18 June 2005. The benefit of caloric restriction (CR) on health and longevity was one of the major themes at the 34th annual meeting of the American Aging Association, held June 3-6 in Oakland, California. CR can increase lifespan in yeast, fruit flies, and mammals by up to 20 percent or more, and it can also lead to healthier living (see Oakland report on CR in primates). So how do fewer calories translate into health benefits?
To answer that question, many people have looked to the mitochondria. These organelles are not only the major power plants in most eukaryotic organisms, consuming most of the calories to keep up with demand for life-sustaining adenosine triphosphate (ATP), but they are also primary polluters, pumping out toxic reactive oxygen species (ROS), which have been considered a causative factor in a variety of age-related disorders, including Alzheimer and Parkinson diseases (see ARF related news story). Mitochondria are also the major instigators of apoptosis, a suicidal spiral in which many damaged cells find themselves caught up.
So what impact does aging have on mitochondrial health, and how might caloric restriction change that? Michal Jazwinski, from the Louisiana State University Health Sciences Center in New Orleans, addressed the first part of the question. Using yeast as a model system, he has studied the relationship between longevity and fitness of mitochondria.
Though it may appear at first blush that yeast can grow ad infinitum, they do age. There is a limit to the number of times a yeast cell can bud or divide to yield a daughter cell, for example. This limit is commonly called the replicative lifespan. However, like humans, not all yeast live to the ripest of old age. Some small yeast cells, or “petites,” always seem to live longer than their larger relatives.
It has been known since the 1950s that these petites are missing some of their mitochondrial DNA, thus linking these organelles to longevity. Jazwinski described how dysfunctional mitochondria set off what is known as the retrograde response, a complicated signal transduction pathway regulating the expression of many genes, which has been traced largely by work from Ron Butow’s lab at the University of Texas Southwestern Medical Center in Dallas. A protein called Rtg2, which triggers the transcription of many genes, is primarily involved in initiating the response. Jazwinski and colleagues found that activation of the retrograde response increases replicative lifespan in normal yeast. The response also seems to increase reproductive fitness—as yeast get older they take longer and longer to produce daughter cells, but if the retrograde response is triggered, this increase in regeneration time is postponed, again suggesting that the retrograde response protects yeast from the ravages of aging.
So might there be something akin to the retrograde response that occurs during normal aging? Jazwinski’s group found that there is an increase in mitochondrial mass as yeasts age, but that the increase in electric potential across the mitochondrial membranes is fivefold less, suggesting that, on average, mitochondrial membrane potential is fivefold lower than in young yeast cells. Concomitantly, there is a proportionate induction of the retrograde response. Life extension is more pronounced the greater the activation of the retrograde response. The results suggest that the mitochondrial membrane potential may be related to longevity—drops in mitochondrial membrane potential have, in fact, been linked to neurodegeneration (see ARF related news story). The extent of the retrograde response also increases with age and is also higher in petite cells, which suggests, said Jazwinski, that the retrograde response is not simply a reaction to the absence of some mitochondrial DNA, but is actually some kind of compensatory mechanism for mitochondrial dysfunction (which may or may not be a result of genomic changes). The nature of the signal that triggers this response is as yet unclear, he said.
Petite cells, though they live longer than regular yeast cells, are also packed with extrachromosomal ribosomal DNA circles (ERC). These circles are normally deleterious to yeast, but petite cells seem to accumulate them at over twice the normal rate. Jazwinski showed that these circles are also related to the retrograde response because Rtg2, the response trigger, can also suppress the formation of ERCs. However, it appears that Rtg2 is not that good at multitasking, so when the retrograde response is set to full steam ahead, as when the mitochondria are not functioning normally, ERCs accumulate faster, Jazwinski reported. These recent findings from his lab, recently reviewed in the journal Gene (see Jazwinski, 2005), link mitochondrial dysfunction, compensatory catabolic mechanisms such as the glyoxylate pathway, genomic instability, and chromatin-dependent gene activation.
Similar mechanisms might be at work in humans. Jazwinski suggested that the transcription factor myc might be a human equivalent of yeast Rtg2 because myc expression is induced in several human cell lines that have naturally occurring losses of mitochondrial DNA (see Miceli and Jazwinski, 2005). Also of interest is the fact that in yeast, Rtg2 complexes with the SAGA/SLIK multisubunit histone acetyltransferases. This links the life-lengthening retrograde response with histone acetylation, and perhaps more interestingly, with polyglutamine diseases such as spinocerebellar ataxia—Sgf73, a homolog of human ataxin-7, is a component of yeast SAGA/SLIK. In fact, it has just been shown that though a polyglutamine expanded ataxin-7 can lead to normal assembly of SAGA, the complex is devoid of acetyltransferase activity and fails to acetylate nucleosomes (see McMahon et al., 2005).
