[Back to Number 11 ToC] [Back to Journal Contents] [Back to Biokhimiya Home page]

Aging is a Specific Biological Function Rather than the Result of a Disorder in Complex Living Systems: Biochemical Evidence in Support of Weismann's Hypothesis

V. P. Skulachev1

1Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, Moscow, 119899 Russia; fax: (095) 939-03-38; E-mail: skulach@head.genebee.msu.su

"All roads are open for young talents."
V. Lebedev-Kumach

"Then you are robbed and naked,
Death does not come quickly,
Life, not yet coming to the end,
Gradually slows down its pace."

A. Blok

Submitted September 6, 1997.

A concept postulating that aging is a specific biological function that promotes the progressive evolution of sexually reproducing species is reviewed. Death caused by aging clears the population of ancestors and frees space for progeny carrying new useful traits. Like any other important function, aging is mediated by several molecular mechanisms working simultaneously. At least three such mechanisms have been postulated thus far: 1) telomere shortening due to suppression of telomerase at early stages of embryogenesis; 2) age-related activation of a mechanism that induces the synthesis of heat shock proteins in response to denaturing stimuli; and 3) incomplete suppression of generation of reactive oxygen species (ROS) with inadequate scavenging of already existing ROS. None of these phenomena can kill the organism, but only weaken it, which becomes crucial under certain extreme conditions. This mechanism of age-induced death can be compensated for (within certain time limits) by several positive traits that greatly increase the evolutionary potential of species capable of performing this function. Similarly to apoptosis (programmed cell death), the programmed death of the body can be called "phenoptosis". Aging presumably belongs to the category of "soft" (extended in time and allowing a certain degree of compensation) phenoptosis, in contrast to "acute" phenoptosis; the death of salmon females immediately after spawning is a good example of the latter.
KEY WORDS: telomere, telomerase, heat-shock proteins, superoxide, mitochondria, aging, apoptosis.

When Weismann formulated his "tough" hypothesis postulating that the death of ancestors frees room for better adapted progenitors [1], he relied especially on considerations of biological expedience. Indeed, every new trait appearing in the offspring would be inevitably diluted by the ocean of old traits carried by parents, grandparents, etc., if they were immortal. In other words, immortality of specimens belonging to any biological species would block the progressive development of that species.

At the cellular level, programmed cell death is now a phenomenon well documented by direct evidence. In this context, the discovery of Hayflick's limit, that is, the existence of a finite number of divisions of certain somatic cells [2-4] is the first to be mentioned. Apoptosis is another good example. Cells infected by a virus appear to activate a special suicide mechanism whereby the cell degrades its own polymers, primarily DNA and specific enzymes, and then dies to burrow the infecting virus. Monomers released in this process are used by other, noninfected cells as construction material. It was also shown that such a mechanism permits the body to get rid of cells and even organs that have become hazardous or simply unnecessary. Apoptosis causes the rejection of the tadpole's tail and shedding of plant leaves (whence the name of the process is derived). Apoptosis plays a key role in immunity, preventing the formation of antibodies to autologous proteins. It was also suggested that apoptosis eliminates cells dangerous to multicellular organisms in many other cases [5-8]. However, certain cells can live for decades and conserve their native functions. Human oocytes are an example of this kind of cells.

A recently proposed hypothesis postulates that a mechanism similar to apoptosis operates at the subcellular level to eliminate mitochondria that produce superoxide in amounts capable of damaging the cell [7-9]. I suggest that this process, similarly to apoptosis, should be called mitoptosis.

Recent data strongly suggest that aging is nothing but programmed death at the supracellular level (the level of the whole organism). In the same terminological system, this event could be called phenoptosis.

Telomeres and Telomerase

A. M. Olovnikov formulated the problem of terminal underreplication of linear DNA molecules in 1971 [10-12]; this phenomenon is caused by the inability of DNA polymerases to replicate several nucleotides at 3´ ends of DNA templates. Olovnikov also suggested that a specific biological mechanism should normally prevent this phenomenon. This mechanism was expected to be active in gametes, cancer cells, as well is in cells of vegetatively reproducing organisms. In most other cases, e.g., in many human somatic cells, this mechanism is suppressed.

