Abstract
Bone loss occurs following chronic ethanol (EtOH) consumption in males and cycling females in part as a result of increased bone resorption. We have demonstrated in vivo that estradiol treatment can reverse this effect. Using osteoclast precursors from bone marrow and osteoblast/preosteoclast coculture, we found that EtOH-induced receptor activator of nuclear factor-κB ligand (RANKL) expression in osteoblasts was able to promote osteoclastogenesis. These effects were blocked by pretreatment of cells with either 17β-estradiol (E2) or the anti-oxidant N-acetyl cysteine (NAC). EtOH treatment of stromal osteoblasts increased the intracellular level of reactive oxygen species (ROS). This was associated with induction of NADPH oxidase (NOX) and a downstream signaling cascade involving sustained activation of extracellular signal-regulated kinase (ERK) and activation of signal transducer and activator of transcription 3, resulting in increased gene expression of RANKL. In the presence of EtOH, sustained nuclear ERK translocation >24 h was observed in calvarial osteoblasts and UMR-106 cells transfected with green fluorescent protein-ERK2 plasmid. This was abolished by pretreatment with either E2 or NAC. NOX subtypes 1, 2, and 4, but not 3, were expressed in stromal osteoblasts. Chemical inhibition of NOX by diphenylene iodonium also reversed the ability of EtOH to phosphorylate ERK and induce RANKL mRNA expression. Down-regulation of EtOH-induced ROS generation in osteoblasts was also observed after treatment with E2 or NAC. These data suggest that the molecular mechanisms whereby E2 prevents EtOH-induced bone loss involve interference with ROS generation and cytoplasmic kinase activation.
Chronic alcohol intake results in toxicity in a variety of tissues. Alcoholic injury in bone eventually results in osteopenia, a disease causing substantial morbidity and mortality in both males and females (Turner, 2000). Such osteopenic bone loss may be initiated by changes in behavior of two bone cell types: osteoblasts and osteoclasts or their precursors. Ethanol (EtOH) is well known to dose-dependently reduce cell proliferation and alkaline phosphatase activity in osteoblasts. Moreover, suppression of osteoblastogenesis is considered to be a major cause of EtOH-inhibited bone growth, bone loss, and deficient bone repair (Chakkalakal, 2005). However, cytokine-mediated stimulation of osteoclastogenesis after EtOH treatment of male mice has been demonstrated previously (Dai et al., 2000), and blood EtOH levels associated with binge drinking have been reported to stimulate bone resorption (Callaci et al., 2006). In agreement with these findings, we have recently demonstrated EtOH-induced bone resorption in cycling female rats through promotion of osteoclastogenesis directly mediated via enhanced receptor activator of nuclear factor-κB ligand (RANKL) expression in stromal osteoblasts (Chen et al., 2006; Shankar et al., 2006). RANKL is a tumor necrosis factor family member, and upon binding to its cognate receptor RANK leads to recruitment of adaptor proteins and tumor necrosis factor-α receptor-associated factors, resulting in the commitment of precursor cells to osteoclastic differentiation (Boyle et al., 2003).
EtOH is metabolized to acetaldehyde by alcohol dehydrogenase (ADH) and CYP2E1. Our previous studies demonstrated that ADH, but not CYP2E1, was highly expressed in stromal osteoblasts (Chen et al., 2006), suggesting that ADH-dependent EtOH metabolism to acetaldehyde may take place locally in osteoblasts and that acetaldehyde may mediate the cellular effects of EtOH. We previously demonstrated that 17-β estradiol (E2) can reverse the effects of EtOH on RANKL expression. In contrast to EtOH, estradiol acts as a ligand for nuclear estrogen receptor α and estrogen receptor β receptors, and each receptor/ligand complex may exert different effects on gene transcription, depending on the presence of tissue-specific coactivators and corepressors. However, in bone cells, estradiol can also have nongenotropic effects as a result of actions on kinase cascades to exert its unique biological functions (Kousteni et al., 2002). It is not yet known whether the inhibitory cross-talk of E2 on EtOH-mediated induction of RANKL in osteoblasts involves actions via genomic or nongenotropic pathways.
