The distinct functions of RPE cells make them susceptible to oxidative stress, due to their high metabolism and exposure to high oxygen, oxidized POS and polyunsaturated fatty acids (PUFAs). In addition, environmental factors, such as visible or ultraviolet (UV) light exposure and cigarette smoking, also pose oxidative stress to the RPE cells. These could lead to elevated levels of ROS and reactive nitrogen species in the cells, which can modify and damage carbohydrates, membrane lipids, proteins and nucleic acids, and eventually lead to pathological consequences. RPE cells are equipped with enzymatic and non-enzymatic antioxidative systems to protect against oxidative stress. However, the antioxidative capabilities diminish with aging, which leads to diminished RPE functions. RPE structural and functional changes associated with aging have been reviewed by Bonilha (Bonilha, 2008) and are summarized below (Table 1; Figure 1): 1) Aged RPE from human donors showed loss of melanin granules on their apical surface and accumulation of age pigment lipofuscin. Accumulation of secondary lysosomes and residual bodies containing lipofuscin, has been observed in post-mitotic and intermitotic cells during aging, and has been used as a universal index for cellular senescence (Schmucker and Sachs, 2002; Kubasik-Juraniec et al., 2004; Morales et al., 2004). 2) Shortening of RPE microvilli, which could affect some RPE key functions, such as phagocytosis of shed POS (Boulton and Dayhaw-Barker, 2001). 3) BrM is an acellular and pentalaminar structure formed by the RPE and choroid. Increased thickness, lipid content and advanced glycation products have been reported in the aging BrM, likely due to the entrapment of proteins and lipids in the extracellular matrix (Zarbin, 2004; Harris et al., 2017). Functionally, the elasticity of human BrM-choroid complex decreases linearly with aging. 4) Drusen are extracellular deposits of biomaterials below the RPE along BrM. Drusen can be originated from blood and/or RPE and is a clinical hallmark for age-related macular degeneration (AMD) and other age-related diseases including Alzheimer’s disease (Bergen et al., 2019). Drusen are ubiquitous in people over 50 years old and are considered as part of normal aging. 5) RPE cell density/size changes during aging. An earlier study using normal human donors showed that RPE density in the fovea decreased significantly by 0.3% per year with increasing age. An increase in RPE cell size and multinucleation has been observed in mice (Chen et al., 2016). 6) Age-related accumulation of iron in RPE/choroid has been observed in rats, while its increase in the retina was modest (Chen et al., 2009a; Chen et al., 2009b). Exposure of ARPE-19 cells (a spontaneously arising RPE cell line derived from the normal eyes of a 19-year-old male) to increased iron markedly decreased their phagocytosis activity, suggesting that iron accumulation with age may impair the phagocytosis and lysosomal functions of the RPE. In human, phagocytosis level in RPE from donors showed a moderate decrease with aging (Inana et al., 2018). 7) RPE secretome changes during aging. Secretome includes soluble proteins and proteins secreted as part of extracellular vesicles which function as mediators of cell-to-cell signaling and modulate cell activities. The secreted proteins could be used as biomarkers of some diseases. It has been reported that RPE secreted over 1,000 proteins, many of which change significantly due to ROS accumulated during aging (Meyer et al., 2019).

Aging, together with genetic and environmental factors, cause RPE dysfunction and degeneration, which significantly contributes to age-related retinal diseases including AMD. The RPE changes involve RPE functional decline, morphological changes, epithelial-mesenchymal transition (EMT), senescence and cell death. AMD is a leading cause of blindness among the elderly population. Late AMD has two forms, “dry” and “wet” AMD. Geographic atrophy (GA) is the advanced form of dry AMD. It is featured by the irreversible loss of the RPE, photoreceptors and choriocapillaris, which eventually lead to vision loss. The features and phenotypes of RPE in AMD have been under extensive studies by numerous groups. Here we summarize the RPE changes in AMD (Table 1; Figure 1): 1) Structural changes of RPE can be visualized by spectral domain optical coherence tomography (SD-OCT), autofluorescence imaging, combined with physiopathological analysis under electromicroscopy. Large soft drusen and hyper- or hypopigmentation in the RPE were identified as key features preceding RPE cell dysfunction and late form of AMD (Wang et al., 2003). Lipofuscin aggregation has been observed almost exclusively in AMD eyes, implying more cellular senescence in AMD (Ach et al., 2015). In an independent longitudinal study, increased fundus autofluorescence was observed preceding GA development, suggesting the involvement of cellular senescence and subsequent cell death in atrophic AMD (Holz et al., 2001; Hwang et al., 2006). 2) RPE layer becomes highly variable and overall thicker in both GA and wet AMD eyes, which could be explained by hypertrophy or cellular senescence (Zanzottera et al., 2015). 3) Shedding, dissociating and sloughing RPE cells were observed in GA, suggesting death, transdifferentiation or emigration of RPE cells (Zanzottera et al., 2015). Apoptosis of RPE cells in AMD was also detected by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, the most widely used method to detect fragmented DNA in apoptotic cells or tissue samples (Dunaief et al., 2002). However, cells may become TUNEL positive during other types of cell death. Therefore, more studies are needed to confirm the RPE cell death mechanisms in AMD, especially during the process of GA. 4) RPE phagocytosis, while showing a moderate decline with age, was significantly reduced in AMD RPE, more than expected during normal aging (Inana et al., 2018) 5) Induced pluripotent stem cell (iPSC)-RPE cells from homozygous ARMS2/HTRA1 risk genotype showed significantly higher complement and inflammatory factors expression, and iPSC-RPEs from complement factor H (CFH) (Y402H) risk genotype showed reduced mitochondrial function and increased inflammation markers (Saini et al., 2017; Ebeling et al., 2021). 6) By analyzing the secretome from RPE cells of AMD and age-matched control donors, RPE cells were found to secrete a variety of extracellular matrix proteins, complement factors, and protease inhibitors that have been reported to be major constituents of drusen (An et al., 2007). AMD RPE cells secrete more galectin 3 binding protein, fibronectin, clusterin, matrix metalloproteinase -2 and pigment epithelium derived factor, but less secreted protein acidic and rich in cysteine than RPE cells from age-matched healthy donors. Overall, significant progress has been made regarding the morphological changes and pathogenic mechanisms of RPE in AMD. More mechanistic studies of RPE in AMD, including molecular mechanism, the involvement of EMT, cell senescence and different types of cell death, are still needed.

