Chapter I

Now the leading pre-Darwinian form of evolution was that ultimate “by-your-bootstraps” model of biological improvement, the inheritance of acquired characteristics. This idea has come to be associated exclusively (and undeservedly) with the Frenchman Jean-Baptiste Lamarck (1744-1829), who was flirting with the transformation of species early in the 19th century, Milner (2009, 268-270). The concept is simple enough: if only you pumped iron thoroughly enough, you might pass on those hefty muscles to your descendants without their actually having to work at it themselves. It was a very attractive idea (Darwin himself dallied with such explanations in some of his work, as when trying to explain the giraffe’s long neck), that hung over general science until modern genetics finally knocked it out of the court. The idea that animals mutate and change because they “want” or “need” to occasionally gets a revival moment, as Sniegowski & Lenski (1995) and Brisson (2003) noted of Cairns et al. (1988) regarding “directed mutation” (more on the directed mutagenesis issue in due course), but in the public education world the idea of intentionality in evolution hangs on as such a common student conceptual error that science educators have to labor at ways to correct it, Colburn (1994) and Gregory (2009c).

Milner (2009, 253-254. 319-320, 322-323) surveys the Neo-Lamarckian crannies of evolutionary history. Though explicit inheritance of acquired characteristics didn’t pan out, the contested experiments of Paul Kammerer (1880-1926) early in the 20th century would have been bumping into the area of epigenetics (whereby gene expression or even physical appearance can be influenced apart from DNA coding) prompting recent reevaluations of Kammerer’s work, Pennisi (2009d) and Randy Moore (2011b). Several genetic systems turn out to function in a “Lamarckian” way, Koonin & Wolf (2009), especially ones relating to how bacteria deal with viral attackers or develop resistance to other organism’s defenses (including aspects of Horizontal Gene Transfer between bacteria, which issue will crop up repeatedly in the antievolutionism debate).

Though the concept of epigenetics had been knocking around for decades, improvement in genetic analysis and measuring gene expression opened the field up to detailed investigation in the 1990s and work in the area has ballooned in the last few years. A measure of how far the field had advanced came in October 2010 when Science devoted a special issue to it: Riddihough & Zahn (2010) re Bonasio & Tu et al. (2010), Bourc’his & Voinnet (2010), Chandler (2010), Feng et al. (2010), Halfmann & Lindquist (2010) & Hemberger & Pedersen (2010).

Epigenetic factors have been found to operate at various levels: why identical twins can grow up far from identical, G. Martin (2005) re Fraga et al. (2005), or P. Miller (2012); the dynamics of skeletal variations, R. Young & Badyaev (2007); genetic conflict between the sexes, Lemos et al. (2010); the mammalian immune system and obesity, Tykocinski et al. (2010) and Ng & Lin et al. (2010); human cognition and behavior, G. Miller (2010d-e) and Nestler (2011). They even run into politically and culturally contentious areas: from the specialization of stem cells, such as Loh & Lim (2012) re Doege et al. (2012) or Sassone-Corsi (2103) re Shimazu et al. (2013) & Shyh-Chang et al. (2013),to the origins of human homosexuality, S. Richards (2013) re Rice et al. (2012). Such research has even spilled over into politics and environmental controversies by suggesting that inherited epigenetic reactions might be triggered by chemical pollutants, such as the work of Michael Skinner noted by Kaiser (2014), though Skinner’s findings have been dogged with difficulties in replication and even some data fabricated by a less than scrupulous postdoc, Hughes (2014).

As for the Lamarck factor, in some cases epigenetic tracers may be inheritable, Mango (2011) re Greer et al. (2011), Crews et al. (2012), and Mattick (2012) re V. Nelson et al. (2012), and Schmitz (2014) re Cortijo et al. (2014) experimentally tracking how the “epialleles” (counterparts of the variant alleles of regular genes that are distinguished not by difference in their underlying DNA but by the specialized epigenetic signatures attached to them) in the plant Arabidopsis contribute to changes in the complex traits of flowering time and root length. The moment the epigenome qualifies as an agent of heritability, of course, it becomes grist for the Darwinian natural selection mill to cull or preserve—though still not in quite the archetypal “Lamarckian” manner of altering the organism’s base DNA, Pennisi (2013h), but instead through attached expression signals that can remain in place through the replication process in descendants.