Chronic EtOH intake results in production of reactive oxygen species (ROS) in liver Kupffer cells and stellate cells and in the lung in part through activation of NADPH oxidase (NOX) (Novitskiy et al., 2006; Polikandriotis et al., 2006; Thakur et al., 2006). ROS is known to modulate the activity of many signal transduction pathways. Production of ROS by plasma membrane-associated NOX in nonphagocytic cells regulates a number of biological processes, including growth and necrosis/apoptosis (Colston et al., 2005). One of the cytoplasmic kinase pathways activated by ROS is the extracellular signal-regulated kinases (ERKs) (Torres, 2003). E2 has been reported to have direct antioxidant properties in bone cells (Lean et al., 2003). E2 has also been demonstrated to inhibit NOX activity through the regulation of p47phox mRNA and protein expression in nonphagocytic cells (Sumi et al., 2003). We have previously reported that the duration of nuclear accumulation of ERK can be altered by treatment of osteoblasts with E2 and that transient versus sustained nuclear accumulation of ERK can determine the cellular fate of bone cells (Chen et al., 2005). Downstream of ERK, there are a number of factors that can be activated or inactivated, resulting in differential gene expression. Signal transducer and activator of transcription (STAT) 3 may be one such factor that has been reported to be required for induction of RANKL in stromal osteoblastic cells (O'Brien et al., 1999).
The current report investigates the hypothesis that E2 antagonizes EtOH-induced bone resorption as a result of inhibited ROS generation. We present evidence that NOX plays a critical role in the effects of EtOH and E2 on the ERK/STAT3/RANKL signaling cascade associated with osteoclastogenesis.
Materials and Methods
Cells, Cell Culture, and Reagents. Animal handling and experimental treatments were conducted in accordance with ethical guidelines for animal research and were approved by the Institutional Animal Care and Use Committee at University of Arkansas for Medical Sciences (Little Rock, AR). Femoral bone marrow cells were aspirated from control cycling female rats, and bone marrow stromal osteoblasts were differentiated in culture using methods described previously (Di Gregorio et al., 2001; Chen et al., 2006). In brief, bone marrow cells were seeded at a density of 3 × 106 cells per well of six-well cell culture plates in the presence of minimum essential medium (Invitrogen, Carlsbad, CA) with 10% fetal bovine serum (HyClone Laboratories, Logan, UT) and 1 mM ascorbic acid 2-phosphate sesquimagnesium salt (Sigma-Aldrich, St. Louis, MO), 4 mM l-glutamine, and 100 U/ml of each penicillin and streptomycin (Sigma-Aldrich). After 2 days, nonadherent cells were collected and frozen in liquid nitrogen. As published previously, these nonadherent cell fractions, considered to contain osteoclast precursors (Chen et al., 2005), were used in later osteoclast development cultures. Half of the medium was replaced every 5 days. Mature differentiated stromal osteoblasts could be observed between day 20 and day 25. In vitro cell culture systems and the EtOH treatments are identical to those described previously (Chen et al., 2006). Cell culture medium remained the same except that FBS was reduced to 2%. After 6 h for medium saturation with oxygen and carbon dioxide in the incubator, cells were treated with 50 mM EtOH and other reagents for periods of time indicated in under Results. Plates were sealed by a tape to prevent evaporation of EtOH from the culture medium in the wells. For osteoclast formation, at day 20, frozen osteoclast precursors described above were added back into each well at a density of 1 × 106 cells per well in the presence of 20 nM 1α,25-dihydroxyvitamin D3 (Gaddy-Kurten et al., 2002). Cell treatment was immediately started, and half of the culture medium was changed every 2 days until mature multinucleate osteoclasts appeared in the wells by 7 to 10 days. The cells were stained for tartrate-resistant acid phosphatase (TRAPase) (Sigma-Aldrich), and TRAPase-positive multinucleate osteoclasts were counted under a microscope at 10× magnification. Neonatal rat calvaria cells were isolated from untreated 4-day-old rats by sequential collagenase digestion using a method described previously (Wong and Cohn 1975). Rat calvaria cells and the rat osteoblast-like cell line UMR-106 (American Type Culture Collection, Manassas, VA) were cultured in α-minimum essential medium supplemented with 10% FBS. When cells were ready to be treated, culture medium was saturated with oxygen and carbon dioxide in an incubator for 2 h, and plates were sealed during EtOH treatment of stromal osteoblasts described above.
