Phase separation in mitochondrial fate and mitochondrial diseases

Abstract

Mitochondria are central metabolic organelles that control cell fate and the development of mitochondrial diseases. Traditionally, phase separation directly regulates cell functions by driving RNA, proteins, or other molecules to concentrate into lipid droplets. Recent studies show that phase separation regulates cell functions and diseases through the regulation of subcellular organelles, particularly mitochondria. In fact, phase separation is involved in various mitochondrial activities including nucleoid assembly, autophagy, and mitochondria-related inflammation. Here, we outline the key mechanisms through which phase separation influences mitochondrial activities and the development of mitochondrial diseases. Insights into how phase separation regulates mitochondrial activities and diseases will help us develop interventions for related diseases.

Phase separation is a phenomenon in which different condensates assemble to form a new condensate with novel functions (1–3). The driving forces behind phase separation in biological systems are mediated by weak interactions between proteins or nucleotides (4). Modular domains or sequences of intrinsically disordered regions (IDRs) mediate the occurrence of phase separation (5, 6).

Recently, phase separation has been reported to play a crucial role in intercellular physiological processes in both membrane-bound and non-membrane-bound compartments. Mitochondria, which resemble parasitic bacteria in the early stages of cell formation, contain a semiautonomous nucleoid encoding electron transport chain (ETC) protein (7). As the energy metabolic center of the cell, they participate in cell fate determination through oxidative phosphorylation (OXPHOS) and the tricarboxylic acid cycle (TCA) (8). Moreover, mitochondria regulate cell functions through mitochondrial activities, including mitochondrial biosynthesis, mitophagy, mitochondrial fusion and fission. Phase separation directly regulates cell fate by driving the concentration of RNA, proteins, or other molecules into lipid droplets. Interestingly, recent studies reported that phase separation also regulates cell functions through the regulation of subcellular organelles, particularly mitochondria (9–11).

We summarize how phase separation affects mitochondrial activity, including mitochondrial nucleoid (mt-nucleoid) assembly and transcription, autophagy, and mitochondrial DNA or RNA leakage. Furthermore, we also discuss the feasibility of targeting the process of phase separation for the treatment of mitochondrial diseases.

Phase Separation in Nucleoid Assembly and Transcription

The mitochondrial nucleoid is a highly compact and spherical suborganelle that serves as the primary site for genetic material storage and mtDNA transcription (12). Unlike the cell nucleus, the mt-nucleoid lacks a membrane structure. It is composed of mitochondrial DNA (mtDNA) and protein complexes, ensuring that mitochondria maintain a semiautonomous and spatially isolated genome (13). The minimal transcriptional machinery of mitochondria comprises mitochondrial transcription factor A (TFAM), TFB2M, and mitochondrial RNA polymerase (POLRMT) (14, 15). As a key structural protein, TFAM condenses mtDNA by compressing it into the core nucleoid structure (11).

Mitochondrial nucleoid assembly and transcription are essential steps in mitochondrial biosynthesis. Mitochondrial nucleoids, like other membraneless structures including P bodies, stress granules, and promyelocytic leukemia nuclear bodies, are assembly through phase separation (16–18). The TFAM-mtDNA droplets undergo compartmentalization via phase transition (Fig. 1) (19). TFAM belongs to the high mobility group (HMG) box family that can bind DNA. The experiments from in vitro and vivo suggest that the intrinsically disordered regions (IDRs) of TFAM, along with the multivalent effects mediated by its multiple DNA-binding sites, drive the phase separation of TFAM-mtDNA (11, 20). The nucleoid droplet promotes the recruitment transcription substrates and transcription machineries, such as TFB2M, POLRMT, mitochondrial transcription elongation factor (TEFM), and mitochondrial transcription termination factor (MTERF1) through cophase separation, thereby dramatically concentrating these enzymes and substrates for efficient transcription (Fig. 1) (11, 21, 22).