In eukaryotic organisms, where the DNA is sequestered in a nucleus and consists of long open ended strings rather than shorter prokaryotic rings, the protein-making machinery of the cell can only get at the DNA to process it into RNA transcriptions to act as a blueprint for the protein when the DNA is unwound from its chromatin packets, illustrated by Nestler (2011, 79). Epigenetic markers like histone acetyltransferase (HAT)—themselves coded by still other stretches of DNA—either keep the DNA wound tight around its histone spacers, preventing the automatic replication machinery from ever getting at it, or relax the spools so that the mechanism can move in to do its job and make active proteins. In that way the level of an active protein can vary in a cell.

Since the DNA doesn’t signal directly whether to tighten or loosen the chromatin for its transcription, it was no shock to learn there were hitherto unknown additional layers involved, and researchers had been actively exploring that trail for some years by the time Marmorstein (2001) specified HAT structure. An extensive exploration of chromatin modification in plants commenced, such as Pandey et al. (2002), and from the reviews by Carrozza et al. (2003) and K. Lee & Workman (2007) a tremendous range of functions were discovered for HATs (for instance, the same systems having taken on highly specialized roles in yeast, flies and humans). Far from isolated from the epigenetic chromatin-histone dance, though, coding variants in the DNA directly influence the epigenetic signaling, Furey & Sethupathy (2013) re Kasowski et al. (2013, Kilpenin et al. (2013) and McVicker et al. (2013). Some modifications to histones appear to be heritable, Ruth Williams (2014b) re Gaydos et al. (2014), Audergon et al. (2015) and Ragunathan et al. (2015).

The result can be quite a complicated dance to follow, with long noncoding RNAs (LncRNA) involved in a “Wild West” tangle of epigenetic signaling explored by Latos et al. (2012), J. Lee (2012) and K. Morris (2012), in turn having to run through the gauntlet of conformation changes chromosomes go through during the cell cycle actively working against preserving epigenetic tracers trying to hitch a ride down to descendants, Kleckner et al. (2013) re Naumova et al. (2013). As Stephanie Keep (2014c-e) summarized, cell processes tend to sweep clean the epigenetic markers, but anything that makes it through can play a continued role (for a while at least). The dynamics of vertebrate sexual reproduction plays a part here too, as the paternal side of cellular DNA methylation is actively swept clean during human embryonic development, Reik & Kelsey (2014) re H. Guo et al. (2014) and Z. Smith et al. (2014).

The deep evolutionary roots of epigenetic regulation are suggested by the way noncoding RNA operates in the very ancient yeast, Stritch (2011) re Lardenois et al. (2011), and evolutionary scientists have begun relating those processes to the broader ones of how the genetic novelties of life come about since by these natural means. As Nina Fedoroff (2012a) summarized in a recent review in Science, the epigenetic system arose deep in the history of life to protect their DNA from external invaders: from bacteriophages to viral retrotransposons (packages of RNA with “copy me” instructions that allow their spectacular proliferation in eukaryotic genomes).

While prokaryotic organisms tend to eventually delete transposons that venture into their closed-loop DNA rings, the long winding nucleated genomes of eukaryotes are less tidy and more vulnerable. The result is that some transposons linger on, shut down but present, accumulating what amounts to a genetic library of untapped possibilities. Should further mutations switch them back on again, they can cause anything from a diseased genetic wreck to sparking novel functions to appear, along with the antievolutionary spin put on such findings by the likes of the Discovery Institute’s Jonathan Wells.

Such an active level of “epigenomic programming” (where the regulatory system has adapted to flexibly permit favorable mutations to get their chance in the selectionist sun) has been identified regarding the F-Box protein superfamily in plants, Hua et al. (2013), where one branch with a high deleterious mutation rate has less epigenetic regulation (natural selection would weed those mutations out regularly without it) than other lines where innovative mutations are more likely to occur, and so benefit from an epigenomic layer that can dampen the expression of unfavorable mutations but relax when good ones emerge.

With all this coverage it’s clear that the regular scientific community was neither unaware of the epigenetic phenomena (they defined and explored it) nor treated it as some antievolutionary news bulletin (as we’ll be seeing in further examples to be covered in due course). And yet that is exactly how antievolutionists have approached epigenetics, when they address it at all. Thus CreationWiki (2011j) opined that epigenetics “has serious implications for creation biology, given the fact that major phenotypic changes can occur without the Darwinian process of genetic mutation and natural selection.”