Numerous studies support the involvement of mitochondria in aging (Table 2). In the 1950s, Denham Harman proposed the free radical theory of aging, which he later expanded it to the mitochondrial free radical theory of aging. In this theory, free radicals produced by mitochondrial activities damage cellular components and lead to aging. Mitochondrial changes during aging include: 1) Morphological changes, such as abnormally rounded mitochondria (Shigenaga et al., 1994), decreased number of mitochondria (Yen et al., 1989), concurrent with decreases in mtDNA copy number and mitochondrial protein levels (Stocco et al., 1977). 2) Many studies have reported mitochondrial functional decline during aging, including lower oxidative capacity, reduced OXPHOS, decreased ATP production (Chistiakov et al., 2014). There is an average decline of 8% per decade in ATP producing capacity (Short et al., 2005). 3) Mitochondrial ROS significantly increases with aging and in aging-related diseases. Moderate ROS helps to maintain cellular energy production and to regulate mitochondrial protective signaling pathways, which increase lifespan in lower organisms (Giorgi et al., 2018). However, excess ROS are pathogenic and may induce cell degeneration or even cell death. For example, ROS and mitochondrial Ca²⁺ overload can cause the opening of the mitochondrial permeability transition pore, which leads to apoptosis (Rottenberg and Hoek, 2017). During normal aging, increased ROS level promotes oxidative damage to mitochondrial DNA, lipids, and proteins (Jarrett et al., 2008). 4) Impaired balance between mitochondrial fission and fusion is related to age-dependent decline in mitochondrial biogenesis. Mitochondrial fusion has beneficial effect to prolong lifespan by increasing bioenergetics efficiency, maintaining ATP production (Gomes et al., 2011; Rambold et al., 2011). Meanwhile, mitochondrial fission is associated with aging due to increased oxidative stress, mitochondrial depolarization, and reduced ATP production (Jheng et al., 2012). Fission maintains mitochondrial quality and integrity by involving in the selection of dysfunctional mitochondria. Mitophagy selectively removes defective mitochondria which have impaired oxidative capacity and declined integrity (Ding and Yin, 2012). With age, mitophagy was observed to decrease (Cavallini et al., 2007), which leads to an accumulation of damaged mitochondria, advanced oxidative stress, and increased apoptosis (Masiero and Sandri, 2010). 5) Mitochondrial density appears to decline gradually during aging, suggesting a decrease in mitochondrial biogenesis during aging, which could result from age-dependent reduction in levels of PGC-1α (Crane et al., 2010). 6) mtDNA damage, mutation and deletion during aging. mtDNA encodes essential components of OXPHOS and protein synthesis machinery (Falkenberg et al., 2007). Thus, oxidative stress-induced mtDNA damage and mutations impair either the assembly or the function of the respiratory chain which then trigger further accumulation of ROS which could be lethal (Hiona and Leeuwenburgh, 2008). Besides point mutations, deletions of mtDNA are detected at higher frequency in aged human and animal tissues (Nekhaeva et al., 2002; Kujoth et al., 2007). In addition, mtDNA abundance also declines with age in various tissues of human and rodent (Wallace, 2001; Miller et al., 2003; Welle et al., 2003). MtDNA containing unmethylated CpG dinucleotides and can trigger inflammation that aggravates tissue injury by activating toll-like receptor 9, inflammasomes, and the stimulator of interferon genes pathway (Nie et al., 2020). It has been reported to be pro-inflammatory in various diseases such as Alzheimer’s disease and heart failure (Wilkins et al., 2014; Nakayama and Otsu, 2018). Intracellular mtDNA has been shown to induce the secretion of inflammatory cytokines IL-6 and IL-8, and the priming of the NLR Family Pyrin Domain Containing 3 (NLRP3) inflammasome in RPE cells (Dib et al., 2015).

Only a few papers have reported the mitochondrial structural and functional changes during RPE aging in humans. Using electron microscopy, mitochondria in young RPE were found to be numerous, most bacillus-like shaped, rich in well preserved cristae, and orientated parallel with the apical-basal axis (Feher et al., 2006). In aged eyes, mitochondria of the RPE were decreased in number, variable in size, usually oval shaped, sometimes with disorganization of cristae (Figure 2). In another study, using isolated primary RPE cells from young (9–20)-, mid-age (48–60)-, and >60 (62–76)-year-old donors, some different morphological changes in mitochondria were observed. Mitochondria from the two younger groups were found to be numerous, regular in size, and with round or oval shapes (He et al., 2010). Cristae were distinctly visible and the outer membranes appeared to be intact. However, mitochondria from the >60 age group RPE cells were sparsely distributed in the cytoplasm, irregular in size, tubular in shape, larger, and with electron-dense matrices, less distinct cristae, and disrupted outer membranes. Length of the mitochondria in this group (Length/width ratio) was almost seven folds greater compared to the other ages. These mitochondrial abnormalities correlated with lower ATP levels, reduced mitochondrial membrane potential, decreased cytoplasmic Ca²⁺ concentration, and increased Ca²⁺sequestration in the mitochondria in cells with advanced aging (He et al., 2010).

