication and expression of Cx43 contribute to alkaline phosphatase activity, as well as osteocalcin, osteopontin, and Cbfa1 expression in osteoblastic cells.
gap junctions are membrane-spanning channels that facilitate intercellular communication by allowing small (≤1.0 kDa) signaling molecules (e.g., ions, cAMP, inositol 1,4,5-trisphosphate) to pass from cell to cell. Gap junctional intercellular communication (GJIC) and expression of connexins have been previously identified in bone tissue by morphological, structural, and functional analyses (1, 6). Our previous studies (7) suggest that although connexin (Cx)43 and Cx45 are expressed in the human fetal osteoblastic cell line hFOB 1.19, Cx43 is the predominant gap junction protein and its expression is most closely correlated with GJIC. In vivo studies suggest that gap junctions may be involved in cell signaling processes important to limb bud differentiation and skeletogenesis in embryonic mice (22) and contribute to cellular differentiation and intramembranous bone formation in developing chick mandible (24, 25). Furthermore, Cx43-null mice display impaired intramembranous bone formation, and osteoblastic cells from these animals express decreased levels of type I collagen, osteopontin, and osteocalcin (18), suggesting a defect in osteoblastic maturation. Several in vitro studies from our laboratory (8, 21, 34) as well as others (5, 17, 26, 28, 33) demonstrated that Cx43 expression and GJIC parallel osteoblastic differentiation and that inhibiting GJIC and Cx43 expression in osteoblastic cells, including MC3T3-E1, UMR-106, ROS 17/2.8, human primary culture osteoblastic cells, and murine calvarial cells, with pharmacological agents or genetic manipulation results in decreased expression of phenotypic characteristics of differentiated osteoblasts, including alkaline phosphatase, osteocalcin, bone sialoprotein (BSP), and parathyroid hormone responsiveness. Additionally, Schiller et al. (27) showed that inhibiting GJIC induces the transdifferentiation of osteoblastic MC3T3-E1 cells and primary culture human osteoblastic cells to an adipocytic phenotype. On the other hand, at least one study suggests that GJIC is related to decreased osteoblastic differentiation (15).
Although the majority of studies suggest that gap junctions contribute to bone cell differentiation, most of these studies examined nonhuman cells with nonspecific pharmacological inhibitors of GJIC. To examine specifically the role of Cx43 in human osteoblastic cell differentiation, we developed and characterized a human fetal osteoblastic cell line (hFOB 1.19) transfected with a Cx43 antisense DNA (hFOB/Cx43−). hFOB 1.19 cells are conditionally immortalized with a temperature-sensitive mutant of the SV40 large T antigen. When cultured at 37°C, these cells proliferate rapidly and display characteristics of an immortalized cell line. However, when cultured at 39.5°C, the cells proliferate more slowly while displaying characteristics of differentiated osteoblastic cells more quickly and to a greater extent than when maintained at 37°C (14). Differentiated hFOB 1.19 cells express phenotypic characteristics of osteoblastic cells, including high alkaline phosphatase activity, 1,25-dihydroxyvitamin D3-inducible osteocalcin expression, parathyroid hormone-inducible cAMP production (8, 14), expression of appropriate gap junctional proteins (8), and responsiveness to mechanical signals (36). Furthermore, hFOB 1.19 cells have minimal chromosome abnormalities, can thus be considered nontransformed, and form mineralized tissue in vivo (32). These cells display several advantages over primary culture human bone cells in that they are more homogeneous, proliferate rapidly until they are forced to differentiate, and do not display phenotypic drift in culture. Therefore, they are an appropriate model in which to study human osteoblastic cell differentiation. We found that hFOB 1.19 Cx43 antisense transfectants have reduced GJIC, and we used these cells to examine the hypothesis that inhibiting Cx43 in hFOB 1.19 cells would decrease expression of phenotypic markers of differentiated osteoblasts.
EXPERIMENTAL PROCEDURE
Reagents.
All reagents were of analytical grade and commercially available. Monoclonal antibody against Cx43 was purchased from Zymed (San Francisco, CA). The oligonucleotides for the probes and primers for Cx43, Cx45, osteopontin, osteocalcin, type I collagen, and core binding factor α1 (Cbfa1) were synthesized by Biosearch Technologies (Novato, CA), and cDNA probes for Cx43 and Cx45 were prepared as previously described (20). hFOB 1.19 cell lines were provided by Dr. Steven Harris (Bayer, West Haven, CT). pcDNA3.1/hygromycin B(+), hygromycin B, ampicillin, and Top10 One Shot kits were purchased from Invitrogen (Carlsbad, CA), BamHI, XbaI, KpnI, and T4 DNA ligase were from Boehringer Mannheim (Indianapolis, IN), and Qiaquick Gel Extraction, EndoFree Plasmid, and RNeasy kits were from Qiagen (Valencia, CA).
Construction of human Cx43 antisense cDNA expressing plasmid and stable transfection.
