Brief outline of the epigenetic theory of evolution, or ETE for busy people. Column for Kompyuterra #128

←
Dmytro Shabanov
→

How do adaptive traits arise during evolution, or which theory of evolution is supported by modern genetics data? A brief outline of the epigenetic theory of evolution, or ETE for short. Selected correspondence with Russian and pro-Russian friends

{"translated_text":"←\nDmytro Shabanov\n→\n\nHow adaptive traits arise in the course of evolution, or Which theory of evolution is confirmed by modern genetics data?\nA brief outline of the epigenetic theory of evolution, or ETE for busy people\nUkraine is a large Vradiivka. Selected passages from correspondence with Russian and pro-Russian friends\n\nColumn for Kompyutterra #127\nColumn for Kompyutterra #128\nColumn for Kompyutterra #129\n\nTwo weeks have passed for me under the sign of sharing the shock of Ukrainian politics and debates about the epigenetic theory of evolution. About politics — not now; here we will discuss ETE. I regularly hear complaints about the absence of its brief outline. This column is my attempt at such an outline, taking into account the fresh experience of discussing ETE on the KT website, in Alexander Markov's LiveJournal, on my website, at the meeting of the \"Evolution\" club in Kyiv (here is an extended presentation of my report), at a round table with Sergey Yastrebov held during the youth conference of the Kharkiv biology faculty.\nYes, don't forget:\n— the fact of evolution is not disputed here; the discussion concerns problems of studying its mechanisms;\n— the views presented here have nothing in common with the ideas of T.D. Lysenko, \"intelligent design,\" \"scientific creationism,\" and other constructs that judge evolution based on ideological or religious dogmas;\n— this reflects my understanding of ETE; its creators and other supporters may have (and often do have) different opinions on many important questions for this theory.\nAnd one more thing. Let's clarify the terms.\nScience (particular) — a developing complex of notions about a certain aspect of reality, technologies for its study and modification, which may include various, including partially contradictory, hypotheses and theories.\nTheory — a holistic system of views in which some propositions are derived from others. A hypothesis can become a theory as it develops, explaining certain phenomena and possessing predictive value.\nEvolution — irreversible changes of biosystems during the historical time of the biosphere. Leads to changes in existing biosystems, including their complication, increase of their adaptation to the environment, growth of their stability, emergence of new properties in them, and appearance of new types of biosystems. Evolution is a multi-level process; populations, species, supraspecific groups, as well as communities and ecosystems evolve.\nEvolution of evolution — changes in the mechanisms of evolution as biosystems evolve.\nEvolutionary biology — the science that studies the mechanisms of evolution. The study of how exactly evolution proceeded is also often included in the competence of this science, but here it is used in the narrow, specified sense.\nOntogeny — individual development of an organism, the totality of its regular and random transformations during its life.\nSelection — preferential reproduction of individuals and their groups, depending on their properties; selection predominantly preserves and reproduces more adaptive individuals.\nAdaptation — the correspondence of the organism to the environment in which it develops, enabling it to successfully complete ontogeny and leave offspring.\nSo, as you understand, evolutionary biology is a science that includes many theories. Its development is not complete, and a complete picture of the mechanisms of the multi-level process of evolution does not exist today. Considering the history of evolutionary biology, we can see that ideas accepted by the majority of scientists spread in it, and times of disagreement followed. To describe them, I will use the scheme proposed by N.N. Vorontsov (filling in the last row of the table, I will get ahead of myself, reflecting what I come to in this column).\n[IMG_1]\nOne of the theories that appeared during the crisis of synthesis II became ETE. It is based on results obtained in the 1940s and 1950s by Soviet zoologist I.I. Shmalhausen (theory of stabilizing selection) and English geneticist C.H. Waddington (epigenetic landscape and \"genetic assimilation of modifications\"). The foundations of ETE were formulated by Moscow paleontologist M.A. Shishkin in works published from 1984 to 1988. Contributions to the development of the theory were also made by his colleagues A.P. Rasnitsyn (metaphor of adaptive compromise) and A.S. Rautian (evolution as maintenance of stability).\nDescribing ETE, it is compared with STE, the synthetic theory of evolution, meaning precisely the relatively holistic theory that took shape by the mid-20th century. Why? Modern evolutionary biology is a rather loose and to some extent internally contradictory complex of concepts. Each of them more or less explains some complex of factors, ignoring other data. However, STE, due to its simplicity, remains the default version to this day: it is what is taught in schools and universities, trying to identify it with evolutionary biology as a whole.\nIt is time to give a brief description of ETE.\n\nThe epigenetic theory of evolution considers evolution as a process of replacement of one stabilized ontogenetic pathway by another. In representatives of highly organized groups, the result of ontogeny is determined by an extremely complex complex of factors and the results of their interaction.