Reverse Transcription-Polymerase Chain Reaction and Real-Time RT-PCR. Total RNA from osteoblastic cells were extracted using TRI Reagent (Sigma-Aldrich) according to the manufacturer's recommendations followed by cleanup and DNase digestion using RNeasy Mini columns (Q1AGEN, Valencia, CA). Reverse transcription was carried out using iScript cDNA synthesis kit from Bio-Rad (Hercules, CA). Real-time RT-PCR was performed using SYBR Green and an ABI 7000 sequence detection system (Applied Biosystems, Foster City, CA). Primers for rat RANKL, NOX1 to 4, GAPDH, and 18S were designed using Primer Express software 2.0.0 (Applied Biosystems). To check for expression of all four subtypes of NOX catalytic subunits in the primary mature stromal osteoblasts, RNA was taken from untreated cells, and cDNA was synthesized using the procedure described above. The as basic PCR amplification conditions were 58°C annealing temperature and 35 cycles. All gene primer sequences used in this study are shown in Table 1.
Flow Cytometric Measurement of ROS. The cell-permeable dye 2,7-dichlorodihydrofluorescein diacetate (2,7-DCF-DA) (Sigma-Aldrich) becomes fluorescent upon reaction with ROS. 2,7-DCF-DA was dissolved in dimethyl sulfoxide and stored as 50 mM stock. Stromal osteoblasts were loaded with 10 μM 2,7-DCF-DA for 30 min, and then they were treated with 10–9 ME2 or 20 mM N-acetylcysteine (NAC) for an additional 30 min before addition of 50 mM EtOH. The cells were continuously treated for 24 h, and then they were washed three times with phosphate-buffered saline before they were harvested. Washed cells were resuspended in 500 μl of phosphate-buffered saline and kept on ice until flow cytometric analysis was started. ROS measurement was immediately carried out by flow cytometry using FACSort (BD Biosciences, Rutherford, NJ) with a 488-nm excitation beam. The signals were obtained using a 530-nm band-pass filter for 2,7-DCF-DA. Each determination was based on the mean fluorescence intensity of 5000 cells.
Western Blotting. Cellular proteins for Western immunoblot analysis were extracted using cell lysis buffer as described previously (Chen et al., 2005). The phosphorylation status of ERK1/2 in rat calvaria osteoblastic cells was examined by a Western blotting using a mouse monoclonal antibody recognizing tyrosine phosphorylated ERK1/2 or rabbit polyclonal antibodies recognizing total ERK1/2 followed by incubation with either an anti-mouse or an anti-rabbit antibody conjugated with horseradish peroxidase (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). The phospho-ERK1/2 and total ERK1/2 blots were reprobed to determine phosphorylation of STAT3 and total STAT3 using a goat polyclonal and a rabbit polyclonal antibody recognizing phosphorylated or total STAT3 followed by incubation with either an anti-goat or an anti-rabbit antibody conjugated with horseradish peroxidase (Santa Cruz Biotechnology, Inc.). Similar Western blotting studies were performed with antibodies to p38 and phospho-p38 (Santa Cruz Biotechnology, Inc.). Blots were developed using chemiluminescence according to the manufacturer's recommendations. Quantization of the intensity of the bands in the autoradiograms was performed using a VersaDoc imaging system (Bio-Rad).
