Mitophagy-promoting agents and their ability to promote healthy-aging

“Art is the elimination of the unnecessary” (Picasso), and by analogy, so is mitophagy for the mitochondrial network.

Healthy mitochondria are essential for many fundamental processes in aerobic cells, including the transformation of biomolecule chemical energy to adenosine triphosphate (ATP) via oxidative phosphorylation [1], synthesizing heme [2], degrading fatty acids through β oxidation [3], regulating calcium [4] and iron [5] homeostasis, and controlling cell death by apoptosis [6]. Moreover, mitochondria generate reactive oxygen species (ROS) that play crucial functions in various cellular and physiological processes, including redox homeostasis and immunity [7,8]. In contrast, damaged mitochondria lead to ATP depletion [9], impaired metal iron homeostasis [10,11], excessive ROS formation [9], and cell death activation through the release of cytochrome C [9,12]. Therefore, eliminating injured mitochondria by a process called mitochondrial autophagy (mitophagy) is crucial for proper cellular activity and health.

Multiple mitophagy pathways involve different signaling molecules, enzymes, and adaptor proteins. Below, we briefly describe three canonical pathways — excellent in-depth and up-to-date reviews about the different mitophagy pathways [13–15].

In healthy mitochondria, the PTEN-induced kinase 1 (PINK1) is imported to the inner mitochondrial membrane (IMM) and thus targeted to degradation by the presenilin-associated rhomboid-like protein (PARL) and the cytosolic proteasome. However, when mitochondrial membrane potential (MMP) drops, PINK1 accumulates on the outer mitochondrial membrane (OMM) and becomes activated through autophosphorylation [16,17]. PINK1 phosphorylates ubiquitin and thus recruits the E3 ubiquitin ligase parkin to the OMM [16,18]. Moreover, it can fully activate parkin through direct phosphorylation on serine 65 (in the ubiquitin-like domain), which is further enhanced if parkin binds to phospho-ubiquitin [19]. Parkin ubiquitinates multiple OMM proteins including, voltage-dependent anion channel-1 (VDAC1), mitofusin-2 (MFN2), and TANK-binding kinase 1 (TBK1) [20]. The ubiquitinated OMM proteins interact with cargo receptors, i.e. proteins that can simultaneously interact with autophagosome-associated Atg8/LC3 family proteins and ubiquitin [20], including optineurin (OPTN) and nuclear dot protein 52 (NDP52) [16,18] (Figure 1). Apart from the described signaling pathway, the mitophagy pathway mediated by Pink1/Parkin may involve other mitophagy receptors, such as the Prohibitin 2 IMM protein [21,22].

FUN 14 domain-containing 1 (FUNDC1) is an OMM protein conserved from the nematode Caenorhabditis elegans (C. elegans) to humans [24]. Studies from recent years show that FUNDC1 plays a vital function in hypoxia-induced mitophagy [25]. FUNDC1 has three transmembrane regions and an amino-terminal LC3 interacting region (LIR) that can interact with Atg8/LC3 family proteins (Figure 1, [25]). This interaction is tightly regulated by phosphorylation [26]. When oxygen is not limiting (normoxic conditions), SRC and CK1 phosphorylate FUNDC1 at tyrosine 18 (Y18) and serine 13 (S13), respectively, thus decreasing its affinity to LC3 [26,27]. However, in hypoxia these residues are dephosphorylated. Hence, the interaction with LC3 increases. In contrast, FUNDC1 phosphorylation at serine 17 (S17, adjacent to the LIR sequence motif) by the Unc-51-like autophagy activating kinase 1 (ULK1) significantly increases its interaction with LC3 [26].

