1. Mitochondrial Complex I
The mammalian complex I is the largest mitochondrial respiratory chain enzyme and it is composed of 45 constituent subunits. Seven subunits are encoded by the mitochondrial DNA, 38 by nuclear genes
[1]. 14 subunits are conserved from procaryotes to eucaryotes and constitute the catalytic core that maintains the minimal structure for its catalytic activity. The others 31 subunits are called supernumerary. The function of the supernumerary subunits is largely unknown. Some are involved in the assembly and regulation of the complex
[1][2][3][4][5]. Complex I is involved in apoptosis
[6] and age-related functional decline
[7] as well as in a number of metabolic
[4], degenerative
[8] and proliferative diseases
[9]. It is the entry point of NADH reducing equivalents in the mitochondrial respiratory chain catalyzing the electron transport from NADH to ubiquinone. This process is coupled to proton pump from the matrix to intermembrane space, thus conserving the free energy as a transmembrane potential
[10]. The transmembrane potential drives ATP formation by complex V. In complex I, several redox centers have been found (FMN, iron-sulfur clusters, and tightly bound ubiquinone), which transfer electrons from the enzyme-bound NADH to the ubiquinone
[10]. The complex I structure has a well-known “L” shape with a membrane hydrophobic arm and a matrix-exposed hydrophilic arm. It can be subdivided into three functional modules named N-module (NADH binding and oxidation), Q-module (electron transfer to ubiquinone) and P-module (proton pumping)
[11]. Beyond the entry point of reducing equivalents, complex I is also considered to be the major source of reactive oxygen species (ROS) production
[12][13] indeed it can also transfer single electrons to oxygen with formation of oxygen superoxide
[7][14]. Complex I has been found to produce ROS at FMN and probably iron sulphur cluster N2
[15][16][17][18][19] in the hydrophilic arm.
Several conditions can generate an increase of complex I-dependent ROS production, for example, an excessive reducing power can result in the overproduction of ROS in the mitochondrial matrix with oxidative damage of mitochondrial DNA and subunits of complex I exposed to this space. Complex I can produce anion superoxide during forward electron transfer
[13] but it becomes the most important site of superoxide production within mitochondria during reverse electron transfer (RET)
[20]. RET is the process, driven by the inner mitochondrial membrane potential, by which electrons from the reduced ubiquinone pool (supplied by succinate dehydrogenase, glycerol-3-phosphate dehydrogenase, electron-transferring flavoprotein or dihydroorotate dehydrogenase) pass through complex I reducing NAD
+ to NADH
[20].
Complex I is a pacemaker of the oxidative phosphorylation system for ATP production
[21][22] and its functional capacity is regulated, at transcription and post-translational levels, in response to a variety of regulatory factors/events
[2][4][23][24]. In mammals, complex I biogenesis depends on the coordinated expression of mitochondrial and nuclear genes and involves nucleus/cytosol/mitochondria protein traffic and the stepwise assembly of a 45 subunits oligomer
[3]. In addition to “de novo” assembly of complex I from all the subunits, an exchange of individual subunits, between the newly imported ones and the older already assembled in the complex, can also occur
[3][25]. This process can play a proof-checking role of the subunits in the complex, in particular those superficially exposed to the matrix space, where they can be oxidatively-damaged by ROS, with depression of the catalytic activity. The content of complex I is determined, as for all proteins, by the balance between protein expression and proteolysis. This turnover is particularly critical for complex I because the subunits of its hydrophilic arm are directly subjected to oxidation in the mitochondrial matrix by ROS produced by the complex itself. The functional capacity of complex I appears also to be linked to mitochondrial dynamics
[26]. Moreover, the complex I can assemble at higher order with complex II, complex III and complex IV to form supercomplexes
[27] in order to modulate the capacity and efficiency
[28][29][30] of respiratory chain and the ROS production
[30].
2. Mitochondrial Complex I in Aging
In mammals, a functional decline of mitochondrial respiratory chain activity and thus of complex I has been reported during the aging process, associated with an increased ROS production
[7]. For example a study of human skeletal muscle of 63 orthopedic patients of age ranging between 17 and 91 years showed a statistically significant decrease with aging of mitochondrial oxygen consumption with pyruvate plus malate, succinate and ascorbate plus TMPD associated with an age-dependent decrease of the enzymatic activity of complex I, II and IV
[31]. It has been suggested that the aging-dependent decline of complex I activity is caused by a decreased expression and increased oxidation of complex I subunits
[32][33]. In agreement, several neurodegenerative diseases, such as Parkinson’s disease, are also associated with a decrease of complex I activity and content and an increase of mitochondrial ROS production and complex I subunit oxidation
[8][26][34][35][36]. The free radical theory of aging postulates that mitochondrial ROS production causes oxidative damage of macromolecules leading to mitochondrial respiratory chain dysfunction which, in turn, increases ROS production
[37]. The complex I, together with complex III, of the mitochondrial respiratory chain is considered the major source of ROS production
[7]. Some redox centers, such as FMN radical and ubiquinone in complex I, can be directly oxidized by dioxygen with the single electron transfer to O
2 and the production of oxygen superoxide
[13][14]. Oxygen superoxide can directly oxidize proteins, lipids and nucleic acids. However, complex I produces ROS mainly during reverse electron transport in which electrons pass through complex I reducing NAD
+ to NADH
[20]. This can happen in the condition of reduced ubiquinone, accumulated succinate and higher proton motive force
[20].
