Gap junctions are membrane-spanning channels that facilitate intercellular communication by allowing passive flux of intracellular signaling molecules, such as Ca2+, cyclic AMP (cAMP), and ATP, to pass from the cytosol of one cell directly into the cytosol of another.1 Gap junction channels are composed of protein subunits termed connexins (Cx), at least 20 of which have been identified in mammalian species.2 Hexamerization of Cx generates a connexon, which can function unapposed in the plasmalemma (Cx hemichannel), or can bind to another connexon in a neighboring cell to generate a functional gap junction. In adult bone, the predominant gap junction protein expressed is connexin 43 (Cx43) but Cx45 and 46 are also expressed,3 although Cx46 is retained in monomeric form in the golgi network and thus does not present itself within the plasmalemma.4
The role of gap junctions and gap junction intercellular communication (GJIC) in skeletal metabolism and homeostasis is well-established. Several studies have presented compelling evidence that GJIC is critical for osteoblastic cell differentiation into mature bone-forming cells both in vitro and in vivo.5–9 Additionally, previous studies demonstrate that GJIC contributes to the responsiveness of osteoblastic cells in vitro to diverse anabolic signals including parathyroid hormone,10 electromagnetic fields,11 and fluid shear stress.12–15 Gap junction channels also allow osteocytes to communicate mechanical signals they detect to osteoblastic cells.16 The critical importance of connexons, especially Cx43, are noted in murine models and in human disease.17 In mice, targeted Cx43 mutagenesis elicits neonatal lethality because of severe cardiovascular malformation18; nonetheless, these mice demonstrate hypomineralization of craniofacial bones, as well as delayed ossification of appendicular and axial skeleton.6 Similar results are observed in mice with osteoblast and osteocyte-specific deletion of Cx43.19 These mice also display altered bone adaptation to mechanical load.20 In humans, mis-sense mutations in Cx43 produce the rare genetic disease oculodentodigital dysplasia (ODDD; OMIM 164200),21 involving, among other pathologies, cranial hyperostosis (reviewed in).22 Taken together these studies indicate that GJIC involving Cx43 significantly contributes to skeletal homeostasis.
Bone formation rates decrease with increasing age,23, 24 suggesting that decreased osteoblastic activity may contribute to age-related osteopenia.25 Osteoblast activity is determined to a large extent by the capacity of osteoprogenitors and osteoblasts to adapt to changes in their extracellular environment. The ability of osteoblastic cells to adapt to hormonal or biophysical signals appears to decrease with advanced age.26–28 Since GJIC contributes to bone cell responsiveness to extracellular signals, we postulated that GJIC may change as a function of age. Furthermore, we have previously reported an age-related decrease in PTH-stimulated cAMP accumulation in osteoblastic cells27 and since PTH stimulation of GJIC is at least partially dependent on cAMP,27, 29, 30 we also postulated that PTH-stimulated GJIC decreases as a function of osteoblast age.
We examined GJIC in primary cultures of rat osteoblastic cells isolated from young (4-month-old), mature (12-month-old), and old (24–28-month-old) rats. In order to examine the mechanisms underlying any age-related changes we found in GJIC, we also examined PTH-stimulated Cx43 mRNA and protein expression, and PTH and cholera toxin (CTX)-stimulated GJIC in ROB. We observed that PTH stimulated GJIC in young ROBs, but not mature or old ROBs, and this occurred independently of PTH-stimulated changes in Cx43 mRNA or protein expression. In agreement with our previous work upon age-related changes in cAMP induction, we observed impaired CTX-stimulated GJIC in osteoblastic cells from mature and old rats compared to those from young rats, suggesting that aging may alter the mechanisms whereby GJIC communication is regulated.
Rat parathyroid hormone fragment 1–34 (rPTH[1–34]) was purchased from Bachem. RNeasy RNA isolation kits were purchased from Qiagen. Reagents for real time RT-PCR were purchased from Applied Biosystems, Foster City, CA. Calcein-AM and 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI) were purchased from Molecular Probes, Grand Island, NY. Dulbecco's modified Eagle's media (DMEM), and penicillin/streptomycin were purchased from Gibco, Grand Island, NY. Fetal bovine serum (FBS) was purchased from Hyclone Laboratories, Rockford, IL. CTX was purchased from Calbiochem, Billerica, MA. All other reagents were purchased from Fisher Scientific (St. Louis, MO) and were of tissue culture grade.