There are several reports regarding age-related mitochondrial changes in animal models of RPE aging. Using monkeys up to 35 years old, Gouras et al. found decreased mitochondrial number and density, increased mitochondrial length and increased clustering of very elongated mitochondria during aging (Gouras et al., 2016). Using C57BL/6 mice (up to 22 months old) and rats (up to 26 months old), Wang et al. found an increased mtDNA damage in aged RPE and choroid (Wang et al., 2008). Yako et al. reported that in the RPE of 12-month-old C57BL/6J mice, mitochondrial number was decreased, but cristae width was increased compared to young mice (Yako et al., 2021). They also found that pharmacological suppression of mitochondrial fission improved POS phagocytosis, suggesting that mitochondrial dysfunction and fission in RPE could impede phagocytosis and cause retardation of the visual cycle. Metabolomics analysis in the RPE/choroid of young (6 weeks) and old (73 weeks) C57BL6/J mice identified 45 significantly changed metabolites, with most of which are involved in the mitochondrial metabolism, glucose metabolism and amino acid metabolism pathways (Wang Y et al., 2018). NAD and riboflavin (a precursor for flavin adenine dinucleotide) levels were reduced to less than half, but nicotinamide (substrate for NAD synthesis) and the substrates for mitochondrial metabolism such as pyruvate and dihydroxyacetone phosphate were accumulated to higher levels in the old mice, suggesting impaired mitochondrial energy metabolism during RPE/choroid aging. Taken together, the research on the mitochondrial changes in RPE aging is still limited, more systematic studies with even older groups, with focus on mitochondrial functions, mitochondrial dynamics and quality control, are still needed.

Several studies have reported RPE mitochondrial changes in AMD (Table 2). A reduction in mitochondrial number has been observed in AMD donor RPE compared to age-matched controls (Feher et al., 2006). A proteomic study revealed decreased expression of ATP synthase subunits in AMD RPE (Nordgaard et al., 2008). ATP synthase complexes participate in OXPHOS and maintain mitochondrial morphology and mitochondrial membrane potential. Decreased expression of ATP synthase subunits with AMD could lead to defects in these critical mitochondrial functions (Arselin et al., 2004; Bornhovd et al., 2006). Another change in the mitochondrial proteome is the decreased expression of the mitochondrial heat shock protein (mtHsp70) in AMD RPE (Nordgaard et al., 2008). mtHsp70 functions as a molecular chaperone that regulates the ATP-dependent import of nuclear-encoded proteins into the mitochondrial matrix (Nordgaard et al., 2006; Nordgaard et al., 2008; Kaarniranta et al., 2009). Therefore, decreased mtHsp70 levels could be detrimental to overall mitochondrial function and limit energy production in AMD RPE. mtDNA damages increase with age, more mtDNA damage in the macula of human AMD RPE has been detected compared to age-matched controls (Karunadharma et al., 2010). The extent of RPE mtDNA damage correlates with AMD severity. Damage to the mtDNA could be mitochondrial genome-wide, including the region encoding the subunit of ETC and the D-loop, the site of initiation for mtDNA transcription and replication of one mtDNA strand (Karunadharma et al., 2010; Terluk et al., 2015). Damage to either the D-loop or region encoding the ETC proteins could lead to negative functional outcomes for the mitochondria (Terluk et al., 2015). The role of mitochondria in AMD was also studied using transmitochondrial cybrids (Miceli and Jazwinski, 2005; Nashine et al., 2017; Nashine et al., 2020). Transmitochondrial cybrids were created by fusing mitochondrial DNA-deficient APRE-19 cell line with platelets isolated from either AMD patients or age-matched normal subjects. Therefore, the cybrids had identical nuclei but different mitochondria. AMD cybrids showed: 1) reduced cell viability, lower mtDNA copy numbers, and downregulated mitochondrial replication/transcription genes and antioxidant enzyme genes; and 2) elevated levels of genes related to apoptosis, autophagy and ER stress along with increased mtDNA fragmentation and higher susceptibility to amyloid-β-induced toxicity compared to control cybrids. These studies support an important role for mitochondria in AMD.

Mitochondrial functions have also been studied using in the cultured RPE or iPSC-RPE cells derived from AMD donors. Ferrington et al. found that mitochondria of RPE from AMD donors have significantly declined functions such as lower basal respiration, ATP production, and maximum respiration compared to age-matched controls (Ferrington et al., 2017). There is no difference in mtDNA content when comparing healthy and AMD donors. In another study by Golestaneh et al., accumulation of lipid droplets and glycogen granules, damaged mitochondria, and increased autophagosomes were observed in cultured AMD RPE cells (Golestaneh et al., 2017). Decreased NAD+ and Sirtuin 1 (SIRT1), increased PGC-1α acetylation (inactive form), lower AMPK activity, and overactive mammalian target of rapamycin (mTOR) pathway were also observed in AMD RPE cells, suggesting RPE metabolic dysregulation in AMD (Zhang et al., 2020). Compared with normal RPE, AMD RPE exhibit increased susceptibility to oxidative stress, produce higher ROS levels under stress conditions, and showed reduced mitochondrial activity with decreased ATP production. Of note, Ferrington et al.‘s study showed that AMD RPE are more resistant to acute oxidative stress (Ferrington et al., 2017). iPSC-RPE cells have been derived from AMD patients, and reduced mitochondrial function has been observed in iPSC-RPE cells from the CFH Y402H risk genotype RPE (Ebeling et al., 2021).