CreationWiki thinks epigenetic inheritance “provides a potential mechanism of the created kinds”—though without going into any details. Which is understandable, since creationists haven’t actually made much progress in working out just how many “created kinds” there were supposed to have been or to what extent they have been modified by natural speciation. To put the point more baldly, were someone to contend that epigenetic inheritance “provides a potential mechanism for the existence of unicorns” don’t you have to show there actually were unicorns first, before being permitted to drag epigenetics (or any other natural system) into your argument?

More fundamentally, though, CreationWiki misses the lesson that the existing scientific research (none of which emanated from any creationist researcher) is already confirming: that these newfound mechanisms do not represent anything other than just more naturalistic and unguided processes operating in an evolving natural world, as reminded from the anticreationist camp by geneticist Jerry Coyne (2011k), or Jack Scanlon (2012a) commenting on the Discovery Institute’s Evolution News & Views (2012d). Indeed, there seems no reason to think that epigenetic signaling is any less prone to natural mutation and variation than the DNA it helps regulate, as evidenced by Becker et al. (2011) or Schmitz et al. (2011), or isolated from the genome doubling process that can lead to hybrid speciation, Paun et al. (2007).

This is the same problem for the Intelligent Design spin on epigenetics. Cornelius Hunter (2009c) claimed it “leaves evolutionists in an awkward position” because “such intelligent adaptation capabilities suggest design, not accident.” Hunter’s only documentation for this was a brief track back online to an entry in his own C. Hunter (2009a) and to the general published survey by Jablonka & Raz (2009), as though any of this buttressed epigenetics as an intelligent response to anything, rather than a totally blind mechanistic reaction to chemical markers that can function usefully or dangerously depending on circumstances.

For example, epigenetic factors have been implicated in cancer for some time, such as Lotem & Sachs (2002), and this appears to be due to inherent instability in the genetic process. As explored by Francis (2011), the many factors that can affect the functionality of coding DNA (from demethylation of the chain at its source to changes in expression as other molecules modify the “control panel” parts of the gene) can easily end up as cancerous systems, an issue of “defective design” as troublesome for ID rationales as other peculiarities of complex biological systems.    This is especially true of addictions, where humans can be ensnared in a destructive dependence by the epigenetic processes running normally, as described by Volkow (2011) re Levine et al. (2011) on nicotine addiction, or the more general review of epigenetic effects on cocaine addiction and depression by Nestler (2011). Similarly some cardiorespiratory complications like hypoxia appear to be aggravated by the epigenetic system doing its thing, Lagercrantz (2012) re Nanduri et al. (2012).

The evolutionary nature of all this becomes even clearer the moment you recognize the Map of Time aspect of it. Some of the epigenetic systems in mammals occur because they have had to deal with invading retroelements for a very long time, roughly since the dinosaurs checked out. Sixty-five million years later Leonova et al. (2013) identified at least three layers involved: the p53 protein to suppress their transcription in the first place, epigenetic factors to prevent their translation into RNA if they do get transcribed, and finally a “suicidal interferon response” to kill the cell as a last ditch defense. The reason why you have to take all this trouble is that the most likely effect of a buried retroelement turning on is cellular malfunction, such as cancerous tumors. From a “design” perspective, why put all that stuff in there to begin with, only to have to build moat after moat to keep the stuff accidentally turning back on and become dangerous? From an evolutionary perspective there is no such option, no neat designer editing device exists to prevent retroelements from intruding every now and then, or to detect them automatically once they are present. That leaves only one natural response: either further mutations enable your systems to manage the invader by a variety of ad hoc means, or in the long run your lineage ends up extinct—and thus excluded from any further analysis by the likes of Katerina Leonova et al.

But design advocates are not prone to explore these deeper issues. If something like epigenetics is “new” and hence not yet integrated into the bigger scientific picture, that’s about as far as the ID treatment is prepared to go. Thus another possible epigenetic effect (this time in redwoods, where the genes appear to vary from bottom to top) drew the attention of Uncommon Descent (2011h) to a secondary account in ScienceDaily (2011b) regarding Christopher Cullis’ experiment work on the possible environmentally induced inheritance of gene insertions in flax. Though nothing in the main paper, C. Johnson et al. (2011), nor the earlier Cullis (1981), Schneeberger & Cullis (1991) and Y. Chen et al. (2009), suggested anything other than a natural process was going on here, Uncommon Descent discerned the hand of Design: “Some forward look must be built into the system in advance.”

