Is it possible to program aging (3)
Ability to develop
Continuation. The beginning of the article is here.
Skulachev (Skulachev, 1997)
Many proponents of the theories of programmed aging believe that an aging species has a breeding advantage due to faster evolution. For example, in 1997, Skulachev wrote: "death caused by aging cleanses the population of its predecessors and frees up space for offspring with new useful properties." This is reminiscent of the idea of Weismann, to which Skulachev refers. The difference is that according to Skulachev, aging eliminates completely healthy individuals from the population as a whole in the hope that this will be compensated by the appearance of newborns with favorable mutations. The problem with this idea is that most mutations are unfavorable, so the next generation is not automatically more adapted. It is unclear how, in the absence of aging (which should be recognized as the original state, if the proposed theory is not cyclical), the mechanisms of death described by Skulachev can actually affect the "predecessors". Even if they could, the elimination of chronologically old individuals can be a process that selectively removes organisms whose fitness is above average, since surviving to a later age is generally a sign that an organism has a genotype that provides increased fitness. Skulachev does not offer any quantitative justification for his hypothesis, and upon closer examination, following his arguments would lead to an unforeseen consequence, manifested by the fact that in reality aging slows down evolution by selectively eliminating individuals with innate increased fitness.
Goldsmith (2008)
Goldsmith proposed another hypothesis based on the idea that programmed aging enhances the ability to develop. He pays attention to species with sexual reproduction and notes that sexual recombination generates high genetic variability in the population, which, according to Goldsmith, should accelerate the process of evolution. If aging reduces the average life expectancy, it also shortens the lifetime of a generation. Thus, sexual recombination provides the opportunity to "test" more genetic variants over the same period of time. This hypothesis considers various selection forces (the negative effect of shortening life expectancy versus the hypothetical positive effect on genetic variability) acting in opposite directions and, accordingly, a mathematical model is required to visualize and understand its adequacy. Unfortunately, Goldsmith (2008) formulated his idea solely in words and the authors decided to analyze its plausibility by developing an agent-oriented computer simulation of its consequences.
Agent-oriented computer simulations are a modern version of a cellular automaton in which agents (in this case, individuals included in the population) live in a 2- or 3-dimensional environment and follow an arbitrarily complex set of rules. This type of modeling is particularly well suited for analyzing population-level issues, since it automatically takes into account spatial effects that are important for studying phenomena based on kinship or group selection.
For their simulations, the authors used MASON, an excellent free software library based on the Java language designed for large–scale, computationally complex simulations. In the model they created, the agents live in a 2-dimensional world (the step of the coordinate grid is 250x250), the faces of which are cyclically connected in such a way that they form the surface of the torus. In this model, agents follow three rules that allow them to move, reproduce and die. The movement rule creates a slow diffusion of agents, simplistically reproducing the movement of animals in the wild. Agents either move to unoccupied vacant fields or swap fields with neighboring agents. Historically, four neighboring agents, located to the north, south, west and east, form the so-called von Neumann neighborhood, and 8 neighboring agents, including four diagonal corners, are Moore's neighborhood. Reproduction depends on reaching the age of maturity and a set of genes that may have significance (alleles) ranging from 0 to 1, additively contributing to the overall probability of reproduction (fertility). As suggested by Goldsmith (2008), sexual reproduction of individuals occurs due to the choice of a mating partner from the neighborhood of Moore, with whom genetic material is exchanged. Thus, each of the parents passes on 50% of the genes to the offspring. Finally, agents may die due to age-dependent mortality from external causes (y) or due to reaching the maximum possible lifespan (simulation of programmed aging). Table 1 summarizes this set of rules.
Table 1. A set of rules describing the behavior of agents in a simulation testing the idea of the ability to develop, expressed by Goldsmith (2008). "A" corresponds to the current agent, "E" corresponds to an empty field, and "X" corresponds to the contents of any field. The index "N" corresponds to a field in Moore's neighborhood, and the absence of an index is the field where the agent is currently located. For example, according to the first rule, the neighborhood field, which can have arbitrary content, is filled with the probability of mProb by the current agent. In addition, the field where agent "A" is currently located is filled with substance "X" (derived from X N).
In an agent-oriented simulation, thousands of individuals compete with each other using the same set of rules, at the same time they have individual sets of genes that control reproduction. If this process of evolutionary selection continues for thousands of time steps (in this case, each step corresponds to 1 year), alleles that poorly contribute to fertility will gradually be lost. To prevent this, the authors reverse the direction of selection at regular intervals (while the low significance of the allele ensures high fertility), which can be done using the graphical user interface (Fig. S1).
The authors believe that this is also consistent with Goldsmith's idea, since his arguments imply that higher genetic variability will accelerate evolution, especially under changing environmental conditions. Competition and selection taking place in the world of agents also provide a natural method for studying the evolution of a genetically programmed maximum lifespan. Each agent has not only its own inherited fertility gene complex, but also its own inherited maximum lifespan. However, when a descendant is formed, its maximum lifespan "mutates" in such a way that it becomes for a short period (epsLifespan) more or less than the lifespan of its parents. Subsequently, selection determines in which direction the programmed maximum life expectancy of the population changes.
Figure 2 shows typical results of such simulations. The diagram represents the results of three simulations that started using agents with a maximum lifespan of 20, 50 or 100 years. Genetically programmed life expectancy always increases during evolution as a result of the process of choosing between individuals having a shorter and longer life. The authors also conducted simulations with different mortality rates from external causes, different numbers of fertility genes, and different time intervals between selection direction switches (data are not provided). As a result, they did not find any combination of parameters that would contribute to shortening the maximum lifespan (maxLifespan). As can be seen in the diagram, the lower the starting value of maxLifespan, the stronger its increase over 50,000 time steps. This is exactly in line with expectations, as the power of natural selection weakens with chronological aging. Therefore, the increase in maxLifespan will slow down, but it will never stop.
Figure 2. According to Goldsmith (2008), aging should evolve, as it enhances the ability to develop. However, computer simulations carried out by the authors demonstrate that selection forces always increase the maximum life expectancy (this figure shows for three different starting values of maxLifespan).
Where is the weak spot in Goldsmith's argument? Evolution is short-sighted, since the possible benefits for the distant future do not provide an immediate beneficial effect. If a change in the environment (in this case, a change in the direction of selection) occurs on a time scale far exceeding the life span of individuals, the selection pressure does not provide preparation for such distant events. At the same time, if changes in the environment occur on a time scale comparable to the life span of representatives of the species, there is not enough time to reduce genetic variability in the population. In any case, the programmed limitation of life expectancy has only disadvantages (manifested by the destruction of agents) and does not have a compensatory advantage.
As the authors have shown, Goldsmith's assumption, according to which programmed aging provides an evolutionary advantage by accelerating evolutionary progress, is untenable in its own framework. However, there is a more general objection to this hypothesis, according to which the rate of appearance of offspring (and, accordingly, the ability to form new adaptations) is determined by the duration of the maturation period, and not by the duration of the period of achieving physiological aging, as well as the rate of genetic recombination and/or the appearance of mutations in germ line cells. Despite the existence of a tendency to correlate between the period of maturation and life expectancy, it is obvious that selection has a stronger effect on the first factor, as well as the fact that aging itself is unlikely to be a large potential consequence. A very intriguing question is whether there is an optimal rate of genetic recombination and/or the appearance of mutations in germ line cells. This issue has been actively discussed in publications devoted to the evolution of the sexes, but practically not considered in the context of programmed aging.
Continuation: Analytical models
Portal "Eternal youth" http://vechnayamolodost.ru
03.11.2016