The BCL2 interacting protein 3 (BNIP3) and BCL2 interacting protein 3 like (NIX) are stress-induced mitophagy receptors originally categorized as proapoptotic BH3 proteins. Like FUNDC1, BNIP3/NIX are mitochondrial transmembrane proteins (with a single transmembrane region) that project their LIR motif into the cytoplasm, where it can directly interact with LC3 and other Atg8/LC3 family proteins [28–30] (Figure 1). This interaction is regulated, in part, by the phosphorylation of serine restudies nearby the LIR motif (serine 17 (S17)) and serine 35 (S35) in BNIP3 and NIX respectively [31,32]). BNIP3/NIX are also regulated at the transcriptional level. One critical transcription factor (TF) that regulates their activity is the Hypoxia Inducible Factor 1 (HIF-1) TF, particularly BNIP3 when O₂ levels drop below 2% [33–35].

Studies from recent years identified alternative mitophagy pathways. This includes the BCL2-13 and lipid (e.g. cardiolipin and ceramide)-dependent mitophagy pathways [23]. Due to space limitations, we do not describe these pathways in detail.

Yet, all mitophagy pathways lead to the same endpoint: the degradation of damaged mitochondrial components by lysosomal enzymes. Therefore, to simplify, we refer to all types of defective mitochondrial clearance pathways as ‘mitophagy.’

Mitophagy progressively declines with age [36,37]. As a result, the accumulation of damaged mitochondria contributes to cellular damage that may lead to the development of various disease states, such as Alzheimer's [38] and Parkinson's diseases [39], idiopathic pulmonary fibrosis (IPF) and aging lung [40], cancer [41], and sarcopenia [42,43]. Therefore, mitophagy enhancement by pharmacological intervention is an emerging strategy for treating various age-associated diseases. The main goal of this review is to provide an up-to-date resource, both for new researchers in the mitophagy field and for established ones, regarding the variety of substances that: (1) Stimulate the mitophagy process in cells/organisms/animals; (2) Extend the lifespan and healthspan of organisms/animals.

For the reader's convenience, the above information is organized in a table (Table 1), categorizing substances based on their biological activity. It is important to acknowledge that numerous substances can fall into multiple categories. To ensure clarity and ease of reading Table 1, we allocated each substance into a single category that best describes its predominant mode of action. We indicate each ‘substance's chemical formula, molecular weight (MW), structure, and effect on lifespan. Also, we indicate many ‘substances’ healthspan effects and the genes/proteins that mediate their activity. Finally, we indicate in the table which mitophagy-activating substances are natural.

An important clarification must be made regarding Table 1. The table showcases substances that activate mitophagy and potentially extend lifespan and/or healthspan. It is important to acknowledge that the data for these two aspects may originate from separate studies. Furthermore, in many instances, a proven causal relationship has not been established between a substance's impact on mitophagy and its ability to prolong life or healthspan. In simpler terms, while strong evidence might suggest that a specific substance induces mitophagy and extends lifespan or healthspan, the direct cause-and-effect link between these functions remains unverified. Given this context, the data presented in Table 1 offers an opportunity for further investigation into the causal relationship between substance activity in mitophagy and life/healthspan extension.

We classify each substance based on its biological activity. This is has been a challenging task, because most substance (or even all) have more than one biological activity. However, for the sake of clarity and readability, we have assigned each substance to a specific category. Additionally, we acknowledge that some readers may be interested in knowing whether a substance is of natural or synthetic origin. Hence, we have explicitly indicated the natural origin for all substances. We classified the mitophagy activating substances into the following categories: antioxidants, mitochondrial uncouplers, complex I inhibitors, redox, metabolic, neuronal, autophagy modulators, antibiotic and ant-fungal agents, and reactive oxygen species (ROS) generators (Table 1 and Figure 2).

