Rediscovering Life

Symptoms: Notes from the Biological Literature (2)

The Limits of Causal Understanding
Stephen L. Talbott

RSS Feed for Rediscovering Life: RSS Feed for this page   

This article is part of “”, a news portal for the Biology Worthy of Life Project. Copyright 2014 The Nature Institute. All rights reserved.
By placing your cursor on scientific terms such as “DNA” (try it here), you may find them to be clickable links into a separate glossary window (or tab, if your browser is set that way). You can in any case open the glossary for browsing by clicking here.

Tags: chromosome/chromatin; DNA; explanation/causation; gene/genocentrism; gene regulation; machine idea; plasticity/genome; transcription/transcription factors

Posted: July 3, 2014   (Article 21)

Viruses, cytoplasmic DNA, and the web of life arrow

The (molecular) days of our lives arrow

In brief:

DNA and RNA brought to life arrow

Functional vs. nonfunctional DNA binding arrow

Viruses, cytoplasmic DNA, and the web of life. Here are the lead sentences of an interesting news release from MIT, although in typical fashion the author mixes newer realizations (the roles or meanings of DNA are substantially provided by its cellular context) with thoroughly disgraced cliches (DNA contains a blueprint for the organism):

In 2003, when the human genome had been sequenced, many people expected a welter of new therapies to follow, as biologists identified the genes associated with particular diseases. But the process that translates genes into proteins turned out to be much more involved than anticipated. Other elements — proteins, snippets of RNA, regions of the genome that act as binding sites, and chemical groups that attach to DNA — also regulate protein production, complicating the relationship between an organism’s genetic blueprint, or genotype, and its physical characteristics, or phenotype. (Hardesty 2014*)

The report concerns some new research in which MIT scientists participated. Working with certain yeast strains, the investigators were looking for traits relating to the presence or absence of particular genes. Far from being consistent, their results, differing greatly between strains, were “all over the map”. What they eventually discovered is that the roles of individual genes depended on two cytoplasmic sources of nonchromosomal DNA:

In the yeast experiment, both these sources greatly influenced the phenotypic consequences following upon the presence or deletion of the various genes being studied. Importantly, however, the various influences were inextricably woven together. As it was put by David Gifford, a professor of computer science and engineering at MIT, who led the quantitative analysis:

You might think that the effect of the chromosomal modification and the effect, for example, of the virus were both important but independent. What we found is that they weren’t independent. They were synergistic.

Of course, mitochondrial DNA has been known for a long time, but in most genetic studies it scarcely even registers as an afterthought — something that is likely to change now. This is particularly significant because, as the researchers state in their technical paper:

Nonchromosomal information is not under the usual constraints of the nuclear genome. These nonchromosomal elements are extremely unstable: they mutate at higher frequencies than the DNA of the chromosomal genome, may be lost at high frequencies without loss of viability, and can vary in copy number from cell to cell. (Edwards et al. 2014*)

The authors seem confident that the interplay between nuclear and cytoplasmic DNA will also be found prominently in humans, as soon as biologists begin looking for it. They are particularly concerned to show that the cytoplasmic factors play a largely overlooked role in heredity. What interests me personally the most, however, is the role of viruses. For example, the researchers remark that “Recent work on a mouse model of Crohn disease supports a combinatorial model of complex disease traits in which the pathology requires the interaction between a specific mutation in the mouse and a specific strain of virus”.

I’ve found it hard to track down to what extent more or less “latent” viruses may be present in human cells, and I’m not at all sure this would be easy to discover. But I can’t help wondering: there’s been a lot of attention lately on the so-called human “microbiome” — the collection of microorganisms in our bodies. This microbiome, whose genetic complement outweighs our own, is now known to play vital roles in our well-being. Might viruses be the next frontier of identity — the next boundary between “self” and “other” — that begins to go all blurry on us?

Moreover, viruses would not need to be transmitted from parents to offspring in order to play an important role. They could be “inherited” as common elements of our environment, just as many microorganisms important for human health remain available from one generation to the next as environmental resources.