Fig. 1.

Phase separation involved in mitochondrial transcription. TFAM mediates mt-nucleoid assembly, mt-transcription initiation, mt-transcription elongation, and termination via phase separation.

Mitochondrial RNA granules (MRGs), which are juxtaposed to nucleoids, contain newly synthesized RNA, RNA processing proteins, and mito-ribosome assembly factors, all of which are involved in mitochondrial RNA processing and ribosome production. Recently, MRGs have been identified as nanoscale liquid phases, distributed across mitochondria through membrane dynamics, displaying characteristics akin to fluid condensates (23–25). As membraneless organelles, the ultrastructure of MRGs comprises compacted RNA integrated within a protein cloud (25). MRGs play a crucial role in mtDNA expression, with mtDNA helicase TWINKLE and single-stranded DNA-binding protein mtSSB, contributing to RNA metabolism (26). FASTKD2, an RNA-binding protein (RBP) within MRGs, rapidly exchanges components and undergoes fusion by live-cell superresolution structured illumination microscopy and fluorescence recovery after photobleaching (25). However, the mechanism of phase separation in MRGs remains unclear, while the multivalent interactions between RBPs and RNA may drive their phase separation.

Phase separation protects mitochondrial transcription by regulating the structure and function of transcriptional units. The prolonged dwell time of proteins associated with the condensate microenvironment favors the assembly of reacting complexes. Feric et al. reconstituted mt-transcription under condensate-forming conditions in vitro and observed the formation of arrested phases of mt-transcription components (19). Moreover, phase separation concentrates transcribed substrates and enzymes, providing a relatively stable compartment that enhances mt-transcription (11). It suggests that phase separation makes mt-transcription more stable, which may explain why phase separation has been maintained throughout evolution (11, 27). In the nucleus, phase separation increases the concentration of local transcription factors (TFs), but it fails to further enhance mt-transcription (28). The actual formation of liquid TF droplets has a neutral or inhibitory effect on transcriptional activation, as demonstrated by the relationship between binding site occupancy, residence time, and coactivator recruitment in the context of multivalent interactions of TFs (28). The distinct transcription mechanisms of the mitochondrial nucleoid and the nucleus lead to different functions of phase separation in transcriptional regulation.

Phase Separation in Mitophagy

Mitophagy is a crucial pathway for mitochondrial quality control. Dysfunctional mitochondria, recognized by autophagosomes, are first labeled with ubiquitin (29, 30). The condensate is then delivered to the mitochondrial membrane and broken down by proteasomes, which interact with cargo receptors to link the targeted cargo to the autophagosome membrane by LC3/ATG8 proteins (31–33). Several cargo receptors, include OPTN, NDP52, p62/SQSTM1, NBR1, and TAXBP1, bind to ubiquitin chains on damaged mitochondria and interact with LC3/ATG8 proteins attached to the inner membrane layer of the autophagosome (9, 34–36). Phase separation within and outside the mitochondria concurrently promotes the phagocytosis of damaged mitochondria. Several mitophagy-related phase separations are discussed below.

p62/SQSTM1 (hereafter referred to as p62) is one of the most well-characterized selective autophagic receptors and facilitates mitochondrial autophagy through phase separation outside mitochondria (35). p62 and its associated proteins form membraneless condensates that isolate dysfunctional mitochondria and tether them to autolysosomes for autophagic degradation via LC3 proteins (9, 37). Orphan nuclear receptor 77 (Nur77) is an interacting protein of p62. Upon mitochondrial damage, Nur77 moves from the nucleus to mitochondria and binds with both the C-terminal and N-terminal of p62, leading to phase separation (9, 38). This is known as a “head-to-head” or “tail-to-tail” interaction between Nur77 and p62 (Fig. 2A). The ubiquitin-carrying C-terminal ligand-binding domain (LBD) of Nur77 interacts with p62 in a manner resembling ubiquitination, forming the Nur77-LBD/p62 condensate that mediates mitochondrial autophagy through phase separation (9). Consequently, phase separation also occurs between the IDR of the Nur77 N-terminal and a Phox and Bem1p (PB1) domain of p62 (35). In contrast, the N-terminal interaction primarily functions to anchor mitochondria (9). Fluidity is a key determinant in the selection of impaired mitochondria by protein condensates (39). The IDR of the Nur77 N-terminal exhibits liquid-like properties, imparting similar fluidity to the condensate after phase separation (9).