Plasmids, Transient Transfection, and Subcellular Localization of ERK2. Wild-type ERK2 fused to green fluorescent protein (GFP-ERK2) was provided by Dr. Cobb (University of Texas Southwestern Medical Center, Dallas, TX) (Khokhlatchev et al., 1998). Wild-type mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (MEK) was provided by N. G. Ahn (University of Colorado, Boulder, CO) (Mansour et al., 1994). Plasmid of red fluorescent protein targeted to the nucleus (nuclear red fusion protein; nRFP) was created in the laboratory of Dr. Bellido (University of Arkansas for Medical Sciences, Little Rock, AR) (Chen et al., 2005). Calvaria and UMR-106 cells were seeded in 24-well plates. Eighty percent confluent cells were transiently transfected using Lipofectamine Plus (Invitrogen) with GFP-ERK2 and wild-type MEK along with nRFP. After transfection, cells were cultured for 24 h. Subsequently, cells were serum-starved by culturing in the presence of 2% FBS for 4 h, and then cells were treated with 10–9 M E2 or the antioxidant 20 mM NAC for 30 min before adding 50 mM EtOH. Plates were sealed with tape as described above. The cells showing nuclear accumulation of GFP-ERK2 were visualized using a fluorescence microscopy by enumerating cells exhibiting increased GFP in the nucleus compared with the cytoplasm.
Data and Statistical Analysis. Data are expressed as means ± S.E.M. One-way analysis of variance (ANOVA) followed by Student-Newman-Keuls post-hoc analysis was used to compare the treatment groups with the vehicle-treated group. Values were considered statistically significant at p < 0.05.
Results
The Antioxidant NAC Inhibits EtOH-Induced Osteoclastogenesis through Suppression of Induced Expression of RANKL mRNA in Stromal Osteoblasts. We measured RANKL mRNA expression using real-time RT-PCR in stromal osteoblastic cells that were treated with 50 mM EtOH or that were pretreated with 20 mM NAC for 30 min before adding EtOH. We found that 50 mM EtOH increased the RANKL mRNA expression (p < 0.05) (Fig. 1A). Strikingly, NAC completely blocked the effect of EtOH on RANKL gene expression in stromal osteoblasts (Fig. 1A). To test the ability of NAC to inhibit EtOH-triggered osteoclast differentiation, we used our previously established in vitro osteoblast/osteoclast-precursor coculture system (Chen et al., 2006). We found that EtOH promoted osteoclastogenesis (Fig. 1D), and the number of TRAPase-positive multinuclear osteoclasts in EtOH-treated samples was significantly higher compared with control (Fig. 1, B and F). These are consistent with our previous in vitro studies where significant RANKL induction was observed in osteoblasts at EtOH concentrations >25 mM (Chen et al., 2006). In concordance with the inhibitory effect of NAC on RANKL mRNA expression, NAC completely blocked EtOH-induced osteoclastogenesis (Fig. 1, C, E, and F). These effects of NAC on EtOH-induced RANKL gene expression and osteoclastogenesis were similar to our previously published effects of E2 on osteoclastogenesis. To test whether the effect of EtOH on RANKL in osteoblasts is transcriptional, we used the same stromal osteoblast cell culture system with the protein synthesis inhibitor cyclohexamide (Chx) added at a concentration of 10–7 M before adding E2 or EtOH. Cyclohexamide did not affect RANKL mRNA expression by itself, but it abolished both the stimulatory effects of EtOH and inhibitory effects of E2 on RANKL gene expression in stromal osteoblasts (Fig. 2). We obtained identical results when GAPDH was used to normalize RANKL gene expression instead of using 18S (data not shown).