It is, in any case, already widely accepted that we are riddled with the results of past viral activity. Over half the human genome is thought to be derived from mobile DNA sequences called “transposons”, some, or many, of which apparently originated from viruses. Certain transposons — and particularly retrotransposons — have a habit of moving around, duplicating themselves, or otherwise helping to remodel our genomes. Testifying to the usual red-in-tooth-and-claw, them-versus-us, struggle-for-survival mindset, biologists have long categorized transposon sequences as “selfish” or “parasitic” — a prejudicial view that is rapidly, if belatedly, being overturned.

In a paper entitled, “Genomic Creativity and Natural Selection”, British medical researcher Frank Ryan reviews the role of viruses along with various other genome remodeling factors. I close this note with a few snippets from his discussion of viruses:

It is becoming increasingly clear that endosymbiotic unions of viruses and hosts have influenced the evolution of life throughout most, if not all, of biodiversity. . . .

[University of California at Irvine molecular biologist, Luis] Villareal believes that “host colonization harnesses the creative power of viruses [that] create new genes from stitching together lots of smaller genes”. . . .

Persistent [as opposed to acute] viruses do not usually follow ... models of predator-prey relationships; they are not associated with high level production of progeny; and, as a rule, they cause unapparent, generally asymptomatic, infections. ... In many cases, the persistent virus becomes incorporated into the host genome. ...

Some 8% of the human genome consists of human endogenous retroviruses, or HERVs, and, approximately one half of our DNA comprises HERV genes, fragments, and derivatives. ... However, the great majority of HERVs are no longer infectious in the exogenous sense [that is, acting as external organisms distinct from ourselves]. ...

Once viruses enter a genome, their capacity for evolutionary innovation remains persistently active. Millions of years after initial genomic incorporation, and despite the policing of natural selection, endogenous retroviruses can interact with newly arrived exogenous viruses or with other genetic components and regulatory mechanisms, thus increasing evolutionary “plasticity”. ... (Ryan 2006*)

And so viruses represent yet another element in the intricate web of life that catches up the genome and shapes its expression.

Tags: DNA; gene regulation; plasticity/genome

The (molecular) days of our lives. If one thing is true of the molecular biological literature today, it is that the biologist’s seemingly unquenchable thirst for unambiguous, controlling (causal) elements in the cell is being frustrated on every hand. In the Winter, 2011 issue of The New Atlantis (Talbott 2011*) I offered an example centering on the circadian (daily) rhythms that figure so prominently in human life. That text serves as a nice preface to some new research:

Biologists, of course, set out to identify the “clock mechanism” that was presumed to “control” these rhythms, and, yes, they found a rhythmical feedback loop involving genes and transcription factors in a certain area of the brain that seemed the perfect candidate. However, ongoing research has revealed distinct “clocks” in different mammalian organs and tissues, and indeed in every cell. These “clocks” are interwoven with each other and, it now seems, with virtually all aspects of the organism’s physiology — metabolism, reproduction, cell growth and differentiation, immune responses, central nervous system functions . . .

In each of these areas the quest for causes and master controllers leads to the usual perplexity about who’s doing what to whom. For example: “Although metabolism is thought to be primarily downstream of the cellular clock, numerous studies provide evidence that metabolic cycles can operate independently from or even influence circadian rhythms” (Kumar and Takahashi 2010*). At the molecular level, one research team remarks that the enzymatic function of a certain clock protein “may be controlled by changing cell energy levels, or conversely, could regulate them” (Doi et al. 2006*). In general, “It seems that connections between the circadian clock and most (if not all) physiological processes are bidirectional” (Yang 2010*).

What we’re gaining from all this research is a wonderful portrait of the organism as a rhythmical being. Investigators have not found controlling mechanisms that single-handedly establish or govern the circadian rhythms of the organism, but rather are discovering how those rhythms come to expression at every level and in every precinct of the organism — perhaps more centrally here and more peripherally there, but altogether in a single, organism-wide harmony. There is no sensible way, as a scientist, to speak of particular mechanisms that explain this harmony. Instead, every isolated “mechanism” is found to be a reflection of the harmony, and we thereby gain further, detailed understanding of how the whole organism functions as a being in time.

