Mitochondrial oxidative stress, cellular damages and stem cell aging in premature aging models with complex II electron transport defect

Abstract

Mitochondria which are the major intracellular reactive oxygen species (ROS) sources produce especially superoxide anion (O₂^•−) as a byproduct of energy production. It has been well known that O₂^•− is converted from oxygen (O₂) and is overproduced by excessive electron leakage from the mitochondrial electron transport chain (ETC), mainly complexes I and III. However we have previously reported that several point mutations (specifically G71E in C. elegans, I71E in Drosophila and V69E in mouse) in succinate dehydrogenase C subunit (SDHC) of complex II cause mitochondrial electron transport defect leading to O₂^•− overproduction from mitochondria. These mutations can cause endogenous oxidative stress resulting in tumorigenesis and apoptosis as well as premature death. Recently, we have also demonstrated that premature aging of hematopoietic stem cell with a mutation in SDHC is developed after the growth phase and normal development. Here, we review cellular damages by complex II electron transport defect-induced endogenous oxidative stress in premature aging models.

Mitochondrial Complex II Functions and Diseases

Energy production in aerobic organisms is the result of glycolysis, the citric acid (tricarboxylic acid, TCA) cycle and electron transport chain (ETC).⁽¹⁾ The ETC is composed of four membrane-bound complexes within the mitochondrial inner membrane that forms the respiratory chain. Its main purpose is to transfer electrons from electron carriers [like nicotinamide adenine dinucleotide (NADH) and flavin adenine dinucleotide (FADH₂)] to molecular oxygen (O₂), which is the final electron acceptor, while also helping to generate a proton gradient that drives ATP synthesis.^(2,3) The eukaryotic mitochondrial ETC is composed of about 80 core subunits and requires about 100 additional proteins for its assembly with accessory subunits.⁽⁴⁾ Mitochondrial complex II activity has a crucial role for energy metabolisms of both the TCA cycle and ETC in mitochondria. Complex II contains the TCA cycle enzyme succinate dehydrogenase (SDH), which is composed of the flavin protein (Fp, SDHA), the iron-sulfur protein (Ip, SDHB) and two other subunits (a large subunit of cytochrome b, SDHC and a small subunit of cytochrome b, SDHD), and catalyzes electron transport from succinate (a substrate for complex II that stimulates complex II activity) to ubiquinone (CoQ).^(5–7) The SDHB subunit contains three iron-sulfur centers that are speculated to be important for mediating electron transfer.^(5–7) The SDHC and SDHD subunits containing a heme are anchored to mitochondrial inner membrane and involved in catalyzing FADH₂ produced by SDH activity (Fig. 1).

Fig. 1.

Mitochondrial ROS production by complex II electron transport defect on aging.

In humans, SDHA (70 kDa, 664 amino acids) and SDHB (27 kDa, 280 amino acids) are probably hydrophilic and are involved in substrate binding and oxidation. SDHC (15 kDa, 169 amino acids) and SDHD (12 kDa, 159 amino acids) subunits are hydrophobic and provide the membrane anchor and binding site of ubiquinone for the electron transfer.^(5–7) The SDHA gene lies on the short arm of chromosome 5 (5p15) and is composed of 15 exons within a genomic region of 38.4 kb. The SDHB (35.4 kb, 8 exons) and SDHC (50.3 kb, 6 exons) genes are located on the short and long arms of chromosome 1 (1p36 and 1q23.3), respectively. The SDHD gene is located on 11q23.1, spans 8.9 kb and contains four exons.

Homozygous germline mutations affecting the SDHA gene cause Leigh syndrome, a sub-acute necrotizing encephalomyelopathy during infancy.^(8,9) SDHB, SDHC, and SDHD heterozygous mutations cause a genetic predisposition to non-chromaffin palagamgliomas (PGLs) and adrenal/extra-adrenal phaeochromocytomas (PHEOs) called ‘PGL/PHEO syndrome’.^(10–12) PGLs are usually benign and slow-growing tumors of the parasympathetic ganglia with an incidence of roughly 1:30,000–1:100,000 in the general population. They are more frequently located in the head and neck region (HNPGLs) at the carotid bifurcation (carotid body tumor), along the vagal nerve, in the jugular foramen and in the middle ear space. Following the discovery of SDHD (OMIM ID: 602690) as the gene responsible for PGL1 in familial HNPGLs, it was subsequently recognized that two other subunits of this mitochondrial enzyme, SDHC (PGL3, OMIM ID: 602413) and SDHB (PGL4, OMIM ID: 185470) were associated with heritable PHEO and/or PGL.^(10–12) SDHB, SDHC, and SDHD gene mutations are responsible for 6% and 9% of sporadic PGLs and PHEOs, respectively.⁽¹³⁾ They also constitute 29% of pediatric cases, 38% of malignant tumors and more than 80% of familial aggregations of PHEO and PGL.⁽¹⁴⁾

The inactivation of SDH activity resulting in succinate accumulation leads to inhibit prolyl-hydroxylation of hypoxia inducible factors (HIF)-1alpha (HIF-1a), which is an essential step for its degradation through the complex VHL-ElonginBC-Cul2.^(15,16) We have hypothesized that high succinate concentration outside normal physiological range induce cellular responses under a hypoxic condition and accumulate histone methylation though inhibited demethylase activities for epigenetic regulation in nucleus.^(17–19) Collectively, SDH defects could trigger changes in not only energy metabolisms but also cellular responses such as cell differentiation and proliferation under hypoxic conditions (Fig. 2).⁽²⁰⁾

Fig. 2.

Mitochondrial ROS impacts on aging.