A 1,817-bp fragment of human Cx43 cDNA was excised from PGF1 vector (12) and ligated into the XbaI and KpnI sites of pcDNA3.1/hygro(+) vector in an antisense orientation. The structure of pcDNA3.1/hygro(+) human Cx43− was analyzed, and the antisense orientation was confirmed by sequence analysis (sequencer ABI model 377). The human Cx43 antisense cDNA expression plasmid was named Cx43−, and its plasmid vector control was named pvc. For transfection, hFOB 1.19 cells were plated at 1 × 104 cells/mm2 in 60-mm dishes containing 5 ml of growth medium (10% FBS, 1% penicillin-streptomycin, and DMEM-F-12) and maintained in culture (37°C, 5% CO2) for 36 h. Two micrograms of Cx43− [pcDNA3.1/hygro(+) human Cx43−] or pvc were precomplexed with 6 ml of Plus reagent and incubated at room temperature for 15 min. Twelve microliters of Lipofectamine were added to 240 ml of DMEM-F-12 before being mixed with precomplex containing either Cx43− or pvc and incubated at room temperature for 15 min to a final volume of 260 ml of transfection medium. Before addition of transfection medium, hFOB 1.19 cells were washed once with DMEM-F-12 and maintained in 2 ml of DMEM-F-12. After the wash, 260 μl of transfection medium were added to each dish containing 2 ml of DMEM-F-12, mixed gently, and incubated (37°C, 5% CO2). After 3 h of incubation, 2.26 ml of 20% FBS-supplemented DMEM-F-12 were added to the cells and cultured for 24 h. After this 24-h incubation, medium was replaced with the selection medium containing 150 μg/ml hygromycin B and cultured for 10–14 days with medium replacement at 3-day intervals until resistant colonies were clearly visible. Resistant colonies (∼100–200 cells) were trypsinized in plastic colonizing cylinders. Cells were passaged and maintained in selection medium containing 100 μg/ml hygromycin B until sufficient numbers of cells (∼1 × 107) were obtained. Routine growth conditions included incubation at 37°C and medium replacement with medium containing 300 μg/ml geneticin instead of hygromycin B at 3-day intervals, with every alternate medium changed to maintain neomycin resistance (14). All clones of hFOB 1.19 cells transfected with Cx43− (hFOB/Cx43−) and hFOB 1.19 cells transfected with pvc (hFOB/pvc) were screened using Western blot analysis with a Cx43 monoclonal antibody. Lucifer yellow dye spread was assessed to quantify functional GJIC. hFOB/Cx43− clones that displayed lower GJIC and expressed no or low Cx43 protein levels and hFOB/pvc clones were selected for further investigation.
Cell culture.
Cell culture was as previously described (7). Briefly, hFOB 1.19, hFOB/Cx43−, and hFOB/pvc cells were cultured in DMEM-F-12 supplemented with 10% charcoal-stripped FBS, 1% penicillin-streptomycin, 10−8 M menadione, 100 μg/ml ascorbic acid, and 10−8 M 1,25-dihydroxyvitamin D3. Cells were plated at 8 × 104 cells in round, 100-mm-diameter tissue culture dishes and cultured to confluence. For quantification of GJIC by Lucifer yellow dye spread, cells were plated on round 25-mm-diameter coverslips at 2 × 104 cells/coverslip and cultured to confluence. To examine osteoblastic cell differentiation markers, 1 × 104 cells were plated onto 24-well plates for alkaline phosphatase activity assays or 8 × 104 cells were plated onto round, 100-mm-diameter dishes for Western blot assays and real-time RT-PCR. Medium was changed, and cells were collected for assay on days 3, 6, 9, and 12. On day 6, one-half of the tissue culture plates were placed in an incubator at 39.5°C. These cells were also collected for assay on days 9 and 12.
Quantification of dye coupling.
Lucifer yellow dye spread was used to quantify GJIC as previously described (21). Confluent cells grown on coverslips were washed with α-MEM and transferred to the stage of an inverted fluorescence microscope. Individual cells were impaled with glass micropipettes backfilled with 10% Lucifer yellow dye dissolved in 1 M LiCl2 solution. Once injected, the micropipette remained in the cell for 2 min. The pipette was then removed, and the number of neighboring cells to which the dye had spread after an additional 3 min was quantified. The dye was excited at 450–490 nm, and emitted light was visualized at 515 nm.
Detection of Cx43 protein.
For detection of Cx43 protein, cells were cultured as described above. The crude membrane protein fraction isolated from cells was subjected to Western blot analysis as previously described (21). Before electrophoresis, protein concentration was quantified using the Bradford method (4). Equivalent amounts of protein (10 μg) from each cell type were loaded onto 10% SDS-polyacrylamide gels, resolved by electrophoresis, and transferred onto nitrocellulose membranes that were blocked by Blotto (5% nonfat milk, 10 mM Tris·HCl pH 8.0, 150 mM NaCl, and 0.05% Tween 20, pH 8.0). The membranes were then incubated for 3 h at room temperature with monoclonal antibodies against Cx43 diluted 1:1,000 in Blotto. The membranes were washed three times and incubated for 1 h with goat anti-mouse IgG linked to horseradish peroxidase (Jackson ImmunoResearch) diluted 1:3,000. After three additional washes with PBS, the membranes were soaked in ECL detection reagents (Amersham, Little Chalfont, UK). The sheet was then air dried and exposed to X-ray film.
Quantification of alkaline phosphatase activity.
Cells were cultured as described above. On days 3, 6, 9, and 12, medium was removed and cellular alkaline phosphatase activity was determined by the conversion of p-nitrophenyl phosphate to p-nitrophenol (21). Briefly, cells were washed twice with PBS and incubated for 30 min in 0.5 ml of 0.75 M 2-amino-2-methyl-1-propanol, pH 10.3, containing 2 mg/ml p-nitrophenyl phosphate substrate. The reaction solution was mixed with an equal volume of 50 mM NaOH and then diluted 1:40 with 20 mM NaOH. Absorption was measured at 410 nm, and conversion to enzyme activity was made with a p-nitrophenol standard absorption curve. Data were normalized to protein levels as determined using the Bradford method (4).
