Previous studies show that caffeine induces various cell responses, including cell death [28
]. However, the effects of caffeine on osteoblasts and osteoporosis are currently unclear. The potential cytotoxicity of caffeine was examined by determining the viability of human osteoblasts treated with various doses of the compound, using the MTT assay. Osteoblasts were incubated in medium containing 0–2 mM caffeine for 24 h. The viability of treated osteoblasts was decreased by approximately 10–35% at concentrations higher than 0.5 mM caffeine in a dose-dependent manner (). We further investigated whether caffeine-induced cell death represents apoptosis or necrosis. The percentage of apoptotic cells increased significantly in cultures exposed to >0.5 mM caffeine, and the necrotic cell population simultaneously increased at higher concentrations ( and ). These results indicate that treatment with caffeine triggers two cell death modes in osteoblasts, primarily apoptosis, and to a smaller extent, necrosis ( and ). In addition, the DNA content of various cell cycle phases was determined by flow cytometry analysis of propidium iodide-labeled cells (). The decline in osteoblast survival ratio following treatment with caffeine was attributed to the simultaneous occurrence of G1 arrest, apoptosis and necrosis ( and ).
Figure 1. Effects of caffeine on osteoblasts. Osteoblasts were incubated with various concentrations of caffeine for 24 h. (A) Cell viability was determined using the MTT assay. (B) The percentages of apoptosis and necrosis were determined by propidium iodide and (more ...)
There is no documented evidence to show that caffeine directly provokes oxidative stress in cells. However, several reports demonstrate that ROS are effective cell injury inducers, leading to apoptosis and necrosis [29
]. Therefore, we examined whether ROS formation occurs in caffeine-treated osteoblasts by immunostaining analysis with DCF-DA as the detection reagent. As shown in , treatment with 0–2 mM caffeine for 24 h enhanced the intracellular ROS content in osteoblasts. ROS generation was additionally measured with DCF-DA and DHR-123 fluorescence dyes using the fluorescence ELISA reader (). In addition, ROS generation can detect when cells treatment with caffeine for more than 1 h (). To our knowledge, this is the first study to show that caffeine directly induces ROS generation in osteoblasts.
Figure 2. Caffeine induces ROS generation in osteoblasts. Osteoblasts were incubated with 20 μM DCF-DA or dihydrorhodamine 123 (DHR 123) for 1 h, and treated with various concentrations of caffeine for another 2 h. (A) Cells were observed using a fluorescence (more ...)
The protein expression ratio of Bax versus Bcl-2 is relevant to apoptosis. Specifically, a high Bax/Bcl-2 ratio is associated with a lower threshold of apoptosis, while a low ratio represents a higher apoptotic threshold [30
]. Here, we investigate whether caffeine induces apoptosis by modulating the Bax/Bcl-2 ratio, the major effectors of mitochondria-mediated apoptosis. Immunoblotting revealed that treatment of osteoblasts with more than 0.5 mM caffeine triggered an increase in Bax and decrease in Bcl-2 protein levels (). Densitometric analysis quantitatively revealed that caffeine-treated osteoblasts have a higher Bax/Bcl-2 ratio, favoring apoptosis ().
Figure 3. Caffeine induces an increase in the ratio of Bax/Bcl-2 protein level and loss of MMP in osteoblasts. Osteoblasts were incubated with various concentrations of caffeine for 24 h. (A) Extracts (40 μg) were analyzed by immunoblotting with anti-Bcl-2 (more ...)
Bax and Bcl-2 regulate changes in the mitochondrial membrane potential (MMP) and permeability, which play important roles in apoptotic processes [32
]. As mitochondrial membrane potential change is directly associated with apoptosis [33
], we further investigated the effects of caffeine on this parameter. Treatment of osteoblasts with caffeine decreased DiOC6(3) and TMRE uptake into mitochondria, indicating that caffeine induces significant loss of mitochondrial membrane potential (). In addition, we also found that loss of MMP was detectable when cells treatment with caffeine for more than 6 h (). Our findings support the theory that caffeine triggers apoptosis via a mitochondria-dependent pathway.