Mitochondrial Oxidative Stress with Complex II Electron Transport Defect

The ETC or oxidative phosphorylation system (OXPHOS) is located within the mitochondrial inner membrane and is the major endogenous source of reactive oxygen species (ROS) such as superoxide anion (O₂^•−), hydrogen peroxide (H₂O₂), and hydroxyl radicals (^•OH).⁽²¹⁾ It has been estimated that generation of O₂^•− and its catalyzed product H₂O₂ through the catalyst action of superoxide dismutases (SODs; Cu/Zn-SOD, Mn-SOD) may constitute as much as 1–2% of total electron flow, although others have placed this value at 0.1% (Fig. 1).^(22,23)

Oxygen is initially converted to O₂^•− by electrons leaked from mainly complex I and III.^(24–27) Complex III generates O₂^•− by the auto-oxidation of ubisemiquinone also known as semiquinone, formed during the Q cycle.⁽²⁸⁾ O₂^•− can be generated at two different sites, Q_o and Q_i, within complex III. The Q_o site releases O₂^•− into the intermembrane space, whereas the Q_i site releases O₂^•− into the matrix. Most of the O₂^•− generated by complex III is generated at the Q_o site.⁽²⁹⁾ In contrast, O₂^•− generated by complex I is probably released into the matrix. The mechanisms of O₂^•− generation in complex I are poorly understood, controversial and dependent on experimental conditions. Furthermore, several studies have reported that exposing cells or tissues to hypoxia leads to an increase in mitochondrial O₂^•− production, and that it is required for the cellular response to hypoxia (Fig. 2).^(30–32) Another study has revealed a much higher rate of O₂^•− production during reverse electron transport from succinate to NAD⁺ than during forward electron transport.⁽³³⁾ On the other hand, it has been shown that O₂^•− is produced from complex II in Saccharomyces cerevisiae with genetic background resulting in a compromised complex II and in Ascaris suum.^(34,35)

We have also demonstrated that O₂^•− is also produced from complex II with a SDHC mutation in Caenorhabditis elegans mev-1, isolated an oxygen hyper-sensitive mutant (Fig. 1).^(36–38) In comparisons of SDHC, the amino acid sequences of Escherichia coli K-12 (GenBank ID: AAA23893.1), Caenorhabditis elegans (GenBank ID: AAA20081.1), Drosophila melanogaster (fruit fly) (GenBank ID: AAF54602.2), Mus musculus (house mouse) (GenBank ID: AAH05779.1), Sus scrofa (pig) (UniProtKB/Swiss-Prot ID: D0VWV4.2), and Homo sapiens (human) (GenBank ID: AAC27993.1) for the gene named Succinate dehydrogenase C (SdhC), cytochrome b large (CybL), CYT-1, or SDHC subunit are retrieved from Entrez Nucleotide of National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/protein).

Consensus patterns in the PROSITE database [64]: R-P-[LIVMT]-x(3)-[LIVM]-x(6)-[LIVMWPK]-x(4)-S-x(2)-H-R-x-[ST], which is a putative a heme ligand “Succinate dehydrogenase cytochrome b subunit signature 1” (IPR018495, SDH_CYT_1, PS01000) and H-x(3)-[GA]-[LIVMT]-R-[HF]-[LIVMF]-x-[FYWM]-D-x-[GVA], which is a putative heme ligand “Succinate dehydrogenase cytochrome b subunit signature 2” (IPR018495, SDH_CYT_2, PS01001) are highly conserved among animals.^(39–50) The SDH_CYT_1 domain sequence is positively related to the activity of succinate-ubiquinone oxidoreductase as a ubiquinone-binding site.^{(42,43,46–50)}

In humans, it has been reported that the SDHC gene mutation caused paraganglioma type 3 (PGL3).⁽¹¹⁾ Since that time, several type mutations of SDHC have been found in paragangliomas, pheochromocytomas, and gastrointestinal stromas.⁽⁵¹⁾ Most of the SDHC gene mutations resulting in amino acid missense, flame shift, nonsense or deletion mutations destroy or delete the quinone-binding site (from 50 to 74 amino acid sequence) or heme-binding site (from 127 to 140 amino acid sequence) in the succinate-ubiquinone oxidoreductase main activity region.⁽⁵¹⁾ Therefore, the mutations in the quinone or heme-binding motifs including some nucleophilic amino acids and basic amino acids could cause an electron transport defect leading to ROS overproduction for tumorigenesis.

The ETC of C. elegans is composed of about 70 nuclear and 12 mitochondrial gene products. It closely parallels its mammalian counterpart in its metabolism and structure. C. elegans mitochondrial DNA (mtDNA) is similar in size and gene content to that of humans.^(52,53) The abnormal methyl viologen sensitivity C. elegans (mev) mutants have been isolated based upon its hypersensitivity to the ROS-generating chemical methyl viologen.⁽³⁶⁾ In addition to its methyl viologen hypersensitivity, mev-1 mutants are oxygen hypersensitive with respect to both development and aging.⁽³⁶⁾ The mev-1 mutation (kn-1), which results in an amino acid substitution at the 71st position from glycine to glutamate (G71E) in the putative ubiquinone-binding motif of SDH cytochrome b large subunit, has been identified as residing in the cyt-1 (a human SDHC gene homologue) gene (Fig. 1).⁽³⁷⁾ The mutation results in a greater than 80% reduction in complex II–III oxidoreductase activity in the mitochondrial membrane fraction.⁽⁵⁴⁾ Using separate assays, it is possible to quantify specifically both SDH activity and complex II–III electron transport activity.⁽³⁷⁾ The SDH activity in the mev-1 mitochondrial fraction was experimentally identical to that of wild-type. However, it dramatically compromised the ability of complex II to participate in electron transport.

