The overall goal of the investigation was to examine the role of AMPK in regulating chondrocyte autophagy. We showed for the first time that chondrocytes expressed the AMPKa1 isoform and that activation of the kinase promoted autophagy. This new observation is in line with a very recent study of autophagy where it was shown that AMPK is a potent activator of the autophagic process (18
). Furthermore, activation of AMPK was dependent on the transcription factor HIF-1, a stimulator of autophagic flux. The study also indicated that induction of autophagy was regulated by mTOR, a second sensor of cellular metabolism (13
). Accordingly, an increase in AMPK, probably related to a change in the chondrocyte energy charge, would inhibit mTOR and promote autophagy; this finding is consistent with the observation that when mTOR activity is blocked, autophagy is induced (20
). A second observation was that when AMPK was silenced and autophagy was suppressed, there was a marked decrease in apoptogen sensitivity, probably mediated by increased Akt-1 signaling. Based on these findings, we suggest that activation of AMPK promotes autophagy as a mechanism of survival. Related to the increase in the rate of the autophagic flux there is enhanced sensitization to local apoptogens, an event that would serve to promote apoptosis and the subsequent deletion of chondrocytes from articular cartilage.
The expression profile of AMPK in the growth plate was examined by immunohistochemistry. We showed that epiphyseal chondrocytes express AMPKa1; AMPKa2 was not detected. We also determined the expression of AMPK in N1511 chondrocytes, a cell line that undergoes rapid maturational changes in culture (23
). The cultured chondrocytes expressed AMPKa1; AMPKa2 was not evident in maturing cells. While AMPKa1 positive cells were seen throughout the growth plate, the most densely stained cells were localized to the pre-hypertrophic and hypertrophic regions of the epiphysis. Remarkably, the location of the maximally positive chondrocytes was similar to that described for both Beclin-1 and LC-3, two major indicators of autophagic vacuole formation (3
). We have also shown that the high HIF-1a immunoreactivity in the maturing cell zone was linked indirectly to the expression of Beclin-1 and correlated well with other determinants of chondrocyte energy status (the adenylate charge ratio, the ATP/ADP ratio and the redox ratio (1
). Since these latter measures indicate that the two regions were under metabolic stress (1
), it would not be unreasonable to conclude that regional HIF-dependent changes in energy metabolism are related to the rise in AMPK activity and the induction of autophagy. AMPK-dependent induction of autophagy would serve to sustain the viability of the cells during the final stages of the terminal differentiation pathway.
We investigated the relationship between AMPK and autophagy using gene-silencing technology and by activating the calcium/ER stress response. Based on earlier observations that thapsigargin-mediated release of intracellular calcium stores triggers autophagy (21
), we treated gene-silenced chondrocytes with low levels of the drug. Our results demonstrated that thapsigargin-mediated activation of AMPK was dependent on HIF-1, a previously described activator of autophagic flux (3
). Thus, when HIF-1 was silenced, AMPK activation was blocked; in contrast, there robust AMPK activation when control chondrocytes were treated with thapsigargin. We next examined this relationship from an autophagic perspective: the distribution of the microtubule-associated protein, LC3, was evaluated after thapsigargin treatment. Drug treatment caused redistribution of LC3 into distinct puncta, indicative of an induction of chondrocyte autophagy. The observation that thapsigargin treatment promoted the expression of LC3II confirmed the finding that the cells were autophagic. These results suggest that AMPK activation by thapsigargin is dependent on HIF-1, and, even more importantly, lends considerable strength to the notion that these two factors regulate the induction of thapsigargin-mediated autophagy.
To test the hypothesis that AMPK was directly linked to the onset of autophagy, we assessed the expression of two key molecules, the autophagy associated BH3 domain only protein, Beclin-1 and the pro-survival protein, Bcl-2. In this study, AMPK silenced cells were treated with thapsigargin. Immunoprecipitation followed by Western blot analysis showed that drug treatment resulted in dissociation between Beclin-1 and Bcl-2; however, when AMPK was silenced, following immunoprecipitation, the proteins remained associated even after thapsigargin treatment. Based upon these findings it is concluded that thapsigargin-mediated induction of autophagy is dependent on the activity of AMPK.
Since autophagy can serve to delay the onset of apoptosis, it raised the following question: is there a direct relationship between the induction of autophagy and apoptosis? To test this notion, we treated the AMPK silenced chondrocytes with a powerful apoptogen, H2O2. We noted that the silenced cells were almost completely refractory to the apoptogen challenge. This result was surprising, as AMPK is an important sensor of the cellular energy status and a loss of activity would be expected to compromise the viability of cells, especially when challenged with peroxide. One explanation for the observed resistance is that AMPK regulates the activity of mTOR, a protein that is responsive to the Akt survival pathway. Thus, the possibility exists that by activating a signaling pathway that promotes cell survival, AMPK suppression promotes apoptogen resistance. Indeed, we found that AMPK suppression led to a significant increase in pAkt expression. When mTOR was silenced, there was extensive reorganization of LC3, indicating that the kinase was regulating development of the autophagic state; AMPK inhibition did not block this activity. Based on this finding, it is concluded that the mechanism by which autophagy is controlled in chondrocytes is through mTOR, which is sensitive to, and regulated by, AMPK. Furthermore, caspase-8 activity and BID cleavage were reduced in the AMPK silenced cells. These findings suggested that while AMPK was required for the induction of autophagy, its suppression results in the stimulation of survival signaling and suppression of apoptosis. Accordingly, in a low energy state while AMPK is activated, inhibition of mTOR leads to decreased Akt survival signaling. Low survival pathway activity and promotion of caspase-8 and BID activities would lead to loss of viability and increased sensitivity to local apoptogens.
In summary, we show that in chondrocytes, AMPK, an enzyme that monitors the energy status of the cell, is activated in a HIF-1 dependent manner. This enzyme responds to the energy status and intracellular calcium fluxes; inhibition of mTOR promotes development of the autophagic state. We also provide evidence that suppression of AMPK activates mTOR and survival signaling mediated by the Akt pathway. The relationship between HIF-1, AMPK, mTOR and autophagy is shown schematically in . The importance of these, key regulators of cell function have hitherto escaped serious study and their detailed role in chondrocyte biology has yet to be determined. However, modulation in the autophagic flux has recently been correlated with the hypertrophic and osteoarthritic phenotype (5
). Furthermore, the autophagic flux in the growth plate has not been studied from the perspective of growth retardation as a result of chronic renal disease. Detailed studies of the relationship among HIF-1, chondrocyte hypertrophy, chronic renal disease and autophagy, and their regulation at the AMPK-mTOR axis is currently under intense investigation.
Schematic of the interrelationships among AMPK, HIF-1, mTOR and autophagy