Chemical arbitrariness and the causal role of molecular adapters

https://doi.org/10.1016/j.shpsc.2019.101180Get rights and content

Highlights

  • Jacques Monod saw “chemical arbitrariness” as critical to certain molecular systems.

  • Modal interpretations of chemical arbitrariness fail to account for this importance.

  • This can be done by citing the key causal-functional features of arbitrary systems.

  • Chemically arbitrary causal relationships rely on one or more molecular adapters.

  • Adapters couple two processes by acting as intermediate, not cooperating, causes.

Abstract

Jacques Monod (1971) argued that certain molecular processes rely critically on the property of chemical arbitrariness, which he claimed allows those processes to “transcend the laws of chemistry”. It seems natural, as some philosophers have done, to interpret this in modal terms: a biological relationship is chemically arbitrary if it is possible, within the constraints of chemical “law”, for that relationship to have been otherwise than it is. But while modality is certainly important for understanding chemical arbitrariness, understanding its biological role also requires an account of the concrete causal-functional features that distinguish arbitrary from non-arbitrary phenomena. In this paper I elaborate on this under-emphasised aspect by offering a general account of these features: arbitrary relations are instantiated by mechanisms that involve molecular adapters, which causally couple two properties or processes which would otherwise be uncorrelated. Additionally, adapters work by acting as intermediate rather than cooperating causes.

Introduction

It's common to hear that certain biological phenomena at the molecular level involve relations that are “arbitrary” or “chemically arbitrary”. This is most commonly attributed to the genetic codeーspecifically, the relationship between the three-base triplets of mRNA and the amino acids they specify. By contrast, other molecular relations are cited as chemically “necessary” or “non-arbitrary”, including those between DNA and complementary RNA bases in copying. This difference has been taken by some to be deeply important to biological functioning and evolution, and yet it has proven difficult to spell out just what this difference amounts to.

The intuition behind this distinction is that non-arbitrary relations are in some sense determined, or at least constrained, by the chemical properties of the molecules involved in a way that arbitrary ones are not. Transcription, for example, occurs by complementary pairing between the bases: a G on the parent strand produces a C on the daughter strand, for example, because one binds to the other stably through their complementary stereochemical structures. This reliance on the “intrinsic” chemical properties of the bases, the thinking goes, constrains the pairings that are possible, suggesting a degree of chemical necessity to the pairings that actually occur.

In contrast, the translation of an RNA transcript into sequences of amino acids seems markedly different: The fact that “AUG” codes for methionine isn't due to any direct contact between the codon and its corresponding amino acid. Instead, this is because AUG binds to specific tRNA molecules that carry methionine at their other end. If the tRNAs were different, then, the genetic code would be different (indeed, in some organisms it is different). Hence, the RNA-amino acid pairings seem more contingentーless determined by the dictates of chemistryーthan the DNA-RNA pairings in transcription. This is the reason typically given for calling the genetic code chemically arbitrary.

Here it is critical to distinguish this notion of chemical arbitrariness from what we might call functional arbitrariness. Francis Crick (1958) famously suggested that the particular assignments of codons to amino acids may be a “frozen accident”ーone that became entrenched due to the severe consequences of deviating from it once it was established. This claim is now widely questioned: there is now a mass of evidence suggesting that the standard genetic code is optimal or near-optimal in several respects relative to other possible codes (see e.g. Butler, Goldenfeld, Mathew, & Luthey-Schulten, 2009; Kumar & Saini, 2016; Maynard Smith & Szathmáry, 1995). For example, point mutations in DNA that lead to amino acid changes are likely to substitute amino acids that are chemically similar to the original, which minimises the chances of loss of function. This suggests that the standard code is not a frozen accident at all, but that it was reached after a period of variation and selection. Understanding how the code was reached is a fascinating scientific puzzle (Koonin & Novozhilov, 2008). Nevertheless, this matter is quite independent of the claim that alternative codes are at least chemically possible. In fact, the claim that the genetic code has been subject to selection implies that variations existed to be selected from in the first place, and chemical arbitrariness is said to be the property that makes this variation possible (Bergstrom & Rosvall, 2009). So the concept of chemical arbitrarinessーthe theme of the present paperーis quite independent of claims about functional arbitrariness and frozen accidents (However, as I will revisit in Section 4, the two are connected in important ways.).