The biochemical pathologies of mev-1 include elevated ROS. Specificity, O₂^•− levels in both intact mitochondria and sub-mitochondrial particles are approximately two times greater in mev-1 mutants as compared to wild-type.⁽³⁸⁾ Given that most O₂^•− generation is thought to occur around complex III, our data has been demonstrated that the mev-1 mutation increases O₂^•− production at another point in succinate-dependent electron transport, perhaps even at complex II. Another of the biochemical pathologies is that of reduced glutathione concentration in mev-1 animals (Fig. 1).⁽³⁸⁾ A number of biochemical pathologies likely derive from the role played by SDH in the TCA cycle. First, the ratio of lactic acid to pyruvic acid is significantly higher in mev-1 mutants, suggesting that a metabolic imbalance known as lactic acidosis occurs. Second, a number of TCA cycle intermediates are present at abnormal concentrations in mev-1 mutants. Conversely, ATP levels are normal in mev-1 mutants. This has been initially surprising but could suggest that mev-1 animals rely more heavily on glycolysis for energy acquisition, thus explaining the elevated lactic acid levels. However, it is also possible that ATP consumption is decreased in mev-1 because of some sort of global decrease in the metabolic rate that acts to counterbalance the compromised ATP generation in mev-1.⁽³⁸⁾

A transgenic mouse embryonic fibroblast NIH3T3 cell line with the mutated SDHC^V69E (changing a neutral amino acid: valine to an acidic amino acid: glutamic acid) has been established. The mutation corresponds to the amino acid substitution of G71E in the C. elegans mev-1 mutant allele kn1.^(37,55) After transfection, the selected SDHC E69 cell lines as mev-1-mimic cells have expressed the equal amounts of sdhc mRNA from the transgene relative to endogenous wild-type allele.⁽⁵⁵⁾ Overexpressed transgene cell lines were not obtained, most likely because cells with more than 80% abnormal mtDNA are inviable.⁽⁵⁶⁾ Consistent with this prediction, RNAi with cyt-1 produced an embryonic lethal in C. elegans.⁽⁵⁷⁾ The addition of succinate dramatically increases O₂^•− production in mitochondria isolated from the mev-1-mimic cells. The enzymatic activity of complex II–III in the mev-1-mimic cells was reduced to 40% whereas the activity of complex I–III was unaffected.⁽⁵⁵⁾ ATP levels were not affected, suggesting that this mutation did not directly compromise cell survival through reduced respiration per se. The mutation results in an excess electron leakage from electron transport by decreasing the affinity between complex II and ubiquinone and thereby uncouples electron transfer.

Finally, mev-1-mimic conditional transgenic (Tet-mev-1) mice with SDHC^V69E mutation that does not constitutively overexpress the mutated SDHC^V69E encoding transgene has been established.⁽⁵⁸⁾ We contemplated developing another mev-1 model mouse using a knock-in approach. However, it has been reported that most of hereditary diseases and genetic disorders have been impacted by genomic imprinting.^(59,60) Therefore, it was feared that the individual phenotypes could be unstable in the same strain, even though it has been unclear whether the imprinting occurs at the mev-1-locus (1q23.3). Given this, we opted to employ a modified tetracycline system (Tet-On/Off system) to construct a mev-1-mimic mouse model. Our modified Tet-On/Off system enables the transgene to be expressed at lower levels such as approximate endogenous expression levels from wild-type allele. The fourteen lines of Tet-mev-1 mice with SDHC^V69E mutation were established using our modified Tet-On/Off construct.⁽⁵⁸⁾ As was the case with C. elegans mev-1 and mice embryonic fibroblast mev-1 cells, these mice were all hypersensitive to methyl viologen, which induces oxidative stress.⁽⁵⁸⁾ In Tet-mev-1 mice treated with doxycycline (Dox) for induction of SDHC^V69E mutated protein, the electron transport ratio from complex II to III through ubiquinone has been diminished, therefore O₂^•− level has been overproduced and accumulated as compared to the control C57BL/6j mice treated with Dox.⁽⁵⁸⁾ The newborn Tet-mev-1 mice which are continuously exposed to Dox before fertilization through fetal development develop a low birth weight and growth retardation through the weaning age.

Cellular Damages with Complex II Electron Transport Defect

Such endogenously generated ROS can readily attack a wide variety of cellular entities, resulting in damage that compromises cell integrity and function (Fig. 2).^(61–63) This can cause or at least contribute to a variety of pathologies, including some in humans.^(63–68) The mean and maximum lifespans of the mev-1 mutant have been shorter than control strain in C. elegans.⁽⁶⁹⁾ The mev-1 mutants have accumulated fluorescent materials (lipofuscin) and protein-carbonyl derivatives that are a specific indicator of oxidized lipid and protein at significantly higher rates than its control during aging.^(70–76) In addition, the oxidative stress by hyperoxia make mev-1 mutants high mutable on nuclear DNA.⁽⁷⁷⁾ Mitochondrially-derived ROS can mutate other genes, including tumor suppressor genes and oncogenes and can lead to cellular transformation, tumorigenesis (Fig. 1 and 2). In a similar fashion, Lemire and colleagues constructed transgenic C. elegans strains with a series of mutations in the succinate dehydrogenase iron-sulfur subunit (SDHB-1) (a human SDHB homologue). These strains also have resulted in reduced lifespans.⁽⁷⁸⁾ They have been also more sensitive to oxygen and methyl viologen with O₂^•− overproduction by impaired succinate-cytochrome c reductase activity.