What brings these comments to mind is a recent paper in Trends in Cell Biology entitled “How Pervasive Are Circadian Oscillations?” (Patel et al. 2014*). The answer is “very”. Or as the authors of the paper put it: many genes and metabolites, whose rhythms of activity and production were “once thought to only be controlled by a master clock”, are now known to be capable of a vast range of rhythmic activity adjusted to their local circumstances. “Each cell can reprogram itself and select a relatively small fraction of this broad repertoire for circadian oscillations, as a result of genetic, environmental, and even diet changes”.

Differences between tissues can be dramatic, and this goes for comparisons between the “master controller” of circadian rhythms (the suprachiasmatic nucleus, or SCN, in the brain’s hypothalamus) and other organs. For example, the researchers found 337 gene products oscillating in the SCN and 335 in the liver, and yet only 28 oscillating in both. A similar difference was found in comparisons of other organs. In sum, “a large fraction of all molecular species is capable of oscillating in a circadian manner under some set of conditions”, but the relevant conditions seem to be as diverse as the makeup of the organism itself.

It is hard to maintain the illusion of a hard-and-fast master controller in the face of such divergent performances, so the question becomes, in the authors’ words, “How does the cell select which molecular species should oscillate in a given situation?” They conclude:

Circadian oscillations are quite different from the oscillations of a simple physical system, such as a spring, because they are non-rigid and highly plastic with respect to perturbations, allowing for instance an organism to change its diet or its time of sleep. Perturbations not only modify existing oscillations, but can also induce new oscillations ... revealing the underlying oscillatory fabric that enables flexible cellular reprogramming.

The reader will perhaps forgive the researchers their gratuitous reference to “programming”, which seems obligatory (if bizarrely so) within the biological guild. Less forgivable is the usual mixing of the language of causation or control with descriptions utterly contradicting that language. Happily, the paper under consideration is relatively free of this confusion. Nevertheless, it is good to keep on one’s toes when reading, for example, that “the cell can rapidly and massively reprogram which species oscillate, such as the metabolite NAD+, which plays a central role in regulating circadian rhythms”.

We can grant the importance of NAD+ while also asking ourselves just how central its role can be when the cellular context so readily “reprograms” that role. This takes us back to the fundamental question: is biological understanding essentially an understanding of causally controlling elements, or of governing context? Clearly, we cannot have understanding of the context without examining all the details — including what can be seen approximately as (local) causes and effects. These must be investigated. But, just as clearly, we cannot make larger sense out of what is going on locally simply by extrapolating and adding together such causal relationships. It is the context that coordinates, continually modifies, and makes biological sense out of them.

Another recent paper (Reddy and Rey 2014*) reinforces this truth. Current models for so-called “circadian oscillators” are centered on the expression of certain genes, with “feedback loops” (involving all the complexities mentioned above) playing a role in establishing the rhythmic character of gene transcription. But now, according to Reddy and Rey, “mounting evidence is questioning the absolute necessity of transcription-based oscillators for circadian rhythmicity”. In particular, “a more fundamental mechanism based on metabolic cycles” could underlie circadian rhythms.

The authors cite various experimental results that undermine the single-minded focus on gene transcription rhythms:

In sum, current models centered on gene expression “cannot possibly account for the multiple lines of experimental data that have revealed circadian oscillations in the presence of inactivated [gene-expression] feedback loops”. The point, however, is not to deny the importance of a “transcriptional oscillator”, but only to note the absence of a single “controller”:

The boundaries between cellular timekeeping and metabolism are becoming blurred as new knowledge is acquired. ... It is difficult to tease apart the relative dependence of so-called [metabolic] accessory loops and the [transcriptional] timekeeping mechanism, if this distinction is actually meaningful in view of the current evidence.

Distinctions, wherever they can be made, are always meaningful. It’s just that in the organism we rarely if ever see unambiguous, one-way relations between “controlling” causes and their effects. Rather, the primary governance always runs from larger context toward local physical interactions, and we cannot speak of any context as a discrete cause. It has its own qualitative and complex character, and its governance is less a matter of causation (in the usual sense) than of character expression.