As the above suggests, the distinction between arbitrary and non-arbitrary phenomena is not thought to be an idle observation: Not only does such a distinction exist in the biological world, but this distinction has been claimed to be extremely important for understanding many aspects of life processes and their evolution. This importance was emphasised by the molecular biologist Jacques Monod in his book Chance and Necessity (1971). Monod argued that chemical arbitrariness (or gratuité) was a critical property of biological phenomena that permitted living things a near-limitless plasticity of function. His central example is the lac operon system that he famously co-discoveredーa segment of the E. coli genome containing genes for a number of proteins involved in the uptake and metabolism of lactose. Transcription of all these genes is controlled by a single repressor protein, LacI. In the absence of lactose, LacI binds to an “operator” region upstream of the lac genes, which prevents them from being read by blocking the transcription machinery. When lactose is present, however, it binds to LacI and causes a conformational change that releases LacI from the operator, allowing the genes to be transcribed. The upshot is that the lac genes are switched on when there is lactose for them to process, and switched off when there isn't.

The key feature of this system that Monod highlights is the particular way that LacI couples the expression of the lac genes with the presence of lactose. While there is no “chemically necessary relationship” (ibid. p. 76) between these two things, LacI achieves this regulation through an allosteric interaction: the binding of lactose at one of its sites distorts its shape at another, making it unable to bind to the operator. Allosteric interactions like these are what Monod calls “gratuitous” (or elsewhere “chemically arbitrary”ーI will use this term), a property he views as critical to understanding the plasticity of biological systems:

“In a word, the very gratuitousness of these systems, giving molecular evolution a practically limitless field for exploration and experiment, enabled it to elaborate the huge network of cybernetic interconnections which makes each organism an autonomous functional unity, whose performances appear to transcend the laws of chemistry if not ignore them altogether." (Monod, 1971, p. 78, emphasis added)

Aside from Monod's claims about the importance of arbitrariness itself, the notion has also been used in philosophical work to explicate other biological concepts. In particular, John Maynard Smith, 2000a, Maynard Smith, 2000b draws on Monod's ideas to elaborate his account of biological information. In his view, biological phenomena that we recognize as “informational”, such as DNA and hormones, differ from non-informational ones in key ways. In particular, DNA and hormones can be said to carry information or “meaning” because what they doーtheir functionーdepends on their “interpretation” by other “evolved receivers” such as tRNAs or cell surface receptors. In contrast, what an enzyme does is “directly determined by its structure” (Maynard Smith, 2000a, 2000b, p. 193), and hence there is no need for anything resembling the interpretation of a signal carried by the enzyme.

Yet despite the importance they attribute to chemical arbitrariness for biological systems and our understanding of them, Monod and Maynard Smith offer little by way of sustained philosophical analysis of the concept itselfーof exactly what the implied lack of a “necessary connection” amounts to, for instance. When we try to clarify the difference between arbitrary and non-arbitrary, as some have done, we run into some difficulties. For example, one suggestion is that the difference is a matter of directness of interaction: the necessary connection between DNA and mRNA bases, or between enzyme and substrate, arises from their direct stereochemical contact, whereas there is no such direct interaction between codons and amino acids in translation. However, Godfrey-Smith (2000) questions whether directness of interaction is able to ground the distinctions typically made between arbitrary and non-arbitrary phenomena:

“Perhaps there is a difference in degree here; an enzyme's catalytic action is strongly constrained by its physical structure, whereas a hormone could have a huge variety of effects depending on the location and structure of the receptors with which it interacts. On the other hand, the hormone's interaction with those receptors is certainly a matter of its physical structure; why look further “downstream” with the hormone but not with the enzyme?” (Godfrey-Smith, 2000, p. 203, emphasis in the original)

In other words, it is trivially true that all “indirect” causal interactions are mediated by a series of steps each of which is directly related to the next. So the idea that arbitrariness is just a matter of indirectness is hard to square with the crucial biological importance that Monod and others have afforded it. In particular, it is hard to interpret it the way Monod does, as allowing living things to “transcend the laws of chemistry”. In summary, it has proven difficult to make explicit why certain phenomena in molecular biology have been described as chemically arbitrary and others not, without deflating the concept to a trivial matter of indirectness and losing sight of why so much has been made of it.

The aim of this paper is to develop an account of chemical arbitrariness that clarifies what it is and why it is thought to be so important for our understanding of molecular biological systems. As we will see, a number of philosophers have aimed to interpret the “non-necessity” of chemically arbitrary phenomena in modal terms; that is, they offer a way of understanding chemical arbitrariness by clarifying the sense in which the dictates of chemistry permit certain alternatives and forbid others. I will argue that, while modal considerations are an important piece of the puzzle, understanding the importance Monod and others have afforded this concept requires some further elaboration, which I offer here.

The paper proceeds as follows: Section 2 outlines the aforementioned attempts to clarify the concept of chemical arbitrariness in modal terms, and explain why they are in need of elaboration. It is this elaboration that I offer in Section 3, which involves an account of the causal-functional features that characterise phenomena that we recognize as arbitrary. Specifically, I argue that arbitrary causal relations rely on molecular adapters. I define adapters as molecules that couple two other properties or processes by acting as intermediate, rather than cooperating causesーa distinction I clarify in terms of the interventionist theory of causation. In Section 4, I discuss how this helps make sense of what is biologically important about chemical arbitrariness; that is, it highlights that the causal relations we typically call arbitrary are problems to which adapters are solutions. Section 5 concludes.