The mutated sdhC gene (a human SDHC gene homologue) encoding succinate dehydrogenase C in Drosophila by using transgenic flies expressing a dominant-negative form, mev-1-mimic SdhC^I71E has been investigated.⁽⁷⁹⁾ Expression of SdhC^I71E significantly reduced the mean lifespan by 22% compared to that of control flies. The amount of protein carbonyl was significantly increased, suggesting that a high level of oxidative stress was induced in these flies with decreasing of complex II activity. Furthermore, Benzer and colleagues performed a genetic screen for mutations that cause decreased survival under hyperoxia (100% O₂).⁽⁸⁰⁾ Among a collection of P-element insertion mutants, they identified a hyperoxia-sensitive line, EY12081, which has an insertion in the sdhB gene (a human SDHB gene homologue) encoding SDHB.⁽⁸¹⁾ Under hyperoxia, the mean survival time of sdhB^EY12081 flies, which have the P-element insertion site within the 5' untranslated region (UTR) of the gene resulting in reducing the expression of sdhB gene, was reduced to 10% of that of normal flies.⁽⁸⁰⁾ Under normoxic conditions, sdhB^EY12081 flies displayed a 66% decrease in mean survival time and a 17% decrease in maximum survival time. There was a 56% decrease in the complex II-specific (succinate-dependent/antimycin A-sensitive) respiration rate, a 40% decrease in the complex II-mediated electron transfer ratio (malonate-sensitive succinate-cytochrome c reductase activity and a 32% increase in mitochondrial hydrogen peroxide production, as compared with wild-type controls.⁽⁸⁰⁾ Both mutants (SdhC^I71E and sdhB^EY12081) mimic the C. elegans mutants; namely, they are very similar in complex II mutation-associated ROS production, highly sensitive to oxidative stress, including hyperoxia, and age precociously. The results from these lower animal models support the consequential damages of ROS production with electron leakage by mitochondrial complex II deficiency, which ultimately lead to accelerated aging and age-associated diseases.

The mev-1-mimic mammalian cells, mouse embryonic fibroblast cells with SDHC^V69E mutation, accumulate cytoplasmic protein carbonyls at a faster rate than control cell line. In addition, the amount of 8-hydroxydeoxyguanosine (8-OHdG), a DNA marker of oxidative stress is two-fold higher in mev-1 cells.⁽⁵⁵⁾ In three-month cultured mev-1 cells, the morphology has been dramatically changed from the typical solid and elongated fibroblasts to smooth and rounded cells. In addition, the cells form multiple layers. The doubling time of one-month cultured mev-1 cells after establishment has been 1.5 to 2 times slower than that of wild-type cells. However, in mev-1 cells cultured for 3 months, the doubling time is fully restored to wild-type upon transformation.⁽⁵⁵⁾ When one-month cultured mev-1 cells are injected under the epithelium of nude mice, they rapidly disappear as compared to wild type. This suggests that these cells have been dying under apoptosis and been shortly phagocytized after injection. Conversely, injecting the three-month cultured mev-1 cells results in the production of tumors.^(55,82) The transformation rate on soft-agar medium for wild-type NIH3T3 cells is less than 1 × 10⁻⁶. On the other hand, the rates are 5 × 10⁻⁴ for the one-month mev-1 cells and 5 × 10⁻³ for the three-month cells. Thus, the mev-1 cells have 100- to 1,000-fold higher transformation rates with high genomic mutation levels under the 6-thioguanine tolerance test and activated AP-1 site-dependent transcriptions through Ras-Raf and Ras-MEKK signal transductions than wild-type cells.^(55,82) Therefore, the mev-1-mimic cell lines have demonstrated the cellular senescence and tumorigenesis like the pathogenesis of paragangliomas, pheochromocytes and gastrointestinal stromas in humans (Fig. 1 and 2).

The mev-1-mimic (Tet-mev-1) conditional transgenic mice with SDHC^V69E mutation have accumulated carbonylated proteins in the intracellular membrane fraction protein levels at newborn mice. Moreover, both TUNEL-stained and c-Caspase-3 immunostained apoptotic cells have been more frequently appeared in brain, lung, liver, kidney (especially in the adrenal region), salivary grand, nasal sinus tissue (especially secretory cells and mucosal cells), and muscles of Tet-mev-1 mice relative to the control C57BL/6J mice.⁽⁵⁸⁾ Interestingly, the programed cell death has been observed in the stomach, intestine, spleen and lymphatic tissues of control C57BL/6j mice with Dox and Tet-mev-1 mice with Dox at roughly equal rates. Thus, the Tet-mev-1 mice have resulted in a significant decrease in body size and weight between the developmental stages leading to the growth retardation, just as in the C. elegans mev-1 mutant.⁽⁵⁸⁾ Then, the neonatal unexplained mortality is increased, and the number of weaning mice is dramatically decreased in Tet-mev-1 mice.^(58,83) These phenomena, including decreasing electron transport ratio, increase in O₂^•− production and accumulation levels, excessive apoptosis, low-birth-weight infants, and growth retardation in Tet-mev-1 mice, are recovered by ubiquinol (CoQ-H₂) supplementation to that of the wild-type C57BL/6j.⁽⁵⁸⁾ This strongly suggests that the excessive apoptosis induction caused by the electron leakage from mitochondria with complex II-ubiquinone oxidoreductase deficiency leading to O₂^•− overproduction during cell proliferation stage. Moreover, complex II deficiency-dependent excessive apoptosis in several tissues have been not observed after the period of growth of body length at about 12-week age. This indicates that the excessive apoptosis induction might be caused by mitochondrial ETC defect-enhanced intracellular oxidative stress at cell growth and proliferation (Fig. 1).