Mitochondrial ROS production, specifically at complex I, is considered to be one of the main causes involved in the aging process and has been proposed as one of the determinants for longevity of species
[7][38][39][40]. Indeed, a comparative approach of complex I in the heart tissue of eight different mammalian species with a longevity ranging from 3.5 to 46 years showed a specific difference in the gene and protein expression of complex I related to longevity, and in particular that longevity phenotype is associated with low protein abundance of matrix hydrophilic subunits NDUFV2 and NDUFS4 of complex I
[41]. These differences were also associated with reduced ROS levels
[41]. Accordingly, a low rate of mitochondrial ROS production has been observed in long-lived organisms, invertebrates and vertebrates, with respect to short lived
[39][40]. At the molecular level, the oxidation of mitochondrial DNA and membrane lipids appears to be the traits that correlate with longevity across species. In fact, a negative correlation between mitochondrial DNA oxidation, lipid peroxidation and longevity has been observed, that means a lower oxidation in long- than in short-lived animals even in the same species
[42][43].
In the last years, interesting evidences have been provided on the role of content of complex I and ROS production, on the lifespan. The content of complex I, and, in particular, that part of complex I responsible for ROS production, appears to positively correlate with ROS production
[44]. Importantly, in mouse and rat, nutritional or pharmacological treatments such as protein or methionine restriction, rapamycin or metformin (inhibitor of complex I) treatment, resulted in higher longevity associated with a decrease of complex I-dependent ROS production
[42][45][46][47][48] in heart, skeletal muscle, liver and brain
[49]. Moreover, dietary restriction has been also associated with a decrease in the content of complex I
[38].
As mentioned above, the longevity-related ROS production by complex I has been ascribed to the peripheral arm
[44], and thus to FMN and/or FeS centers
[15][16][17][18][19]. In line with this, a study showed that mice under dietary restriction activate an adaptive response leading to a decrease of respiratory chain subunits and, in particular, those that belong to the hydrophilic peripheral arm of complex I
[44]. This change has been associated again with a decreased superoxide production at complex I and the improvement of the complex assembly
[44]. On the contrary, the inhibition of complex I assembly increased mitochondrial ROS production accelerating aging
[44]. The data of the abundance of complex I hydrophilic peripheral subunits have been associated with ROS level
[44]. In particular, it has been shown that young longer living mice had a low level of subunits of the hydrophilic arm of complex I compared to old or shorter-living mice, while the hydrophobic arm subunits levels remained unaltered
[44]. The authors propose that if the amount of hydrophilic arm subunits is higher than the amount of the hydrophobic arm subunits, the hydrophilic subunits can also assemble into a subcomplex that is able to oxidize respiratory substrate without coupling the proton pumping. This process leads to ROS production that, in turn, accelerates the aging process
[44]. This is in agreement with the metabolic labeling study in mammals showing that hydrophilic arm subunits of mitochondrial complex I have a shorter half-lives than hydrophobic arm subunits in liver and heart
[50] and in HEK cell line
[51]. This supports the hypothesis that the hydrophilic arm subunits of complex I might exist as free
[52] or less stable monomer
[44] in according also to the finding showing that the new-imported subunits of the hydrophilic peripheral arm of complex I can be assembled in the complex I in exchange with the “older oxidated” ones already assembled in the complex
[25][53]. Considering the increase of oxidative stress and the oxidation of complex I subunits during aging
[32][33], the continuous import and exchange of the new-synthesized subunits of peripheral hydrophilic arm with the old oxidized ones could represent a mechanism of scavenger of complex I. The levels of hydrophilic arm subunits of complex I appear to correlate positively with the chaperon function of prohibitin proteins (PHB1 and PHB2). Interestingly, in
C. elegans, the depletion of PHB increases lifespan in mutants with compromised respiratory chain
[54] and, in cell with a damaged assembly of complex I, the PHB knockdown restores the assembly lowering the complex I-dependent ROS production
[54]. All data are consistent with the possible dangerous effect due to the abundant hydrophilic arm subunits of complex I that can assemble to form a sub-complex producing superoxide, without contributing to proton pumping but contributing to age-dependent mitochondrial decline
[44]. In this scenario, it would be very interesting to understand if the specific complex I subunits of the hydrophilic arm are mainly involved. In higher eukaryotes, both NDUFV2 and NDUFS4, a core and a supernumerary subunits, respectively, have been found to negatively correlate with increased longevity
[41], instead, the down-regulation of NDUFS2, a core subunit, did not appear to affect longevity
[55].