Caspases play important roles in apoptosis, and their activation is critical in biochemical changes for cell death analysis [9
]. For identifying the signaling pathways involved in caffeine-induced cell death, we used an immunoblotting assay to monitor the levels of caspases-9 and -3, which are activated during the apoptosis of multiple cell types triggered by various stimuli [36
]. Treatment of osteoblasts with caffeine stimulated caspase-9 and -3 activities (–). The activation of caspase-9 and caspase-3 were detectable when cells treatment with caffeine for more than 9 h ( and ). Poly(ADP-ribose)polymerase (PARP) is a key participant in DNA base excision repair. Apoptosis is accompanied by early and unique cleavage of PARP. The formation of a 89 kDa cleavage fragment resulting from the specific PARP cleavage pattern is a hallmark of apoptosis. Here, we show that PARP, a specific caspase-3 substrate, is cleaved in caffeine-treated osteoblasts (). Accordingly, we propose that caffeine-triggered apoptosis occurs via caspase-9 and caspase-3 activation.
Figure 4. Activation of caspase-9 and caspase-3 in caffeine-treated osteoblasts. Osteoblasts were incubated with various concentrations of caffeine for 24 h. Cell extracts (40 μg) were immunoblotted using anti-caspase-9 (A), anti-caspase-3 (C), or anti-PARP (more ...)
To further clarify the role of ROS in caffeine-induced apoptosis, we assessed the effects of two commonly used ROS scavengers, N-acetyl cysteine (NAC) and α-tocopherol, on caffeine-treated osteoblasts. Pretreatment of cells with α-tocopherol (300 μM) or NAC (500 μM) attenuated caffeine-induced intracellular ROS generation, loss of MMP, and caspase-3 activation (, and ). Based on the results, we propose that caffeine triggers ROS generation, which in turn activates mitochondria-dependent apoptotic processes in osteoblasts.
Figure 5. Effects of ROS scavengers on caffeine-treated osteoblasts. Osteoblasts were treated with or without the indicated concentrations of α-tocopherol (Toc; 300 μM) or N-acetyl cysteine (NAC; 500 μM) for 30 min, followed by incubation (more ...)
Previous studies show that p21-activated protein kinase 2 (PAK2) is activated during cell death, and may be involved in apoptotic signaling events, possibly via caspase-3-directed proteolysis [18
]. We investigated whether PAK2 cleavage/activation occurs in caffeine-treated osteoblasts. Immunoblotting and immunoprecipitation kinase activity assays revealed that caffeine induces PAK2 cleavage () and activation (), respectively. Caspase-3 is activated during apoptosis of several cell types triggered by a variety of stimuli, including anti-Fas antibodies, TNF-α, and the chemotherapeutic agent, etoposide [36
]. To elucidate the relationship between caspase-3 activation and PAK2 during caffeine-induced apoptosis, we examined the effects of the specific tetrapeptidic caspase inhibitors, Ac-DEVD-cho and Ac-YVAD-cmk [18
], on caffeine-treated osteoblasts. Caffeine-induced activation of caspase-3 was markedly suppressed upon pretreatment with either of the inhibitors (). Pretreatment with caspase inhibitors additionally blocked caffeine-induced PAK2 activity (), indicating that caspase-3 functions upstream of PAK2 in the apoptotic pathway.
Figure 6. Caffeine induces the cleavage/activation of PAK2. Osteoblasts were incubated with various concentrations of caffeine for 24 h. (A) Cell extracts (60 μg) were immunoblotted with anti-αPAK (C19) antibody. (B) PAK2 was immunoprecipitated, (more ...)
Activation of JNK is essential for apoptotic induction in some cell types [13
], and in earlier investigations, we demonstrate that UV irradiation-, methylglyoxal-, and high glucose-induced apoptosis are mediated by JNK [38
]. Here, we use immunoblotting and ELISA assays to examine JNK activation during caffeine-induced apoptosis. JNK activity was stimulated in osteoblasts treated with caffeine (). To further determine the relationship between JNK and PAK2 and the functional role of PAK2 in caffeine-induced apoptosis, we incubated osteoblasts with antisense or sense oligonucleotides against PAK2 for 3 days, subjected cells to caffeine treatment, and analyzed cell extracts by immunoblotting with anti-PAK2 (N17) antibodies and immunoprecipitation. Pre-incubation of osteoblasts with an antisense oligonucleotide against PAK2 led to a significant decrease in PAK2 protein levels (by ~40%), compared to untreated controls, whereas the sense oligonucleotide had no such effect (). Similarly, cells treated with the antisense PAK2 oligonucleotide displayed PAK2 activation levels that were approximately 40% those of control cells (). These reductions in PAK2 protein expression and activation were associated with a significant decrease in caffeine-induced activation of JNK and apoptosis ( and ). Furthermore, activation of JNK was markedly reduced in the presence of the caspase-3 inhibitors (). However, caspase inhibitors had no effects on ROS generation in caffeine-treated osteoblasts (data not shown). Our findings support the theory that ROS generation is an upstream regulator of sequential apoptotic biochemical changes, such as caspase activation. The data strongly suggest that PAK2 activation plays an important role in caffeine-induced JNK activation and apoptosis of osteoblasts.
