Updating the mitochondrial free radical theory of aging: an integrated view, key aspects, and confounding concepts

Abstract

An updated version of the mitochondrial free radical theory of aging (MFRTA) and longevity is reviewed. Key aspects of the theory are emphasized. Another main focus concerns common misconceptions that can mislead investigators from other specialties, even to wrongly discard the theory. Those different issues include (i) the main reactive oxygen species (ROS)-generating site in the respiratory chain in relation to aging and longevity: complex I; (ii) the close vicinity or even contact between that site and the mitochondrial DNA, in relation to the lack of local efficacy of antioxidants and to sub-cellular compartmentation; (iii) the relationship between mitochondrial ROS production and oxygen consumption; (iv) recent criticisms on the MFRTA; (v) the widespread assumption that ROS are simple "by-products" of the mitochondrial respiratory chain; (vi) the unnecessary postulation of "vicious cycle" hypotheses of mitochondrial ROS generation which are not central to the free radical theory of aging; and (vii) the role of DNA repair concerning endogenous versus exogenous damage. After considering the large body of data already available, two general characteristics responsible for the high maintenance degree of long-lived animals emerge: (i) a low generation rate of endogenous damage: and (ii) the possession of tissue macromolecules that are highly resistant to oxidative modification.

FIG. 1.

Endogenous antioxidants and rates of mitochondrial ROS production are low in long-lived animals. (A) Generalized relationship between maximum longevity of different animal species and the levels of endogenous tissue antioxidants (antioxidant enzymes or low-molecular-weight antioxidants). (B) Negative correlation between rates of mtROSp and longevity in mammals. These relationships have been observed in different tissues (for references see text). mtROSp, mitochondrial reactive oxygen species production.

FIG. 2.

Scheme of the respiratory chain showing sites of ROS production at complexes I and III. AA, antimycin A; c, cytochrome c; Q, ubiquinone; ROS, reactive oxygen species; ROT, rotenone.

FIG. 3.

Oxidative damage in mtDNA is low in long-lived animals. Oxidative damage to heart mtDNA decreases as longevity increases when comparing different mammalian species [reproduced with permission from Ref. (16); r=−0.92, p<0.001]. The same kind of relationship has been observed in brain mtDNA [Ref. (16); r=−0.88, p<0.016). mtDNA, mitochondrial DNA.

FIG. 4.

The membrane fatty acid unsaturation degree is low in long-lived animals. The double bond index, indicating the membrane fatty acid unsaturation degree, decreases as longevity increases in the mammalian heart (124), r=−0.78, p<0.02. The same has been observed in other tissues and kinds of animals (for references see text).

FIG. 5.

Mitochondrial ROS production, oxidative damage, DNA mutations and aging: schematic relationships suggested by available information. This scheme summarizes the main known causes and mechanisms of oxidative damage, finally leading to aging, that are known to be associated with animal longevity or DR in rodents. a* mtROS are produced at rates related to longevity, especially at complex I, which is inserted into the inner mitochondrial membrane. There is close vicinity or even contact between the site of ROS generation and mtDNA; so, antioxidants cannot interfere with ROS-induced final forms of irreversible damage in mtDNA and, therefore, cannot modify longevity; b* the well-known capacity of DR to decrease mtROSp and 8-oxodG in mtDNA is known to be exclusively due to the lower methionine content (MetR) of the DR diet. Around 50% of the longevity extension effect of DR is due to MetR and seems to work through decreases in mtROSp; c* Protox., protein oxidative modification (including protein glyco- and lipo-oxidation); d* LPox products, lipid peroxidation products (such as malondialdehyde, hydronenonenal) which are highly toxic and mutagenic; e* 8-oxodG, this DNA adduct does not accumulate extensively during aging, as it is repaired by mtBER, but can contribute to producing somatic mtDNA point mutations; f* these add to point mutations, deletions, and fragments of mtDNA directly generated by mtROS, as well as by other mechanisms; g* genomic instability: including nDNA point mutations, deletions, insertions of mtDNA fragments, changes in nDNA (e.g., TE activation), and modifications in intergenic, promoter, intron or exon gene sequences, chromosome rearrangements, modifications in gene expression, and cancer promotion. Solid arrows describe processes for which evidences in favor of a cause–effect relationship are abundant, whereas dotted arrows describe processes for which there is a logical mechanism but there are few data available supporting it. 8-oxodG, 8-oxo-7,8-dihydro-2′deoxyguanosine; BER, base excision repair; DBI, double bond index of fatty acids of cellular and subcellular membranes in tissues; DR, dietary restriction; MetR, methionine restriction; MLSP, maximum species-specific longevity; nDNA, nuclear DNA; TE, transposable elements.

FIG. 6.