\nOntogeny is influenced by the interaction of the following factors and their effects:\n\nhereditary determinants:\ngenetic (sequences of nucleotides in nucleic acids — NA);\nepigenetic (chemical and spatial modifications of NA macromolecules);\nothers (related to the organization of the cytoskeleton, the set of RNA and regulatory molecules, protein conformation, etc.);\nvarious environmental influences;\nchance.\n\nThe result of ontogeny cannot be predicted unambiguously. It can only be characterized by a distribution of probabilities of various outcomes, among which the norm (the state maintained by stabilizing selection) and various morphoses (deviations, aberrations) should be distinguished. The metaphor describing the distribution of possible outcomes of ontogeny is C.H. Waddington's epigenetic landscape. From this point of view, possible ontogenetic pathways can be described as a set of stabilized sections (creodes), bifurcation points, and sets of improbable and unstable states separating creodes. Stabilizing selection — the preferential preservation and reproduction of individuals whose ontogeny led to the norm — increases the stability of the development of the norm (increases its probability). This stability grows both due to the increase in the equifinality of the development of the norm (the ability to realize the norm in increasingly different individuals) and due to the increase in the autonomy of such development (the ability to realize the norm in increasingly different environmental conditions). This is ensured by the fact that selection rebuilds the entire system of ontogenetic control (and the genotype in particular). In the epigenetic landscape model, the action of stabilizing selection looks like the deepening of the corresponding creode.\n[IMG_2]\nIf the nature of selection changes and it ceases to support the former norm, its development destabilizes and various morphoses appear. If some morphosis turns out to be adaptive, selection selectively preserves those systems of ontogenetic control that led to such an adaptive state. Offspring of such individuals will more likely be adaptive if their ontogeny leads to the same result. Therefore, selection will support those offspring in which the development of the state adaptive under these conditions becomes increasingly stable (increasingly probable). The result becomes an increase in the stability of the development of the morphosis supported by selection, that is, an increase in its heritability. Thus, the very phenomenon of heredity turns out to be a result of selection.\nA typical case of norm replacement in the course of evolution should be considered one in which an adaptive morphosis arises as an adequate response of the ontogenetic control system to changed developmental conditions. If selection supports such a morphosis, it becomes a new norm, its development becomes autonomous and acquires independence from specific external influences.\nIn the course of evolution, the system of ontogenetic control becomes more complex and the mechanisms for finding adaptive morphoses when the nature of selection changes improve. The emergence of genetic inheritance, sexual reproduction, cultural inheritance, complex social organization — some stages of this process.\n\nHow does the described approach differ from the STE approach? For STE, evolution is a rebuilding of the genotype as a result of selection based on the results of gene activity reflected in organism traits. That is why STE tries to describe ontogeny as a set of relatively independent cause-and-effect relationships. Any interactions that complicate the transmission of information from genotype to phenotype are, for STE, simply obstacles that hinder the rebuilding of the genotype based on its phenotypic effects.\nIf information from the genotype is reflected in the phenotype directly, the STE mechanism works quite satisfactorily. The Hardy-Weinberg equation describes how the ratio of alleles (alternative versions of one gene) in offspring depends on that in ancestors. The mathematical apparatus of STE is based on the fact that each allele makes a constant contribution to the final fitness of its bearer. If an allele increases the fitness of the organism, selection will increase its frequency and over time the favorable allele will displace its alternatives. There are cases when such a model works. Consider two bacterial strains. One grows faster but is unstable to antibiotics. The other is resistant and pays for this with slower growth. Their traits unambiguously reflect their genotype. Their dynamics in media with different antibiotic contents are well described by the selection equations of STE.\nIf the STE model is correct, the evolution of evolution should lead to the phenotype reflecting the genotype more and more effectively, more \"transparently.\" In such organisms, selection will rebuild the genotype particularly efficiently. They will begin to develop adaptations faster and will gain an advantage in a variable environment. And indeed, organisms with complex interactions of various factors in ontogeny should evolve slowly. Particularly \"hard\" — species consisting of long-lived and low-fertility individuals. Is this prediction borne out?\nNo! I wrote about this, remember?\nIn the world we observe, the most complex, the fastest-evolving organisms in terms of changes in their structure and behavior turn out to be completely improbable from the STE perspective. Take humans, for example. Our genotype contains very little information, fitting in unarchived form on a CD. The actual genes constitute a small part of it; about 25,000 of our genes require less than 10 megabytes to record (and an archiver will compress them much more). I recall the well-known joke that a file with a detailed description of the shape of the kneecap (one of our simplest bones) in AutoCad will take more space. The information necessary to describe our structure is immeasably greater than the capacity of our genome. Even more amazing is the array of information in our psyche. This means that in our ontogeny, a dizzying number of choices of possible developmental pathways occur with the memorization of their results.