What researchers are discovering about circadian “clocks” is another illustration of this truth.

Tags: explanation/causation; gene/genocentrism; machine idea

DNA and RNA brought to life

“Nucleic Acid Movers and Shakers”, by Karolin Luger and Simon E. V. Phillips*

These two prominent structural/molecular biologists have put together a special issue of Current Opinion in Structural Biology dealing with some of the molecular complexes responsible for the cell’s use of its DNA and RNA sequences. Their editorial introduction reads as if part of the special issue were expressly designed to underscore themes of the Biology Worthy of Life project.

The general idea is that DNA and RNA — those highly complex but, in and of themselves, relatively inert molecules that headlined the age of informational and molecular biology — are brought to life and movement by a remarkably dynamic range of protein activities. (I could have said “are brought to life and movement and meaning, if the word were not banned from biology. But, of course, the meaning is still invoked, as indicated by such words as “task”, “defective”, “good”, “well-placed”, “regulated”, and “function” in the following quotation.)

Daunting tasks such as the recombination of two DNA molecules, regulated gene transcription, the protection of DNA ends from being shortened each time the genome replicates, the degradation of “defective” RNA and the splicing of introns from “good” RNA are all performed by large multi-subunit molecular assemblies. Adding to this complexity, many of these assemblies are highly dynamic and variable with respect to subunit composition and orientation even while performing their tasks. Thus, the most interesting macromolecular complexes are perhaps the least amenable to conventional approaches in structural biology due to their shape-shifting characteristics. ... The importance of well-placed regions of intrinsic disorder [in proteins] is being recognized as an indispensable component of function, allowing regions of the molecule to move and flex just the right amount to perform its task ...

The authors go on to speak of a “veritable army” of molecules, each of which may require “the concerted action of many domains and subunits” — all toward the end of remodeling chromatin, that unfathomably complex “sheathing” of DNA, which mediates between DNA and the rest of the cell.

All this — involving as it does the qualitative re-molding of form, whether dramatic or subtle, whether in a small molecule or a massive macromolecule, and involving the infinitely intricate interplay of the forces of molecular bonding, which are inseparable from a chemical transformation of substances that remains in many ways as mysterious today as it was in the age of alchemy — all this carries us far away from the digital / informational vision of the organism that ruled biology during the past several decades.

Tags: chromosome/chromatin; DNA; protein

Functional vs. nonfunctional DNA binding

“The Uncommon Roles of Common Gene Regulatory Factors in the Genomes of Differentiating Cells”, by Eric H. Davidson*.

A Caltech developmental biologist, Davidson reviews work on the binding of transcription factors to DNA in two distinct mammalian cell types during differentiation. Transcription factors are key regulators — or, I should say, are among the dozens of key regulators — of gene expression. Two points made by Davidson are particularly noteworthy.

First, transcription factors are notorious for being capable of binding to huge numbers of DNA sequences, and it remains a great mystery how they manage to get to the right places at the right times in order to meet the cell’s immediate needs — and how they do this differently in different cell types. No one has a very good answer as yet, but a common view has been that we can ignore most transcription factor binding: it is “opportunistic” — undirected and without functional significance for the cell1.

But the work under review comes down strongly in favor of functional binding. The problem has been too narrow a focus: researchers haven’t adequately reckoned with the way multiple cellular elements work together in determining which transcription factors bind which DNA sequences. A given transcription factor might be capable of binding 10,000 different DNA sequences, but which sequences it actually binds in any given cell or cell type depends on the particular context of that cell.

(I’ve always wondered how any biologist at all familiar with organisms could have the nerve to claim that this or that dimly understood element or process is a manifestation of “randomness”, “noise” or “junk”, or is “nonfunctional”. In any case, you only need to hold onto your seat for a few years to see the general outcome of such claims, since they are rapidly being put to rest on all fronts.)