Quantitative real-time RT-PCR.
Total RNA was isolated from cells with a Qiagen RNeasy Mini kit. RNA was subjected to real-time RT-PCR with a PerkinElmer ABI Prism 7700 sequence detection system as previously described (7). The sequences for primers and probes used are shown in Table 1.
| Genes | Forward Primer | Reverse Primer | Probe | GenBank Accession No. |
|---|---|---|---|---|
| Connexin43 | 5′-GCT CCT CAC CAA CCG CT-3′ | 5′-TTG CGG CAG GAG GAA TTG-3′ | 5′-TCT CAC CTA TGT CTC CTC CTG GGT ACA A-3′ | NM_000166 |
| Connexin45 | 5′-AGC AGA GCC AAC CCA AAC C-3′ | 5′-ATG AGC CCA TCT TCC CGA A-3′ | 5′-AAG CAT GAT GGC CGA CGA CGG-3′ | NM_005497 |
| Osteopontin | 5′-TTG CAG CCT TCT CAG CCA A-3′ | 5′-CAA AAG CAA ATC ACT GCA ATT CTC-3′ | 5′-CGC GGA CCA AGG AAA ACT CAC TAC CA-3′ | X13694 |
| Type I collagen | 5′-CGC ACG GCC AAG AGG A-3′ | 5′-ACG CAG GTG ATT GGT TGG G-3′ | 5′-CAA GTC GAG GGC CAA GAC GAA GAC A-3′ | X06269 |
| Cbfa-1 | 5′-TGC TTC ATT CGC CTC ACA AA-3′ | 5′-TGC TGT CCT CCT GGA GAA AGT T-3′ | 5′-AAC CAC AGA ACC ACA AGT GCG GTG C-3′ | AF053949 |
| Osteocalcin | 5′-GCA GGT GCG AAG CCC A-3′ | 5′-TCC TGC TTG GAC ACA AAG GC-3′ | 5′-CGG TGC AGA GTC CAG CAA AGG TGC-3′ | NM-000711 |
Primers and probes were designed with sequence data from GenBank and the real-time RT-PCR probe/primer design software Primer Express (version 1.0, PerkinElmer), which optimized the sequences for use in RT-PCR. These sequences were synthesized, and PCR conditions were optimized with respect to concentrations of Mg2+, probe, and both primers to maximize signal. For real-time RT-PCR analysis, 2.5 μl of RNA (20 ng/ml) sample were added to a reverse transcription reaction mixture consisting of 0.5 μl of RNase inhibitor (40 U/μl), 2.0 μl of 10× TaqMan universal master mix buffer, 3.6 μl of MgCl2 (25 mM), 2.0 μl (10 μM) of reverse primers for Cx43 (same procedures and methods used for Cx45, osteopontin, type I collagen, osteocalcin, and Cbfa1), 1.0 μl of 18S reverse primer, 1.0 μl each of ATP, CTP, GTP, and UTP (10 mM), 0.44 μl (50 U/μl) of murine leukemia virus reverse transcriptase, and 5.5 μl of diethyl pyrocarbonate-treated deionized H2O. This mixture was placed in the thermocycler for 1 h at 42°C, 5 min at 72°C, and 2 min at 25°C. Eight microliters of this mixture were then added to the PCR mixture consisting of 5 μl of 10× TaqMan universal master mixture buffer, 4.0 μl of MgCl2 (25 mM), 2.0 μl of each forward and reverse Cx43 primer (10 μM), 1.0 μl of forward and reverse 18S primer (0.25 μM), 1.0 μl of 18S probe (1.0 μM), 1.0 μl each of ATP, CTP, GTP, and UTP (10 mM), 0.25 μl of TAQ Gold (5 U/μl), and 18.35 μl of deionized H2O and subjected to 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C in the PerkinElmer ABI Prism 7700. The fluorescence time course was recorded, and the number of amplification steps required to reach an arbitrary intensity threshold (Ct) was counted. As an example, the ratio of Cx43 to 18S mRNA was calculated from the formula: ACx43/A18S = 2Ct,18S/2Ct,Cx43 = 2(Ct,18S − Ct,Cx43), where ACx43 and A18S are the initial concentrations of Cx43 and 18S rRNA, respectively, and Ct,Cx43 and Ct,18S are the number of cycles necessary to reach threshold intensity for Cx43 and 18S rRNA, respectively. Using this procedure, Cx32, Cx45, osteocalcin, osteopontin, type I collagen, and Cbfa1 were also quantified.
Statistical analysis.
All data are reported as means ± SE. One-way ANOVA, followed by a Student-Newman-Keuls test, was used to analyze all data. P < 0.05 was considered significant.
RESULTS
Vector construction and stable transfection of hFOB 1.19 cells with Cx43 antisense DNA.
