Effect of mitochondrial function on the proteasome

In addition to the effect of proteasome dysfunction on mitochondria, some reports demonstrate effects of mitochondrial dysfunction on proteasome function. Mitochondria are the powerhouse of the cell and proteasomes are a huge ATP-consuming machine. Hence, it seems plausible that mitochondrial metabolism can affect proteasome activity. Metabolic sensors such as adenosine monophosphate-activated protein kinase (AMPK) or protein kinase A (PKA) can induce posttranslational modifications of the proteasome and thereby modulate its activity (Ronnebaum et al., 2014; Zhang et al., 2007b). AMPK is the major energy sensor of the cell and is activated upon low ATP availability. AMPK activation subsequently stimulates ATP producing processes and inactivates ATP consuming processes (Hardie et al., 2012). Indeed, AMPK activation was shown to decrease proteasome activity, while in turn AMPK inhibition increased proteasomal activity (Ronnebaum et al., 2014; Viana et al., 2008; Xu et al., 2012). Two potential mechanisms how AMPK influences proteasomal activity have been described. On the one hand, AMPK activation induces O-GlcNAc transferase mediated O-GlcNAcylation of the proteasome promoting 20S-19S disassembly (Xu et al., 2012). On the other hand, AMPK interacts with Rpn6 and was shown to directly phosphorylate this subunit (Moreno et al., 2009). However, how this modification affects proteasome activity is still unknown. In contrast to AMPK-induced inactivation, PKA induces phosphorylation of the 19S proteasome subunits Rpt6 and Rpn6 thereby increasing proteasome activity (Lokireddy et al., 2015;

Zhang et al., 2007a). However, although inhibition of mitochondria can for example induce AMPK activation (Distelmaier et al., 2014), a direct connection between mitochondrial and proteasomal function via AMPK or PKA has not yet been described.

In addition to metabolic control of the proteasome, it was reported that an increased need for mitochondrial quality control by the proteasome can lead to proteasomal activation. Induction of the unfolded protein response of the mitochondrial intermembrane space by overexpression of an unstable IMS protein induced an increase in proteasome activity (Papa and Germain, 2011). Likewise, a recent report showed that a defect in the mitochondrial protein import machinery and therefore accumulation of mitochondrial proteins that cannot be correctly imported, induced upregulation of proteasome activity (Wrobel et al., 2015). These results further underline the importance of the proteasome in mitochondrial quality control and suggest an adaptive regulation of proteasomal activity by mitochondrial dysfunction.

The direct effect of mitochondrial dysfunction on proteasome activity was analyzed by some other studies by testing if inhibition of the mitochondrial respiratory chain could induce alterations in proteasome function. Indeed, treatment of HEK cells with rotenone (complex I inhibitor), antimycin A

ubiquitin fused GFP (Chou et al., 2010). Similarly, in NT2 human carcinoma cells, the complex I inhibitors MPP+ or rotenone diminished proteasome activity measured by degradation of small peptidic proteasome model substrates (Domingues et al., 2008). These effects were attributed to an increase in mitochondrial ROS production upon respiratory chain inhibitor treatment (Chou et al., 2010; Domingues et al., 2008). Additionally, a recent siRNA screen in C. elegans expressing ubiquitinated GFP as a proteasome reporter showed that knockdown of respiratory chain subunits or enzymes involved in mitochondrial metabolic processes decreases proteasome activity. This effect was attenuated by treatment with antioxidants such as N-acetyl cysteine (NAC) (Segref et al., 2014) showing its dependency on cellular ROS. A more detailed description of proteasomal alterations in response to respiratory chain inhibitors was provided by Livnat-Levanon et al. (2014). Treatment of yeast cells or hamster kidney cells with antimycin A led to disassembly of 26S proteasomes thereby decreasing proteasome activity and inducing accumulation of ubiquitinated proteasome substrates.