Fig. 2.

(A) Phase separation involved in mitophagy. Phase separation between Nur77 and p62 promotes mitophagy. (B) Once leaked from damaged mitochondria, TFAM starts phase separation with LC3 to eliminate itself. (C) Phase separation between mitochondrial nucleoids leads to mitochondrial fission.

Another molecule that interacts with p62 to mediate mitochondrial phagocytosis through phase separation is NF-κB effector molecule (NEMO) (40). NEMO and p62 are corecruited to damaged mitochondria, where they are compressed into condensates along with the PINK1/Parkin complex via phase separation (40). Damaged mitochondria are phagocytosed by these phase-separated condensates (40).

TFAM has also reported as an autophagy receptor. Liu et al. demonstrated that TFAM interacts with the autophagic protein LC3 through autolysosomal pathway to remove leaked mtDNA via autophagy and thereby limits inflammation (41). TFAM contains a molecular “zip code” known as the LC3-interacting region (LIR) motif, which facilitates this binding (Fig. 2B) (41). Thus, TFAM-mtDNA leaked by autophagy mitochondria is eliminated.

Intramitochondrial phase separation also promotes damaged mitochondrial mitophagy. Damaged mitochondria contain the larger nucleoid condensates than normal mitochondria, due to the formation of gel-phase-like condensates generated by high membrane curvature (42). Enlarged nucleoid condensates trigger peripheral mitochondrial fission, which acts as a mechanism for sorting damaged nucleoids from intact ones to facilitate mitophagy (Fig. 2C) (42).

Phase Separation in Mitochondria Related Inflammation

The inflammation caused by mitochondrial dysfunction primarily results from RNA or DNA leakage. TFAM dysfunction and cyclic GMP-AMP synthase (cGAS) pathway were reported as the trigger of mitochondrial dysfunction-depended inflammation. TFAM dysfunction and the cyclic GMP-AMP synthase (cGAS) pathway have been identified as key triggers of mitochondrial dysfunction-dependent inflammation. As mentioned earlier, TFAM functions as an autophagy receptor by binding to LC3 to eliminate leaked mtDNA, a process essential for activating inflammatory signaling pathways (43). Regarding the cGAS–STING pathway, which is regulated by phase separation, it plays a crucial role in the mechanisms by which DNA damage induces inflammation (44). Researchers have shown that cGAS, localized to mitochondria, is activated upon the release of double-stranded DNA (dsDNA) and triggers an immune response through the stimulator of interferon genes (STING) pathway (44, 45). cGAS proteins exist in dimeric conformations, exhibiting an internal ladder-like structure (46). Two cGAS proteins bind to two sandwiched DNA molecules, promoting oligomerization and forming condensates that facilitate phase separation (47). It has been reported that DNA induces the formation of droplets with activated cGAS, and phase separation between cGAS and DNA is the primary mechanism through which the cGAS–STING pathway resists negative regulation, enabling efficient sensing of immunostimulatory DNA (45, 47, 48). On the other hand, TFAM has been shown to strongly stimulate long DNA sensing by cGAS (46). TFAM induces U-turns and structural bends in DNA, which precede the binding of cGAS to mtDNA, forming a ladder-like structure that stabilizes the condensates (46). The leakage of mtDNA into the cytosol activates inflammation via the cGAS–STING pathway, suggesting that phase separation between cGAS and mtDNA may trigger mitochondrial dysfunction, a critical element in kidney inflammation (49). When mtDNA leaks from mitochondria, it undergoes phase separation with cGAS, leading to the release of NF-κB and triggering immune activation (Fig. 3A).