Cell Death Signaling with Complex II Electron Transport Defect

Apoptosis induction pathways are classified largely into the extrinsic and intrinsic pathways activated by receptor-ligand binding signals through caspase-8 activation and mitochondria-dependent machinery through caspase-9 activation.⁽⁸⁴⁾ The extrinsic apoptotic pathway induces apoptosis mediated through the receptor-ligand binding signals of the tumor necrosis factor (TNF) receptor superfamily: FasL (Fas ligand)/Fas (also called CD95 or APO-1), TNF-α/TNFR1 (TNF receptor 1), and TRAIL (TNF-related apoptosis-inducible ligand)/TRAILR-R1 and 2 (TRAIL-receptor 1 and 2; also called DR4 and 5).^(85–88) The FasL/Fas binding signal requires Fas-associated DD (FADD) and procaspase-8 resulting in activation of the initiator caspase-8 with the endocytosed death-inducing signaling complex (DISC).^(89–91) The caspase-8 activation exists at the DISC and determines Type 1 or Type 2 mechanisms; significant caspase-8 activation directly activates caspase-3 (Type 1), whereas low caspase-8 activation mediates caspase-3 activation through an amplification loop in mitochondria (Type 2).⁽⁹²⁾ In Type 2, the activated caspase-8 cleaves pro-apoptotic Bid leading to outer mitochondrial membrane permeabilization through the interactions of cleaved Bid (tBid) with Bax/Bak, resulting in apoptogenic cytochrome c release from mitochondria, involved in intrinsic pathway as described below. On the other hand, TNFα/TNFR1 binding signal requires the TNFR1-associated death domain (TRADD)-composed complex I and II formation that activates distinct downstream survival and apoptotic signaling pathways, respectively.^(93,94) Apoptosis-inducible complex II is composed of TRADD, FADD, and caspase-8, and is formed after TNFR1 receptosome endocytosis leading to caspase-8 activation as mentioned below.^(95–97) Given this, the both TNFα/TNFR1 and FasL/Fas-mediated apoptosis are similarly induced by the caspase-8-dependent modulations.

The intrinsic apoptotic pathway, which is activated by intracellular stimuli such as mitochondrial oxidative stress, depolarization, and DNA mutation is mediated by promoting mitochondrial outer membrane permeabilization and cytoplasmic translocation of mitochondrial apoptogenic or pro-apoptotic proteins: cytochrome c, Smac/DIABLO, and HtrA2/Omi, which are caspase-dependent factors, and apoptosis-inducing factor (AIF), EndoG, and CAD, which are released from mitochondria to directly go to the nucleus for induction of nuclear chromatin condensation and DNA fragmentation.^(98–101) Cytochrome c normally resides in the intermembrane space of mitochondria for ETC reaction. The released cytochrome c from mitochondria forms an apoptosome complex with apoptotic protease-activating factor-1 (Apaf-1) and procaspase-9, producing active caspase-9 which is a crucial factor to induce apoptosis by directly activating caspase-3, whereas Smac/DIABLO and HtrA2/Omi inhibit inhibitors of apoptosis (IAPs) and thereby indirectly activate caspase-3.^(88,99,102)

The mitochondrial permeability transition pore (mPTP) is composed of cyclophilin D (cypD), voltage-dependent anion channel (VDAC), and adenine nucleotide translocase (ANT). It plays a key role in mitochondrial permeabilization and release of mitochondrial pro-apoptotic proteins.^{(87,103–106)} The anti-apoptotic (Bcl-2, Bcl-XL, and Bcl-w) and pro-apoptotic (Bax, Bak, Bad, Bim, and Bid) Bcl-2 superfamily are major players to modulate mitochondrial outer membrane permeabilization.^(107,108) Upon receiving apoptotic stimuli, Bad is activated by phosphorylation and inhibits Bcl-2 and Bcl-XL that maintains Bax/Bak inactivation.^(109–111) Activated Bax and Bak then directly interact with VDAC leading to cytochrome c release from mitochondria.⁽¹¹²⁾ In addition, the tBid converted from Bid is degraded by caspase-8. This promotes Bax and/or Bak oligomerization resulting in megapore formation, a highly orchestrated and active process.^(113–115) The contribution of pro-apoptotic tBid in both mitochondria and the death receptor-ligand binding signals has a crucial role for cross talk between the intrinsic and extrinsic apoptotic pathways.

In addition, it has been reported that oxidative stress directly plays an important role in the mPTP modifications through ROS-mediated modification of the thiol group of ANT or ROS-mediated oxidation of cardiolipin for binding between tBid and VDAC.^(111,116) Another interesting report indicated that ROS-activated c-Jun-N-terminal kinase (JNK) can be induced by extrinsic or intrinsic apoptotic signaling.^(117–119) TNFα is a potent activator of the MAPK cascade, and the apoptosis signal-regulating kinase 1 (ASK1)-JNK pathway plays an important role in TNFR1-mediated apoptotic signaling in various cell types.⁽¹²⁰⁾ TNFα induces pro-apoptotic effects via ASK1 signaling with a prolonged and robust JNK activation by ROS; however, a transient and modest JNK activation mediates cell survival via NF-κB-induced antiapoptotic gene expression.^(121–123) Furthermore, the mitochondrial ASK1-dependent apoptotic signaling pathway reportedly activated both JNK-dependent and JNK-independent apoptosis.⁽¹²⁴⁾ Nuclear translocation of activated JNK promoted activator protein-1 (AP-1)-mediated expression of pro-apoptotic TNFα, FasL, and Bak, whereas mitochondrial JNK translocation promoted cytochrome c release.^(125,126) The above-mentioned cross talk between intrinsic and extrinsic apoptotic pathways via caspase-8 or JNK activated by intracellular and extracellular stimuli with oxidative stress has not been adequately clarified.