Further studies revealed an enzyme called telomerase [13, 14] (whose existence had been predicted by Olovnikov) that compensates for DNA shortening in the mentioned cell types. The function of telomerase is to add a repeated sequence (in humans, the hexamer TTAGGG), which forms the so-called telomere, to ends of nuclear DNA. After this, underreplication of the linear DNA molecule only shortens this nontranscribed sequence of the telomeric fragment of the chromosome without damaging the genetic information or the mechanism that reads it.

At certain stages of development in early embryogenesis, the gene encoding telomerase in the majority of human somatic cells is switched off, thereby making the genome susceptible to shortening. The telomere shortens at a low but appreciable rate which impairs the functioning of the chromosome. This impairment begins long before the disappearance of the whole telomere, which removes protection from genetic information contained in transcribed regions. The telomere, in addition to protection from the loss of genetic material during replication, plays a structural role in the spatial arrangement of chromosomes in the nucleus and their correct functioning [15-17]. This role is still poorly understood.

There is close correlation between shortening of telomeric DNA regions and Hayflick's limit [4, 16]. To surpass this limit and continue reproduction, the cell should activate its telomerase gene [4, 16].

The most remarkable feature is that switching off the telomerase gene is an ontogenetic stage that occurs at a distinct time point in the life of an organism and involves only some of its cell types. This event seems to perform a specific function and cannot be regarded as a disorder in the living system or a kind of unpredictable age-related defect, although it clearly promotes aging. In this context, we should be reminded of an observation made in experiments with barley germs. During the development of the germ, the telomere suddenly loses 50 kb. It loses an additional 20 kb during growth of the spike [17]. The mechanism responsible for this event remains unknown. If the arrest of telomerase synthesis is considered as an accidental fault, telomere shortening in cells containing no functional telomerase appears to be an act of deliberate damage to the organism. An alternative interpretation is that the inhibition of the telomerase gene and the telomere shortening are biologically important events.

To understand the meaning of this process, we should remember that telomerase genes of somatic cells are switched off only in sexually reproducing organisms, but not in vegetatively reproducing organisms [17]. It is exactly in this case that the appearance of a new trait, which can result from a combination of parental genomes, becomes the most probable event. However, in vegetatively reproducing organisms, the appearance of a new trait results from random mutations occurring in the same cell. Data collected in studies of telomerase could probably be explained by Weismann's hypothesis referred to in the beginning of this paper.

Induction of Heat-Shock Proteins

All living cells contain proteins used for repairing other proteins should they adopt an incorrect conformation. These repair proteins are called heat-shock proteins because their contents increases considerably upon changes in the ambient temperature that denature protein molecules [18, 19]. In addition to thermal stimuli, a similar effect can be caused by any other condition resulting in denaturation of cell proteins, e.g., by oxidative stress [13]. This phenomenon is mediated by the following mechanism: in response to a denaturing stimulus, the cell activates the trimerization of heat-shock factor I (HSF I), a specific protein normally existing in the cytosol in monomeric form. The HSF I trimer is transported to the nucleus, where it recognizes and activates heat-shock protein genes.

The inducibility of heat-shock proteins [19-21] and the ability of HSF I monomers to undergo trimerization [22, 23] under stressful conditions were found to be inversely proportional to the number of past divisions of fibroblast cell cultures in vitro. Further, the induction of heat shock and the inducing activity of HSF I in cells taken from aged animals are always inhibited in comparison to cells taken from young animals [24, 25], and the activity (but not the content) of HSF I decreases during aging [19, 26]. The aging effect is reversible in vivo after switching to a reduced-calorie diet [25]. These observations suggest that reactive oxygen species (ROS) are involved in this process because their levels depend on the amount of food substances oxidized in the body.

Reactive Oxygen Species

Reactive oxygen species (ROS) include superoxide anion (O2·-), singlet oxygen, H2O2, and hydroxyl radical (OH·). Superoxide is the primary ROS generated by one-electron reduction of molecular oxygen in humans and other animals. Superoxide dismutase converts superoxide to H2O2, whose further nonenzymatic conversion in the presence of Fe2+ or Cu2+ generates OH·, an extremely strong oxidant (its redox potential is approximately +1.35 V) capable of degrading nearly all types of organic compounds of biological origin.