It is worthwhile noting that many of the natural substances that trigger the mitophagy pathway are polyphenols. Polyphenols are a diverse group of naturally occurring compounds found abundantly in plants and are considered to be the most widely distributed phytochemicals among all plant-based sources. Collectively, they have gained significant attention due to their health-promoting properties, including lifespan extension [164,165]. The longevity-promoting effects of polyphenols depend on the structural characteristics of their carbocyclic rings and the number of hydroxyl groups present on the ring [68]. Studies in C. elegans suggest that polyphenols act through specific longevity pathway proteins, i.e. the DAF-16/FoxO3 and SKN-1/NRF2 TFs and the SIR-2.1/SIRT1 sirtuin [166]. These proteins appear to mediate the life-extension activity of the polyphenol mitophagy activators baicalein [53], catechinic acid [167], curcumin [58], myricetin [76], naringin [78], resveratrol [87], and urolithin A [95].

Intriguingly, DAF-16 and SKN-1 regulate the activity of the mitophagy receptor DCT-1 (the C. elegans homolog of the mammalian BNIP3 and BNIP3L/NIX) [168]. In this regard, it is important to indicate that mitophagy activating via DCT-1 protects from oxidative injury [155]. Many dietary polyphenols are potent antioxidants [165]. However, whether their bioavailability through the digestive system and effective transport to cells facilitate such function in vivo is still a matter of controversy [169,170]. Thus, a plausible hypothesis is that the antioxidant activity of polyphenols is partly mediated indirectly, such as through the activation of mitophagy [171].

Many of the synthetic substances are drugs, with some being FDA-approved., i.e. aspirin, doxycycline hydrochloride, ivermectin, metformin, and rapamycin. Other substances are potent ROS inducers (paraquat and sodium arsenite), Uncouplers of oxidative phosphorylation (2,4-dinitrophenol and CCCP), Metabolism modulators (dichloroacetate and metformin), and Synthetic diamines (VL-004 and VL-850).

Importantly, as indicated for the natural substances, some compounds could be classified in more than one group. For example, CCCP is a potent mitochondrial uncoupler and (thus) inducer of oxidative stress [172]. However, we chose to assign each material to a single category for simplicity.

Similar to the natural substances, several synthetic ‘substances’ activities are mediated by DAF-16 and SKN-1 TFs; for instance, ivermectin, sodium arsenite, aspirin, rapamycin, rilmenidine, 1,8 diaminooctane (VL-004), and others. Our studies on the synthetic diamine VL-004 show that, at least in C. elegans, additional TFs are needed for mitophagy-dependent antioxidant activity. That is, the hypoxia-inducible factor 1 (HIF-1), HLH-30/TFEB, and PHA-4/FOXA. Notably, these TFs are known to regulate/modulate lifespan and healthspan in C. elegans [173–176], further supporting the causative function of mitophagy in lifespan regulation. A major challenge is understanding how these TFs activities are coordinated to extend life- and healthspan.

Previous studies suggest that drug combinations may elicit a therapeutic effect greater than the sum of the individual drugs [177]. Indeed, there are evidence that targeting different components of the same cellular/physiological process may be advantages [178]. An example of such synergistic activity between two drugs acting on the same target is the anticancer combination of cisplatin and trabectedin. Trabectedin interacts with DNA and DNA repair systems differently than cisplatin does. Specifically, trabectedin inhibits DNA replication through a different mechanism of action, which decreases the antagonistic activity against cisplatin, resulting in a synergistic effect [178]. To the best of our knowledge, no systematic research has been conducted to examine the combined impact of mitophagy-activating substances on lifespan and health in any model system.

Nevertheless, despite the medical potential inherent in the MAC cocktail strategy, we would like to point out a possible limitation. A recent article we published showed that the synthetic diamine VL-004 extends the health and lifespan of C. elegans [155]. Moreover, we demonstrated that VL-004 protects C. elegans and human-derived cell lines from oxidative injury. This protective activity depends on the TF DAF-16, which DAF-2 negatively regulates; the C. elegans sole ortholog of the mammalian insulin and IGF-1 receptors (IIR) [179]. DAF-2 inhibits DAF-16 from entering the nucleus. As a result, when the activity of DAF-2 is suppressed, DAF-16 can move into the nucleus, where it activates genes crucial for stress resistance and increased lifespan. Among these genes are antioxidant genes (i.e. catalases 1 and 2 and superoxide dismutase 3 [180]), and the TF SKN-1 activates multiple antioxidant genes [181]. Therefore, it is unsurprising that worms bearing loss-of-function mutations impairing DAF-2 activity (daf-2 mutants) are more resistant to oxidative stress [182].