Judd Aiken, University of Wisconsin, Madison, also addressed the impact of mitochondrial fitness on normal aging—this time in mammalian skeletal muscle tissue. Muscles, by their very nature, consume vast amounts of energy and are packed with mitochondria. But some muscle fibers, such as in the soleus which flexes the foot, and the vastus lateralis which extends the leg, undergo a gradual atrophy as animals age. Could this be related to dysfunctional mitochondrial?
To test this theory, Aiken and colleagues looked for changes in activity of mitochondrial electron transport chain (ETC) components in and around sites of sarcopenia, the age-related muscle wasting that affects many elderly. The researchers found that atrophy correlates with loss of cytochrome c oxidase (COX) and hyperactive succinate dehydrogenase (SDH).
During aging, the cross-sectional area of the muscle fiber shrinks, but this shrinkage is not uniform. Aiken showed how the COX/SDH phenotype was normal in fibers that are of normal size, but is abnormal in areas where the fiber has thinned out most. Intriguingly, in muscles that do not undergo age-related atrophy, for example, the adductor longus, which helps squeeze your thighs together, there appears to be no age-related change in COX or SDH activity, suggesting that some muscles are protected against electron transport changes that might bring on sarcopenia.
Could spontaneous mitochondrial mutagenesis explain muscle atrophy? Aging has been shown to have a profound effect on the quality of mitochondrial DNA (see, for example, ARF related news story on the effect of aging on mitochondrial promoters and how mitochondrial fitness may impact nuclear DNA). Using laser capture microdissection, Aiken and colleagues have isolated individual fibers from human vastus lateralis and have found that they harbor abnormal ETC genes. Point and deletion mutations accrue in this muscle, but can be different from fiber to fiber, he said.
One question that does need to be addressed is the true level of mitochondrial mutations in any given fiber. Remember, each muscle cell is packed with mitochondria, which may have both qualitative and quantitative differences in DNA content. Aiken and colleagues are still working on this, but have found that near sarcopenia-related fiber breaks, up to 90 percent of mitochondrial DNA can be mutated in some form.
Might a slowdown in age-related mitochondrial changes explain how caloric restriction increases longevity? Brian Merry, from the Institute of Human Ageing at the University of Liverpool, England, addressed this possibility.
One theory linking caloric restriction to mitochondria suggests that decreased levels of reactive oxygen species might explain the increase in lifespan elicited by eating fewer calories. But the evidence supporting this is weak, said Merry. Experiments show, for example, that the metabolic rate in CR animals is no different from controls once they have become accustomed to their lower calorie diet (though see Sohal et al., 1994). And Merry and colleagues (see Lambert et al., 2004) showed that state IV respiration is normal in mitochondria isolated from the various organs of CR animals, including brain, liver, and heart (state IV respiration, measured in the absence of ADP, is when production of ROS by the ETC is at its highest). So what explains such findings that superoxide and hydrogen peroxide are lower in mitochondria isolated from the heart and liver, respectively, of CR animals?
To investigate this question, Merry and colleagues focused on the proton leak across the mitochondrial membrane. They found that for a given inner mitochondrial membrane potential, the proton leak across the membrane was higher in organelles isolated from CR animals than from controls. This was unexpected, suggested Merry, because CR has been shown to decrease ROS, which are generated by leaks in the electron transport chain. And, in fact, Merry acknowledged that other researchers, including Jon Ramsey and Richard Weindruch at the University of California, Davis, just showed the opposite, that leak across the membrane is lower in CR-derived mitochondria (see Hagopian et al., 2005).
But though the CR mitochondria have no change in stage IV metabolic rate, Merry did find that the resting membrane potential is actually lower across these mitochondria membranes than across those from control mitochondria. It is this lower membrane potential that explains the reduced production of ROS in CR, suggested Merry, because it is the extent to which the ETC components are reduced, rather than the flow through the chain, that determines the rate of formation of ROS, he contends.
Whether this is the correct explanation for reduced ROS in CR remains to be seen. However, Merry did have one other interesting observation. He showed that insulin, which is dramatically reduced in the plasma of CR animals (see related Oakland news), lowers leakage across the membrane, but at the same time increases the membrane potential. So if the overall reduction rather than flux through the ETC turns out to be the key to ROS production, then CR, insulin, and ROS may be inextricably linked.—Tom Fagan.