Figure 7. Active PAK2 is required for JNK activation in caffeine-induced apoptosis. (A) Osteoblasts were incubated with various concentrations of caffeine for 24 h. Cell extracts (60 μg) were prepared and immunoblotted with the anti-p-JNK antibody. JNK/AP-1 (more ...)
Analysis of the results in reveals that the decreased ratios of PAK2 (~57%) and JNK activity (~54%) are higher than that of apoptosis inhibition (~36% decrease) in caffeine-treated cells. This finding indicates that other regulatory mechanisms are additionally involved in caffeine-induced apoptosis in osteoblasts.
Survival signaling processes protect against apoptosis induced by specific stimuli [41
]. Our group and others have demonstrated that some apoptotic stimuli inhibit the Ras→ERK survival signal pathway by decreasing the protein levels of various survival components [31
]. Based on these results, we evaluated the effects of caffeine on the cell survival signal pathway, including activities of components critical for the Ras→ERK-dependent and PI3 Kinase→Akt survival signal pathways. Our results show that caffeine triggers a decrease in the protein levels of HSP90, Ras, Raf-1, and Akt, and inhibition of ERK and Akt activities (). In addition, inhibition of protein expression or activities of these survival signal components by caffeine could be effectively blocked by pretreatment with lactacystin, a specific proteasome inhibitor, and NAC, a potent ROS scavenger (). Moreover, lactacystin and NAC prevented caffeine-induced cell apoptosis, which correlates the blocking of survival signal components decreases ( and ). These data imply that Ras, Raf-1, and Akt are degraded by ROS and proteasome-dependent pathways in caffeine-treated osteoblasts, subsequently reducing ERK and Akt activities for cell apoptosis.
Figure 8. Effects of caffeine on HSP90 and components of survival signaling. Osteoblasts were incubated with or without lactacystin (LC; 10 μM) or NAC (500 μM) for 1 h, and treated with caffeine (2 mM) for another 24 h. (A) The protein levels of (more ...)
Lastly, we analyzed whether the apoptotic effect of caffeine on osteoblasts induces bone density injury in vivo. Wistar rats were fed a standard diet supplemented with or without caffeine (10 or 20 μM) in drinking water for 8 months. Initial body weights were similar among the control and caffeine-treated groups, and did not differ significantly at the end of the 8-month experimental period (data not shown). However, the bone mineral densities (BMD) of whole femurs and distal femurs from rats in the caffeine-treated groups were significantly lower than those in the control (caffeine-free) group (). These results suggest that elevated caffeine levels, resulting from continuous long-term consumption of high caffeine-containing foods, negatively affect BMD, an important osteoporosis indicator.
Figure 9. Effects of caffeine intake on BMD of rats. DXA was used to assess BMD in whole femurs and distal femurs of rats fed a controlled diet with or without caffeine (10 or 20 μM) in the drinking water for 8 months. Values are presented as means ± (more ...)
Substantial amounts of caffeine are ingested by people drinking coffee, tea or caffeinated soft drinks. While the consumption of caffeine-containing beverages is associated with significantly increased osteoporosis risk, its precise role and regulatory mechanisms in bone loss and fracture risk are currently unclear [44
]. Our current results provide evidence that the apoptotic and necrotic properties of caffeine contribute to bone loss and osteoporosis in osteoblasts. The majority of caffeine-treated cells die via apoptotic processes through ROS generation and a mitochondria-mediated pathway, while a minor proportion of cell death occurs through necrosis (–).
Oxidative stress is a stimulator of cell responses such as apoptosis [29
]. Importantly, caffeine directly induced oxidative stress in osteoblasts in our study (). In addition, pretreatment with antioxidants effectively prevented ROS-induced apoptotic biochemical changes (). Several reports show that ROS is an important upstream regulator of JNK and caspase activation during UV-, methylglyoxal-, and osmotic shock-induced apoptosis in various mammalian cells [38
]. These results, coupled with our findings, strongly imply that oxidative injury plays a pivotal role in caffeine-induced apoptosis. However, the underlying mechanisms of caffeine-triggered ROS formation remain to be established. The effects of caffeine on target cells may depend, at least in part, on cell type and treatment protocol (i.e., treatment period and dosage).