Meanwhile, the oxidative stress-sensitive, short-lived mev-1 mutant of C. elegans that exhibits growth retardation and small body size has supernumerary embryonic apoptosis mediated by the oxidative stress-stimulated ced-9 (Bcl-2 family)/ced-3 (caspase)/ced-4 (Apaf-1) apoptotic pathway, especially under hyperoxia.^(36,127,128) The abnormal apoptosis is suppressed by mutations in either ced-3 or ced-4, indicating that the inappropriate signal in mev-1 embryos stimulated induction of the normal ced-9/ced-3/ced-4 apoptotic pathway in C. elegans.⁽¹²⁸⁾ Furthermore, the mev-1;ced-3 double mutant lives longer than mev-1, which suggests that the supernumerary apoptosis contributes to the phenotype of life shortening in mev-1.^(128,129) Recently, some repots described that the C. elegans p53 homolog, CEP-1, mediates also the apoptotic pathway required for all normal cell deaths in germline.^(130,131) Inactivating cep-1 in mev-1 mutant restored their lifespan.⁽¹³²⁾

Interestingly, the mev-1-mimic cells have maintained caspase-3 activity through caspase-9 activation of both caspase-9-dependent intrinsic and caspase-8 and -9-dependent extrinsic pathways.⁽¹³³⁾ The inhibition of apoptosis induction with caspase-8 inhibitor do not affect the survival rate in mev-1-mimic cells, although that increased survival rate in control NIH3T3 cells and transformed mouse embryonic fibroblast mev-1-mimic cells (transformed mev-1-mimic cells). On the other hand, the inhibition of apoptosis with caspase-9 inhibitor decreased survival rate with necrotic cell death induction in both control and mev-1-mimic cells. From these results, it has been suggested that the caspase-9 is affecting the caspase-8-mediated extrinsic apoptotic pathway and mitochondrial deficiency-mediated intrinsic pathway plays a crucial role in survival in both control and mev-1-mimic cells excepting transformed mev-1-mimic cells.⁽¹³³⁾ Moreover, JNK activation patterns have been dramatically changed in each cell type: phosphorylated JNK pattern has been completely changed between control and mev-1-mimic cells and the JNK activity has been robustly enhanced in transformed mev-1-mimic cells.⁽¹³³⁾ This JNK functional change might affect or at least reflect different apoptosis induction machinery via in each intrinsic and extrinsic pathway or both, as mentioned above. In the transformed mev-1-mimic cells, benign tumor’s characterizations have been confirmed under the epithelium of nude mice, because the cells maintain active caspase-3 through caspase-8 and -9 with cytochrome c release from mitochondria and p53–p21 activation.^(55,82) These data support the notion that oxidative stress from mitochondria play an important effect on apoptosis during both precocious aging and tumor suppression (Fig. 2).

Stem Cell Aging with Complex II Electron Transport Defect

Premature cell death and apoptosis induction in reproductive organs

The Tet-mev-1 females have developed symptoms such as ovarian hemangioma leading to endocrine-related disorders and have shown decreased ovulations with non-synchronously developed follicle maturation.⁽⁸³⁾ It is estimated that only 12% of the maximum pre-birth non-growing follicles (NGF) population is present in 30-year old women and by the age of 40 years only 3% remain.^(134,135) The ATP levels increase during oocyte maturation. The oocytes with higher mtDNA copy number and higher ATP levels have greater fertilization rates and embryo development than those with lower ATP levels.^(136–140) The mtDNA bottleneck theory suggests that the number of homoplastic mtDNA molecules are transmitted to progeny are restricted in order to select high quality oocytes with healthy mitochondria.^(141–145)

Tet-mev-1 male mice induce excessive apoptosis in only spermatogonia on the basal compartment of seminiferous tubules.⁽⁸³⁾ The increased activation of caspase-2 in the mitochondria of germ cells plays an important role in apoptosis induction through cytochrome c release from mitochondria during the process of spermatogenesis.⁽¹⁴⁶⁾ Bax, one of Bcl-2 family members is also crucial factor for the apoptosis induction through the mitochondrial pathway in dividing spermatogonia.^(147–149) Mitochondrial ROS might have a part in the mitochondria-dependent apoptosis induction pathway in dividing spermatogonia. It has been well known that cell death via apoptosis seems to be a survival strategy for some male germ cells, allowing survival of remaining healthy spermatozoa under various stresses.^(150,151) The apoptosis induction in spermatogonia does not affect the total number of spermatozoa in epididymis of Tet-mev-1 mice, indicating that spermatogenesis and spermatozoa morphology are successfully normal after the meiotic phase.

Spermatozoa mitochondria are rearranged in elongated tubular structures and are helically arranged around the anterior portion of the flagellum.^(152–154) The outer membranes of sperm mitochondria are enclosed in a keratinous structure and formed by disulfide bonds between cysteine- and proline-rich selenoproteins, including the sperm-specific glutathione peroxidase, catalases and SOD.^(155,156) These formations contribute to protect the sperm against ROS generated by surrounding defective sperms in the ejaculate.^(157,158) The sperm motility of Tet-mev-1 mice is just slightly decreased to 50–80% of wild type depending on culture time in an in vitro fertilization assay.⁽⁸³⁾ It has been reported that the activation of caspase-3 and -9 decreases sperm motility at ejaculation.^(159–162)

It has been reported that the mitochondria of zygotes are barely active through the blastocyst stage before implantation. The mitochondrial shapes are spherical and vacuolated with a dense matrix and few cristae before ovulation,^(163–165) and then become elongated dumb-bell shapes with decreases in matrix density before changing to numerous transversally-supported cristae from concentrically located cristae after fertilization through implantation.⁽¹⁶⁶⁾ This is logical given that energy metabolism is mainly from glycolysis in the enclosed environment during the blastomere stage.⁽¹⁶⁷⁾ After that, the blastocyst stage oocyte, which starts cell division for the involved differentiation in both inner cell mass (ICM) and trophectoderm (TE), prepares an increase in OXPHOS in mitochondria with glycolysis and oxygen consumption.^(168–170) Given this, it is logical that the mitochondrial complex II SDHC^V69E mutation does not affect the fertilization and early embryogenesis but rather exert its effects in later development in Tet-mev-1 mice.⁽⁸³⁾