Although no AMD mouse models can recapitulate human AMD phenotypes due to their lack of macula, AMD mouse models can provide some insight into the mechanism of AMD. Here we present several mouse models with connection to mitochondria. RPE-specific knockout of mitochondrial antioxidant enzyme MnSOD (or SOD2) resulted in reduced RPE function with age (Brown et al., 2019a). Less electron dense and swollen mitochondria, disorganized mitochondrial cristae, reduced ATP production, and a compensatory increase in glycolytic metabolism have been observed in the SOD2 ^−/− mice, supporting that mitochondrial oxidative stress can lead to mitochondrial malfunction, RPE metabolic reprogramming and RPE dysfunction. In another model, global knockout of NRF2 and PGC-1α, the master regulators of antioxidant production and mitochondrial biogenesis, led to disturbed autophagy, an accumulation of drusen-like deposits, and the infiltration of Iba-1 positive immune cells mimicking clinical features of the dry AMD phenotype (Felszeghy et al., 2019). In the NRF2/PGC-1α ^−/− mice, a decline in viable mitochondria and higher mitochondrial damage were observed in the RPE cells. Damaged mitochondria were marked by Pink1/Parkin and autophagosomes with mitochondrial cargo. These data support that defective mitochondrial antioxidative system and biogenesis in the RPE could lead to AMD-like phenotypes (Wang K et al., 2018; Wang et al., 2019). For more information about RPE mitochondria in AMD, refer to reviews by Kaarniranta (Kaarniranta et al., 2020) and Blasiak (Blasiak et al., 2020).

Cellular senescence and cell death are important processes involved in aging and age-related diseases. Mitochondria have been established as important regulators of senescence and cell death. In this and next section, we summarize the role of mitochondria in RPE cellular senescence (Figure 3) and cell death (Figure 4).

Deregulation of mitochondrial homeostasis, shown by impaired mitochondrial biogenesis, metabolism and dynamics, has emerged as a hallmark of cellular senescence, which also drives the senescent phenotypes (Vasileiou et al., 2019). Defects in mitochondrial OXPHOS, reduced ATP production, and increased mitochondrial ROS production have been reported in cells undergoing senescence. In senescent cells, mitochondrial ROS can further aggravate cellular senescence by increasing telomere shortening, DNA damage and sustaining DNA damage response signaling pathway (Passos et al., 2010). In addition, mtDNA damage induced by ROS can impair OXPHOS function and further increase ROS release, forming a vicious cycle in aggravating senescence. Mitochondrial dynamics (including biogenesis, fusion, fission and mitophagy) are also linked to senescence. In aging endothelial cells, mitochondrial membrane potential, mitochondrial fusion and fission are reduced (Jendrach et al., 2005). In stress-induced senescent cells, highly elongated mitochondria with enhanced cristae structure have been reported (Yoon et al., 2006). Depletion of mitochondrial fission 1 protein (Fis1) caused sustained elongation of mitochondria and senescent phenotypes, whereas reintroduction of Fis1 protein restored mitochondrial fission and partially reversed the senescent phenotypes (Lee et al., 2007). Depletion of both Fis1 and mitochondrial fusion protein OPA1 resulted in mitochondrial fragmentation phenotypes and rescued senescence-associated changes. Intriguingly, sustained mitochondrial elongation is associated with reduced mitochondrial membrane potential, increased ROS level and DNA damage. These indicate that sustained mitochondrial elongation can trigger senescence-associated changes. However, in a cellular senescent model induced by cigarette smoke extract (CSE), more mitochondrial fragmentation was observed. In addition, mitochondrial fragmentation induction by knockdown of fusion proteins, OPA1 or MFN, increased mitochondrial ROS production and cellular senescence. For in vivo studies, inhibition of mitochondrial fission by deleting a mitochondrial fission protein DRP1 or maintenance of the fused mitochondrial network is necessary for longevity in yeast or C. elegans (Scheckhuber et al., 2007; Chaudhari and Kipreos, 2017). However, disruption of either mitochondrial fission or fusion significantly reduces medium lifespan in C. elegans, while in another study, promoting mitochondrial fission in midlife prolongs healthy lifespan of D. melanogaster (Rana et al., 2017; Byrne et al., 2019). Together, these conflicting data warrant more research to study the role of mitochondrial fusion or fission, or the imbalance of mitochondrial fusion/fission, in cellular senescence. The other mitochondrial quality control processes include mitochondrial biogenesis that propagates pre-existing pool of mitochondria, and mitophagy that eliminates malfunctional mitochondria. Mitochondrial biogenesis occurs constantly at basal levels and increases during cell renewal, development and stress conditions. Although in general mitochondrial biogenesis decreases during aging, expression of mitochondrial biogenesis regulators is upregulated in models of cellular senescence, which may reflect a compensatory response (Lee et al., 2002; Moiseeva et al., 2009). Mitophagy generally positively impacts lifespan and healthspan. Decreased mitophagy was observed during aging, while overexpression of Parkin and Pink1 extended lifespan in flies (Todd and Staveley, 2012; Rana et al., 2013; Bakula and Scheibye-Knudsen, 2020). Regarding mitophagy in senescence, mitophagy activities have been shown to be reduced in senescent cells in vitro and in vivo (Dalle Pezze et al., 2014; Garcia-Prat et al., 2016). This may be the consequence of reduced mitochondrial fission, autophagy and lysosomal activities, since mitochondrial fission is required for the separation of dysfunctional mitochondrial from the mitochondrial network, while lysosome is required for the terminal events of both autophagy and mitophagy.