\nA typical path of evolutionary change for ETE is one that corresponds to the logic of G.K. Waddington's experiments. Waddington induced morphoses in experimental animals using external influences. Offspring from crossing carriers of morphoses were again subjected to similar influences, and again carriers of the same morphosis were selected for reproduction. After a small (first tens) number of generations, these morphoses began to develop without specific influences. An unstable, environment-dependent pathway of development became stable. To verify that this is not about the inheritance of acquired traits, compare these experiments with Weismann's classic experiments.\n[IMG_3]\nIn the language of STE, Waddington's experiments are described with strain. Selection for the ability to develop a certain modification (a non-heritable trait) led to the change in many modifier genes and ultimately to the \"genetic assimilation of the modification,\" the transfer of the control of development of this trait to the genotype. This explanation presupposes some modifier genes, not found by genomics. It presupposes the rapid evolution of these modifiers, not corresponding to the models of selection developed in STE itself. If we accept that it is not about virtual modifiers but about other structural genes, it becomes unclear why the \"genetic assimilation\" of the control of one trait does not lead to chaos in the development of others.\nIn the language of ETE, Waddington's experiments are described simply. By supporting the morphosis, stabilizing selection leads to an increase in the stability of its development. And — note! — no strained attempts to present the organism as a sum of traits and the genotype as a sum of genes. Waddington's experiments describe not the transformation of \"non-heritable\" traits into \"heritable\" ones, but the influence of selection on the stability of development. And, by the way, the example of humans shows that often \"non-heritable\" traits (what we learn) are no less important for us than some others.\nNow we can discuss the variety of assessments of ETE. They are very different. One pole is that ETE is a complete alternative to STE, and no compromise between them is possible. At the other pole are those who declare ETE pseudoscience and even try to prohibit mentioning the names of its supporters in the presence of students, so as not to plant doubts about the unalternative truth of STE. Believe me, I write about such attempts not speculatively, but based on sad experience... My assessment is closer to the first pole, although it differs from it. I believe that ETE can become the core of synthesis III: only this theory has the potential to explain what happens at the organismal level.\nOf course, ETE retains many insufficiently developed questions. One of them is the description of the diversity of traits from the point of view of the regulation of their development in ontogeny. Probably, even in complex organisms there are relatively simple traits that almost unambiguously depend on the state of individual alleles. In these cases, STE models will describe the evolution of such traits relatively adequately. The breakdown of normal developmental pathways is probably regulated more simply than the functioning of existing gene mechanisms in the tissues where they usually work. But the emergence of fundamentally new traits cannot be explained by such mechanisms...\nSo, in my opinion, ETE is a broader generalization than STE, and the cases when the STE approach proves applicable can be considered (in the apt expression of S. Yastrebov), as degenerate (simplified) cases of the applicability of ETE.\nAn important advantage of ETE is, in my opinion, its ability to explain the rapid (by evolutionary standards) appearance of adaptive innovations that harmoniously fit into the complex of organism traits. For STE, new adaptive traits are the result of a happy chance, a mutation of a structural or regulatory gene that turned out to be useful. The more complex the organism, the more rarely such chances should occur. For ETE, new traits arise as a response of the whole organism to changed conditions of its development. The entire experience of preceding evolution is reflected in the formation of this response, the results of selection in the evolutionary past. The chances that such a response will be adequate to the new conditions are much greater.\nThe more complex the organism, the more connections manifest in its ontogeny, the more significantly the preceding history of the species, the adaptive compromise it has achieved, will direct its possible changes. ETE allows us to understand that life \"proceeds by touch\" (© P. Teilhard de Chardin), rather than drifts at the mercy of chance.\nI thank Alexander Pavlovich Rasnitsyn for the criticism of this text. I was able to partially take into account his remarks, but I want to emphasize that he is in no way responsible for the shortcomings of my explanations. Did I manage to mention everything important for understanding ETE? Of course not. Some of what I omitted can be understood from the presentation. The introduction (discussion of terms, disclaimer) in this column is constructed roughly the same as in the presentation, but there are differences in the presentation of the characterization of ETE itself. If you really want to understand, try to grasp the other version of the same explanation as well.\nThinking about these things is interesting to me. And you?"}