Premature cell death in placental, embryonic and neonatal developments

The number of functionally active mitochondria based on membrane potential per cell is higher in the TE than in the ICM of mouse blastocyst.^(168,170) The existence of two types of mitochondria in the mouse blastocyst has been reported: spherical mitochondria in the ICM and elongated mitochondria in the TE.⁽¹⁶⁹⁾ It has been anticipated with those mitochondrial shape’s characterizations that the TE cells become the placenta and extraembryonic tissues, are highly polarized and very active at producing ATP and oxygen consumption.^{(163,164,168,171)} In Tet-mev-1 mice, the natural pregnancy and safe delivery rates at primiparity are decreased to about 55% and 65%, respectively, before they first experience a successful delivery.⁽⁸³⁾ The Tet-mev-1 mice exhibit placental inflammation, thrombocytosis and splenomegaly with megakaryocytic differentiation by intracellular oxidative stress-activated Nrf-2 signaling.^(83,133,172) This causes placental angiodysplasia with decreasing Flt-1/VEGFR-1 protein, occasionally leading to placental inflammation in Tet-mev-1 mice.⁽⁸³⁾ The VEGF signaling is important in regulating vascular cell recruitment and proliferation for placental formation.^(173,174) Flt-1, also known as VEGFR-1, is well known as the most abundant and active member of the VEGF receptor family. This binds to VEGF-A and PLGF-1 as a key factor promoting angiogenesis in the placenta. Disruptions contribute to the pathogenesis of female infertilities such as preeclampsia.⁽¹⁷⁵⁾ We have suggested that these placental phenotypes would lead to placental inflammation and hypoxic conditions in embryos. Given this, it is not surprising that the embryos of Tet-mev-1 mice frequently have a developmental arrest that mimics human embryopathy and a randomly abnormal angiogenesis leading to decreasing progeny (Fig. 1).⁽⁸³⁾

Premature cell death in mature tissues: cornea, liver, and brain

The oculus which has contact with atmosphere is more sensitive to mitochondrial oxidative stress with age compared to other tissues.⁽¹⁷⁶⁾ The Tet-mev-1 mice have increased levels of carbonylated protein and 8-OHdG nucleotide with mitochondrial O₂^•− accumulation in the eyes, particularly with age.⁽¹⁷⁷⁾ In the corneal epithelium of Tet-mev-1 mice, the proliferation of epithelial basal cells has been decreased, resulting in delayed epithelialization with keratitis and dry eyes relative to control C57BL/6j mice.^(177–179) Furthermore, both the age-dependent decrease of post-mitotic corneal endothelial cells and the increase in thickness of Descemet’s membrane in Tet-mev-1 mice are consistent with the pathological phenotypes of Fuchs’s corneal dystrophy (FCD).^(177,180) The corneas have age-dependently been thinning owing to decreased numbers of corneal stromal cells with catalase activation as pathogenesis of keratoconus in Tet-mev-1 mice.^(177,181) The Tet-mev-1 mice have demonstrated accelerated age-dependent corneal pathophysiological changes: FCD, delayed epithelialization with keratitis, decreasing endothelial cell number, thickened Descemet’s membrane, and keratoconus with corneal stromal thinning with mitochondrial oxidative stress by complex II electron transport defect on aging (Fig. 1).^(177–179)

In another our studies, age-dependent pathophysiological changes with mitochondrial oxidative stress were analyzed for abnormal fatty acid metabolism (metabolic dysfunction associated steatohepatitis, MASH) and for neurodegeneration and neurocognitive disorders using Tet-mev-1 mouse model.^(182,183) Mitochondrial oxidative stress by complex II electron transport defect has been age-dependently and excessively accumulated in the liver and brain of Tet-mev-1 mice. The liver fibrosis with inflammation has been accelerated in Tet-mev-1 mice under the excessive feeding condition such as high fat/high calorie intake, though premature cell death has not been appeared.⁽¹⁸²⁾ It has been suggested that the tissue cells such as hepatocytes which very activate the recovery rate and redox activity with inflammation could be quickly recovered in organs with liver sinusoidal endothelial cells (LSECs), when the intracellular oxidative damages are occurred.⁽¹⁸⁴⁾

The neuronal activation and plasticity through Ca²⁺ signal transduction for learning and memory have been age-dependently diminished with damaged dendritic outgrowth of astrocytes in Tet-mev-1 mice, though premature cell death in neurons and glial cells has been identical relative to control.⁽¹⁸³⁾ From this result, it has been suggested that Tet-mev-1 mice have developed age-dependent neurocognitive disorders with the impaired neuron-astrocyte networks by mitochondrial oxidative stress (Fig. 1).

We have considered that complex II activity could be important in cell proliferation and differentiation during developmental stage, because premature cell death has been significantly occurred in several tissues of fetal and infant individuals (Fig. 1).⁽⁵⁸⁾ After reproductive maturation, the event of premature cell death has been identical in many tissues relative to control C57BL/6J mice. Therefore, the oxidative damages by mitochondrial complex II electron transport defect could be occurred, when complex II activity is more activated for the cell proliferation and differentiation.

Premature cell death in hematopoietic stem cells (HSCs)

HSCs have the potential for multipotency and self-renewal.⁽¹⁸⁵⁾ HSCs replenish all lineages of mature blood cells throughout life or even after transplantation-induced replicative stress among patients with hematologic disorders. HSCs preferentially reside in the specialized hypoxic bone-marrow (BM) niches,⁽¹⁸⁶⁾ which maintain HSCs under low ROS levels and prevent commitment and differentiation.^(187,188) Nevertheless, long-term hematopoiesis inevitably causes ROS accumulation and DNA damage. These factors then trigger the functional defects of HSCs and hematopoietic senescence.⁽¹⁸⁹⁾

Considering that HSCs rely on glycolysis for the metabolism and maintenance of low ROS levels,^(190–192) mitochondria in HSCs do not play an essential role in metabolism. However, recent studies revealed the functional indispensability of mitochondria,^(193–195) despite the low energy requirements in quiescent HSCs.^(190,191) In particular, an imbalance in ETC components in HSCs (a large amount of complex II and few complex I) indicates that mitochondrial complex II can be a key regulator of HSC metabolism.⁽¹⁹⁶⁾ Conditional knockout of succinate dehydrogenase complex subunit D (Sdhd) resulted in a lower number of HSCs and progenitors and mature blood cells.⁽¹⁹⁷⁾ However, the previous research was based on observation at the steady state and short-term spleen colony-forming unit,⁽¹⁹⁷⁾ or even in vitro colony replating capacity.⁽¹⁹⁶⁾ Therefore, the effects of complex II dysfunction on long-term hematopoiesis and hematopoietic reconstitution capacity had not been elucidated.

