The present study demonstrates that the knockout of Drp1 enhances cellular resilience against ferroptotic cell death by attenuating ferroptosis-mediated accumulation of mitochondrial ROS production and through rescue of mitochondrial membrane integrity. Interestingly, Drp1-deletion also prevented excessive mitochondrial iron uptake caused by the ferroptosis inducers. Further, in one Drp1 KO cell line, the resilience against ferroptosis was further enhanced through metabolic effects, i.e. reduced mitochondrial respiration and a compensation of ATP production through glycolysis.
We previously reported that Drp1 inhibition preserved cellular integrity in conditions of glutamate toxicity by preventing mitochondrial fragmentation and mitochondrial membrane depolarization [16]. The present study significantly extends our previous findings applying an erastin- and RSL3-mediated ferroptosis model in Drp1 knockout cell lines and now highlights the role of mitochondrial metabolism and iron homeostasis in these model systems of ferroptosis.
It is well established that inactivation of Drp1 function leads to the elongation of mitochondria [31, 32], whereas oxidative stress causes excessive mitochondrial fragmentation [25, 33]. Our results are in line with these earlier findings and, in addition, demonstrate that Drp1 deficiency inhibits mitochondrial fission caused by erastin, but not by RSL3 (Fig. 1D, E). However, mitochondria remain significantly elongated in the Drp1 KO cell lines also during RSL3 treatment when compared to the HT22 Drp1 expressing cell lines (Fig. 1D, E).
Notably, we observed that BAX protein localization at mitochondria was attenuated in the Drp1 KO #1 cell line (Fig. 3C, D). Previous studies have highlighted the role of the Bcl-2 protein family members BAX and BID in regulated cell death and, in particular, their interaction with Drp1 at the mitochondrial membrane [29, 34]. For example, it has been reported that Drp1 and BAX accumulate at mitochondria forming the so called hemi-fission intermediates thereby facilitating mitochondrial outer membrane permeabilization (MOMP) [34–36]. These intermediate membrane pores lead to the release of cytochrome c, which is an established mechanism in mitochondrial apoptosis pathways [37, 38]. The reduced BAX expression levels in the Drp1 KO cell lines suggests that depletion of Drp1 also attenuates BAX-induced MOMP. Further, the mitochondrial elongation in the Drp1 KO cells may reduce the formation of mitochondrial hemi-fission intermediates and therefore prevent an accumulation of pro-apoptotic Bcl-2 proteins in the outer mitochondrial membrane. In line with the observations of the current study, our previous studies have shown that BID, another member of the Bcl-2 family, can mediate mitochondrial fission and ultimately lead to mitochondrial demise in conditions of oxidative cell death [13, 16, 39, 40]. Remarkably, mitochondrial BID-translocation and associated mitochondrial membrane permeabilization was also abrogated by pharmacological inhibition of Drp1 [16].
Based on the data obtained in this study in the models of ferroptosis, Drp1 induces mitochondrial fission within 6 hours of ferroptosis-induction (Fig. 1D, E), and concomitantly, iron is taken up into the mitochondria (Fig. 2A, B). Previous studies have found that iron overload causes excessive mitochondrial fission, which is then mediated by Drp1 [41, 42]. How Drp1 deficiency and the associated mitochondrial elongation alters mitochondrial iron homeostasis, however, is still elusive. Our results show that Drp1 deficiency abrogated mitochondrial iron uptake that was caused by the ferroptosis inducers (Fig. 2A, B). Notably, erastin or RSL3-mediated mitochondrial iron uptake occurred between 2 to 6 hours of treatment, preceding ferroptosis mediated decline of mitochondrial integrity in the neuronal HT22 cells. This time course of ferroptosis-induced mitochondrial fission is in line with findings in our previous studies [14, 25], suggesting that increased mitochondrial iron uptake and concomitant mitochondrial fission are upstream events, initiating the devastating damage of the mitochondria during ferroptosis. This is supported by data from the present study showing that deferoxamine, which is established to chelate iron and protect against ferroptosis, inhibited erastin- or RSL3-mediated mitochondrial iron uptake (Fig. 1S C), impairment of mitochondrial integrity and ultimately ferroptotic cell death (Fig. 6C, E).
Further, Drp1 deficiency led to increased cytosolic iron levels between 2 to 6 hours of erastin and RSL3 treatment, which was not seen in the HT22 controls (Fig. 1S A, B). Several studies have shown the interconnection of mitochondria on iron metabolism through iron-sulfur clusters [43–45] that serve as binary switch on Iron Regulating Proteins (IRPs) [44]. IRPs and their binding to the Iron Responsive Element (IRE) tightly regulate protein expression of either iron storage proteins like ferritin or iron uptake proteins like the transferrin-receptor [44, 46, 47]. To which extent Drp1 deficiency or mitochondrial elongation influences iron-sulfur cluster assembly and the regulation of IRPs, is not clarified. Our data show increased cytosolic iron uptake up to 6 hours within induction of ferroptosis in the Drp1 KO cell lines (Fig. 1S A, B). After 8 hours however, cytosolic iron levels decline in erastin and RSL3 treated cells. These results suggest that Drp1 depletion might lead to the expression of iron uptake proteins, which was triggered by erastin and RSL3 treatment.
