The living and determinism: what place is there for chaos?

Is the living predictable in its evolution? What is the connection between physical individuation and biological individuation?

Denis Pompon’s point of view Emeritus Director of Research at the CNRS

In 1936, Turing described his vision of the limits between the “calculable”, i.e. what is rationally deducible, and the ”non-calculable”. He hence defined a “predictive determinism” which he also extended to the field of nature sciences through two concepts, the Turing “machine” and “structure” which aimed at defining the limits between

Earlier, the mathematician Henri Poincaré had considered the

, demonstrating that a very simple physical system, although perfectly described by mathematical equations, could become totally unpredictable, seemingly chaotic, for an external observer. Such a system remained formally deterministic in the sense that the knowledge of its state at a given point in time allowed one to reasonably predict its state a short time later, but this prediction very quickly lost any relevance in the sense that the slightest precision deviation in the description of the initial point in time translated into a major difference of evolution on the longer-term. In 1927, The German physicist Werner Heisenberg set forth another principle, the

, this time derived from studies in quantum mechanics. This principle, largely verified experimentally since, defined that at the quantum scale, that of atoms and molecules, the increase in precision of a position measurement of an object translates into an increasing uncertainty on its velocity, and therefore on its energy and vice versa, and that this rule is inescapable whatever the experimental device. In this case, as in Poincaré’s three-body system, the evolution of simple systems perfectly described by known physical laws is or quickly becomes unpredictable at a given level of precision. The living, which defines us and surrounds us, is a much more complex system than the preceding simple systems. Nevertheless, we often increasingly consider it as simply defined by its inherited properties largely encoded in the genetic information.

, such as previously described? In fact, at this stage it becomes important to introduce another notion which is that of scale. At the level of a population of a billion cells for example, the behaviour of a population of genetically identical microorganisms is fortunately (for all our industrial processes) relatively predictable and reproducible. Nonetheless, if we now extract individual cells from this population and characterise them, we will observe that behind the homogeneous and predictable aspect of the population, lies an extraordinary diversity of individual behaviours. At the level of each cell, the transcription and translation of each gene appear as a chaotic phenomenon, unpredictable at a given point in time, even if it remains predictable in average value. When the man in the profession designs a multi-stage engineering, these chaotic behaviours will have to follow on from each other in a more or less efficient manner for the synthesis of a product of interest within each cell. The result is that the macroscopic behaviour (the performance) will finally be impacted by the anarchy of these stochastic behaviours at the cellular scale. However, is the

This question leads us to two other questions: those of adaptability and evolution. These two questions have in common the problem of the search for a solution within a surrounded, complex and multiparametric system. A complex system can be characterised by the combinatorics of its possible states. The search for an adaptive solution must scan this combinatorics in search for the best responses. However, this approach becomes impossible if the combinatorics are too large. For example, if we take the case of an enzyme with 500 amino acids, the combinatorics of the possible sequences 20500 far exceeds the number of atom vibrations since the big-bang at the origin of our universe. Usefully exploring the combinatorics is impossible except if the number of solutions is itself very large or if the preexisting ones are memorised. This leads to the concept that the living would benefit from memorising a repertoire of potential solutions not necessarily exploited, and not only the optimal solution of the moment. There are then two options: the first exploits chaos to create a population of states of which only the average value is genetically inheritable, adaptability or evolution then involves the selection and subsequent amplification of the preexisting state variant best responding to a new constraint. The other possible mechanism involves memorising potential solutions during the course of evolution itself. For example, the classic description of a metabolic pathway often overlooks a plethora of alternative biosynthesis pathways exploiting enzyme activity promiscuities (“parasite” activities). The application of a metabolic constraint or the accidental inactivation of the “normal” pathway can then quickly activate a secondary metabolic network, which will thereafter be optimised. Such a mechanism presupposes in return that the choice of a sequence solution within the combinatorics of possibilities for an enzyme responds not only to the constraints of an efficient “principal” activity but maintains also a maximum number of alternative potentialities. Do these mechanisms constitute one of the secrets of the living? This still remains to be demonstrated. At all events,

Vincent Grégoire-Delory’s point of view Manager of the TWB Ethics platform

. One cannot go without the other: no genetic information, no metabolism. This reciprocal interweaving however poses serious problems for pre-biotic chemistry specialists since understanding the emergence of the living is tantamount to understanding the emergence of the structure of self-replicable molecules carrying the information necessary for metabolic functioning. It is the eternal ”chicken and egg” problem. Indeed, it seems well established that

. From one replication to the next, the structure of the DNA molecule is preserved and the combination of the four nitrogenous bases is copied… with a few errors. There is both identicality and novelty in the molecular replication phenomenon. Moreover, if we examine the various metabolic reactions at the scale of a living organism, it is possible to reveal feedback loops characteristic of the reciprocal effect of the metabolism properties and those of the genetic material. Recent discoveries in epigenetics for example clearly show the non-deterministic character of the DNA molecule, capable of expressing different properties according to the evolution of the physico-chemical conditions of its “nano-environment”. In other words, the DNA molecule is not in a simple relationship with its environment, it is expressed in an individuating relationship. Yet if we consider that such an individuating relationship goes on from generation to generation through self-replication, it clearly appears that

. In other words, the question is to know whether the self-replication phenomenon in itself emerges from physical individuation and/or biological individuation. The molecular replication of the genetic material displays a plasticity unknown to the crystallization phenomenon for example but the question is to know whether this phenomenon is observable outside of the spaces where metabolic reactions occur. The question is whether this type of phenomenon can occur spontaneously within a solution composed of organic molecules self-organising into RNA or DNA. Such experiments are underway since the 1990s

and show that it is

. This research is extremely interesting since it provides new insights into the possibility of considering living organisms as self-organised from molecules with self-replicating properties. The question is therefore to specify whether these spontaneous self-replicating properties participate in themselves in biological individuation. For this to be the case, it is indeed not sufficient that some organic molecules have self-replicating properties, they should also be able to carry and transmit imperfectly encoded information. In other terms, self-replication is comparable to physical individuation if the molecules replicated in this way have the sole ability to participate in chemical reactions (reversible or not). In this case indeed, the process is purely stochastic and comparable to a complex system of molecular elements connected with each other. On the other hand, if the self-replicating properties participate in biological individuation, they carry a potential inchoative dynamic capable of perpetuating self-assembled organisations. One of the keys to understanding the physico-chemical properties of the living could then lie in the forms of transition to these new dynamics. Since this key is not yet available to us, it is still possible to speculate on the fact that such self-replicating molecules are already ”living” or, on the contrary, that they represent pure physical individuations.

. The “imperfect memorisation-transmission” phenomenon could then be a constitutive property of the biological individuating relationship.

.

References

LEHN, Jean-Marie, La chimie supramoléculaire : concepts et perspectives. De Boeck (1997), 288 pp.

i.e. information capable of participating in the organisation of chemical reactions.

More information

• http://www.pourlascience.fr/ewb_pages/a/article-le-determinisme-et-le-vivant-21414.php
• https://fr.wikipedia.org/wiki/Alan_Turing
• Turbulence et déterminisme. Marcel Lesieur (Presse universitaire de Grenoble) : extrait https://books.google.fr/books?isbn=2759801411
• https://fr.wikisource.org/wiki/Henri_Poincar%C3%A9,_le_probl%C3%A8me_des_trois_corps
• James E. Skinner et al. Application of chaos theory to biology https://link.springer.com/article/10.1007%2FBF02691091