E2 Is Able to Block EtOH-Induced ROS Generation in Stromal Osteoblasts. We have previously demonstrated that either endogenous or exogenous estrogen treatment in vivo can antagonize EtOH effects on bone resorption (Chen et al., 2006; Shankar et al., 2006), and we have shown that RANKL was the key molecule either induced or suppressed by alcohol or E2. However, the mechanisms underlying the inhibitory cross-talk of EtOH and estrogen signaling pathways was unclear. ROS has been shown to stimulate RANKL expression in osteoblasts (Bai et al., 2005). Therefore, we tested the hypothesis that E2 prevents EtOH-induced bone loss through inhibition of EtOH-induced ROS formation. Mature stromal osteoblasts were treated with E2 or NAC with or without EtOH. After 24 h, intracellular ROS production was measured by flow cytometry. EtOH treatment resulted in increased numbers of osteoblasts with ROS accumulation (Fig. 3D) compared with cells from control wells. The 2,7-DCF-DA fluorescence was increased 4-fold in osteoblasts from EtOH-treated wells (p < 0.05) (Fig. 3G). Pretreatment with either E2 or NAC completely blocked ROS accumulation and 2,7-DCF-DA fluorescence (Fig. 3, E and F). At the concentration we used in this study, neither E2 nor NAC alone significantly changed ROS production in stromal osteoblasts (Fig. 3, B, C, and G). These data suggest that the action of E2 to antagonize EtOH effects on osteoblasts may be due to its ability to block oxidative stress.
NOX Is the Target of Both EtOH and Estradiol in Stromal Osteoblasts. NOX-dependent production of ROS has been implicated in EtOH-induced liver injury (Thakur et al., 2006). Surprisingly, the nature of the NOX in nonphagocytic cells, including osteoblasts, is largely unknown. Therefore, we examined gene expression of four subtypes of NOX catalytic subunits in stromal osteoblasts using RT-PCR. We found that NOX1, -2 and -4, but not NOX3, were abundantly expressed in stromal osteoblasts (Fig. 4A). To determine whether EtOH, E2, or NAC regulate NOX in differentiated stromal osteoblasts, we performed real-time RT-PCR, and we looked at all three subtypes of NOX mRNA. As shown in Fig. 4, B and D, EtOH treatment up-regulated NOX1 and -4 gene expression (p < 0.05), whereas NAC treatment by itself down-regulated NOX4 mRNA expression (p < 0.05). NAC and E2 both reversed the inductive effects of EtOH on NOX4 gene expression. A similar pattern of NAC and E2 effects was observed on NOX1 gene expression (Fig. 4B), but we did not see any difference in NOX2 gene expression (Fig. 4C). These findings suggested that effects on NOX expression and activity may be the predominant mechanism whereby E2 and NAC antagonize EtOH-induced ROS generation in nonphagocytic osteoblasts.
Effects of NOX Inhibition on Suppression of EtOH-Induced RANKL mRNA Expression and Phosphorylation of ERK in Stromal Osteoblasts. If the NOX is a key molecule activated by EtOH in osteoblasts to produce ROS, blocking NOX activity should mitigate EtOH action, at least in part. To examine this possibility, we treated differentiated stromal osteoblasts with diphenylene iodonium (DPI), a specific inhibitor for the flavoprotein that is the major constituent of the NOX complex. Addition of DPI 30 min before EtOH treatment abolished the induction of RANKL mRNA expression by EtOH at a concentration of 10 nM, and its inhibitory effects were concentration-respondent (Fig. 5A). NOX has been implicated in activation of several signaling cascades, including the ERK-signaling cascade (Jackson et al., 2004). Therefore, we conducted Western immunoblot analysis in stromal osteoblasts to determine whether blocking NOX activity would eliminate the sustained phosphorylation of ERK1/2 by EtOH that we reported previously (Chen et al., 2006). DPI treatment blocked phosphorylation of ERK1/2 by EtOH at concentration of 100 nM. Surprisingly, DPI itself in the absence of EtOH did not affect either RANKL mRNA or ERK phosphorylation.