Regarding the role of mitochondrial in RPE senescence, PGC-1α, a key protein for mitochondrial biogenesis and redox control, has been hypothesized to repress RPE senescence (Kaarniranta et al., 2018). We have recently reported that mitochondrial hyperfusion drives RPE senescence (Yu et al., 2020). We found that deletion of mitochondrial phosphatase phosphoglycerate mutase 5 (PGAM5) leads to accelerated RPE senescence in vitro and in vivo. Mechanistically, PGAM5 is required for mitochondrial fission through dephosphorylating DRP1. PGAM5 deletion leads to increased mitochondrial fusion and decreased mitochondrial turnover. As results, cellular ATP and ROS levels are elevated, mTOR and IRF/IFN-β signaling pathways are enhanced, leading to cellular senescence. Overexpression of DRP1-K38A mutant overexpression attenuated mitochondrial fission and elongated mitochondrial branches, which phenocopies PGAM5 ^−/− senescent phenotypes, while overexpression of DRP1-S37A mutant dramatically inhibited mitochondria fusion and rescued mTOR activation and senescence in PGAM5 ^−/− cells. These data support that mitochondrial dynamics can regulate signaling pathways linking to RPE senescence. In hydrogen peroxide (H₂O₂)-mediated premature RPE senescence model and natural passage-mediated replicative RPE senescence model, levels of ROS and mitochondrial membrane potential were increased, while coculture of RPE cells with embryonic stem cells reversed RPE senescence phenotypes in vitro (Wang et al., 2020). Cigarette smoke is one of the major risk factors for AMD. CSE can induce RPE senescence (Marazita et al., 2016). CSE increased superoxide while decreasing ATP production in RPE cells, suggesting an uncoupling of OXPHOS and mitochondrial dysfunction. Mice with intravitreal injection of CSE showed decreased TOM20 protein and ATP levels in the injected eyes (Cano et al., 2014). The exact function of mitochondria in CSE-induced RPE senescence is yet to be established.

Mitochondria also have significant involvement in different types of cell death. The role of mitochondria in apoptosis is well-established. Mitochondrial outer membrane permeabilization (MOMP) induced by B-cell lymphoma 2 (BCL-2) antagonist/killer (BAK) and BCL-2–associated X (BAX), causes the release of intermembrane space proteins including cytochrome c, second mitochondria-derived activator of caspase (SMAC) and Omi to the cytoplasm. Omi and SMAC promote caspase-9 activation by cleaving or binding to inhibitor of apoptosis proteins, while cytochrome c binds to apoptotic peptidase activating factor-1. These lead to apoptosome formation and caspase recruitment, caspase cascade activation and dismantling of the cellular contents (Vringer and Tait, 2019). Although apoptosis is generally believed to be immunotolerant, MOMP was recently shown to induce potent pro-inflammatory signaling. The role of mitochondria in necroptosis, a regulated form of necrosis, has been controversial (Marshall and Baines, 2014). Mitochondrial fragmentation has been observed in necroptosis (Wang et al., 2012). BAX and BAK function as the outer membrane component of the mitochondrial permeability pore (MPTP) and regulate MPTP opening through oligomerization during necrotic cell death (Karch et al., 2013; Karch et al., 2015). Numerous studies support the role of mitochondrial dysfunction, including mitochondrial ROS, activation of mitochondrial phosphatase PGAM5, and induction of mitochondrial permeability transition in necroptosis. However, recent studies suggest none of these are involved in necroptosis. Necroptosis was still functional in cells with mitochondria depleted by mitophagy (Tait et al., 2013). Further work is needed to confirm whether mitochondria are fully deleted in those cells and whether necroptosis is mitochondrial independent. During pyroptosis, rounded and fragmented mitochondria and mitochondrial outer membrane permeabilization were observed after the opening of gasdermin D (GSDMD)-dependent pores before the sudden rupture of the plasma membrane, and there is little evidence that mitochondria are involved in pyroptosis (de Vasconcelos et al., 2019). However, mitochondrial ROS was recently shown to mediate pyroptosis (Evavold et al., 2021). In that study, Ragulator-Rag complex controls mTORC1-dependent events to promote mitochondrial ROS, which operates downstream of GSDMD cleavage to promote GSDMD oligomerization and pore formation. The role of mitochondria in ferroptosis is still controversial (Gan, 2021). Diverse metabolic activities in mitochondria can facilitate ferroptosis, but mitochondria are also equipped with strong anti-ferroptosis defense systems. For example, ubiquinol (CoQH2) generated in mitochondria can defend against ferroptosis, but mitochondrial ROS and ATP can promote ferroptosis. Ferroptotic cells do not have typical necrotic morphological features, but mainly display mitochondrial shrinkage, increased mitochondrial membrane density and reduced mitochondrial cristae (Yagoda et al., 2007; Yang and Stockwell, 2008). However, cells lacking mitochondria are as sensitive to ferroptosis as parental cells with intact mitochondria, supporting the existence of mitochondrial independent ferroptosis (Dixon et al., 2012). Ferroptosis Suppressor Protein 1 (FSP1) was recently shown to exert strong anti-ferroptosis function on plasma membrane in the absence of glutathione peroxidase 4 (GPX4), supporting the existence of mitochondrial-independent ferroptosis (Bersuker et al., 2019; Doll et al., 2019).

Different kinds of RPE cell death have been reported depending on the types of stressors (Hanus et al., 2015; Tong and Wang, 2020). Mitochondrial changes have been associated with RPE cell death. N-retinylidene-N-retinyl-ethanolamine (A2E) is a by-product of the visual cycle formed by the reaction of two trans-retinal molecules with phosphatidyl-ethanolamine (Sparrow et al., 2003). It is a major fluorophore in lipofuscin and is accumulated in aging RPE (Sparrow and Boulton, 2005). A2E has been shown to induce RPE apoptosis by inhibiting cytochrome C oxidase, associated with declined mitochondrial activity and release of cytochrome c (Shaban et al., 2001). Blue light could potentially produce retinal toxicity leading to the development of degenerative eye diseases, such as AMD (Contin et al., 2013; Kuse et al., 2014). Blue light could lead to RPE cell necrosis or apoptosis, possibly depending on the intensity and duration of the light (Pang et al., 1999; Seko et al., 2001). In A2E-loaded RPE cells, blue light reduced cell viability, associated with decreased ATP and increased ROS levels (Marie et al., 2018; Alaimo et al., 2019). Blue light caused the imbalance of mitochondrial fusion/fission towards mitochondrial fragmentation in both non-loaded and A2E-loaded cells, which is correlated with decreased OPA1 and increased DRP1 expression levels. It was also showed that A2E treated cells led to an increase in both mitochondrial fusion and fission. Blue light induced mitochondrial fragmentation in A2E-loaded cells, consistent with their increased propensity to die. Exposure to UV radiation can induce oxidative stress and is associated with ocular pathologies (Marchitti et al., 2011). UV has been shown to induce RPE apoptosis (Yao et al., 2013). In UV-treated RPE cells, upregulation of cytochrome c protein and increased ROS production were observed, suggesting that mitochondrial membrane integrity was compromised (Balaiya et al., 2010). Bidirectional movement of short and elongated mitochondria was noted in control RPE cells. Immediately after UV irradiation, mitochondria became shorter and no longer moving, with very few branched mitochondria remaining within the cells (Bantseev and Youn, 2006; Youn et al., 2010). Loss of mitochondrial membrane potential and the number of mitochondria were also reported in UV-treated RPE cells (Youn et al., 2007; Hsieh et al., 2018). Menadione (or Vitamin K3) induces RPE apoptosis, associated with mitochondrial depolarization and cytochrome C release (Zhang et al., 2003). Exposure of RPE cells to a lethal dose of H₂O₂ (1 mM) has been shown to induce BAX translocation to the mitochondria and the release of apoptosis-inducing factor from the mitochondria (Ho et al., 2006).