Fig. 3.

(A) Phase separation affects cGAS–STING and MAVS pathway. Phase separation triggers cGAS–STING pathway to touch off immune activation. (B) Phase separation triggers MAVS pathway leads to inflammation.

Meanwhile, inflammation caused by mtRNA is possibly driven by the mitochondrial antiviral signaling protein (MAVS) pathway. RNA produced upon binding of external RNA viruses probably triggers the phase separation of tripartite motif-containing protein 25 (TRIM25), an E3 ubiquitin ligase, which may subsequently recruit retinoic acid-inducible gene I (RIG-I) and mediates the RIG-I-MAVS pathway (50). However, Tigano et al. discovered a previously uncharacterized retrograde signaling pathway. They reported that mtDNA breaks lead to the release of mtRNA into the cytosol, which activates RIG-I and ultimately triggers MAVS (51). MAVS plays a critical role in the nucleotide-binding oligomerization domain-like receptor pyrin domain-containing 3 (NLRP3) inflammasome, revealing that mtRNA may be a key target for inflammation treatment (52). While the exact mechanism remains unknown, it is hypothesized to resemble the TRIM25-RIG-I pathway, where phase separation enhances ubiquitination. If mtDNA breaks into mRNA, the MAVS pathway would be activated. RIG-I, in turn, binds with mtRNA through phase separation, leading to an increase in NLRP3 and, ultimately, inflammation (Fig. 3B). This event remains a potential occurrence that needs further investigation.

The Role of Mitochondria Related Phase Separation in Diseases

Mitochondrial dysfunction, including impairments in mitochondrial biogenesis and defects in mitophagy, contributes to the pathogenesis of mitochondrial diseases (53). Phase separation plays a pivotal role in regulating mitochondrial biosynthesis and autophagy. Consequently, dysregulated phase separation is a critical factor contributing to mitochondrial dysfunction and the onset of mitochondrial diseases. In this review, we summarize the mechanisms underlying age-related mitochondrial diseases associated with disordered phase separation in different mitochondrial compartments, offering new insights into potential therapeutic strategies for these conditions.

Age-related macular degeneration (AMD) is a progressive degenerative ocular disease that primarily affects the retinal pigment epithelium (RPE) (54, 55). RPE damage in AMD patients is predominantly caused by mitochondrial dysfunction (56, 57). Mitochondrial dysfunction triggers phase separation of p62, a key protein involved in mitochondrial phagocytosis, leading to mitochondrial rupture and accumulation of damaged mitochondria (10, 58). Other proteins containing reactive cysteine residues, similar to p62, may also participate in this process (10). Mitochondrial injury induces phase separation of apolipoprotein E (ApoE) via redox-sensitive cysteine interactions, resulting in the formation of biomolecular condensates that contribute to the nucleation of drusen (10). These ApoE condensates subsequently lead to epithelial cell damage and contribute to the pathogenesis of AMD (Fig. 4A) (10). Collectively, these findings support the involvement of mitochondrial damage and the aggregation of ApoE in the RPE of AMD patients.

Fig. 4.

Phase separation drives mitochondrial diseases. (A) Damaged mitochondria drive the redox state-sensitive, cysteine-mediated phase separation of ApoE, leading to AD and nucleated drusen in the RPE. (B) The incorrect position of MRGs caused by phase separation impairs mitochondrial respiration and leads to mitochondrial diseases. (C) Impaired NRF2 transcriptional activity under stress enlarges mt-nucleoids through phase separation, which inhibits mitochondrial respiration. Increased ROS causes chromatin damage leading to HGPS.