After the ligation of human Cx43 cDNA into pcDNA3.1/hygromycin(+), sequence analysis indicated that a 1,817-bp fragment of human Cx43 was inserted into the plasmid in an antisense orientation. After stable transfection with an antisense expressing plasmid (Cx43−) or pvc, single hygromycin-resistant colonies of hFOB/Cx43− and hFOB/pvc were produced. hFOB/Cx43− cells and hFOB/pvc cells attached to the culture plates within 8–10 h of plating and displayed a growth pattern similar to that observed in nontransfected hFOB 1.19 cells. All clones were subjected to screening by Western blot analysis and Lucifer yellow dye microinjection. Clones 1 and 2 (hFOB/Cx43− 1 and hFOB/Cx43− 2) expressed very low levels of Cx43 protein compared with hFOB/pvc (Fig. 1). Similarly to parental hFOB 1.19 cells, Cx43 protein was increased in hFOB/pvc cells as a function of time in cells cultured at 39.5°C but was inhibited in clones of hFOB/Cx43− (Fig. 2). Lucifer yellow dye spread experiments revealed that the two hFOB/Cx43− clones exhibited similar GJIC coupling ratios of 2.1:1 (coupling cell/microinjected cell). There was a 50% reduction in GJIC coupling ratio compared with hFOB/pvc cells and hFOB 1.19 cells, in which the GJIC coupling ratio was 4.5:1 (Fig. 3). Therefore, hFOB/Cx43− 1 and hFOB/Cx43− 2 were selected as candidate cell lines for this investigation.
Fig. 1.Western blot for connexin (Cx)43 protein in human fetal osteoblasts (hFOB). A: hFOB 1.19, hFOB/Cx43− clone 1, hFOB/Cx43− clone 2, and hFOB/plasmid vector control (pvc) were cultured at 37°C for 3 days. Proteins were isolated from cell cultures and subjected to Western blot analysis with a Cx43 monoclonal antibody as described in text dehydrogenase. B: densitometric analysis indicated that the abundance of Cx43 protein in hFOB/Cx43− was significantly decreased relative to hFOB/pvc (average of 3 gels from 2 clones of hFOB/Cx43− and hFOB/pvc). Data are means ± SE. *Significantly less than hFOB/pvc, P < 0.05 (n = 3).
Fig. 2.Western blot analysis of Cx43 protein in hFOB/Cx43− and hFOB/pvc cells. A: hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Proteins were isolated and subjected to Western blot analysis with a monoclonal antibody against Cx43. Cx43 protein in hFOB/pvc cells was increased as a function of time but inhibited in hFOB/Cx43−. B: densitometric analysis indicated that the abundance of Cx43 protein in hFOB/Cx43− was significantly decreased relative to hFOB/pvc at each time point (average of 3 gels from 2 clones of hFOB/Cx43− and plasmid control hFOB/pvc). Data are means ± SE. *Significantly less than hFOB/pvc, **significantly greater than hFOB/pvc on days 3, 6, and 9, P < 0.05 (n = 3).
Fig. 3.Quantitative analysis of functional gap junctional intercellular communication (GJIC) in hFOB. Individual cells were loaded with Lucifer yellow by microinjection. The number of neighboring cells containing dye was quantified 5 min after loading. Values are plotted as means ± SE of the number of coupled cells (n = 40). *Significantly less than hFOB/pvc and hFOB (P < 0.05).
Expression of Cx43 and Cx45 mRNA in hFOB/Cx43− cells.
To examine connexin expression in hFOB/Cx43− cells, cells were cultured for 3, 6, 9, and 12 days. On day 6, one-half of the dishes were cultured at 39.5°C to stimulate differentiation. Samples were collected and RNA was isolated for real-time RT-PCR. Because our previous data (7) showed that hFOB 1.19 expressed predominantly Cx43 and Cx45 but not other connexins, in this study we focused on Cx43 and Cx45. Quantitative real-time RT-PCR analysis indicated that steady-state levels of Cx43 mRNA in hFOB/pvc cells were increased as a function of time when cells were cultured at 39.5°C, which was consistent with our previous study (7) on parental hFOB 1.19 cells. Cx43 mRNA levels in hFOB/Cx43− were significantly decreased relative to hFOB/pvc at 39.5°C and did not increase with time in culture (Fig. 4). No significant changes in Cx45 mRNA levels were found in hFOB/Cx43− and hFOB/pvc cells during culture periods at either 37°C or 39.5°C, and no significant differences in Cx45 mRNA levels were observed between hFOB/Cx43− and hFOB/pvc cells (Fig. 5).
Fig. 4.Quantitative real-time RT-PCR analysis of steady-state levels of Cx43 mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Cx43 mRNA levels in hFOB/Cx43− decreased relative to hFOB/pvc on day 12 at 39.5°C. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. *Significantly greater than hFOB/Cx43− on day 12 and same cell type, hFOB/pvc, on days 3, 6, and 9 (P < 0.05).
Fig. 5.Quantitative real-time RT-PCR analysis for steady-state levels of Cx45 mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. No significant difference between Cx45 mRNA in hFOB/Cx43− and hFOB/pvc was detected.
Expression of osteoblastic differentiation markers in hFOB/Cx43− cells.