Our lab has shown that H₂O₂ or tert-butyl hydroperoxide (tBHP) treatment induces RPE necrosis, which can be prevented by necrosis inhibitors necrostatins but not caspase inhibitor z-VAD (Hanus et al., 2013). Fragmentation and degeneration of mitochondrial network were observed in the treated cells. Sodium iodate (NaIO₃) injection has been extensively used as a pre-clinical model of RPE dystrophy and GA (Wang J., 2013). Similar necrotic phenotypes were observed in NaIO₃-treated RPE cells, which was associated with fragmented and clustered mitochondrial network in the perinuclear region (Hanus et al., 2016). Auranofin, an inhibitor of redox proteins TrxR1 and TrxR2, induced pyroptosis in RPE cells, which was repressed by NLRP3 and Caspase-1 inhibitors (Yumnamcha et al., 2019). In these cells, reduced ATP production and mitochondrial membrane potential, increased ROS, accumulation of damaged, fragmented and vesiculated mitochondria, and mitophagic flux to lysosomes, were observed. NaIO₃ and high glucose have also been shown to induce RPE ferroptosis, which can be repressed by ferroptosis inhibitors (Liu et al., 2021; Tang et al., 2022). In an independent study, high glucose promoted RPE apoptosis and inhibited cell proliferation and mitophagy by inactivating ROS/Pink1/Parkin signal pathway (Zhang et al., 2019). Reduced mitochondrial membrane potential, increased ROS production were observed in NaIO₃-treated cells, and reduced mitochondrial size and ridge, disrupted mitochondrial membrane were observed in high glucose-treated cells. Further study indicated that high glucose increased the expression of thioredoxin-interacting protein (TXNIP), which is associated with mitochondrial membrane depolarization, fragmentation and mitophagic flux to lysosomes. Elongated mitochondrial network was observed under low glucose, while fragmented mitochondria were observed in high glucose. TXNIP knockdown by shRNA prevented mitochondrial fragmentation and mitophagic flux under high glucose (Devi et al., 2019). Taken together, different stressors could induce different types of RPE cell death, mitochondrial dysfunction including reduced mitochondrial membrane potential, decreased ATP and increased ROS levels, and in many cases mitochondrial fragmentation and/or mitochondrial fission/fusion imbalance, were generally observed, supporting a role for mitochondria in RPE cell death.

Given the critical role of mitochondrial function and dynamics in senescence and cell death, extensive efforts are being made to target mitochondria for RPE aging and age-related diseases, especially AMD. Here we summarize the current preclinical experiments and clinical studies on this topic (Table 3).

1) Humanin is a mitochondrial-derived peptide with cytoprotective function in different disease models. Using cybrid containing AMD mitochondria, Nashine et al. found that humanin has a pivotal role in protecting cells with AMD mitochondria, reducing pro-apoptosis gene expression and increasing protection against amyloid-β-induced damage (Nashine et al., 2017). Mechanistically, humanin acts via both intracellular (BAX) and extracellular (gp130) pathways and prevents the loss of AMD mitochondria. In an independent study, humanin was shown to protect oxidative-stress induced RPE cell death and senescence (Sreekumar et al., 2016). Humanin treatment prevented oxidative stress-induced decrease in mitochondrial bioenergetics, increased mtDNA copy number and upregulated the expression of mitochondrial transcription factor A, a key biogenesis regulator protein. These studies suggest the potential for humanin therapy for prevention of retinal degeneration, including AMD.

2) Resveratrol is a phytoalexin synthesized by numerous plants including vines with strong antioxidative properties. Resveratrol improved cell viability and decreased ROS levels in AMD cybrid model and stimulated mitochondrial bioenergetics in RPE cells (King et al., 2005; Sheu et al., 2010; Sheu et al., 2013; Nashine et al., 2020; Neal et al., 2020). It also induced autophagy, pro-survival and specific anti-inflammatory response in RPE cells (Josifovska et al., 2020). In addition, resveratrol suppressed choroidal neovascularization in an animal model of wet AMD (Nagai et al., 2014). Other studies showed that resveratrol activates SIRT1, a key regulator of cellular senescence, aging and longevity (Borra et al., 2005). Based on these, resveratrol has great potential as therapeutic targets for RPE aging and AMD.