One-electron reduction of oxygen can, in principle, occur by oxidation of any substance whose redox potential is lower than or equal to -0.15 V (the redox potential of the O2/superoxide pair). Compounds with high kinetic barriers of reaction with O2 were selected by evolution. Highly reactive coenzymes and prosthetic groups of enzymes operating at the initial and middle steps of the respiratory chain, such as coenzyme Q semiquinone (CoQH·) are exceptions to this rule. Being a one-electron transporter, CoQH· can probably commit errors, donating an electron to molecular oxygen rather than to cytochrome b1, which is its natural oxidant.

Cells have several mechanisms minimizing the side-reaction that generates superoxide and preventing its conversion to the extremely dangerous OH·. Among these mechanisms, there are a multilevel system of defense against ROS (which includes antioxidant substances), mechanisms preventing the accumulation of CoQH· and similar reductants, enzymes that decrease the intracellular oxygen concentration to reduce the rate of generation of superoxide, and systems eliminating mitochondria and cells overproducing superoxide because of various conditions (ROS-dependent mitoptosis and apoptosis) [8, 27]. In addition to these mechanisms, higher animals have physiological supracellular systemic effects, such as a decrease in lung ventilation and capillary constriction upon the transition from work to rest, when the demand for oxygen decreases abruptly.

Nevertheless, human and animal cells consume approximately 2% of oxygen by means of the reaction O2 --> O2·-, which is hazardous and senseless in terms of energy, rather than by the energy-producing and safe reaction of four-electron reduction of oxygen to water by cytochrome oxidase. Because of this, cells are not fully protected against ROS-induced damage. The estimated mean daily rate of oxidative damage to nuclear DNA is 10,000 in human cells and 100,000 in rat cells, the later having a higher respiratory rate. The rate of damage to mitochondrial DNA is at least one order of magnitude higher because it is located in close proximity to the respiratory chain, the major source of superoxide anion radical [28, 29].

Such a situation could be explained by imperfection of living systems. However, the actual situation seems to be far more complex. Indeed, how can we explain the existence of xanthine oxidase, an enzyme that oxidizes xanthine by molecular oxygen and produces superoxide in cells? This process is responsible for sterility of milk, because ROS are potent bactericidal agents. This is the function of extracellular xanthine oxidase. However, what about intracellular xanthine oxidase found in a number of tissues? General considerations (such as Kozma Prutkov's "Turpentine would be certainly good for something") suggesting that ROS are used by cells as second messengers in transduction of regulatory signals, and it is exactly for this purpose that they are generated by cells [29], will hardly explain anything. In most cases, ROS are signals of the same event--oxidative stress, that, is the generation of themselves. This is no surprise: ROS are too dangerous to be entrusted to any biological function except one--the suicide of a living system, such as a mitochondrion, cell, or organism.

Mitochondria may play a fatal role in all these events. It is the main site of generation of superoxide, and the production of these ROS can develop as a self-accelerating process. The more superoxide generated in the cell, the higher the probability of damage to mitochondrial DNA located in the mitochondrial matrix, that is, near the site of superoxide generation on cristae of the inner mitochondrial membrane. Damage to mitochondrial DNA impairs the synthesis of proteins that carry electrons in the respiratory chain. Inhibition of the respiratory chain, in turn, accelerates the production of superoxide, etc. Ultimately, the amount of superoxide generated becomes potentially dangerous to nuclear DNA, although it is located farther from the site of superoxide generation (the inner mitochondrial membrane) in comparison to mitochondrial DNA [30].

The existence of age-related increases in ROS production is a well-established fact [31]. In addition to damaging DNA, this process can affect the state of proteins, which are also targets of ROS. An increase in the extent of oxidative denaturation of proteins should be expected to be further accelerated in aging cells because this denaturation is no longer compensated by induction of the repair system of heat-shock proteins.