Primary Culture of Rat Osteoblastic Cells
Rat osteoblastic cells were isolated from the tibiae and femorae of 4-, 12-, and 24–28-month-old male Fisher 344 rats as previously described.27, 31 Briefly, sub-periosteal osteoblastic cells were isolated by sequential collagenase digestion at 37°C. Second digestion cells were collected by centrifugation and placed into T-25 tissue culture flasks and cultured in DMEM supplemented with 20% FBS, 100 IU/ml penicillin and 100 µg/ml streptomycin. Cells were allowed to reach confluence and then subcultured into culture dishes or glass slides for each experiment. We have previously demonstrated that cells isolated in this manner express phenotypic characteristics of osteoblastic cells.31
Assessment of GJIC
GJIC between osteoblastic cells was assessed by a double fluorescent labeling technique.32 “Acceptor” cells were plated at a density of 4 × 104 cells/cm2 on 25 mm glass and were allowed to reach 90% confluence, over a period of 24–48 h. Simultaneously, “donor” cells were plated at a density of 4 × 104 cells/cm2 in 35 mm tissue culture dishes and maintained at 37°C over the same time period. To measure GJIC, donor cells were washed twice with phosphate-buffered saline (PBS), incubated with 10 µM calcein-AM, 10 µM DiI, 1% pluronic acid, and 20 mg/mL BSA in Hank's Balanced Salt Solution, and incubated for 20 min at 37°C. Following incubation, donor cells were washed with PBS, collected via trypsin, and 500 donor cells were added to confluent monolayers of unlabeled acceptor cells and incubated for 20 min at 37°C. Thereafter, coverslips were removed, washed in PBS and inverted onto microscope slides; GJIC was evaluated using a Nikon epifluorescence microscope (Nikon EFD-3, Optical Apparatus Co., Ardmore, PA) and visualized to locate the calcein and DiI loaded cells, respectively. Calcein-AM, because of its small size (MW 994.87), can be transferred to neighboring cells when functional (open) gap junctions are established between donor and acceptor cells. The fluorescent dye DiI intercalates within cell membranes and does not transfer to neighboring cells via GJIC and is used to identify donor cells.
Both donor cells and acceptor cells were exposed for 20 min at 37°C to vehicle control, 10−10–10−6 M rPTH [1–34] or 10−10–10−6 M CTX. Each vehicle or drug was also present during the formation of GJIC between donor and acceptor cells. For each condition, the number of calcein-positive acceptor cells coupled to each of 40 DiI positive donor cells was recorded. The average number of acceptor cells per donor cell was then calculated. The number of donor cells counted was 40 as it was the maximum number of coupled donor cells under any condition.
Western Blotting and Quantitative Real Time RT-PCR
Osteoblastic cells were plated at 5,000/cm2 in 10 cm2 tissue culture dishes and cultured to 80% confluence (3–4 days). The cells were then exposed for 2 h at 37°C to rPTH [1–34] (10−10–10−6 M) or CTX (10−10–10−7 M) or vehicle control, after which whole cell protein lysate or RNA was isolated. Western immunoblotting and qtPCR was subsequently performed as described previously.33
Basal Gja1 and Cx43 protein expression and basal GJIC (Fig. 1a–c) are reported as mean ± SEM. The remaining data are presented as fold-change in coupled cells compared to age-matched, vehicle control in order to minimize inter-animal variability and are presented as mean ± SEM. Data were analyzed by ANOVA followed by Tukey or Dunnet's test. p < 0.05 was considered statistically significant. For CTX studies, osteoblastic cells from one young animals showed no responsiveness to CTX, and were deemed suitable for exclusion from analysis by Q-test.34
Basal Cx43 Expression and GJIC Is Not Altered as a Function of Age of Animal From Which Osteoblastic Cells Were Isolated
We first examined the influence of donor animal age on Cx43 mRNA, protein, and GJIC expression in osteoblastic cells isolated from young, mature, and old rats. We found no significant influence of donor animal age on basal Gja1 (Cx43 mRNA; Fig. 1a) or protein (Fig. 1b). Similarly, we observed no age-related change basal GJIC in osteoblastic cells (Fig. 1c).
Age of Donor Animal Influences PTH-Stimulated GJIC in Osteoblastic Cells
Osteoblastic cells isolated from young rats demonstrated a statistically significant increase in the number of coupled cells in response to PTH treatment (Fig. 2a). Exposure to PTH (10−11–10−8 M) revealed a trend for increased GJIC in osteoblastic cells isolated from young rats, which reached statistical significance at 10−9 and 10−8 M PTH relative to vehicle controls. In contrast, PTH at any dose used had no effect upon GJIC in osteoblastic cells isolated from mature or old rats (Fig. 2a). We found no effect of PTH upon Cx43 mRNA (Fig. 2b) or protein (Fig. 2c), indicating that PTH increases GJIC via post-translational mechanisms in osteoblastic cells from young rats.
