Thursday, May 7, 2009

Manifesto for a Kuhnian Revolution

The book "The Structure of Scientific Revolutions" by Thomas S Kuhn was something of a "paradigm shift" in thinking about how science operates. It took more than a decade to become widely accepted as the dominant paradigm for its subject, and has never been free from controversy and tendentious re-definition.

Of course, in addition to being a paradigm shift, it talked about paradigm shifts, especially in dividing the scientific process between "normal science" and "revolutionary science" (paradigm shift) when many of the underlying tacit assumptions and expectations of most scientists become open to question, challenge, and potential replacement (including substantial modification).

Of course, Kuhn's original thesis was pointed at instances where science was mostly an amateur pastime, generally unfunded by government or industry.

One of the classic examples of a Kuhnian paradigm shift is Plate tectonics as an explanation for many geological observations that had previously simply been taken for granted without explanation. It should be noted, however, that there are two obstacles to considering this as a "classic Kuhnian paradigm shift": a) much of the work that led to it was heavily funded by industry, government, or both, as part of the strategically important oil exploration efforts, and b) participants in the process had access to Kuhn's book, leading to something analogous to the "observer effect" in quantum dynamics: scientists leading the "plate tectonics" charge could use Kuhn's work as something of a blueprint for their own revolutionary efforts.

Among the key elements of a condition leading to a "Kuhnian revolution" is the accumulation of significant anomalies against a current paradigm, such as observations that require tortuous manipulations of current theory to explain, or simply don't fit at all. All paradigms (per Kuhn) have anomalies, the question is how significant they are.

Another point is that the current questions popular for research tend to be at a tangent to many of the unanswered questions raised by those anomalies.

This brings us to a paper published in the April issue of PLoS Genetics: The Genetic Signatures of Noncoding RNAs by John S. Mattick. I briefly mentioned that in my last post, but here I want to expand on its nature as almost a manifesto for a Kuhnian revolution in the study of genetics.

This paper demonstrates that there is a recent upsurge of previously unsuspected influences on development, "of non–protein-coding RNAs (ncRNAs), whose incidence increases with developmental complexity." It goes on to "show that this is explicable by an historic emphasis, both phenotypically and technically, on mutations in protein-coding sequences, and by presumptions about the nature of regulatory mutations."

Here are the typical ingredients of a "Kuhnian Revolution" in the making. The old paradigm can be reasonably described as "gene-focused", that is on the DNA that codes for proteins, and very secondarily on the regulatory DNA that interacts to control transcription in preparation for translation. The origin of that paradigm can certainly be based in the demonstration of the genetic code by James Watson, Francis Crick, Maurice Wilkins and Rosalind Franklin. With this code in hand, it became easy for researchers to focus on protein structures and coding, while neglecting the influences that regulated the conversion of that code into active proteins.

This also led to the classic definition of the "gene" as being the DNA that coded for a protein. Many textbooks, even today, distinguish between the "gene" and the regulatory sequences that control its transcription. (It's important to note the role assigned by Kuhn to textbooks: they tend to codify the current paradigm and train new scientists to think of it as fundamental, and often unquestioned.) A better way of defining "gene" from the start would have been the DNA sequences that code for proteins, and directly participate in regulating their translation together. This would have focused much more attention on regulation.

But "Genetic Signatures" is going beyond that. The current paradigm includes "a strong expectation that mutations that have phenotypic effects will map to protein-coding genes or cis-regulatory elements that interact with regulatory proteins." In this case the term "cis-regulatory" means that the sequence directly regulates transcription because it is on the same strand of DNA, somewhere near the promoter region for the protein-coding sequences in question.

But recent years have seen a rising wave of discoveries outside that paradigm. These include the role of DNA methylation, histone marking, and other forms of epigenetic influences. Key among these are other roles for the sequences of DNA besides protein coding and cis-regulation. "Genetic Signatures" points out that:

In recent years it has also become evident that the vast majority of the mammalian genome, and that of other complex organisms, is transcribed, apparently in a developmentally regulated manner, to produce large numbers of ncRNAs that are antisense, intergenic, interleaved, or overlapping with protein-coding genes [refs]. In addition, there are increasing reports (Figure 1) of the functionality of individual ncRNAs in mammals (Table 1), other animals ([refs]), plants ([refs]), and fungi ([refs]), particularly in relation to developmental processes [ref]. These include the involvement of ncRNAs in the regulation of the expression of homeotic genes [refs], oncogenes [ref], and metabolic genes [ref], as well as in the regulation of skeletal development [ref], eye development [ref], epithelial-to-mesenchymal transition [ref], and subcellular structures [refs], among many others ([refs]). [Many of the references are open access.]

To some extent, the revolution is already under weigh:

It is widely accepted that animals have a relatively common set of protein-coding genes and that, notwithstanding lineage-specific innovations and splice variants, the primary basis of phenotypic, especially morphological, radiation and higher complexity has been the variation and expansion of the regulatory architecture that controls the deployment of these protein components and their isoforms during differentiation and development [71].

Reference #71 in the quote is to Evo-Devo and an Expanding Evolutionary Synthesis: A Genetic Theory of Morphological Evolution by Sean B. Carroll. At the same time, "Evo-Devo" is pushing the cis-Regulatory Hypothesis: that a very complex network of regulatory elements located on the same strand of DNA as the protein-coding sequences (cis-regulatory elements, or CREs) is fully responsible for the control and evolution of morphological development.

In attempting to demonstrate "The Sufficiency of CRE Evolution", Carroll begins:

More than two dozen case studies of evolutionary changes in morphological traits have been attributed to changes in gene CREs.

He finishes up the demonstration with:

Most significantly, these studies reveal that CRE evolution is sufficient to account for changes in gene regulation within and between closely related species. Given that differences at higher taxonomic levels are the product of the accumulation of species-level divergences, then, in principle, CRE evolution is sufficient to account for the rewiring of regulatory networks at all taxonomic levels ([refs]).

The entire section is intended to demonstrate that CRE evolution can explain a number of cases that the previous paradigm based on evolution by protein changes cannot. Indeed, it is followed by a section titled "The Necessity for CRE Evolution in GRN Evolution", demonstrating that "protein evolution is not sufficient to establish novel linkages within GRNs (Gene Regulatory Networks)."

Perhaps, like one wave being overtaken by another, we should think of two revolutions, the "CRE Evolution" based revolution against the traditional idea that evolution is about changes to the protein-coding regions, and an even newer revolution against "CRE Evolution", focusing on ncRNA and other epigenetic influences.

"Genetic Signatures" gives a long laundry list of observations that simply cannot be explained by CREs, or their evolution. Clearly, there is much more that must be considered before we fully understand how life works.

The ultimate focus of any "Kuhnian Revolution" must be the expectational and perceptual factors underlying the current paradigm. "Genetic Signatures" spends a good deal of focus on the practical obstacles to a search within the proposed new paradigm, but those obstacles will likely prove much more easily removed as more people focus on the "hidden world of regulatory architecture underpinning the development of complex organisms that we have yet to explore, both genetically and functionally."

No comments:

Post a Comment