Estradiol Attenuates Sustained Activation of ERK/STAT3 in Rat Calvaria Osteoblasts. Calvaria osteoblastic cells were isolated from calvariae of 5-day-old neonatal rats. Calvaria osteoblasts were treated with E2, NAC, and the ERK-specific inhibitor PD98059 for 30 min before addition of 50 mM EtOH. Cell lysates were collected as untreated cells and 1, 6, 12, 24, and 48 h after treatment. Western blotting data shown in Fig. 6A demonstrate that EtOH treatment results in chronic phosphorylation of ERK over 48 h. Consistent with previous findings that estradiol could phosphorylate ERK rapidly in osteoblasts/osteocytes, in calvaria osteoblasts, E2 itself transiently phosphorylated ERK after 1-h treatment (Fig. 6B). However, in these cells ERK activation was back to normal after 6 h. With a combination of E2 and EtOH, sustained phosphorylation of ERK was no longer observed (Fig. 6C). NAC at concentration of 20 mM and the ERK-specific inhibitor PD98059 at a concentration of 50 μM were also able to abolish EtOH-stimulated sustained ERK phosphorylation (Fig. 6, E and G). Interestingly, ERK phosphorylation was lower than control after 6 h in cells treated with NAC alone (Fig. 6D), but we did not see the changes of ERK phosphorylation status in the cells treated with PD98059 alone (Fig. 6F) compared with control. It has been suggested that STAT3 is essential for gp130-mediated osteoclast formation and that the target of STAT3 during this process is induction of RANKL transcription (O'Brien et al., 1999). To see whether STAT3 is the downstream target for sustained ERK activation by EtOH and whether E2 can block this EtOH action, we carried out Western blots using antibodies against phospho-STAT3 and total STAT3. Strikingly, the patterns of STAT3 phosphorylation paralleled the patterns of ERK phosphorylation from all different treatments (Fig. 7). In contrast, no effects of EtOH were observed on p38 phosphorylation (data not shown). This suggests that EtOH specifically activates the ERK1/2-STAT3 pathway in osteoblasts.
Prolonging Nuclear Accumulation of Activated ERKs in Osteoblasts by EtOH. The mitogen-activated protein kinase ERK targets proteins in multiple cell compartments after an activation stimulus. The location of ERK2 is a significant factor in determining its biological functions. Therefore, we examined whether EtOH is able to trigger and prolong the nuclear accumulation of phosphorylated ERKs, and whether E2 or NAC could prevent EtOH-induced nuclear accumulation. Transient transfections were performed using both UMR-106 cells and calvaria osteoblastic cells using LTX and Plus reagent (Invitrogen, Carlsbad, CA). GFP-ERK plasmid was cotransfected with wild-type MEK1 and nRFP into both cell types. The transfection efficiency was much better in UMR-106 cells compared with calvaria osteoblasts. However, the overall results were similar that is in both UMR-106 and calvaria osteoblasts the majority of transfected GFP-ERK was in cytoplasm before any treatment (Fig. 8). After 24 h of cell treatment, we found that in EtOH treatment, more cells show nuclear accumulation of GFP-ERK compared with control-, E2-, or NAC-treated cells. Both E2 and NAC blocked the EtOH-induced GFP-ERK nuclear accumulation in both cell types (Fig. 8). Data not shown here, the GFP-ERK nuclear localization was confirmed by cotransfection of nRFP.
Discussion
Despite recent data showing that low levels of alcohol consumption may have beneficial effects on bone (Jonsson et al., 2007; Wosje and Kalkwarf, 2007), it has been known for many years that chronic alcohol abuse can cause a variety of harmful effects on hematopoiesis (Heermans, 1998; Prakash et al., 2001), with consequent bone loss in both females and males, representing a significant risk factor for the development of osteoporosis (Chakkalakal, 2005). The molecular mechanisms underlying alcohol toxicity on bone cells, including both osteoblasts and osteoclasts, are not well understood. We recently published data from in vivo and in vitro studies demonstrating that estrogens can prevent alcohol-induced bone loss (Chen et al., 2006; Shankar et al., 2006). We showed that EtOH is able to promote osteoclastogenesis in the presence of osteoclast precursors and osteoblasts due to enhanced expression of the osteoclast differentiation factor RANKL in osteoblasts. In this study, we focused on identification of the molecular targets for both alcohol and E2 and their cross-talk converging on control of RANKL gene expression.