The epigenetic theory of evolution views evolution as a process of changing one stabilized ontogenetic pathway into another. In representatives of highly organized groups, the outcome of ontogeny is determined by an extremely complex set of factors and the results of their interactions. Ontogeny is influenced by the interaction of the following factors and their effects:

of the following hereditary predispositions:

genetic (nucleotide sequences in nucleic acids - NA); epigenetic (chemical and spatial modifications of NA macromolecules); other (related to cytoskeleton organization, RNA and regulatory molecule sets, protein conformation, etc.);

various environmental influences; chance.

The outcome of ontogeny cannot be predicted unequivocally. It can only be characterized by a probability distribution of different outcomes, among which one should distinguish the norm (a state maintained by stabilizing selection) and various morphoses (deviations, aberrations). A metaphor describing the distribution of possible ontogenetic outcomes is C. H. Waddington's epigenetic landscape. From this perspective, possible ontogenetic pathways can be described as a set of stabilized areas (creodes), bifurcation points, and dividing creodes of sets of improbable and unstable states. Stabilizing selection—the preferential preservation and reproduction of individuals whose ontogeny has led to the norm—increases the stability of the norm's development (increases its probability). This stability increases both due to the increasing equifinality of the norm's development (the ability to realize the norm in increasingly diverse individuals) and due to the increasing autonomy of such development (the ability to realize the norm under increasingly diverse environmental conditions). This is ensured by the fact that selection restructures the ontogenetic control system as a whole (and the genotype in particular). In the epigenetic landscape model, the action of stabilizing selection appears as a deepening of the corresponding creode. If the nature of selection changes and it ceases to support the previous norm, its development becomes destabilized, and various morphoses appear. If a morphosis proves to be adaptive, selection selectively preserves those ontogenetic control systems that led to such an adaptive state. Offspring of similar individuals are more likely to be adaptive if their ontogeny leads to the same outcome. Therefore, selection will support those offspring in whom the development of the adaptive state in the given conditions becomes increasingly stable (more probable). The result is increased stability of the development of the selected morphosis, i.e., increased heritability. Thus, the phenomenon of heritability itself turns out to be a result of selection. A typical case of norm change during evolution should be considered one in which an adaptive morphosis arises as an adequate response of the ontogenetic control system to changed developmental conditions. If selection supports such a morphosis, it becomes a new norm, its development becomes autonomous and independent of specific external influences. As evolution progresses, the ontogenetic control system becomes more complex, and the mechanisms ensuring the search for adaptive morphoses when the nature of selection changes are improved. The appearance of genetic inheritance, sexual reproduction, cultural inheritance, and complex social organization are some stages of this process.