In this regard, evidences have been produced in
C. elegans, in which the knock-down of complex I subunits affected the lifespan
[56][57]. In particular the RNA interference of nuo-2
[56][58], the
C. elegans homolog of mammalian NADH dehydrogenase iron-sulfur protein 3 (NDUFS3) and D2030.4, the homolog of mammalian NADH dehydrogenase 1 beta subcomplex subunit 7 (NDUFB7)
[57] extended lifespan leading to a 50% reduction of the ATP production
[56][57]. However, the relationship between electron transport chain function, ROS production and longevity appears to be not so simple. Indeed, the timing of RNAi knockdown is fundamental. For example, the reduced subunit expression during adult stages, after completion of the critical L3/L4 larval stages, caused a reduced ATP level that is not, however, associated with an increase of longevity
[56]. Moreover, other studies in
C. elegans have produced evidence showing that worms with reduced electron transport chain activity, induced by the RNAi knockdown of Nuo-1, orthologous to human NDUFV1, of nuo-6, orthologous to human NDUFB4, have extended longevities associated, this time, with an increase of ROS levels
[59][60][61].
3. cAMP-Dependent Regulation of Mitochondria and ROS Production
cAMP is generated from ATP by adenylyl cyclases (ACs) and degraded by phosphodiesterases. In mammals, there is one gene encoding for multiple isoforms of soluble adenylyl cyclase (sAC) and nine genes encoding for transmembrane adenylyl cyclases (tmAC). tmACs are localized to the plasma membrane, while sAC is present in the mitochondrial matrix and in other organelles
[62]. cAMP-dependent protein kinase A (PKA) is a cAMP downstream effector. It is a tetramer that consists of two cAMP-binding regulatory (R) subunits and two catalytic (C) subunits, which dissociate once cAMP binds R-PKA. R-PKA are associated to sub-cellular membranes by binding to specific A-kinase anchor proteins (AKAP). AKAP protein can also bind phosphatases and also phosphodiesterases
[63]. In this way, these complexes accomplish activation/deactivation of the signal and phosphorylation/dephosphorylation of substrate bound or close to the complexes modulating the activity of specific proteins
[63]. All these proteins involved in reversible cAMP-dependent protein phosphorylation are also present in mitochondria
[62][64][65][66].
First evidence on the cytosolic cAMP-dependent regulation of complex I comes from Papa’s laboratory showing that the activation of cAMP pathway by widespread addition of a permeant derivative of cAMP (dibutyryl-cAMP or 8-br-cAMP) or by the specific activation of tmAC by β-adrenoceptor agonist isoproterenol or by cholera toxin, results, in a variety of murine and human cell cultures, in the elevation of the NADH-ubiquinone oxidoreductase activity of mitochondrial complex I
[2][25][67][68]. This was also associated with a strong reduction of the of ROS level
[25][67][69][70][71]. At the same time, the activation of cAMP signaling did not affect the ROS scavenger systems
[70][71]. Thus, the cAMP-dependent decrease of ROS level appears to be associated with changes in the enzymatic activity of complex I. The cAMP-dependent promotion of the complex I activity and decrease ROS production are associated to the serine phosphorylation in the conserved RVS site in the carboxy terminus of the subunit of complex I encoded by the nuclear NDUFS4 gene
[68][72]. Further insights have been produced on the molecular mechanism of the cAMP-dependent increase of complex I activity. Once in mitochondria, the new-imported NDUFS4 assembles in the complex I by replacing the pre-existing, already assembled, oxidized NDUFS4 subunit restoring the activity of oxidatively damaged complex I
[25]. Recently, using ex vivo and in vitro models, it has also been shown that the activation of cAMP/PKA cascade resulted in an increase of supercomplex formation associated with an enhanced capacity of electron flux and ATP production rate
[29]. This is also associated with the phosphorylation of the NDUFS4 subunit of complex I
[29], and at the same time, the formation of supercomplexes has been associated with the decrease of complex I-dependent ROS production
[30][73].
The cAMP signal also impacts the mitochondrial dynamics by different G proteins. Among these, the Drp1 protein is inactivated by its PKA-dependent phosphorylation. The inactivation of Drp1 results in the inhibition of mitochondrial fission and thus in the promotion of fusion
[74][75]. At the same time, the activity of the complex I, as well as of the all respiratory chain, is affected by the balance between mitochondrial fusion and fission, indeed, the inhibition of complex I results in mitochondrial fission
[74].
Regarding the mitochondrial pool of the cAMP it has, so far, been found that sAC-dependent cAMP production in mitochondria is involved in the regulation of cytochrome
c oxidase activity, by direct PKA-dependent phosphorylation of complex IV subunits
[76], complex I activity, by regulating the degradation of new-imported subunits of peripheral arm
[52] including NDUFV2 and NDUFS4
[52], FoF1 ATP synthase activity and its structural organization
[77] and ATP production
[78], and in the activation of the mitochondrial pathway of apoptosis
[79].