The inhibitory effects of NAC on EtOH-mediated osteoclastogenesis in the current report were similar to those demonstrated for E2 in our previous study (Chen et al., 2006). These findings imply that mitigation of oxidative stress may play a crucial role in prevention of alcohol-induced bone loss by estrogens. We have demonstrated that EtOH treatment results in ROS generation in osteoblasts. This led us to hypothesize that free radicals generated as a result of alcohol metabolism creates oxidative stress in bone cells. Indeed, it has been shown previously that EtOH can produce significant amount of ROS in the liver in hepatocytes, Kupffer, and stellate cells and also in the lung (Purohit and Brenner, 2006). In those tissues, EtOH-induced ROS involves the enzymes CYP2E1 and NOX. Although it has been suggested that CYP2E1-associated radical production is responsible for DNA adduct formation in hepatocytes, it is becoming clear that NOX-derived free radicals play a key role in chronic alcohol-induced tissue damage (Kono et al., 2000; Thakur et al., 2006). We previously demonstrated that class 1 alcohol dehydrogenase, which is an enzyme specifically responsible for the first step in metabolizing EtOH to acetaldehyde, is highly expressed in osteoblasts, whereas CYP2E1 was not expressed (Chen et al., 2006). Moreover, we showed that acetaldehyde was also able to induce RANKL mRNA expression in osteoblasts, indicating a role for downstream alcohol metabolites in this process (Chen et al., 2006). We have shown for the first time in the current study that NOX1, -2, and -4, but not -3, are also highly expressed in stromal osteoblasts. In addition, we have demonstrated that expression of NOX4 and NOX1 catalytic subunit mRNAs are upregulated by EtOH in osteoblasts. Furthermore, induction of NOX expression by EtOH was shown to be blocked by E2 and by NAC. Previous studies of EtOH-induced NOX activation in hepatic Kupffer cells have suggested that EtOH increases GTP binding to the NOX regulatory subunit Rac-1 and enhances recruitment of Rac-1 and other regulatory subunits, such as p67phox, to the cell membrane (Thakur et al., 2006). In contrast, in the lung, chronic EtOH ingestion resulted in increased expression of the major NOX catalytic subunit (gp91phox) at the mRNA and protein level (Polikandriotis et al., 2006). The latter observation is consistent with our current data. Multiple isoforms of NOX have recently been described in bone marrow-derived hematopoietic stem/progenitor cells (Piccoli et al., 2007). It has been suggested that NOX4, which is a constitutively active isoform producing superoxide in the absence of recruitment of coactivators, may act as an oxygen sensor producing ROS, which further signal to activate other NOX isoforms, such as NOX1 via activation of Rac-1 (Piccoli et al., 2007). If this is correct, increased NOX4 expression as the result of ethanol treatment will increase ROS production in osteoblasts by itself, and it may also stimulate further activation of NOX1 and more ROS as the result of coactivator recruitment. This remains to be determined. Recent studies of NOX activation by EtOH in hepatic stellate cells suggest that formation of acetaldehyde is required for NOX activation and superoxide production (Novitskiy et al., 2006). This is consistent with our previous data showing that acetaldehyde treatment of osteoblasts can induce RANKL and osteoclastogenesis and that RANKL induction can be blocked by treatment with the ADH inhibitor 4-methylpyrazole (Chen et al., 2006). Previous in vivo studies in the liver and in in vitro cell culture have shown that estrogens are able to either inactivate NADPH oxidase activity or to inhibit its gene expression (Sumi et al., 2003; Xu et al., 2004). Confirmatory data presented in the current study are that DPI, a specific inhibitor of NOX, not only abolished alcohol-induced RANKL gene expression but also blocked alcohol-induced sustained ERK phosphorylation in osteoblasts. This provides us a molecular explanation of how EtOH and estrogens can antagonize each other in osteoblasts.