The Tet-mev-1 mice cause excessive ROS accumulation and DNA damages in HSCs, leading to myeloid skewing, thus they exhibit leukocytopenia and anemia, both of which are indicative of senescent hematopoiesis.⁽¹⁹⁸⁾ Furthermore, Tet-mev-1 mice with mitochondrial fragmentation in HSCs, which is attributed to a lower mitochondrial membrane potential, have exhibited excessive apoptosis in BM cells resulting in severe lymphopenia. Because the stem cell exhaustion or impaired self-renewal capacity are considered to be the hallmark of HSC aging, these results imply the accelerated aging hematopoiesis in Tet-mev-1 mice.⁽¹⁹⁹⁾

The bone-marrow transplantation (BMT)-induced replicative stress which causes self-renewal and proliferation with mitochondrial activation, leading to exacerbation of electron leakage and ROS accumulation severely accelerates premature senescent hematopoiesis in Tet-mev-1 mice (Fig. 1).⁽¹⁹⁹⁾ Considering that the transplanted SDHC^V69E-mutated HSCs into control C57BL/6J recipients exhibit myeloid-skewed hematopoiesis and impaired long-term hematopoietic reconstitution capacity, the SDHC^V69E mutation in HSCs rather than stromal or other organ cells would play a crucial role in the observed phenotype in Tet-mev-1 mice.⁽¹⁹⁹⁾

Taken together, complex II plays an essential role in long-term HSC maintenance, differentiation, and function, and the complex II electron transport defect on aging or replicative stresses severely causes excessive ROS accumulation and DNA damage in HSCs, leading to premature senescence (Fig. 1).

Conclusion and Future Perspective

A mutation in SDHB, SDHC, or SDHD of complex II was found in patients of PGL/PHEO syndromes.^(10–15) In general, it has been well known that the defect in electron transport causes electron leakage from the complexes and consequently increases ROS production (Fig. 2). Furthermore, it has been suggested that reverse electron transport under hypoxic conditions or excessively accumulated succinate cause the ROS production under a series of hypoxic responses that include transcriptional alterations leading to changes in cellular metabolism, angiogenesis, metastasis, and cell proliferation in tumorigenesis.^(16,20,200)

The chronic elevation in ROS levels generally results in damage to the various cellular components, which in turn results in the production of ROS at an even higher rate leading to oxidative stress. It has been well known that oxidative stress with mitochondrial dysfunctions can result in the breakdown of cellular homeostasis functions and intercellular networks leading to hyperglycemia and chronic inflammation on aging.^(201,202) And also we have recently proposed that age-related stress; oxidative stress, replicative stress, and inflammatory stress (inflammative stress) has mainly caused dynamic changes on aging (Fig. 3). As suggested by phenomena of Tet-mev-1 mice, competitively expressed SDHC^V69E animal models with defects in complex II are suitable for the study of not only tumorigenesis and carcinogenesis such as PGL/PHEO syndromes but also some age-related diseases (Fig. 2). These would include chronic inflammation-related diseases and lifestyle-related diseases such as cardiovascular and pulmonary diseases as well as excessive apoptosis-related diseases such as neurodegenerative diseases.

Fig. 3.

Hallmarks of aging and age-related stress.

Abbreviations

AA

arachidonic acid

AIF

apoptosis-inducing factor

ANT

adenine nucleotide transporter

Apaf-1

apoptotic protease-activating factor-1

ASK1

apoptosis signal-regulating kinase 1

CAD

caspase activated DNase

CoQ

ubiquinone

CoQ-H₂

ubiquinol

cypD

cyclophilin D

DAG

diacylglycerol

DISC

death-inducing signaling complex

Dox

doxycycline

ETC

electron transport chain

ETF

electron transferring flavoprotein

FADH₂

flavin adenine dinucleotide

HE

hematoxilin-eosin

HIF

hypoxia inducible factors

8-OhdG

8-hydroxydeoxyguanosine

ICAD

inhibitor of caspase activated DNase

ICM

inner cell mass

JNK

c-Jun-N-terminal kinase

mPTP

mitochondrial permeability transition pore

NADH

nicotinamide adenine dinucleotide

NGF

non-growing follicles

OXPHOS

mitochondrial oxidative phosphorylation

PGE2

prostaglandin E

PGF2

prostagladin F2

PGLs

paragangliomas

PHEOs

phaeochromocytomas

PKC

protein kinase C

PLGF

placenta growth factor

ROS

reactive oxygen species

SDH

succinate dehydrogease

SDHC

succinate dehydrogenase C subunit

SOD

superoxide dismutase

TCA

tricarboxylic acid

TE

trophectoderm

TNF

tumor necrosis factor

TNFR1

TNF receptor 1

TRADD

TNFR1-associated death domain

TRAIL

TNF-related apoptosis-inducible ligand

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labeling

VDAC

voltage-dependent anion channel

VEGF

vascular endothelial growth factors

VEGFR-1

VEGF receptor-1

Conflict of Interest

No potential conflicts of interest were disclosed.