Intriguingly, VL-004 does not increase the resistance of daf-2 mutants to oxidative stress [155]. On the contrary, it decreases the resistance to wild-type worms level [155]. A hypothesis that may explain the above observation is that an optimal level of mitophagy maintains an efficient balance of mitochondrial activity and ROS level. Beyond this level, the activation of mitophagy is harmful, perhaps because it eliminates intact mitochondria and/or interferes with healthy ROS signals (Figure 3). Indeed, direct mitophagy measurements show that the basal level of mitophagy in daf-2 mutants is high, similar to that of wild-type worms treated with VL-004 [155]. As predicted by the hypothesis above, VL-004 increases mitophagy in daf-2 mutants, driving it into the unhealthy mitophagy zone. In line with this observation, there are evidence that excessive mitophagy can be harmful. For instance, increased mitophagy can lead to the development of heart failure condition due to cardiac myocytes loss [183]. Moreover, it has been suggested that exaggerated mitophagy has a causative role in the pathophysiology of Huntington's diseases [184,185], as well as in stroke and multiple sclerosis [186]. In conclusion, mitophagy is like walking the tightrope, i.e. it requires precise balance. Hence, a treatment plan for the MAC cocktail should be designed not to surpass the healthy mitophagy threshold.

To address this question, we explored the https://clinicaltrials.gov/ database, focusing on studies that met the following criteria. Firstly, we solely considered trials involving individuals aged 60 years and above, both male and female. The choice of this population section is driven by the article's topic, namely, improving life expectancy and health. Secondly, our inclusion criteria encompassed both ongoing and completed trials. In situations where multiple studies (>3) were available, priority was given to completed and published trials to ensure comprehensive data availability. Thirdly, our analysis was limited to compounds used in their original form within the study. This criterion effectively eliminated any confounding effects caused by extracts or compound combinations, thus enhancing the clarity of the results and simplifying interpretation. Fourthly, we detailed potential limitations associated with specific treatments, as indicated in the https://clinicaltrials.gov/ database. These potential limitations include factors such as the number of patients participating in the trial, the duration of the treatment, among others. Additionally, we highlighted any observed side effects. Lastly, our research focused on diseases predominantly affecting the elderly and exhibiting a strong correlation with mitophagy decline. By adhering to these criteria, we aimed to offer reliable insights into the existing dataset concerning mitophagy-activating substances within clinical contexts. The results of these studies are outlined in Table 2.

Dysfunctional mitophagy is implicated in the pathophysiology of multiple age-associated maladies, including Alzheimer's and Parkinson's diseases, cardiovascular diseases, and sarcopenia [193,194]. Therefore, enhancing mitophagy is a promising strategy for future disease treatment [195]. However, there are cases in which mitophagy can contribute to disease development. For example, functional mitophagy appears to be necessary for tumor progression. In tumors with oncogenic KRAS mutations, mitophagy increases malignancy [196]. Indeed, Selective inhibition of autophagy/mitophagy by liensinine, a natural alkaloid, increases the cell death activity of doxorubicin through dynamin 1-like (DNM1L)-dependent mitochondrial fission [197]. Similarly, mdivi-1, a DRP1-mediated mitophagy inhibitor, sensitizes hepatic cancer cells to cisplatin and enhances cancer cell death [198].

Apart from cancer, several pathogenic viruses (e.g. CSFV, HBV, HCV, and MeV) activate mitophagy to escape cell death, thereby promoting virus infectivity. For instance, measles virus vaccine strain Edmonston B (MV-Edm) infection activates mitophagy, leading to a decrease in the release of somatic cytochrome C (CYCS). As a consequence, apoptotic cell death is inhibited in non-small cell lung cancer cells [199]. In conclusion, while using small-molecule substances to activate mitophagy presents promising therapeutic opportunities, this approach should be carefully considered due to the multiple roles of mitophagy in cellular health and disease.