Increasing cAMP Accumulation Does Not Stimulate GJIC in Osteoblastic Cells Isolated From Older Animals
We have previously demonstrated an age-related impairment in cAMP synthesis in response to CTX.27 CTX exerted a dose-dependent increase in GJIC in osteoblastic cells from young animals, but had no effect upon osteoblastic cells from mature or old rats (Fig. 3).
Gap junctions are critical regulators of cell–cell communication and are involved in diverse developmental and homeostatic function. Within cardiac muscle, they rapidly propagate action potentials generated in the sinoatrial (SA) node to enable the myocardium to generate synchronous myocardial contraction.35 Within the vasculature, GJIC between the endothelium and the underlying smooth muscle contributes to endothelium-dependent dilation.36 Since the discovery of gap junctions within the skeleton,37 tremendous effort has been exerted to understand the role and importance of Cx, GJIC, and hemichannels in skeletal development. Ample data have demonstrated a role for gap junctions and Cx, particularly Cx43, in osteoblast differentiation and adult bone homeostasis. Chemical inhibition of GJIC promotes transdifferentiation of osteoblasts into adipocytes 9 and attenuates osteoprogenitor differentiation in vitro.7, 38 Similar findings are reported in vivo.6 What has been studied less extensively, however, is the influence of aging upon Cx43 expression and GJIC in osteoblastic cells. Within, we examined Cx43 expression in osteoblastic cells derived from young, mature, and old rats, and further examined the capacity for GJIC in response to hormonal and chemical signals.
We observed no influence of donor animal age upon Gja1 or Cx43 protein expression, nor was basal GJIC significantly different among these age groups (Fig. 1a–c). In contrast, age-related decreases in Cx43 expression have been observed in the SA node,39 endothelium,40 fibroblasts,41 dental pulp,42 and the bladder.43 While we observed no significant difference in basal Cx43 expression or GJIC among the groups, we did find that PTH was able to stimulate GJIC only in osteoblastic cells from young animals. (Fig. 2a). PTH stimulates GJIC by increasing Cx43 mRNA and protein expression, although this occurs over a longer time-scale (hours rather than minutes). We examined whether there was an age-related impairment of Cx43 mRNA or protein in response to PTH. While there was an age-related impairment in rapid activation of GJIC by PTH (likely independent of de novo protein synthesis; Fig. 2a), no cells from any population demonstrated changes in Cx43 (Gja1) mRNA (Fig. 2b) or protein (Fig. 2c) after 2 h of PTH treatment.
Increasing cAMP positively correlates with increased GJIC, and it is through cAMP signaling that PTH acutely stimulates GJIC.44 We have previously demonstrated that osteoblastic cells from mature and old rats are less capable of generating cAMP in response to PTH or CTX compared to osteoblastic cells from young rats.27 Therefore, we examined whether an age-related inability to generate cAMP in response to hormonal or chemical signals also prevents formation of gap junctions and GJIC in mature and aged mice. PTH dose-dependently increased GJIC in osteoblastic cells from young, but not mature or old, rats. These populations of cells have similar levels of PTH/PTHrP receptor mRNA and numbers of PTH receptors per cell,27 indicating that age-related defects in PTH signaling, but not receptor number or binding affinity, are a likely cause. Since PTH-stimulated cAMP accumulation decreases as a function of age 27 and cAMP mediates PTH stimulation of GJIC,45 the age-related decrease in PTH-stimulate GJIC reported could result from defective cAMP accumulation. CTX, which ADP-ribosylates Gαs to synthesize cAMP, mimicked the capacity of PTH to promote GJIC in young, but, as was the case with PTH, did not increase GJIC in osteoblastic cells from mature and old rats. This suggests that an age-related defect in Gαs protein coupled adenylate cyclase activity at least partially contributes to decreased PTH-stimulated GJIC.
In summary, our results suggest osteoblastic cells isolated from mature (12-month-old) and old (24–28-month-old) rats display decreased PTH- and CTX-stimulated GJIC. The mechanism underlying this decreased responsiveness to PTH involves uncoupling of Gs. protein to adenylate cyclase. Given the critical role GJIC plays in osteoblastic differentiation and bone adaptation to mechanical load, this age-related defect in PTH regulation of cAMP may contribute to age-related bone loss.
This work was supported by Award Number R03AR057547 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (D.C.G.) and Award Number AG013087 from the National Institute on Aging (H.J.D.).