How does the described approach differ from the TGE? For TGE, evolution is the restructuring of the genotype as a result of selection based on the outcomes of gene activity reflected in the organism's traits. This is why TGE attempts to describe ontogeny as a set of relatively independent cause-and-effect relationships. Any interactions that complicate the transmission of information from genotype to phenotype are simply obstacles for TGE, hindering genotype restructuring based on its phenotypic effects. If information from the genotype is directly reflected in the phenotype, the TGE mechanism works satisfactorily. The Hardy-Weinberg equation describes how the ratio of alleles (alternative variants of a gene) in offspring depends on that in ancestors. The mathematical apparatus of TGE is based on the fact that each allele makes a constant contribution to the overall fitness of its carrier. If an allele increases an organism's fitness, selection will increase its frequency, and over time, the favorable allele will displace its alternatives. There are cases where such a model works. Consider two strains of bacteria. One grows faster but is resistant to antibiotics. The other is resistant and 'pays' for it with slower growth. Their traits unambiguously reflect their genotype. Their dynamics in media with different antibiotic content are well described by TGE selection equations. If the TGE model is correct, the evolution of evolution should lead to the phenotype reflecting the genotype more and more effectively, more 'transparently'. Such organisms will adapt faster and gain an advantage in a changing environment. Organisms with complex interactions of various factors in ontogeny, however, should evolve slowly. Species consisting of long-lived and few-offspring individuals are particularly 'difficult'. Is this prediction justified? No! I wrote about this, remember? In the world we observe, the most complex organisms, evolving fastest in terms of changes in their structure and behavior, turn out to be absolutely improbable from the TGE perspective. Take humans, for example. Our genotype contains very little information, fitting unarchived on a CD. Genes themselves form a small part of it; about 25,000 of our genes require less than 10 megabytes to record (and archivers compress them even further). Recall the famous joke that a file with a detailed description of a kneecap (one of our simplest bones) in AutoCad would take up more space. The information needed to describe our structure is incomparably greater than the volume of our genome. Even more amazing is the amount of information in our psyche. This means that in our ontogeny, a dizzying number of choices of possible developmental pathways occur, with the memorization of their results. The typical path of evolutionary change for ETE turns out to be one that corresponds to the logic of G. K. Waddington's experiments. Waddington induced morphoses in experimental animals using external influences. The offspring from crossing carriers of morphoses were again subjected to similar influences, and again, carriers of the same morphosis were selected for reproduction. After a small (first tens) number of generations, these morphoses began to develop without specific influences. An unstable, environment-dependent developmental pathway became stable. To be convinced that this is not about the inheritance of acquired characteristics, compare these experiments with Weismann's classic experiments. In TGE terms, Waddington's experiments are described with difficulty. Selection for the ability to develop a certain modification (an unstabilized trait) led to a change in a set of modifier genes and, ultimately, to the 'genetic assimilation of the modification,' transferring the control of development of this trait to the genotype. This explanation assumes the existence of specific modifier genes not detected by genomics. It predicts rapid evolution of these modifiers, which does not align with the selection models developed within TGE itself. If we assume that we are not talking about virtual modifiers but about other structural genes, it becomes unclear why the 'genetic assimilation' of the control of one trait does not lead to chaos in the development of others. In ETE terms, Waddington's experiments are described simply. By supporting a morphosis, stabilizing selection increases the stability of its development. And—note this!—there are no strained attempts to represent the organism as a sum of traits, and the genotype as a sum of genes. Waddington's experiments describe not the transformation of 'unstabilized' traits into 'stabilized' ones, but the influence of selection on developmental stability. And, by the way, the example of humans shows that often 'unstabilized' traits (what we learn) are no less important for us than other traits. Now we can discuss the variety of ETE assessments. They are very different. One pole is that ETE is a complete alternative to TGE, and no compromise between them is possible. At the other pole are those who declare ETE pseudoscience and even try to ban mentioning the names of its proponents in the presence of students, so as not to sow doubt in the indisputable truth of TGE. Believe me, I am writing about such attempts not abstractly, but based on sad experience... My assessment is closer to the first pole, although it differs from it. I believe that ETE can become the core of a Third Synthesis: this theory alone has the potential to explain what happens at the organismal level. Of course, ETE still has many underdeveloped issues. One of them is the description of trait diversity in terms of regulating their development in ontogeny. Probably, even in complex organisms, there are relatively simple traits, almost unambiguously dependent on the state of individual alleles. In these cases, TGE models will describe the evolution of such traits quite adequately. Disruptions of normal developmental pathways are likely regulated more simply than the activation of existing gene mechanisms in tissues where they normally function. But the emergence of fundamentally new traits cannot be explained by such mechanisms... Therefore, in my opinion, ETE is a broader generalization than TGE, and cases where the TGE approach is applicable can be considered (in S. Yastrebov's apt expression) as degenerate (simplified) cases of ETE applicability. An important advantage of ETE, in my opinion, is its ability to explain the rapid (on evolutionary timescales) emergence of adaptive innovations harmoniously integrated into the organism's trait complex. For TGE, new adaptive traits are the result of a lucky accident, a mutation of a structural or regulatory gene, which turned out to be beneficial. The more complex the organism, the rarer such accidents should occur. For ETE, new traits arise as a response of the whole organism to changed developmental conditions. This response reflects the entire experience of previous evolution, the results of selection in the evolutionary past. The chances that such a response will be adequate to new conditions are much higher. The more complex the organism, the more interconnections are manifested in its ontogeny, the more significantly the species' past history and its achieved adaptive compromise will guide its possible changes. ETE allows us to understand that life 'gropes its way' (© P. Teilhard de Chardin), rather than drifts by the will of chance. I thank Oleksandr Pavlovych Rasnitsyn for his critique of this text. I was able to partially take his remarks into account, but I want to emphasize that he is not responsible for the shortcomings of my explanations. Have I managed to mention everything important for understanding ETE? Of course not. Several points I omitted can be understood from the presentation. The introduction (discussion of terms, disclaimer) in this column is structured similarly to the presentation, but there are differences in how ETE itself is presented. If you want to truly understand, try to delve into another version of the same explanation. It is interesting for me to think about these things. And you?

←
Dmytro Shabanov
→