To examine the effect of inhibiting Cx43 expression on hFOB 1.19 cell differentiation we cultured the cells and quantified osteoblastic differentiation markers. Samples collected were assessed for alkaline phosphatase activity. Steady-state mRNA levels for osteopontin, type I collagen, osteocalcin, and Cbfa1 were quantified by real-time RT-PCR. Alkaline phosphatase activity in hFOB/pvc cells significantly increased as a function of time in cells cultured at 39.5°C and peaked on day 12 (Fig. 6). However, alkaline phosphatase activity was almost completely inhibited in hFOB/Cx43− cells cultured at both 37°C and 39.5°C and did not increase with time in culture. Osteopontin mRNA levels in hFOB/Cx43− and hFOB/pvc cells were decreased as a function of time in culture at both 37°C and 39.5°C. Osteopontin in Cx43-deficient hFOB/Cx43− cells was significantly elevated compared with hFOB/pvc counterparts at all time points (Fig. 7). Type I collagen mRNA levels decreased as a function of time in hFOB/Cx43− and hFOB/pvc, but no significant differences between hFOB/Cx43− and hFOB/pvc cells were observed (Fig. 8). Osteocalcin mRNA levels in hFOB/pvc were increased as a function of time, in a manner similar to parental hFOB 1.19 cells cultured at 39.5°C (7). In hFOB/Cx43− cells osteocalcin levels were lower than in hFOB/pvc on day 9 but rebounded and were higher, although not significantly, on day 12 (Fig. 9). Similar to osteocalcin expression, Cbfa1 mRNA levels in hFOB/pvc were increased as a function of time in culture at 39.5°C. Cbfa1 levels in hFOB/Cx43−, although starting out higher than those in hFOB/pvc, did not increase with time in culture and at day 12 were lower than in hFOB/pvc (Fig. 10).
Fig. 6.Alkaline phosphatase (AP) activity in hFOB/Cx43− and hFOB/pvc cells. Cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Alkaline phosphatase activity in hFOB/Cx43− cells at 37°C and 39.5°C was significantly lower than that in hFOB/pvc at all time points. Alkaline phosphatase activity in hFOB/pvc increased as a function of time cultured at 39.5°C. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. *Significantly greater than hFOB/Cx43− on days 3, 6, 9, and 12 at 37°C or 39.5°C (P < 0.05); #significantly greater than same cell type on days 3 and 6 (P < 0.05); **significantly higher than same cell type on days 3, 6, and 9 (P < 0.05).
Fig. 7.Quantitative real-time RT-PCR analysis for steady-state levels of osteopontin (OPN) mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Osteopontin decreased as a function of time in culture in both hFOB/Cx43− and hFOB/pvc cells at 37°C and 39.5°C. Osteopontin in hFOB/Cx43− cells significantly elevated relative to that in hFOB/pvc cells cultured at both 37°C and 39.5°C (except day 12 at 37°C). Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. *Significantly greater than hFOB/pvc at 37°C or 39.5°C on corresponding days (P < 0.05); #significantly less than same cell type on day 3 (P < 0.05).
Fig. 8.Quantitative real-time RT-PCR analysis for steady-state levels of type I collagen mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Type I collagen mRNA decreased as a function of time in culture in both hFOB/Cx43− and hFOB/pvc cells. However, there was no significant difference between hFOB/Cx43− and hFOB/pvc. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. *Significantly less than same cell type on day 3 (P < 0.05).
Fig. 9.Quantitative real-time RT-PCR analysis of steady-state levels of osteocalcin (OC) mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. Osteocalcin in hFOB/pvc cells increased as a function of time in culture at 39.5°C. Osteocalcin in hFOB/pvc was significantly higher than hFOB/Cx43− on day 9 at 39.5°C. *Significantly greater than hFOB/Cx43− on day 9 at 39.5°C (P < 0.05); #significantly greater than same cell type on days 3 and 6 (P < 0.05).
Fig. 10.Quantitative real-time RT-PCR analysis for steady-state levels of core binding factor α1 (Cbfa1) mRNA. hFOB/Cx43− and hFOB/pvc cells were plated at 1 × 103 cells/mm2 and assayed on days 3, 6, 9, and 12. Starting on day 6, one-half of the cells were cultured at 39.5°C. Cbfa1 increased in hFOB/pvc cells as a function of time in culture, whereas it decreased in hFOB/Cx43− cells. Values represent the averages of 2 clones of hFOB/Cx43− and 5 experiments. *Significantly greater than same cell type on days 3 and 6 and greater than hFOB/Cx43− on day 12 at 39.5°C (P < 0.05); **significantly greater than same cell type on days 3 and 6 and greater than hFOB/Cx43− on day 12 at 37°C (P < 0.05); #significantly lower than same cell type on days 3, 6, and 9 (P < 0.05).
DISCUSSION
The phenotypic characteristics of differentiated osteoblastic cells have been well characterized. Alkaline phosphatase, osteocalcin, and Cbfa1 are upregulated after the proliferative stage of osteoblastic development. Alkaline phosphatase activity peaks during matrix maturation; thereafter, osteopontin and osteocalcin increase to maximal levels during the mineralization phase. Cbfa1 plays a role as a key transcription factor in the regulation of these differentiation markers (13). Studies from our laboratory (8, 21, 34) and the work of others (5, 17, 26, 28, 33) suggest that expression of Cx43, the predominant gap junction protein in bone, parallels human osteoblastic cell differentiation and contributes to expression of alkaline phosphatase activity, an osteoblastic differentiation marker. However, these studies relied on the use of pharmacological agents such as the gap junctional communication inhibitor 18α-glycyrrhetinic acid and oleamide. To remedy this shortcoming, we decreased Cx43 expression by transfection of a Cx43 antisense DNA in hFOB 1.19 cells. Thus we were able to examine the role of a specific connexin, Cx43, in differentiation of a nontransformed human osteoblastic cell line. This is the first study to demonstrate the importance of Cx43 and GJIC in the differentiation of nontransformed human osteoblasts.