3) Chrysoeriol is a flavonoid compound that is commonly found in plants of the genus Perilla frutescens. This compound possesses several health-beneficial properties, including antioxidant (Mishra et al., 2003; Nickavar et al., 2016), anti-inflammatory (Choi et al., 2005) and anti-tumor activities (Min et al., 2020). Chrysoeriol protected H₂O₂-induced RPE cell death by diminishing mitochondrial dysfunction, preventing ROS accumulation and enhancing the expression of anti-oxidative genes including NRF2 (Kim et al., 2021a). It attenuated oxidative stress-induced MMP loss and upregulated mitochondrial related gene expression, including OXPHOS genes, mitochondrial process genes, and mtDNA replication and transcription genes. It also significantly increased TOM20, MFN2, and OPA1 expression and decreased H₂O₂-induced DRP1 expression and phosphorylation (at Ser 616 position). Further work is needed to directly test the role of chrysoeriol in mitochondrial biogenesis and quality control, as well as RPE aging and degeneration in vivo.

4) Necrostatins are inhibitors of necroptosis (Degterev et al., 2005). Necrostatin-1 (Nec-1) is the most common inhibitor of necrosis that targets receptor Interacting Serine/Threonine Kinase 1 (RIPK1); Nec-5 is the necrosis inhibitor that inhibits RIPK1 indirectly (Wang et al., 2007); while Nec-7 targets RIPK1-independent necrosis pathways (Zheng et al., 2008). Nec-1 has been shown to block necroptosis and ameliorate inflammatory response in multiple disease models (Cao and Mu, 2021). Our lab has shown that necrostatins protect oxidative stress-induced RPE cell death in vitro and in vivo (Hanus et al., 2015; Hanus et al., 2016). Other studies showed that Nec-1 can recover mitochondrial dysfunction and reduce ROS production in response to necroptosis inducer TNF-α or acetaminophen (Ye et al., 2012; Takemoto et al., 2014). Based on these, necrotstatins could have potential in regulating mitochondrial function and RPE degeneration, which awaits future work to confirm.

5) PGC-1α enhancers. PGC1-α is a critical regulator of mitochondrial biogenesis and redox control. PU-91, an FDA-approved mitochondrion-stabilizing drug, was shown to improve cell survival, mitochondrial health and anti-oxidative potential by upregulating PGC-1α in AMD cybrid model (Nashine et al., 2019). It increases mitochondrial DNA copy number, upregulates the genes involved in mitochondrial biogenesis pathway including PGC-1α, NRF-1, NRF-2, peroxisome proliferator-activated receptors (PPAR)-α, and PPAR-γ, increases mitochondrial membrane potential, decreases mitochondrial superoxides levels, upregulates SOD2 expression level and increases the production of mitochondrial derived peptides. Triphenylphosphonium (TPP) is a well-known mitochondrial targeting moiety. Vitamin B3 (niacin) is a powerful antioxidant with lipid lowering functions (Ganji et al., 2009; Boden et al., 2014). TPP-conjugated Niacin (TPP-Niacin) has been shown to improve cell viability, reduce ROS generation, and increase the antioxidant enzymes in H₂O₂-treated ARPE-19 cells (Kim et al., 2021b). It ameliorated H₂O₂-induced mitochondrial dysfunction and mitochondrial membrane potential reduction. It also markedly enhanced the expression of transcription factors (PGC-1α and NRF2) and antioxidant-associated genes (especially heme oxygenase-1 and NAD(P)H Quinone Dehydrogenase 1). ZLN005, a selective PGC-1α transcriptional regulator, protected RPE from cytotoxic oxidative damage (Satish et al., 2018). ZLN005-treated ARPE-19 cells showed robust upregulation of PGC-1α and its associated transcription factors, antioxidant enzymes, and mitochondrial genes, and enhanced mitochondrial function shown by increasing basal and maximal respiration rates, and spare respiratory capacity. In addition, ZLN005 protected ARPE-19 cells from cell death caused by H₂O₂, oxidized low-density lipoprotein, and NaIO₃ without any cytotoxicity under basal conditions. ZLN005 protection effect is PGC-1α-dependent as it was lost in PGC-1α-silenced cells. Taken together, PGC-1α regulators, including PU-91, TPP-Niacin and ZLN005, could serve as novel therapeutic agents for RPE degeneration.

6) Autophagy boosters (AMPK activators). As autophagy (including mitophagy) has significant implications in aging and age-related diseases, autophagy regulators have been tested in RPE cells for protective response to oxidative stress. Metformin is the first-line anti-type 2 diabetes drug and has been known to stimulate autophagy in many tissues. In RPE cells, metformin conferred protection against H₂O₂-induced oxidative damage by activating AMPK pathway (Zhao et al., 2020). It also protected photoreceptors from light damage, delayed inherited retinal degeneration, and protected RPE from NaIO₃-induced injury in vivo (Xu L et al., 2018). The protection was associated with decreased oxidative stress, decreased DNA damage, and increased mitochondrial energy production. A retrospective study indicates that metformin use is associated with decreased odds of developing AMD (Brown et al., 2019b). iPSC-RPE cells derived from AMD patients have been established to test the efficacy of drugs in AMD. Three AMPK activators AICAR (5-Aminoimidazole-4-carboxamide ribonucleotide), metformin, trehalose, have been tested for maintaining mitochondrial function in AMD iPSC-RPE cells (Ebeling et al., 2022). Rapamycin is a drug used to prevent organ transplant rejection. It inhibits mTOR and activates autophagy. Increased mTORC activity by RPE-specific deletion of mTOR suppressor tuberous sclerosis 1 led to RPE degeneration (Huang et al., 2019; Go et al., 2020). Inhibition of mTORC1 by rapamycin partially rescued RPE degeneration, supporting the potential therapeutic role for rapamycin in RPE degeneration.