Second, Davidson writes: One idea that “has outlived its usefulness is the concept of the single ‘master regulator’”. Readers of this website will surely appreciate the correctness of that judgment. But wait. Davidson goes on to say, speaking of how transcription factors occupy regulatory enhancer sequences of DNA: “There are no ‘masters’ here; there are specific combinatorial enhancer occupancies that function as logic gates”. So, having dismissed the master regulator, does he find himself so bereft of tools for understanding that he must immediately embrace something like a (never identified) master computer program, opening and closing logic gates?

Tags: gene regulation; transcription/transcription factors


1. “The sheer number of sites bound by a given transcription factor, often more than 10,000, has made it difficult to gain a global understanding of transcription factor-mediated control of cell type identity. The question has been raised therefore whether a large proportion of binding events are ‘opportunistic’ rather than ‘functional’, where ‘functional’ would refer to those binding events that are relevant in terms of transcriptional control processes” (Calero-Nieto et al. 2014*).

Tags: chromosome/chromatin; DNA; explanation/causation; gene/genocentrism; gene regulation; machine idea; plasticity/genome; transcription/transcription factors


Calero-Nieto, Felicia S. Ng, Nicola K. Wilson et al. (2014). “Key Regulators Control Distinct Transcriptional Programmes in Blood Progenitor and Mast Cells”, EMBO Journal vol. 33, no. 11 (June 2), pp. 1212-26. doi:10.1002/embj.201386825

Davidson, Eric H. (2014). “The Uncommon Roles of Common Gene Regulatory Factors in the Genomes of Differentiating Cells”, EMBO Journal vol. 33, no. 11 (June 2), pp. 1193-4. doi:10.1002/embj.201488693

Doi, Masao, Jun Hirayama and Paolo Sassone-Corsi (2006). “Circadian Regulator CLOCK Is a Histone Acetyltransferase”, Cell vol. 125 (May 5), pp. 497-508. doi:10.1016/j.cell.2006.03.033

Edwards, Matthew D., Anna Symbor-Nagrabska, Lindsey Dollard et al. (2014a). “Interactions Between Chromosomal and Nonchromosomal Elements Reveal Missing Heritability”, PNAS Early Edition (April 22). doi:10.1073/pnas.1407126111

Hardesty, Larry (2014a). “It’s in the Genes — But Whose?”, MIT News (May 12). Downloaded May 15, 2014 from

Kumar, Vivek and Joseph S. Takahashi (2010). “PARP around the Clock”, Cell vol. 142 (Sep. 17), pp. 841-3. doi:10.1016/j.cell.2010.08.037

Luger, Karolin and Simon E. V. Phillips (2014). “Editorial Overview: Nucleic Acid Movers and Shakers”, Current Opinion in Structural Biology vol. 24, pp. 1-3. doi:10.1016/

Patel, V. R., K. Eckel-Mahan, P. Sassone-Corsi and P. Bald (2014). “How Pervasive Are Circadian Oscillations?”, Trends in Cell Biology vol. 24, no. 6 (June), pp. 329-31. doi:10.1016/j.tcb.2014.04.005

Reddy, Akhilesh B. and Guillaume Rey (2014). “Metabolic and Nontranscriptional Circadian Clocks: Eukaryotes”, Annual Review of Biochemistry vol. 83, pp. 191–219. doi:10.1146/annurev-biochem-060713-035644

Ryan, Frank P. (2006a). “Genomic Creativity and Natural Selection: A Modern Synthesis”, Biological Journal of the Linnean Society vol. 88, pp. 655-72. doi:10.1111/j.1095-8312.2006.00650.x

Talbott, Stephen L. (2011b). “What Do Organisms Mean?”, The New Atlantis no. 30 (winter), pp. 24-49. Original version published in NetFuture #182 (Feb. 22). Latest version, entitled “From Physical Causes to Organisms of Meaning” is available at

Yang, Xiaoyong (2010). “A Wheel of Time: The Circadian Clock, Nuclear Receptors, and Physiology”, Genes and Development vol. 24, pp. 741-7. doi:10.1101/gad.1920710

Further information:

You can find other articles relating to the themes presented here by clicking on any of the “Tags” listed after each article above.

This document:

Steve Talbott :: The Limits of Causal Understanding