We found that decreasing connexin expression in hFOB/Cx43− greatly attenuated GJIC. However, significant GJIC remained in these cells. Because osteoblastic cells express Cx45 and Cx46 as well as Cx43 (16), it is possible that these connexins might be upregulated in hFOB 1.19 cells lacking Cx43 as a partial compensatory mechanism and that this accounted for GJIC in Cx43-deficient cells. This concept is consistent with what was observed in a Cx43-null mice model (18). We found no changes in Cx45 mRNA in hFOB/Cx43− compared with hFOB/pvc and no changes in Cx45 mRNA in hFOB/Cx43− cells during differentiation. Cx46 is not expressed by hFOB 1.19 cells (unpublished data) and so was not examined in this context. These findings suggest that transfection of Cx43 antisense cDNA did not alter endogenous Cx45 in hFOB 1.19 cells. Therefore, there was not a compensatory increase in Cx45. However, there was detectable Cx45 in hFOB/Cx43−, and this, the residual Cx43, or both, may have contributed to GJIC in hFOB/Cx43−.
Despite a 50% reduction in GJIC in hFOB/Cx43− cells, alkaline phosphatase activity was almost completely inhibited during the entire differentiation period. This suggests that Cx43 expression, and to some degree GJIC, contributes to osteoblastic differentiation. These findings also support the possibility that a GJIC-independent function of Cx43, e.g., Cx43 expression per se or Cx43 hemichannels, influences alkaline phosphatase activity and bone cell differentiation. However, the data obtained do not rule out the possibility that changes we observed were at least in part due to changes in GJIC. We did not examine mineralization in this study. However, because alkaline phosphatase activity is critical to mineralization (11, 35) and correlated with differentiation in several osteoblastic cell lines (2, 3), including hFOB 1.19 cells (Ref. 14 and unpublished data), our results suggest that mineralization would also be decreased in Cx43-deficient osteoblastic cells.
During bone development, osteopontin is expressed at an early stage of bone formation (23), laid into unmineralized matrix before calcification, and localized at matrix-matrix and matrix-cell interfaces as well as between collagen fibrils of fully matured hard tissue. hFOB/Cx43− cells expressing higher levels of osteopontin suggest that Cx43 is a negative regulator of osteopontin in hFOB 1.19. These findings are consistent with our previous work (7) and that of others (18) suggesting higher osteopontin expression in Cx43-deficient bone cells. This inverse relationship between Cx43 and osteopontin has also been demonstrated in nonbone cells (19). In a previous study (19), MDA-MB-435 breast cancer cells, which do not normally express Cx43 but express abundant osteopontin, were transfected with Cx43. We found that MDA-MB-435 cells forced to express Cx43 had reduced osteopontin expression. Additionally, MDA-MB-435 cells expressing the metastasis suppressor gene BRMS-1, and which do express Cx43, had reduced osteopontin expression (19). However, the mechanism of Cx43 downregulation of osteopontin is unknown.
The expression patterns of type I collagen in hFOB/Cx43− and hFOB/pvc cells are similar to those reported in other studies (31), in which expression of type I collagen was higher in the early cell proliferation phase, peaked by the end of the first week, and was followed by a sharp decline in later differentiation. In this study, we did not observe changes in expression of type I collagen in hFOB/Cx43−. These findings suggest that Cx43 is not critical to type I collagen regulation during osteoblastic differentiation.
Osteocalcin has been considered an important marker in osteoblastic differentiation. Interference with gap junction expression has been shown to decrease the activity of a rat osteocalcin promoter-driven reporter gene construct in osteoblastic cells (17). Consistent with these data, our study suggests that osteocalcin increases as a function of time in culture and is inhibited in hFOB/Cx43− cells. Because of the large variability in osteocalcin mRNA levels in hFOB/Cx43− cells at 39.5°C sampled on day 12, the difference between osteocalcin mRNA levels in hFOB/Cx43− and hFOB/pvc on day 12 were not statistically significant. However, it is interesting that at day 9, osteocalcin mRNA levels are significantly lower in hFOB/Cx43− than in hFOB/pvc. This suggests that the hFOB/Cx43− cells are able to compensate for the lower Cx43 expression such that by day 12, osteocalcin mRNA levels have rebounded to “normal” levels. Our previous studies (21) and those of others (17) demonstrate that genetically modifying GJIC, either by expression of Cx45, which is not normally expressed, or by expression of antisense Cx43, resulted in decreased osteocalcin synthesis in ROS 17/2.8 cells. Together these findings suggest that GJIC and Cx43 are important regulators for osteocalcin synthesis both in well-differentiated rat osteoblastic cells and in less well-differentiated human fetal osteoblastic cells.
Cbfa1 has been suggested to regulate the expression of all the major genes expressed by osteoblasts and has been identified as a key regulator of osteoblastic differentiation in vivo (9, 10). As a specific regulator of osteoblastic gene expression, in intramembranous ossification Cbfa1 may act as a differentiation factor acting first on type I collagen and then on other noncollagenous proteins such as osteopontin, bone morphogenetic protein, BSP, and later osteocalcin during differentiation and mineralization (9, 10). In this investigation, cells were stimulated to differentiate by culture at 39.5°C. Cbfa1 mRNA increased as a function of time in culture in hFOB/pvc cells but was inhibited in hFOB/Cx43− cells. These data support the hypothesis that Cx43 contributes to a fundamental mechanism of differentiation by affecting the central transcription factor Cbfa1 in regulating other osteoblastic genes such as osteopontin and osteocalcin.
