Phosphorylations and Acetylations of Cytochrome c Control Mitochondrial Respiration, Mitochondrial Membrane Potential, Energy, ROS, and Apoptosis

Using the double Cytc knockout system, cells transfected with T58E Cytc showed a similar phenomenon to the other phosphorylations when expressed in cells: 68% decreased respiration rates in intact cells, reduced ΔΨm levels and mitochondrial ROS production, and 66% decreased ATP levels [43]. The cells expressing T58E Cytc also underwent less cell death. Overall, these data establish the role of T58 phosphorylation as an anti-apoptotic modification, supporting the model that Cytc phosphorylation sites maintain optimal mitochondrial functioning to limit ROS generation [22,59]. It still has to be shown in what renal cell type this modification is found and what kinase/phosphatase enzymes regulate this modification.

Y97 was the first phosphorylation site of Cytc to be discovered and characterized. It was initially mapped on Cytc purified from bovine heart [40]. Using the in vivo phosphorylated purified protein, the Km of the Cytc-COX reaction was found to be increased from 2.5 μM of WT to 5.5 μM for the phosphorylated protein whereas the Vmax was unchanged. This alteration of the Cytc-COX kinetics is likely due to a perturbation of the heme group, which was seen on the UV–Vis spectra as the classic 695 nm peak, which assesses heme status via the M80-heme iron coordination, was shifted to 687 nm. Interestingly, another publication using phosphomimetic Y97E found no change of the redox potential and a significant decrease in melting temperature, implying that protein stability may be affected by alteration of this residue [49].

A third publication examined phosphorylation using phosphomimetic Y97pCMF replacement [51]. In contrast with the results obtained using the in vivo Y97 phosphorylated Cytc, Y97pCMF demonstrated increased Cytc-COX activity. The Y97pCMF did demonstrate some reduced pro-apoptotic functionality, with a decrease in downstream caspase-3 activity, although there was no change in cardiolipin binding or cardiolipin peroxidase activity using this phosphomimetic. Recently, a O-sulfotyrosine has become available as a novel mimetic, which should be tested as a model for studying tyrosine phosphorylation [89].

In a different study, Y97 phosphorylation was found to be rapidly induced by neuroprotective insulin treatment after induced ischemia-reperfusion in porcine and rat brains [90]. Induction of Y97 phosphorylation protected the CA-1 hippocampal neurons from cell death following the ischemia-reperfusion injury, supporting the anti-apoptotic role of Y97 phosphorylation seen using the phosphomimetic. Additionally, less Cytc was released after 24 h post-ischemia when Y97 phosphorylation was induced by the insulin treatment, suggesting that phosphorylation may be able to prevent Cytc release into the cytosol.

In vitro lysine acetylation of Cytc has been long studied. However, detection of lysine acetylation of Cytc in vivo has only occurred recently and is a new area of research (Table 2). The first acetylation site discovered was lysine 8 (K8) in fasted mouse liver mitochondria via high-throughput acetyl-omics [91]. Later, lysine 53 (K53) was identified in prostate cancer [92]. This was followed by lysine 39 (K39) in ischemic skeletal muscle [93].

Lysine acetylation is the most abundant post-translational modification in mitochondria [94,95]. Lysine acetylation is believed to occur largely non-enzymatically in the mitochondrial matrix, due to the basic pH and high concentration of acetyl-CoA [96]. However, lysine acetylation of Cytc, which resides in the IMS, was found to occur selectively on specific residues tied to a particular tissue or pathology. This suggests that specific acetyltransferase enzymes mediate these modifications. Unlike phosphorylations of Cytc, which are present under basal conditions and are generally lost during ischemia, these acetylations tend to be associated with a specific pathophysiological state. Lysine acetylations suffer from the same methodological problems as phosphorylations: yields are typically low and the modification itself is transient. Therefore, it is common to study acetylation using glutamine as an acetylmimetic. In contrast to the positively charged side chain of lysine, acetyl-lysine and glutamine both have uncharged, polar side chains that contain amide groups [97]. Lysine acetylation of Cytc, which leads to the loss of the positive charge of the ε-amine group, is generally predicted to interfere with binding to COX and Apaf-1, due to making Cytc less positively charged. In this manner, acetylation, which removes a positive charge from a lysine residue, is similar to phosphorylation, which introduces a negative charge on a neutral S/T/Y residue, because both of them result in Cytc becoming more negatively charged.