Aging as a Specific Form of Phenoptosis

In the above sections, I defined phenoptosis as programmed death of an organism. The most noticeable examples of phenoptosis are the death of salmon soon after spawning and the death of bamboo, the species that can undergo vegetative reproduction for decades, but, once flowering, dies in the same season. Suppose that aging belongs to the same class of events, but this kind of phenoptosis is extended in time because it is caused by an age-increased weakening of vital function, rather by their total arrest.

Such a mechanism of "soft" phenoptosis should be expected to have considerable advantages over the cases of "acute" phenoptosis (salmon and bamboo). The function of soft phenoptosis is to reduce the pollution of the population by long-living ancestors, thereby stimulating progressive evolution, facilitating the same function by an additional, very effective method. The appearance of a useful trait allows compensation of the effect of aging within certain time limits. A large-bodied deer, even after reaching an old age, has better chances to win a spring battle for a female or escape from a group of wolves in comparison to a younger but smaller conspecific animal.

The three cases of apparent imperfection of body defenses considered above (protection against DNA shortening, anti-denaturation protection of proteins, and protection of cellular matter from ROS) can be explained if we assume that all these cases are different mechanisms of phenoptosis. In light of this concept, the switch-off of telomerase genes occurring in early embryogenesis in specific tissues does not appear to be a random disorder in a complex biological system. Intracellular xanthine oxidase, a superoxide-generating enzyme, is no longer considered as an enzyme-offender or a dangerous but equally random error of evolution. Most probably, these events do not occur simply by chance. These awkward faces of nature appear be deliberate acts. By analogy, an experienced psychiatrist will never diagnose kleptomania in a patient who steals only money.

A common feature of all the three mechanisms considered above is that they only weaken life processes but never terminate any of them. Telomerase shortening impairs the functioning of chromosomal machinery. Decreased induction of heat-shock proteins impairs renaturation of other proteins. A small increase in ROS production increases the probability of damage to mitochondrial and then to nuclear DNAs.

The example of heat-shock proteins is especially illustrative. The amounts of these proteins and the regulatory protein HSF I do not decrease with aging, that is, aging does not abolish the current repair of cellular proteins started in young cells. However, the response to stress, that is, the process whereby the environment challenges the body under critical conditions, weakens. And it is exactly these situations that serve as important factors of natural selection.

Having assumed that aging is a particular case of phenoptosis, an important biological function, we should never be surprised by the ambiguity of responses to the question of what exactly is the causative factor of this process. Although strong evidence supports the hypothesis of a relationship between the length of telomeres and aging at the cellular level, proponents of the telomeric theory of aging surrender in disputes about aging of the whole organism. Consideration of the oxidative (mitochondrial) theory of aging often raises counter-arguments. These would certainly appear if someone defended the primary role of heat-shock proteins in the development of the senescent syndrome.

Ozawa observed in his recent studies that 89% of mitochondrial DNA in the heart mitochondria of a 97-year-old subject contained extensive deletions which were critical to replication and transcription [30]. Nevertheless, the subject died of gastric cancer rather than of heart failure. Clearly, further progression of mitochondrial DNA dysfunction would eventually lead to the death of this subject, and this would hardly be related to shortening of nuclear DNA. This death could never be explained by telomerase going on strike.

The multiplicity of mechanisms of aging helps species to dispose of "garbage" immortal specimens whose presence would prevent the manifestation of genetic polymorphism caused by sexual reproduction. In other words, Ozawa's senile patient would ultimately die because of shortening of his nuclear DNA even if he were cured of gastric cancer and conserved his native mitochondrial DNA until the 97th year of life.

Maslov [32] was right to state that all important biological functions are mediated by several independent mechanisms. This parallelism considerably increases the stability of living systems. Characteristically, a culture of immortal cells can be produced only by breaking at least three genetically different mechanisms. The last of these mechanisms is that of switching off telomerase [33]. The actual number of such barriers in the body can be even greater. However, the mere fact that their number should be finite can make those seeking human immortality optimistic.

I wish to thank A. M. Olovnikov with whom many years of discussions have made this article possible. I also thank M. N. Kondrashova, K. Lewis, and M. Yu. Sherman for their valuable advice and criticism on the concept of phenoptosis.