In summary, our results suggest that genetically modifying Cx43 and GJIC by transfection with Cx43 antisense cDNA alters phenotypic characteristics of osteoblasts, which include alkaline phosphatase activity and gene expression of osteopontin, osteocalcin, and Cbfa1, suggesting that Cx43 expression contributes at least partially to osteoblastic differentiation. The mechanism by which this occurs is unknown. However, a connexin response element has recently been identified that is regulated by the transcription factors Sp1 and Sp3 and regulates the expression of osteoblastic genes (30). Furthermore, the extracellular signal-regulated kinase pathway regulates recruitment of Sp1 to the connexin response element (29). Therefore, GJIC may contribute to osteoblastic differentiation by facilitating, via the transfer of signaling molecules, the expression of osteoblastic genes associated with differentiation. In any case, modulation of GJIC and expression of connexins may represent a novel target for altering regulation of osteoblastic cell differentiation.
GRANTS
This work was supported by National Institute on Aging Grant AG-13087 and the Pennsylvania Department of Health, using Tobacco Settlement Funds.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
- 1 Akisaka T, Yoshida H, Kogaya Y, Kim S, Yamamoto M, and Kataoka K. Membrane modifications in chick osteoclasts revealed by freeze-fracture replicas. Am J Anat 188: 381–392, 1990.
Crossref | PubMed | Google Scholar - 2 Aubin JE. Advances in the osteoblast lineage. Biochem Cell Biol 76: 899–910, 1998.
Crossref | PubMed | ISI | Google Scholar - 3 Aubin JE and Triffitt JT. Mesenchymal Stem Cells and Osteoblast Differentiation. San Diego, CA: Academic, 2002.
Google Scholar - 4 Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248–254, 1976.
Crossref | PubMed | ISI | Google Scholar - 5 Chiba H, Sawada N, Oyamada M, Kojima T, Nomura S, Ishii S, and Mori M. Relationship between the expression of the gap junction protein and osteoblast phenotype in a human osteoblastic cell line during cell proliferation. Cell Struct Funct 18: 419–426, 1993.
Crossref | PubMed | ISI | Google Scholar - 6 Donahue HJ, Guilak F, Van der Molen M, McLeod RJ, Rubin CT, Grande DA, and Brink PR. Chondrocytes isolated from mature articular cartilage retain the capacity to form functional gap junctions. J Bone Miner Res 10: 1359–1364, 1995.
Crossref | PubMed | ISI | Google Scholar - 7 Donahue HJ, Li Z, Zhou Z, and Yellowley CE. Differentiation of human fetal osteoblastic cells and gap junctional intercellular communication. Am J Physiol Cell Physiol 278: C315–C322, 2000.
Link | ISI | Google Scholar - 9 Ducy P. Cbfa1: a molecular switch in osteoblast biology. Dev Dyn 219: 461–471, 2000.
Crossref | PubMed | ISI | Google Scholar - 10 Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, Geoffroy V, Amling M, and Karsenty G. A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev 13: 1025–1036, 1999.
Crossref | PubMed | ISI | Google Scholar - 11 Fedde KN. Human osteosarcoma cells spontaneously release matrix-vesicle-like structures with the capacity to mineralize. Bone Miner 17: 145–151, 1992.
Crossref | PubMed | Google Scholar - 12 Fishman GI, Spray DC, and Leinwand LA. Molecular characterization and functional expression of the human cardiac gap junction channel. J Cell Biol 111: 589–598, 1990.
Crossref | PubMed | ISI | Google Scholar - 13 Franceschi RT, Iyer BS, and Cui Y. Effects of ascorbic acid on collagen matrix formation and osteoblast differentiation in murine MC3T3-E1 cells. J Bone Miner Res 9: 843–854, 1994.
Crossref | PubMed | ISI | Google Scholar - 14 Harris SA, Enger RJ, Riggs BL, and Spelsberg TC. Development and characterization of a conditionally immortalized human fetal osteoblastic cell line. J Bone Miner Res 10: 178–186, 1995.
Crossref | PubMed | ISI | Google Scholar - 15 Huang W, Rudkin GH, Carlsen B, Ishida K, Ghasri P, Anvar B, Yamaguchi DT, and Miller TA. Overexpression of BMP-2 modulates morphology, growth, and gene expression in osteoblastic cells. Exp Cell Res 274: 226–234, 2002.
Crossref | PubMed | ISI | Google Scholar - 16 Koval M, Harley JE, Hick E, and Steinberg TH. Connexin46 is retained as monomers in a trans-Golgi compartment of osteoblastic cells. J Cell Biol 137: 847–857, 1997.
Crossref | PubMed | ISI | Google Scholar - 17 Lecanda F, Towler DA, Ziambaras K, Cheng S, Koval M, Steinberg TH, and Civitelli R. Gap junctional communication modulates gene expression in osteoblastic cells. Mol Biol Cell 9: 2249–2258, 1998.
Crossref | PubMed | ISI | Google Scholar - 18 Lecanda F, Warlow PM, Sheikh S, Furlan F, Steinberg TH, and Civitelli R. Connexin 43 deficiency causes delayed ossification, craniofacial abnormalities, and osteoblast dysfunction. J Cell Biol 151: 931–944, 2000.