K8 acetylation was discovered in a high-throughput study of fasted mouse liver [91]. This residue is known to interact with cytochrome c1 of Complex III [99,100], and it was found that acetylmimetic K8Q Cytc had a perturbed dissociation constant from cytochrome c1 [98]. K8 of Cytc is also known to interact with D139 of COX subunit II [72] as part of the proximal Cytc-COX binding site, explaining the observed decrease in electron transfer to COX using acetylmimetic K8Q protein [98]. While this residue seems crucial for the electron transfer activity of Cytc, not much is known regarding the effects of this modification on the other functions of Cytc.

K39 acetylation was mapped in Cytc purified from ischemic porcine tibialis anterior muscle after 45 min of ischemia [93]. The modification was not found in the muscle under basal conditions, making it the first reported gain of a post-translational modification during ischemia, suggesting that the response to ischemia may be muscle-specific. It is known that muscles tend to be more resilient to ischemia than other tissues such as brain, heart, kidneys, or liver [101,102]. K39 acetylation may provide a unique mechanism to potentially explain this resilience. Whereas other phosphorylations and acetylations of Cytc tend to be present under basal conditions, lost during ischemia, and inhibitory to the Cytc-COX reaction, K39 acetylation of tibialis anterior muscle follows a unique metabolic and apoptotic strategy by stimulating respiration to meet the cell’s energy demand and at the same time inhibiting apoptosis (Figure 5). K39 acetylation was found to be removed by sirtuin5, which is known to co-localize in the IMS with Cytc [103,104]. Sirtuins are known to be downregulated in response to hypoxia, suggesting that preventing the removal of this acetylation during ischemia facilitates the protective effects [105]. The in vivo acetylated Cytc purified from ischemic tibialis anterior muscle showed a 58% increase in Cytc-COX Vmax, compared to the unacetylated Cytc purified from non-ischemic tibialis anterior muscle [93]. Acetylmimetic K39Q Cytc produced a similar increase of 38% compared to WT. Nuclear magnetic resonance studies demonstrated perturbation of residues interacting with or near the heme group such as the M80-containing loop and the K55-to-W59 stretch, possibly explaining the increased electron transfer rate by facilitating more efficient electron transfer. Interestingly, a separate publication studying K39L did not observe any changes in Cytc-COX kinetics [106].

Additionally, this acetylation, which was triggered as a response to ischemia, blocks the pro-apoptotic functions of Cytc, thereby acting as a protective mechanism during ischemia-reperfusion. Downstream caspase-3 activity using the in vivo acetylated protein purified from ischemic tibialis anterior muscle was 45% decreased, and the K39Q Cytc showed a reduction of 90% in the same assay [93]. K39 of Cytc interacts with the carbonyl carbon of F1063 of Apaf-1 [17,56], and the acetylation likely perturbs this interaction, decreasing downstream caspase-3 activity. Furthermore, cardiolipin peroxidase activity, another proapoptotic action, was decreased with the acetylmimetic K39Q Cytc.

Interestingly, acetylmimetic K39Q Cytc had a decreased rate of reduction by superoxide and an increased rate of oxidation by H2O2 compared to WT [93]. This may be because superoxide is a known signaling molecule produced during muscle contraction [107,108,109]. Therefore, reduced scavenging of superoxide could be an adaptative response to short periods of hypoxia that occur during normal muscle exercise, which becomes mal-adaptive in the context of ischemia-reperfusion injury. The crystal structure of the in vivo K39 acetylated Cytc purified from ischemic tibialis anterior was solved at 1.5 Å and had an occupancy of 0.75 for the acetyl group, showing that a large proportion of the protein pool was acetylated following ischemia [93].