In a similar display of expectations derailing data, Evolution News & Views (2012b) waxed hyperbolic when they declared the epigenome to be “Evolution’s Newest Nightmare.” They took note of how their Center For Science & Culture Fellow Richard Sternberg was making epigenetics “the focus of his research lately”—though his concrete technical accomplishments in this area has yet to advance much beyond a general allusion to it in Sternberg (2002). But the bulk of the posting was directed at another target, a string of quotes from a trio of Harvard biologists sketching out approaches to integrate the pile of epigenetic findings into the regular evolutionary framework, Ben Hunter et al. (2012). Evidently Hunter’s team was unaware they weren’t supposed to be able to do that, and EN&V deemed their perfectly innocuous suggestions as trying to “look brave” in the face of a problem the scientists were supposedly stubbornly failing to acknowledge. “At this point evolutionists do not know which human instinct to follow: fight or flight,” but EN&V peaked into their skulls in absentia to concede, “The look on their faces is curiosity instead of terror.”

An excess of scientific curiosity is a defect Intelligent Design authors can seldom be accused of, for a couple of days later their doppelganger colleagues at Uncommon Descent (2012i) simply reprised the EN&V take on the Hunter paper, dismissing the scientists’ work as one merely trying “to pretend that nothing much has happened.” Sure.

As the scientific literature piles up, the ID cooptation machine ratchets up in response, so that Evolution News & Views (2013aa) declared “The lesson is clear: intelligent design is in the best position to promote scientific discovery, and to deliver the understanding sought by science.” And their evidence for this amazing accomplishment by proxy: yet another parade of quotes from some recent epigenetic work in embryology and other areas, such as Pennisi (2013f), with sections in bold whenever the complexity or organizational structure of the regulatory systems were alluded to. No claim that any of the scientists responsible for this work either were ID advocates or paid the slightest attention to “design thinking” as ID conceives of it (non-evolutionary interventions in life), or that they imagined their work would ever have such implications. Pennisi’s summary alluded to the XIST pseudogene, for example, which the anonymous Evolution News & Views called attention to, but EN&V’s curiosity stalled when it came to exploring any of the relevant background papers, such as Duret et al. (2006) or Chaumeil et al. (2011) on the natural origin and evolution of XIST.

Instead, Evolution News & Views proposed that “While evolutionists scramble to deal with the unprecedented complexity, intelligent design is not surprised by it,” and offered Stephen Meyer’s 2009 book Signature in the Cell as exemplar, which displayed its deep prescience and utility in this department by omission: “epigenetics” didn’t even show up as an index topic. Fortunately, Meyer (2013a, 271-287) has remedied this oversight retroactively in Darwin’s Doubt.

Along with Meyer’s Darwin’s Doubt, Evolution News & Views (2013am) continued to ride the epigenetic hobbyhorse off the design cliff, grasping at increasingly peripheral straws in their campaign to undermine the Darwinian menace, this time contending that a paper on how exercise could alter the DNA methylation pattern in humans, Rönn et al. (2013), had “implications for whether or not evolution is the only guiding force in how man came to be as he is today. If genes are the focus of the Darwinian mechanism, then what does it mean that man can change his genetics by changing his behavior? Perhaps Darwinian evolution explains less than previously thought, particularly in the context of human evolution—as you know we’ve suspected all along.”

The fundamental mistake in the design line of reasoning here is that it turns on a trick of definition: arbitrarily parsing “Darwinian evolution” so that it applies only to one narrow track of evidence (coding genes for example), thereby allowing any new discovery like epigenetic regulation to be seen as somehow contradicting a “Darwinian” dogma of their own contrivance. The issue is actually whether such epigenetic processes are any less natural and mechanistic as Darwin’s natural selection, or have any lasting effect on the genetic structure in question, apart from a healthier body through exercise-stimulated methylation perhaps enhancing the likelihood of successfully getting a date and hence improving the odds of reproducing and thus passing on those epigenetically burnished but otherwise unchanged genes to one’s progeny.

Nothing in the Rönn paper suggested anything design-friendly in this sense. Indeed, the exercise-stimulated methylation changes could only work their purely naturalistic magic on genes whose regulation was prone to methylation in the first place (the paper specifically noted the presence of “gene islands” where the absence of methylation precluded that) and no changes to the underlying genes themselves were ever indicated. This shell game of “aha, something non-Darwinian is happening!” will spool out repeatedly in the design campaign.