Crossref | PubMed | ISI | Google Scholar - 19 Li Z, Zhou Z, Welch DR, and Donahue HJ. Restoration of connexin 43 alters breast cancer cell metastatic potential to bone by down regulation of osteopontin and connexin 32 (Abstract no. 0186). Transactions of the 48th Annual Meeting of the Orthopaedic Research Society, vol. 27, 2002.
Google Scholar - 20 Li Z, Zhou Z, and Daniel EE. Expression of gap junction connexin 43 and connexin 43 mRNA in different regional tissues of intestine in dog. Am J Physiol Gastrointest Liver Physiol 265: G911–G916, 1993.
Link | ISI | Google Scholar - 21 Li Z, Zhou Z, Yellowley CE, and Donahue HJ. Inhibiting gap junctional intercellular communication alters expression of differentiation markers in osteoblastic cells. Bone 25: 661–666, 1999.
Crossref | PubMed | ISI | Google Scholar - 22 Lo CW. The role of gap junction membrane channels in development. J Bioenerg Biomembr 28: 379–385, 1996.
Crossref | PubMed | ISI | Google Scholar - 23 Mark MP, Butler WT, Prince CW, Finkelman RD, and Ruch JV. Developmental expression of 44-kDa bone phosphoprotein (osteopontin) and bone gamma-carboxyglutamic acid (Gla)-containing protein (osteocalcin) in calcifying tissues of rat. Differentiation 37: 123–136, 1988.
Crossref | PubMed | ISI | Google Scholar - 24 Minkoff R, Bales ES, Kerr CA, and Struss WE. Antisense oligonucleotide blockade of connexin expression during embryonic bone formation: evidence of functional compensation within a multigene family. Dev Genet 24: 43–56, 1999.
Crossref | PubMed | Google Scholar - 25 Minkoff R, Rundus VR, Parker SB, Hertzberg EL, Laing JG, and Beyer EC. Gap junction proteins exhibit early and specific expression during intramembranous bone formation in the developing chick mandible. Anat Embryol (Berl) 190: 231–241, 1994.
Crossref | PubMed | Google Scholar - 26 Schiller PC, D'Ippolito G, Balkan W, Roos BA, and Howard GA. Gap-junctional communication mediates parathyroid hormone stimulation of mineralization in osteoblastic cultures. Bone 28: 38–44, 2001.
Crossref | PubMed | ISI | Google Scholar - 27 Schiller PC, D'Ippolito G, Brambilla R, Roos BA, and Howard GA. Inhibition of gap-junctional communication induces the trans-differentiation of osteoblasts to an adipocytic phenotype in vitro. J Biol Chem 276: 14133–14138, 2001.
Crossref | PubMed | ISI | Google Scholar - 28 Schiller PC, Roos BA, and Howard GA. Parathyroid hormone up-regulation of connexin 43 gene expression in osteoblasts depends on cell phenotype. J Bone Miner Res 12: 2005–2013, 1997.
Crossref | PubMed | ISI | Google Scholar - 29 Stains JP and Civitelli R. Cell-cell interactions in regulating osteogenesis and osteoblast function. Birth Defects Res C Embryo Today 75: 72–80, 2005.
Crossref | PubMed | Google Scholar - 30 Stains JP, Lecanda F, Screen J, Towler DA, and Civitelli R. Gap junctional communication modulates gene transcription by altering the recruitment of Sp1 and Sp3 to connexin-response elements in osteoblast promoters. J Biol Chem 278: 24377–24387, 2003.
Crossref | PubMed | ISI | Google Scholar - 31 Stein GS, Lian JB, Stein JL, Van Wijnen AJ, and Montecino M. Transcriptional control of osteoblast growth and differentiation. Physiol Rev 76: 593–629, 1996.
Link | ISI | Google Scholar - 32 Subramaniam M, Jalal SM, Rickard DJ, Harris SA, Bolander ME, and Spelsberg TC. Further characterization of human fetal osteoblastic hFOB 1.19 and hFOB/ER alpha cells: bone formation in vivo and karyotype analysis using multicolor fluorescent in situ hybridization. J Cell Biochem 87: 9–15, 2002.
Crossref | PubMed | ISI | Google Scholar - 33 Upham BL, Suzuki J, Chen G, Wang Y, McCabe LR, Chang CC, Krutovskikh VA, Yamasaki H, and Trosko JE. Reduced gap junctional intercellular communication and altered biological effects in mouse osteoblast and rat liver oval cell lines transfected with dominant-negative connexin 43. Mol Carcinog 37: 192–201, 2003.
Crossref | PubMed | ISI | Google Scholar - 34 Vander Molen MA, Rubin CT, McLeod KJ, McCauley LK, and Donahue HJ. Gap junctional intercellular communication contributes to hormonal responsiveness in osteoblastic networks. J Biol Chem 271: 12165–12171, 1996.
Crossref | PubMed | ISI | Google Scholar - 35 Whyte MP. Hypophosphatasia and the role of alkaline phosphatase in skeletal mineralization. Endocr Rev 15: 439–461, 1994.
PubMed | ISI | Google Scholar - 36 You J, Yellowley CE, Donahue HJ, Zhang Y, Chen Q, and Jacobs CR. Substrate deformation levels associated with routine physical activity are less stimulatory to bone cells relative to loading-induced oscillatory fluid flow. J Biomech Eng 122: 387–393, 2000.
Crossref | PubMed | ISI | Google Scholar