Mitochondrial ROS production is not necessarily proportional to mitochondrial O₂ consumption: the reverse occurs in many physiological situations including exercise. In resting mitochondria respiring in state 4 (a), the electron flow rate is slow and the reducing potential of the respiratory chain (dark area inside respiratory complexes) is relatively high, which stimulates ROS generation; the low rate of O₂ consumption leads to high local pO₂, which, in turn, also contributes to increase ROS production because the K_M for O₂ of the ROS generator/s, contrary to that of cytochrome oxidase, is high and situated within the physiologic range of tissue pO₂. When saturating ADP is added (b), electron flow is strongly accelerated (state 3, active phosphorylating respiration) and the reducing potential of the respiratory chain decreases (smaller dark area inside respiratory complexes); in addition, the strong increase in O₂ consumption lowers the local pO₂; these two changes collaborate toward decreasing the mtROSp rate, which is barely detectable in state 3. For further explanation, see text. S, substrate; pO₂, partial oxygen pressure; ROS, mitochondrial ROS production (mtROSp) rate. The arrow thickness indicates the intensity of flux. Cx I to IV, respiratory complexes I, III, or IV. Reproduced with permission from Ref. (10).

FIG. 7.

The root of the degenerative diseases is the basic aging process: only factors causing aging can cause all these diseases. Degenerative diseases (by definition) are caused by the basic aging process as depicted in the figure by the lines connecting the “aging tree” with them. Therefore, if ROS are causal with regard to degenerative diseases, they should also be main causes of aging. The “minus” (−) symbol emphasizes that geriatrics tries to inhibit the diseases, while a main final goal of gerontology is to decrease aging rate.

FIG. 8.

The gene cluster hypothesis of aging and longevity. This scheme represents the gene cluster hypothesis of aging and longevity. Target genes producing proteins affecting the endogenous aging rate, degenerative diseases, resistance to external stress, fecundity, and other traits can be organized in clusters working through transcriptional cascades and complex interactions. The controller master genes (m) situated at superior hierarchical levels in space or time produce regulatory proteins (ovals) that control the graded expression of other genes. Regulatory proteins that would contain similar DNA binding sequences are depicted with the same shading. The grouping of genes controlled by similar regulatory proteins shown in the figure is only one of the many possible combinations, and is arbitrarily shown only as an example. Gene expression would also be influenced by other actors (e.g., upstream promoters, enhancers) not shown in the figure. The graded activation/repression of the target structural genes will finally affect aging rate as well as other traits needed for final expression of a high longevity (low incidence of degenerative diseases and high resistance to external sources of stress). Interrelations among genes in the cluster are expected to be much more complex than depicted, and would include crossed regulations both at horizontal level, and at vertical levels spanning more than one level per relationship. The real number and kinds of final target genes should be much greater than shown in the figure, and the master genes at the upper control level can be multiple, although their number should be much smaller than the number of target genes. This is most interesting for future possible manipulations that are aimed at greatly increasing maximum longevity. According to present knowledge, the target genes included in the figure should be present in the real cluster, although not necessarily in the sub-clustering tandem positions shown, which were arbitrarily chosen as one among many possible combinations. “a”: horizontal, multilevel, or single-level hierarchical interactions, and overlapping of regulatory elements; “b”: hypothetical example of two genes clustered in tandem in the same region. Reproduced with permission from Ref. (11).

FIG. 9.

Free radical leak in long-lived birds and in dietary restricted rodents. The %FRL in the respiratory chain of heart mitochondria is lower in pigeons than in rats [left, Ref. (59)] as well as in DR compared with ad libitum-fed (AL) rats [right, Ref. (50)]. %FRL, percent free radical leak.

FIG. 10.

Mortality and aging. Mortality increases exponentially with age. Such increase occurs much earlier in mice than in men.

FIG. 11.

A progressive, mainly linear process (aging), can lead to an exponential final result. Essentially linear decreases in physiological functions (F) lead to exponential increases in mortality (decreased survival) during aging, due to organism architecture including decreases in redundancy of components and existence of minimum thresholds to reach detrimental effects (see text). Therefore, there is no need to postulate vicious cycle hypotheses for the basic causes of aging.

FIG. 12.

Relationship between mitochondrial ROS production, steady-state oxidative DNA damage, and its repair, in species with different longevities and dietary restricted rats. Both long-lived animal species and dietary-restricted rodents have low mtROSp rates and low 8-oxodG levels in mtDNA in main internal organs including those containing port-mitotic cells. In agreement with this, it has been found that their base excision repair (BER) activities that repair DNA damage coming from endogenous origin are also low in internal organs (B). The contrary is true in short-lived species and ad libitum-fed rodents (A). This was predicted using a similar model in Fig. 2 of Ref. . Now there are supporting published data for this model both in species with different longevities and in DR rats. In contrast to BER, the repair of exogenous damage after UV irradiation in mitotic skin fibroblasts is known to be higher (instead of lower) in long-lived than in short-lived animal species (reviewed in Ref. 34) For references, see text.