As mentioned, the selective elimination of damaged mitochondrial components is essential for the normal activity of the mitochondrial network. However, this step must be accompanied by the supply of new ones in a process called mitochondrial biogenesis [200]. Indeed, precise synchronization of mitochondrial biogenesis and mitophagy is necessary for the cells to adapt to different metabolic states and stress conditions, and a lack of coordination is closely associated with various pathological conditions (e.g. ‘Parkinson's disease) [200].

Since the process of mitochondrial biogenesis is not the subject of this article, we would like to focus on only one point related to the subject. That is, the activation of mitochondrial biogenesis by the same molecules that activate mitophagy and common signaling pathways/molecules.

Resveratrol induces mitochondrial biogenesis in several cell types, including brain cells [201,202]. Moreover, it has been shown to have beneficial activity in several disease models, including models associated with accelerated aging. For instance, administering 20 mg of resveratrol per kg of body weight one day before the monocular derivation procedure prevented the decrease in AMP-activated protein kinase (AMPK) phosphorylation, peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1α), and nuclear respiratory factors 1 (NRF-1) in the visual cortex of rats [203]. Moreover, resveratrol (60 µM, 24 h pretreatment) enhances mitochondrial biogenesis through the functions of PGC-1α and mitochondrial TF A in a cellular model system (PC-12 cells) for ‘Parkinson's disease [204]. Notably, NRF-1 (as well as NRF-2), the estrogen-related receptors (α, β, and γ), and PGC-1α are key TFs that control mitochondrial biogenesis in mammalian cells [200]. Lastly, resveratrol (160 mg/kg per day, for 8 weeks) improves cognition and mitochondria function in a mouse model of accelerated aging (senescence-accelerated mouse prone 8, SAMP8) subjected to a high-fat diet [205]. Interestingly, resveratrol increases the level of the mitochondria respiratory complexes without affecting the levels of PGC-1α and SIRT1, however increasing AMPK-phosphorylation.

Metformin (2 mM) administration attenuates human endothelial cell senescence by promoting mitochondrial biogenesis through the activation of AMPK and by increasing the levels of SIRT3 and PGC-1α [206]. Moreover, metformin enhances the trimethylation of H3K79 (H3K79me3) in the SIRT3 promoter region through an SIRT1-DOT1L signaling axis. This, in turn, increases SIRT3 expression, leading to enhanced mitochondrial biogenesis [206]. Furthermore, metformin (1700 mg/day, for at least one year) increases PGC-1α levels in peripheral blood mononuclear cells (PBMCs) derived from type 2 diabetic patients [207], as well as in liver and skeletal muscle cells [208,209].

At the molecular level, the activation of AMPK serves as a junction between the pathways of mitophagy activation and mitochondrial biogenesis. Specifically, by activating PGC-1α, AMPK stimulates mitochondrial biogenesis while also triggering autophagy and mitophagy through the unc-51-like autophagy activating kinases 1 and 2 (ULK1/2) [200]. Notably, AMPK is activated by a variety of external/internal stimuli. For instance, elevated calcium (Ca²⁺) levels can activate the Ca²⁺/calmodulin dependent kinase (CaMK), which further activates Ca²⁺-Calmodulin-dependent protein kinase kinase beta (CaMKKβ), and thus AMPK [200,210,211]; CamK can also activate PGC-1 directly.

Despite the abundance of existing knowledge regarding mitophagy and biogenesis processes, there remains a significant knowledge gap in comprehending their interrelationships, especially concerning their control at the organismal level and drug treatment that combines the activation of these processes. Therefore, it would be fascinating to investigate how the combined activation of mitophagy and biogenesis affects the healthspan and lifespan of various biological models.