Similar to the results for the Cytc-COX kinetics using purified proteins, cells expressing the acetylmimetic K39Q Cytc showed an increase in respiration with a 70% increase in basal oxygen consumption rate. Accordingly, the basal ATP levels were increased 51% over WT. ΔΨm and mitochondrial ROS production also experienced increases compared to cells expressing WT. These experiments demonstrate that single point mutations on Cytc can both inhibit and stimulate overall mitochondrial activity, supporting the concept that the Cytc-COX interaction is the rate-limiting step of the ETC. After exposure to OGD/R, cells expressing K39Q Cytc did not show a statistically significant increase in mitochondrial ROS production, unlike WT. These data show that K39Q Cytc increases respiration, ΔΨm, and ROS under baseline conditions compared to WT, but does not lead to a further increase following OGD/R, potentially protecting the cells from additional oxidative stress upon reperfusion. After exposure to H2O2, OGD/R, and thapsigargin treatment, cells expressing K39Q Cytc experienced reduced levels of cell death. Using cells expressing K39Q Cytc, reduced levels of Cytc were co-immunoprecipitated with Apaf-1, suggesting impaired interaction of the two proteins. Furthermore, this translated into reduced cleavage of procaspase-9 and procaspase-3 after induction of apoptosis in the presence of K39Q Cytc. Overall, skeletal muscle is a unique tissue type with drastic changes in perfusion depending on usage [110,111]. Cytc K39 acetylation may be a skeletal muscles-specific adaptation during exercise or other short-term energy deficiencies, meant to allow the muscle to meet the high energy demand while protecting from cell death (Figure 5) [93].

Lastly, we will discuss K53 acetylation, which was mapped in Cytc purified from human prostate cancer xenografts and prostate cancer specimens but was not found in normal prostate tissue [92]. Overall, this modification was found to be advantageous to cancer cells by driving two hallmarks of cancer: a metabolic switch from OxPhos to glycolysis and apoptosis evasion. It is known that acetyltransferase enzymes are upregulated in prostate cancer, which may help explain increased K53 acetylation [112,113]. Acetylmimetic K53Q Cytc showed 35% decreased Vmax for the Cytc-COX reaction. K53 of Cytc is within 5 Å of K58 residue of COX subunit VIIa [52], suggesting that K53 is within the Cytc-COX binding domain and acetylation may spatially interfere with the interaction. This inhibition of respiration would assist the prostate cancer cells in metabolic reprogramming toward the glycolytic pathway.

Additionally, cancer cells often manage to evade apoptosis, which K53 acetylation facilitates. Downstream caspase-3 activity was 80% decreased using acetylmimetic K53Q Cytc and may be explained because K53 is a known regulatory epitope for Apaf-1 binding [17]. The ability of acetylmimetic K53Q Cytc to scavenge ROS was also significantly increased, with large increases in rate of oxidation by H2O2 and rate of reduction by superoxide, compared to WT. These may further facilitate prostate cancer cell survival by detoxifying ROS, blocking a strong pro-apoptotic signal. The pro-apoptotic cardiolipin peroxidase activity was also greatly reduced with acetylmimetic K53Q Cytc. K53 is conserved in all mammals, further pointing to an important regulatory role of K53 for the interactions with both COX and Apaf-1 [71]. Future work should expand these studies to other cancers to see if this modification is more broadly involved in cancer pathology and potentially in resistance to cancer therapies relying on apoptosis induction.