On the other hand, it is well established that excessive labile iron within mitochondria contributes to mitochondrial ROS production [48, 49], leading to devastating damage of mitochondrial function and integrity. Recent studies confirmed that mitochondrial lipid peroxidation was critical in ferroptosis by using a fluorescent mitochondria-targeted dye that is sensitive to lipid peroxidation [50]. Our results show that Drp1 deficiency prevented ferroptosis-mediated accumulation of mitochondrial ROS and mitochondrial lipid peroxidation (Fig. 2C-F). Furthermore, this protection from excessive mitochondrial ROS formation also preserved mitochondrial integrity in the Drp1 knockout cell lines as indicated by the measurements of the mitochondrial membrane potential.
Strikingly, our data show that mitochondrial and cellular integrity was preserved in the Drp1 knockout cell lines despite the depletion of GPX4 (Fig. 4C) and the accumulation of overall lipid peroxides (Fig. 4A, B). These findings suggest that preventing mitochondrial lipid peroxidation (Fig. 2E, F) and preserving mitochondrial integrity is far more important for cellular protection against ferroptosis than interfering with early features of ferroptosis such as GSH depletion, GPX4 inhibition or moderate lipid peroxidation. In fact, our data from the present study are in line with earlier findings [14, 16, 49] and impose Drp1 mediated mitochondrial damage as a key event in ferroptosis that amplifies the initiated oxidative stress and marks the “point of no return” in this form of oxidative cell death [25, 39, 51].
In addition to Drp1-dependent mitochondrial fission, formation of mitochondrial hemi-fusion sites as membrane binding spots of pro-apoptotic BAX and BID, the attenuation of mitochondrial respiration, may a contribute into cellular resilience against oxidative stress through the shift of mitochondrial metabolism to glycolysis [14, 28, 49]. The same effect is seen in the Drp1 KO #1 cell line, but not in the Drp1 KO #2 cell line (Fig. 3F-I). The observed difference in effects on mitochondrial metabolism between the Drp1 KO cell lines were also correlated with other observations of this study, since the Drp1 KO #2 cell line did not completely abrogate erastin and RSL3-mediated accumulation of mitochondrial ROS and mitochondrial lipid peroxidation (Fig. 2C-F). In contrast, we observed that the basal and maximal mitochondrial respiration were significantly reduced in the Drp1 KO #1 cell line (Fig. 3F), and this cell line was far more resistant to increasing concentrations of erastin or RSL (Fig. 6A, B), showed stronger effects on mitochondrial ROS formation compared to the Drp1 KO #2 cell line (Fig. 2C, D), and also attenuated the overall lipid peroxidation upon ferroptosis induction (Fig. 4A, B). In contrast to our data obtained from neuronal cells, basal and maximal oxygen consumption was not significantly different from controls in Drp1 mutant fibroblasts of patients with EMPF1, while these cells also showed an upregulation of glycolysis [53]. Apparently, the de novo missense mutations examined in these patients maintain the functionality of the Drp1 protein, thereby preserving mitochondrial respiration and leading to extended life expectancy into childhood or early adolescence. This differs from the existing knockout model, which, although tolerated in vitro, exhibits lethality in global Drp1 knockout mouse models [19] and also mostly leads to lethality in humans [10, 54]. How Drp1 depletion affects energy metabolism is unknown and requires further investigation. Our findings suggest that the observed metabolic impact on mitochondrial respiration may significantly contribute further to the protective effects of Drp1 knockout on mitochondrial pathways of ferroptosis.
Overall, this study underlines the importance of mitochondrial metabolism and function in the ferroptosis pathway. Our results highlight the detrimental role of mitochondria to amplify ROS production under conditions of oxidative stress. Many studies accentuate mitochondria as the major source for cellular ROS [55, 56]. The present study shows that through preserving mitochondrial integrity during ferroptosis, the detrimental mitochondrial iron accumulation, ROS production, and ultimately cellular demise can be prevented despite erastin- or RSL3-mediated inhibition of GPX4 activity or accumulating lipid peroxidation. The involvement of mitochondria in the ferroptosis pathway is an essential mechanism of ROS amplification and further death signaling, and only in extreme conditions of ferroptosis, mitochondria lose their significance to prevent cell death [57]. In summary, this study leads to the conclusion that Drp1-mediated mitochondrial disintegration plays an essential role in the course of ferroptotic cell death.