Section snippets

Modal accounts of chemical arbitrariness

This section outlines two existing attempts to characterise the notion of arbitrariness. I call both of these “modal” accounts because, while they differ in significant ways, they both tie arbitrariness to the possibility of alternatives within certain constraints imposed by theory. I will argue that while these accounts are not incorrect, they are in need of elaboration with causal-functional detail that I offer in Section 3.

The first and most dedicated discussion of chemical arbitrariness

Molecular adapters and intermediate causes

The previous section outlined two existing analyses of the concept of chemical arbitrarinessーboth modal in that they appeal in some way to the chemical possibility of alternatives. I argued that while tying arbitrariness to modality per se is not misguided, more is needed for an adequate understanding of which phenomena are arbitrary and why it matters. What is needed, in short, is an account of the essential or characteristic mechanistic features that make certain relations arbitrary and

The biological importance of adapters

Suppose we accept the above proposalーthat a molecular mechanism, or the causal relation it sustains, is “arbitrary” when and because it relies on a molecular adapter. What, then, is biologically important about arbitrariness understood in this way? How does understanding the role of adapters in arbitrary causal relations lead to a fuller understanding of living processes and their achievements? Recall that Monod's view of the importance of arbitrariness was that it freed biological processes

Conclusion

In this paper, I've proposed an elaboration of the definition of arbitrariness as the chemical possibility of alternatives. This elaboration serves to bring out the biological importance that Monod and others have granted this property. My proposal has been that what makes chemical arbitrariness functionally important is that arbitrary causal relations are mediated by molecular adapters, which couple a cause to an effect by acting as intermediates in causal chains between them. This establishes

Acknowledgements

I'm grateful to Marc Ereshefsky, Samir Okasha, Ulrich Stegmann, C. Kenneth Waters, colleagues at the University of Calgary, and two anonymous referees for invaluable comments on this paper. This work was supported by the European Research Council Seventh Framework Program (FP7/2007–2013), ERC Grant Agreement No. 295449, and by the John Templeton Foundation (“From Biological Practice to Scientific Metaphysics,” grant no. 50191).

References (24)

  • M.B. Hoagland et al.

    A soluble ribonucleic acid intermediate in protein synthesis

    Journal of Biological Chemistry

    (1958)
  • W. Bechtel

    Constructing a philosophy of science of cognitive science

    Topics in Cognitive Science

    (2009)
  • C.T. Bergstrom et al.

    The transmission sense of information

    Biology and Philosophy

    (2009)
  • C.T. Bergstrom et al.

    Response to commentaries on “the transmission sense of information

    Biology and Philosophy

    (2011)
  • T. Butler et al.

    Extreme genetic code optimality from a molecular dynamics calculation of amino acid polar requirement

    Physical Review E - Statistical, Nonlinear and Soft Matter Physics

    (2009)
  • C.F. Craver

    Role functions, mechanisms, and hierarchy

    Philosophy of Science

    (2001)
  • F.H. Crick

    On protein synthesis

    Symposia of the Society for Experimental Biology

    (1958)
  • R. Cummins

    Functional analysis

    The Journal of Philosophy

    (1975)
  • F.I. Dretske

    Explaining behavior: Reasons in a world of causes

    (1988)
  • P. Godfrey-Smith

    Information, arbitrariness, and selection: Comments on maynard Smith

    Philosophy of Science

    (2000)
  • L.Q. Gu et al.

    Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter

    Nature

    (1999)
  • T. Horng et al.

    TIRAP: An adapter molecule in the Toll signaling pathway

    Nature Immunology

    (2001)
  • Cited by (2)

    • Reflexivity, coding and quantum biology

      2019, BioSystems
      Citation Excerpt :

      Nevertheless, there are many natural variants of the consensus code (Elzanowski and Ostell, 2019) and it has proved possible to evolve artificial variants of the code which cause non-canonical or even non-biological amino acids to be inserted into proteins (Bartholomae et al., 2018). Thus Barbieri (2019) pitches the code’s arbitrariness against any notion of stereochemical influence while Lean (2019) pushes the discussion toward the question “Arbitrary with respect to what?”. It is notable that codon reassignment has not been achieved by design alone: it has always been necessary to allow a translation system using a poorly functioning reassigned aaRS to mutate subject to appropriate selection constraints before an efficacious solution to the reassignment problem has been found.

    • Notions of arbitrariness

      2023, Mind and Language
    View full text