Declining mitophagy contributes to cellular damage and the development of conditions such as Alzheimer's, Parkinson's, and cancer. Therefore, stimulating mitophagy via pharmacological intervention presents a potential means to improve cellular health and combat these diseases.

Various types of small molecules have been identified as mitophagy activators, including antioxidants, mitochondrial uncouplers, complex I inhibitors, redox regulators, metabolic modulators, neuronal modulators, autophagy modulators, antibiotic and antifungal agents, and reactive oxygen species (ROS) generators.

In many cases, the activities of these substances are mediated by TFs that orchestrate the cell's response to stresses, such as DAF-16/FOXO, PHA-4/FOXA, HLH-30/TFEB, and SKN-1/NRF-2. However, in many cases, the specific cellular receptors that bind these substances and how this binding triggers the activation of a particular mitophagy pathway remain unclear. Furthermore, the mechanism by which mitophagy activation extends longevity and healthspan is not fully understood. A mechanistic understanding of these processes is crucial for developing more effective and safer molecules to treat various age-related diseases.

Using a mitophagy-activating substances cocktail (MAC) for therapeutic purposes holds great promise in enhancing medical benefits. The MAC cocktail could potentially improve health and lifespan by targeting different components of the same cellular process. However, it is crucial to consider potential limitations, as excessive mitophagy may lead to harmful effects. Maintaining the optimal level of mitophagy is crucial for balancing mitochondrial activity and ROS levels, and exceeding this threshold might have detrimental consequences. Therefore, the design of the treatment plan for the MAC cocktail should carefully consider the delicate balance of mitophagy to ensure it does not surpass the healthy threshold, maximizing its beneficial effects for therapeutic purposes.

Following the previous paragraph, exercising caution in utilizing small-molecule substances to activate mitophagy is essential, as its effects can be context-dependent. Mitophagy's involvement in tumor progression and viral infection demonstrates the dual nature of its impact on cellular processes, where both beneficial and harmful outcomes are observed. Therefore, any therapeutic approach to manipulate mitophagy should be carefully tailored and thoroughly researched to ensure the best possible outcomes in different disease contexts.

Finally, the mitochondrial biogenesis process balances mitophagy to ensure proper mitochondrial network homeostasis. Interestingly, mitochondrial biogenesis may be activated by the same molecules that induce mitophagy, such as resveratrol and metformin. Despite progress in understanding these processes, much remains to be uncovered, particularly in comprehending their interrelationships at the organismal level and exploring combined drug treatments targeting mitophagy and biogenesis. Future research in this area could shed light on how such combined activation may impact healthspan and lifespan in different biological models, potentially paving the way for novel therapeutic strategies in the future.

The authors declare that there are no competing interests associated with the manuscript.

This work is funded by the ISRAEL SCIENCE FOUNDATION (grant No. 989/19).

Open access for this article was enabled by the participation of Hebrew University of Jerusalem in an all-inclusive Read & Publish agreement with Portland Press and the Biochemical Society under a transformative agreement with MALMAD.

V.S. and E.G. wrote the manuscript.

We apologize to our colleagues whose work could not be directly referenced due to space limitations. We appreciate your contributions and regret any inconvenience caused by our inability to include all relevant citations. The illustrations in Figures 1–3 were generated by combining BioRender and pen drawing.

AMPK

AMP-activated protein kinase

BNIP3

BCL2 interacting protein 3

CaMK

Ca²⁺/calmodulin dependent kinase

FUNDC1

FUN 14 domain-containing 1

HIF-1

Hypoxia Inducible Factor 1

IMM

inner mitochondrial membrane

LIR

LC3 interacting region

MAC

mitophagy activating compound

NDP52

nuclear dot protein 52

OMM

outer mitochondrial membrane

PINK1

PTEN-induced kinase 1

ROS

reactive oxygen species

TBK1

TANK-binding kinase 1

TFs

transcription factors

VDAC1

voltage-dependent anion channel-1