We used rat osteoblast cell lines and rat primary osteoblasts from neonatal calvariae to study the signaling cascade from activation of NOX by EtOH to RANKL gene expression. These studies revealed a significant and prolonged nuclear accumulation of ERK. Interestingly, by Western blot analysis, phosphorylation of STAT3 coincided with the pattern of ERK activation, suggesting that STAT3 activation is downstream of ERK. This pathway is supported by previous data in other cell types such as cardiomyocytes (Sauer et al., 2004). More strikingly, the ERK-specific inhibitor PD98059, but also E2 and NAC, were all able to attenuate ERK signaling and RANKL induction by EtOH. These findings strongly indicate that ERK1/2 kinase activation and downstream phosphorylation of STAT3 are critical steps for EtOH to exert its biological effects in osteoblasts. Other research on ERK signaling supports our findings. First, the duration of intracellular ERK signaling is associated with distinct biological responses (Marshall, 1995; Murphy et al., 2002). Second, the location of ERK is also a significant factor in determining its ability to phosphorylate key substrates and thereby influence cellular behavior (Whitehurst et al., 2002). A previous report in bone cells has revealed that transient versus sustained phosphorylation and nuclear accumulation of ERK in response to E2 underlie anti-versus proapoptotic effects in osteoblasts and osteoclasts (Chen et al., 2005).
In summary, we have studied molecular signaling cascades in bone cells affected by EtOH and we have identified a potential mechanism whereby estrogens and antioxidants may antagonize the effects of alcohol. EtOH-stimulated RANKL gene expression in mature osteoblasts results from ROS generation associated with induction of NOX expression, and further activation and nuclear accumulation of ERK, followed by phosphorylation of STAT3. This results in increased gene transcription of RANKL and enhanced osteoclastogenesis (Fig. 9). In the presence of E2 or the antioxidant NAC, the activated cascade of ROS/ERK/STAT3/RANKL in osteoblasts is attenuated, apparently as the result of inhibition of EtOH-mediated NOX induction. However, additional effects on post-translational activation of NOX or direct scavenging of cellular ROS cannot be ruled out. Our findings provide a potential molecular explanation for the preventive effects of estrogens on alcohol-induced bone loss. Whether the activated cascade of ROS/ERK/STAT3 by alcohol has other biological consequences, such as bone cell apoptosis and altered proliferation/differentiation, will require further study.
Footnotes
-
This study was supported by in part by National Institutes of Health Grant R01 AA12928 (to M.J.J.R.).
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.107.130351.
-
ABBREVIATIONS: EtOH, ethanol; RANKL, receptor activator of nuclear factor-κB ligand; ADH, alcohol dehydrogenase; ROS, reactive oxygen species; NOX, nicotinamide adenine dinucleotide phosphate oxidase; ERK, extracellular signal-regulated kinase; E2, 17-β-estradiol; STAT, signal transducer and activator of transcription; FBS, fetal bovine serum; TRAPase, tartrate-resistant acid phosphatase; RT-PCR, reverse transcription-polymerase chain reaction; Chx, cycloheximide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 2,7-DCF-DA, 2,7-dichlorodihydrofluorescein diacetate; NAC, N-acetyl cysteine; GFP, green fluorescent protein; MEK, mitogen-activated protein kinase kinase; nRFP, nuclear red fusion protein; ANOVA, analysis of variance; DPI, diphenylene iodonium; P-, phosphorylated; T-, total.
- Received August 17, 2007.
- Accepted October 2, 2007.
- The American Society for Pharmacology and Experimental Therapeutics