7) Mitochondrial metabolism regulators. Reductive carboxylation is a major metabolic pathway in RPE cells (Du et al., 2016). It regulates cell viability, redox balance, mitochondrial function, and response to oxidative stress. Nicotinamide mononucleotide (NMN) is a key intermediate of NAD⁺, which is required to support reductive carboxylation and ATP production. NAD⁺ decreases with age. Oxidative stress depletes NAD⁺, and supplementation with NMN completely prevented RPE cell death induced by H₂O₂. Using RPE cells from AMD and control donors, only RPE cells from AMD donors show improvements in mitochondrial functions, including basal respiration, maximal respiration, spare respiratory capacity, and ATP production, after NMN treatment (Ebeling et al., 2020). Of note, NMN has been shown to mitigate age-related physiology decline in mice and is being pursued as anti-aging molecule in humans (Mills et al., 2016; Shade, 2020). These support a potential role of NMN in RPE aging and AMD.

8) Elamipretide is a mitochondria-targeted tetrapeptide that has been evaluated in mitochondrial diseases including primary mitochondrial myopathy and Barth syndrome (Sabbah, 2021). It acts by stabilizing cardiolipin and therefore increasing cellular ATP production and reducing mitochondrial ROS (Birk et al., 2014; Nickel et al., 2014; Szeto, 2014). In RPE-specific SOD2 knockout mice, daily topical Elamipretide treatment led to the prevention of RPE cell size increase, suggesting reduced RPE cell death and RPE senescence. Based on its mechanisms of action, phase 1 clinical trial has been conducted to evaluate its potential in dry AMD and noncentral GA after daily subcutaneous injection (Mettu et al., 2022). It appeared to be well tolerated without serious adverse effects. Phase 2A clinical trial is underway.

9) α-Lipoic acid (LA) is a mitochondria-targeted antioxidant and mitochondrial nutrient (Packer et al., 1997; Liu and Ames, 2005). Jia et al. reported that LA protected against acute acrolein (a toxicant present in cigarette smoke)-induced RPE cell death and mitochondrial membrane potential decrease. It also inhibited acrolein-induced generation of intracellular oxidants, prevented the intracellular SOD decrease, protected mitochondrial complex I, II, and III activity and increased intracellular total antioxidant power in RPE cells (Jia et al., 2007).

10) Melatonin protects mitochondria by scavenging ROS, inhibiting the MPTP, and activating uncoupling proteins. Thus, melatonin maintains the optimal mitochondrial membrane potential, preserves mitochondrial functions, and regulates mitochondrial biogenesis and dynamics. In most cases, melatonin reduces mitochondrial fission and elevates their fusion. It also has been found to promote mitophagy and improve homeostasis of mitochondria (Tan et al., 2016). In the retina, melatonin is released mainly by photoreceptor cells but can be also produced by other cell types in pathological conditions (Sakamoto et al., 2004). It has been reported that melatonin protects human RPE cells against the cytotoxic effects of H₂O₂ (Rosen et al., 2012). It was also shown to be effective in the protection of mtDNA of ARPE-19 cells against H₂O₂-induced damage (Liang et al., 2004). However, high exogenous concentrations of melatonin increase light-induced damage to photoreceptor (Tan et al., 2003). The antioxidant effect of melatonin may indicate its protective role in AMD.

11) Plastoquinonyl-decyl-triphenylphosphonium (SkQ1) is another mitochondrial targeted antioxidant. It has been reported that the treatment with SkQ1 (250 nmol/kg body weight) during the period of active disease progression (from 12 to 18 months of age) significantly prevented the progression of retinopathy and suppressed atrophic changes in the RPE cells in the senescence-accelerated OXYS rats (Muraleva et al., 2014; Muraleva et al., 2019; Telegina et al., 2020).

RPE cells are critical for the metabolism and homeostasis of retina. Due to high metabolism, high exposure to light, oxidized POS and PUFAs, RPE cells are vulnerable to oxidative stress and other relevant stresses which make them more susceptible to aging and age-related disease. Mitochondria are the powerhouse of cells and can be a major source of cellular ROS that contribute to mtDNA damage, cell death, senescence, and age-related diseases. Mitochondria directly participate in cell death and senescence processes, and can undergo dynamic changes including fission/fusion, biogenesis and mitophagy for quality control in response to stresses. In this minireview, we described the RPE changes during aging and in AMD. The role of mitochondria in RPE aging and AMD was also discussed. Particularly, the involvement of mitochondria in RPE cellular senescence and death, two processes critical for RPE degeneration, and the current translational approaches to prevent RPE aging and degeneration, were summarized. Changes in the mitochondria of RPE, including mtDNA deletion and mutation, decreased ATP production, mitochondrial fission/fusion imbalance, decreased mitochondrial biogenesis and mitophagy and et al., were commonly observed during RPE aging and degeneration. More longitudinal studies, especially in vivo studies, are required to confirm some of the findings, since mitochondria could undergo dynamic changes during aging and in response to stresses, with adaptive response in the short term and pathological response in the long term. Based on the current research, although RPE senescence and cell death are involved in RPE aging and degeneration, the extent of their contribution and whether we can target RPE senescence and/or cell death for RPE aging and/or degeneration are still a matter of debate. Elimination of senescence using a senolytic approach increases mouse lifespan and has shown promise in human trials (Xu M et al., 2018; Justice et al., 2019). If RPE senescence is proved to be the major mechanism for RPE aging and degeneration, senolytic approach could be used to treat these conditions. Restoring mitochondrial function for preventing RPE aging and degeneration is an exciting idea, and many compounds listed above have shown promise in preclinical and clinical models. However, the understanding of mitochondria, especially mitochondrial dynamics and quality control, in RPE cell death and senescence, as well as RPE aging and degeneration, is still incomplete and awaits future studies. Stringent and new technologies, including genetic functional study, genetic lineage tracing, and single cell multi-omics studies, would be powered to answer these questions. New drug targets and drug candidates, more preclinical and clinical studies are also needed to test their efficacy and safety for RPE degeneration and AMD. Work on AMD-derived iPSC-RPE cells has shown considerable variability in drug response across patient cell lines (Ebeling et al., 2022). Therefore, a personalized medicine approach, including stratifying patients based on genotyping and more clinically relevant features, is needed in the future.