Again, this effect could be partially reverted with DTT showing a dependency of oxidative modifications induced upon respiratory chain inhibition (Livnat-Levanon et al., 2014). This ROS-dependent decrease of proteasomal activity in response to respiratory chain inhibition is well in accordance with the already known influence of ROS on the proteasome system (Aiken et al., 2011;

Breusing and Grune, 2008; Wang et al., 2010). However, one caveat remains as it is unclear to which extent this effect is physiologically relevant. It was shown that under unstressed or mildly stressed conditions mitochondrial ROS acts as a signaling molecule in cells and is required for cellular homeostasis (Sena and Chandel, 2012). Pharmacologically inhibiting the respiratory chain, however, strongly and acutely increases mitochondrial ROS production to a level which is probably highly above normal ROS production in vivo (Murphy, 2009). Furthermore, much of the effect of mitochondrial ROS in vivo might be dependent on the balance between ROS and cellular antioxidant defenses (Sena and Chandel, 2012) and hence might not be well reflected by an acute outburst of ROS production.

As the dependency of 26S proteasomes on ATP is well documented, decreased ATP levels in response to mitochondrial damage were suggested as another mechanism contributing to diminished proteasome activity. Indeed, Huang et al. showed that treatment of primary rat cortical neurons with antimycin A, rotenone or the complex V inhibitor oligomycin induces a decrease in 26S proteasome assembly and subsequent accumulation of ubiquitinated substrates. This correlated with decreased ATP levels in treated cells (Huang et al., 2013). However, as the treatment conditions in this study were associated with induction of necrosis and substantial loss in cell viability (Huang et al., 2013) it is not clear whether loss of proteasome function is causally linked to reduced ATP levels or rather to the induction of cell death or other confounding factors. Another study in primary rat neurons showed that rotenone or MPP+-induced decrease in proteasome activity could be reversed by

Introduction

increased glucose supplementation to the cell culture medium which also partially reversed the decrease of ATP in response to inhibitor treatment. Importantly, in contrast to glucose supplementation, the antioxidant NAC was not able to rescue proteasome activity in this study (Höglinger et al., 2003). However, as ATP levels are usually tightly regulated in the cell, treatment with respiratory chain inhibitors induces an acute decrease in ATP production which probably might not reflect the physiological regulation.

In summary, the effect of mitochondrial damage and dysfunction on the proteasome is differentially regulated depending on the nature of mitochondrial damage (Figure 1-5). Mitochondrial protein misfolding in the IMS or defective mitochondrial protein import is associated with increased proteasome activity. On the other hand inhibition of the respiratory chain impairs proteasomal function, which most probably due to highly increased mitochondrial ROS production or decreased ATP availability in the cell. However, the studies are sometimes contradictory, e.g. on whether antioxidants are protective or not. This might be due to differences in cell types or even organisms used in the studies or also due to different concentrations of the applied compounds. Additionally, while these results are strongly dependent on the acute chemical inhibition of the respiratory chain, the effect of more physiological triggers of respiratory chain dysfunction as it might occur in disease or in aging are not well characterized. Therefore, these results give important insights in potential mechanisms which can link mitochondrial dysfunction with proteasome function but they are not sufficient to fully explain the connection between these two systems, e.g. in situations without elevated ROS production.

Figure 1-5: Effect of mitochondrial dysfunction on proteasome activity

Mitochondrial dysfunction has diverse effects on proteasome function. Mitochondrial protein misfolding in the mitochondrial intermembrane space or dysfunctional mitochondrial protein import induces an increase in proteasome activity. On the other hand, inhibition of the respiratory chain is associated with decreased proteasome activity and lower levels of assembled 26S and 30S proteasomes. Treatment of cells with respiratory chain inhibitors is signaled to the proteasome by acutely elevated mitochondrial ROS production or diminished ATP availability.

Altogether, proteasomal and mitochondrial function are clearly interconnected as evident by the role of the proteasome system in mitochondrial quality control and also by the observed regulation of proteasomal activity in response to mitochondrial dysfunction. An additional important link is given by a close correlation of mitochondrial and proteasomal function during aging, as both mitochondrial and proteasomal function were described to decline during aging (López-Otín et al., 2013) and in some age-related diseases such as Alzheimer’s disease (Ross et al., 2015).

Introduction