Mitochondria act as important conduits for signals during programmed cell death, and loss of mitochondrial integrity can be promoted or inhibited by several key regulators of apoptosis [47
]. For instance, various cellular stress conditions, including heat shock, DNA damage and oxidative stress, result in caspase activation through cytochrome c release from the mitochondrial intermembrane space into the cytoplasm [48
]. We examined Bax/Bcl-2 protein expression and MMP changes, with a view to further elucidating the precise mechanisms of caffeine-induced apoptosis. Caffeine triggered an increase in Bax/Bcl-2 protein expression and loss of MMP in a dose-dependent manner (–). Moreover, caspases-9 and -3 were activated in the presence of caffeine (). Interestingly, caffeine induced ROS generation and activation of caspase-9 and caspase-3 at concentrations higher than 0.5 mM, but had no effects on MMP changes or the Bax/Bcl-2 ratio at 0.5 mM (, and ). These ambiguous results may be affected by the different assay sensitivities of various methods or assay kits. To elucidate the underlying reasons for these conflicting results, we further analyzed Bax and Bcl-2 mRNA levels by real-time PCR. Our analyses revealed that caffeine treatment of osteoblasts led to increased and decreased Bax and Bcl-2 expression levels at concentrations higher than 0.5 mM, respectively (data not shown). Thus, apoptotic biochemical changes, such as ROS generation, MMP changes, and activation of caspases appear to be correlated. In addition, inhibition of ROS generation by antioxidants was associated with a decrease in caffeine-triggered mitochondria-dependent apoptotic changes (). Accordingly, we propose that ROS act as important regulators of mitochondria-dependent apoptotic processes promoted by caffeine.
Recently, coworkers of our laboratory and other researchers showed that PAK2 is a target substrate for caspases activated by various apoptotic stimuli [17
]. However, the functional role of the caspase-generated C-terminal active fragment of PAK2 remains obscure. Our group showed that decreases in PAK2 protein expression and activation are associated with significant inhibition of methylglyoxal-induced apoptosis in human osteoblasts [30
], strongly suggesting that PAK2 plays an important role in apoptosis by methylglyoxal. Here, we demonstrate the PAK2 is an important upstream regulator of JNK activation in caffeine-directed apoptosis of osteoblasts ( and ). JNK evidently plays critical roles in apoptosis [14
]. Moreover, inhibition of caspase-3 blocks activation of PAK2 and JNK (, , and ). Based on these results, we propose that caffeine triggers caspase-3 activation, which in turn induces cleavage/activation of PAK2 and sequential activation of JNK, consequently leading to apoptosis.
Heat-shock proteins (HSP) protect proteins against proteasome- and ubiquitin-dependent degradation [51
]. HSP90, the most abundant molecular chaperone protein in the intracellular system, is involved in maintaining the correct conformation of intracellular proteins and kinases, such as Raf-1 and Akt [53
], which regulate cell proliferation and survival. Here, we show that caffeine-induced apoptosis is associated with reduced expression of survival components, including Ras, Raf-1, and Akt. Moreover, inactivation of ERK-1, ERK-2 and Akt/PKB, and pretreatment with lactacystin, a proteasome inhibitor, may prevent these decreases in protein level or activity (). Caffeine suppresses HSP90 expression, thus promoting the degradation of client proteins (). Accordingly, we hypothesize that the caffeine-induced reduction of HSP90 stimulates Ras, Raf-1 and Akt targeting for degradation, leading to their downregulation and changes in the related signal pathways. In view of these findings, the following pathway is proposed: caffeine decreases HSP90 expression, leading to increased degradation of Ras, Raf-1 and Akt/PKB, with decreased Raf-1 levels resulting in ERK-1 and ERK-2 downregulation and subsequent cell apoptosis. ROS generation is additionally involved in caffeine-induced suppression of the levels of survival signal proteins ( and ). Thus, it appears that ROS act as upstream negative regulators of Ras→ERK and PI3K→Akt/PKB survival signals in caffeine-induced apoptosis. However, the precise regulatory mechanisms require further investigation.
Lastly, given the ability of caffeine to induce apoptosis or cell injury in osteoblasts in vitro, we examined its effects on bone density in an animal assay model. Rats administered 10 or 20 μM caffeine in the drinking water for 8 months displayed significantly decreased bone mineral density (BMD), compared to untreated control rats, further signifying that caffeine consumption negatively affects bone density ().