The tissue and disease specific regulation of Cytc has interesting implications. It has been long debated which step of the ETC is rate-limiting. We have previously proposed that the interaction between Cytc and COX is rate-limiting [22]. Cytc and COX are both highly regulated via allosteric ATP binding, post-translational modifications, and tissue-specific isoforms, and they are the only known mammalian components of the OxPhos that are regulated by all three mechanisms. Cytc acts at the intersection point between pro-survival respiration and pro-death apoptosis, which depends on the subcellular localization of the protein, and, therefore, lends itself as an ideal regulatory target. Several reports employing traditional metabolic flux analysis experiments have identified COX as the rate-limiting step of the ETC [114,115,116,117,118]. For these experiments, an inhibitor of a specific complex is added in increasing amounts, and total flux through the ETC is measured, usually via oxygen consumption of COX as the terminal step of the ETC. The less inhibitors necessary to affect electron flux globally through the entire ETC, the more rate-limiting a particular step is. However, the possibility that Cytc and its post-translational modifications are rate-limiting was overlooked in the past, perhaps due to the lack of a specific inhibitor of Cytc.

Cytc, which is crucial for physiological mitochondria functioning, has been shown to possess tissue-specific post-translational modifications that react to ischemia-reperfusion injury. Patterns within these post-translational modifications of Cytc help explain different tissue sensitivities to ischemia-reperfusion injury. Specifically, various phosphorylations of Cytc, which are present under basal conditions and lost during ischemia, sensitize the tissue to ischemia-reperfusion injury, while K39 acetylation of Cytc in skeletal muscle, which is not present under basal conditions and is gained during ischemia, protects skeletal muscle and contributes to the resilience of this tissue type to ischemia-reperfusion injury. In this review, we have discussed that single post-translational modifications and point mutations of Cytc are able to affect flux through the entire electron transport chain. When cells expressed T28E, S47E, T58E, or K39Q Cytc, oxygen consumption rate by the cells was altered compared to cells expressing WT. While the phosphorylations of Cytc and most acetylations are inhibitory to Cytc-COX kinetics, K39 acetylation of Cytc is stimulatory. This highlights that a change in Cytc-COX activity, whether inhibitory or stimulatory, can decrease or increase total ETC flux, respectively.

This review has focused primarily on phosphorylations and acetylations of Cytc, specifically those which have been detected in vivo in mammals and for which clear evidence exists that they are functionally important. Other post-translational modifications of Cytc exist (reviewed in [23]). However, they were largely detected through high-throughput studies or are only found in vitro, therefore their biological significance is unknown. In most but not all cases, the eight modifications discussed here serve to support healthy cell functioning by blocking pro-apoptotic functions of Cytc and by lowering mitochondrial activity to minimize ROS production. Furthermore, it has long been observed that different organs and even different tissue types within an organ have different sensitivities to ischemia and reperfusion injury [119]. For the phosphorylations, which are present in their respective tissues under basal conditions, their loss tends to drive ischemia-reperfusion injury by stimulating respiration, increasing ROS production, and sensitizing the cell to apoptosis. Overall, the phosphorylations occur in ischemia sensitive tissues, such as brain, heart, kidney, and liver, which may explain the susceptibility of these tissues to ischemia-reperfusion injury. K39 acetylation follows a different pattern, where acetylation is a response to ischemia, stimulating respiration and blocking apoptosis to enforce maximum muscle activity output. The presence of this acetylation in skeletal muscle, which is notoriously less sensitive to ischemic stress, results in skeletal muscle protection and resilience to ischemia-reperfusion injury. Altogether, the diversity of Cytc regulation by post-translational modifications underscores the crucial nature of the protein for both the ETC and the cell as a whole.

Conceptualization, P.T.M. and M.H.; writing, review, and editing, P.T.M. wrote the manuscript with support of all authors; visualization, T.A. and P.T.M.; resources, funding acquisition, and supervision, M.H. All authors reviewed and edited the manuscript. All authors have read and agreed to the published version of the manuscript.

This work and some of the work discussed in this article was supported by the Office of the Assistant Secretary of Defense for Health Affairs through the Prostate Cancer Research Program under Award No. HT9425-24-1-0073, NIH grant R01 NS120322, NIH grant U44 NS125160, the Michigan Prostate SPORE Developmental Research Program, NSF grant MCB-2329629, and the Michigan Translational Research and Commercialization (MTRAC) Award. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the funding agencies including the Department of Defense, the National Institutes of Health, or the National Science Foundation.