1.Weismann, A. (1882) Ueber die Dauer des Lebens, Fischer, Jena.
2.Hayflick, L., and Moorhead, P. S. (1961) Exp. Cell Res., 25, 585-621.
3.Hayflick, L. (1965) Exp. Cell Res., 37, 614-636.
4.Hayflick, L. (1997) Biochemistry (Moscow), 62, 1380-1393 (Russ.).
5.Kuchino, Y., and Mueller, W. E. G. (eds.) (1996) Apoptosis, Springer, Berlin.
6.Kroemer, G., and Martinez, A. C. (eds.) (1996) Apoptosis in Immunology, Springer, Berlin.
7.Skulachev, V. P. (1994) Biochemistry (Moscow), 59, 1910-1912.
8.Skulachev, V. P. (1996) Quart. Rev. Biophys., 29, 169-202.
9.Zorov, D. B., Kinnaly, K. W., and Tedeschi, H. (1992) J. Bioenerg. Biomembr., 24, 119-124.
10.Olovnikov, A. M. (1971) Dokl. Akad. Nauk SSSR, 201, 1496-1498.
11.Olovnikov, A. M. (1973) J. Theor. Biol., 41, 181-190.
12.Olovnikov, A. M. (1996) Exp. Gerontol., 31, 443-448.
13.Greider, C. W., and Blackburn, E. H. (1985) Cell, 43, 405-413.
14.Greider, C. W., and Blackburn, E. H. (1989) Nature, 337, 331-337.
15.Kurenova, E. V., and Mason, J. M. (1997) Biochemistry (Moscow), 62, 1453-1466 (Russ.).
16.Counter, C. M. (1996) Mutation Res., 366, 56-63.
17.McKnight, T. D., Fitzgerald, M. S., and Shippen, D. E. (1997) Biochemistry (Moscow), 62, 1432-1441 (Russ.).
18.Ritossa, H. (1962) Experientia, 18, 571-573.
19.Liu, A. Y.-C, Lee, Y.-K., Manalo, D., and Huang, L. E. (1996) Stress-Induced Cellular Responses (Felge, U., Morimoto, R. I., Jahara, I., and Polfa, B., eds.) Birkhauser Verlag, Basel, pp. 393-408.
20.Liu, A. Y.-C., Bae-Lee, M. S., Choi, H. S., and Li, B. (1989) Biochem. Biophys. Res. Commun., 162, 1302-1310.
21.Liu, A. Y.-C., Lin, Z., Choi, H. S., Sorhage, F., and Li, B. (1989) J. Biol. Chem., 264, 12037-12045.
22.Choi, H. S., Lin, Z., Li, B., and Liu, A. Y.-C. (1990) J. Biol. Chem., 26, 18005-18011.
23.Liu, A. Y.-C., Choi, H. S., Lee, Y. R., and Chen, K. Y. (1991) J. Cell Physiol., 149, 560-556.
24.Fagnoli, J., Kunisada, T., Fornace, A. J., Scheider, E. L., and Holbrook, N. J. (1990) Proc. Natl. Acad. Sci. USA, 87, 846-850.
25.Heydari, A. T., Wu, B., Takahashi, R., Strong, R., and Richardson, A. (1993) Mol. Cell. Biol., 13, 2902-2918.
26.Fawcett, T. W., Sylvester, S. I., Saryn, K. D., Morinoto, R. I., and Helbrook, N. J. (1994) J. Biol. Chem., 269, 32272-32278.
27.Skulachev, V. P. (1997) Biosci. Rep., in press.
28.Ames, B. N., and Shigenaga, M. K. (1992) Ann. N. Y. Acad. Sci., 663, 85-96.
29.Peskin, A. V. (1997) Biochemistry (Moscow), 62, in press.
30.Papa, S., and Skulachev, V. P. (1997) Mol. Cell. Biochem., in press.
31.Ozawa, T. (1977) Biosci. Rep., in press.
32.Maslov, S. P. (1980) in Levels of Organization of Biological Systems (Molchanov, A. M., ed.), [in Russian], Nauka, Moscow, pp. 8-19.
33.Duncan, E. L., and Reddel, R. R. (1997) Biochemistry (Moscow), 62, 1477-1490 (Russ.).