Alzheimer’s disease (AD), an age-related neurodegenerative disorder, is also associated with the p62-ApoE pathway (Fig. 4B) (59, 60). AD is primarily driven by genetic factors, with the APOE gene ε4 allele being the most significant genetic risk factor, which notably increases the risk of AD (61). This allele induces mitochondrial network fragmentation in a p62-dependent manner, suggesting that mitophagy dysfunction is essential in AD pathogenesis (61, 62).

The abnormal aggregation of mitochondrial genetic material in clusters is a key contributor to mitochondrial diseases. Many of these electron-dense granules have been observed to be closely opposed to membranes, providing further evidence for the role of membrane dynamics in maintaining a uniform distribution of MRGs along the mitochondria. This observation implies that positioning changes could be independent (25). They may not be related to alterations in the biophysical characteristics of RNA subcompartments (25). Inadequate positioning of mitochondrial genetic material can disrupt the proper synthesis of the respiratory chain and oxidative phosphorylation, ultimately leading to the development of mitochondrial diseases (63, 64).

Hutchinson–Gilford Progeria Syndrome (HGPS) is a rare, age-related disorder caused by mitochondrial abnormalities. It is characterized by multitissue dysfunction, affecting organs such as bone, muscle, skin, and the cardiovascular system (65–67). Approximately 70% of advanced HGPS patient cells have a subpopulation of mitochondria that are swollen, spherical in shape, and isolated from the surrounding mitochondrial network, compared to the typical tubular, elongated mitochondrial networks in control cells (19). This contrasts with the typical tubular and extended mitochondrial structure observed in control cells (19). Impaired NRF2 transcriptional activity under stress conditions leads to the expansion of mt-nucleoids via phase separation. Disrupted phase separation in these enlarged mt-nucleoids may result in elevated levels of reactive oxygen species (ROS) and heightened oxidative stress (68, 69). Excessive ROS accumulation is linked to the degradation of mitochondrial genetic material, potentially driving mitochondrial dysfunction and contributing to the pathogenesis of HGPS (Fig. 4C).

Conclusions and Future Perspectives

Phase separation plays an important role throughout the lifecycle of mitochondria, from their generation to their elimination. Despite its emerging significance, phase separation remains a relatively new area of research, and several key questions remain unresolved. One such question is the effect of mitochondrial temperature, which is elevated to nearly 50 °C due to cellular respiration, on condensate formation and mitochondrial function (70). Whether and how temperature regulates phase separation within mitochondria warrants further investigation.

Additionally, phase separation has been implicated in mediating cellular communication. For instance, cell surface receptors facilitate immune signaling through phase separation in multiple immune cells, such as T cells and B cells (50). However, immune cell function is highly dependent on the coordinated transitions between intracellular organelles, and communication between these organelles relies on membrane contact sites (MCS). Thus, understanding how phase separation coordinates interorganelle communication presents an intriguing avenue for future research.

Certain drugs have been identified that can specifically prevent the formation of condensates, offering potential therapeutic strategies for cancer treatment. For example, the anticancer adjuvant melatonin has been shown to inhibit phase separation in cancer cells by targeting the unstructured N-terminal domain, a segment of prion proteins. (71). In the context of mitochondrial dysfunction, drugs can target components, such as Nur77 or TFAM, which undergo phase separation to mediate dysfunction. Ikeda et al. reported that increased Ser349-phosphorylation in the p62^{S403E S407E} mutant enhances p62 phase separation, suggesting that compounds capable of inhibiting or activating p62 phosphorylation may serve as novel targets for treating mitophagy-related diseases (72, 73). Furthermore, the TFAM-LC3 autophagy pathway described above provides a new avenue for targeting inflammation associated with mtDNA accumulation, highlighting the role of TFAM in inflammatory diseases (41). As such, there is a pressing need for the development of drugs specifically targeting phase separation. However, further verification is required to assess the clinical efficacy of phase-separated targeted therapies.

Data, Materials, and Software Availability

There are no data underlying this work.