Siddhartha Mukherjee's recent article on epigenetics has drawn heavy criticism—rightly?by Philip Ball / June 7, 2016 / Leave a comment
Read more: There is no intelligence gene… singular
When I first saw the avalanche of anger and complaint from scientists directed at an article in the New Yorker written by Siddhartha Mukherjee, I couldn’t help wondering: what could it possibly be about this young, Pulitzer-prize winning Ivy League professor and one-time Rhodes scholar that has set people’s backs up so? But the truth is that the criticisms of Mukherjee’s article on epigenetics (a term that I will go on to explain more fully, but which means “above genetics”) can’t be dismissed as jealous carping. In fact, they have a great deal of validity. The article can easily be read as yet another contribution to the hype about epigenetics—which, some have misleadingly alleged, rewrites the standard picture of how genes affect biological development and evolution.
Mukherjee—who won the 2011 Pulitzer for his book on cancer, The Emperor of All Maladies: A Biography of Cancer (at Columbia he is a professor of medicine specializing in cancer)—laid out clearly enough the basic question to which his piece was addressed. What is it that distinguishes a neuron from a muscle cell from a white blood cell, given that all of them, in a single individual, share the same genome? Or as he put it, “Why doesn’t a liver cell wake up one morning and find itself transformed into a neuron?”
The answer has been long recognized. During development, some genes get turned off while others might be turned on or ramped up, so that the different cell types execute different functions. The cells of a very early-stage embryo, just a few days old, possess all of their genetic potential: these “embryonic stem cells” are pluripotent, meaning that they can potentially develop into any of the tissues of the body. As development proceeds, they “differentiate” into specific cell types by the activation and deactivation of genes.
This is why embryonic stem cells are of such potential value to medical science. From them, it might be possible to grow any kind of tissue, given the appropriate cues and nudges—and thus to regenerate damaged tissues and organs, to replace a faulty kidney say, or to create nerve cells to repair spinal-column injury or neurons to combat neurodegenerative disease. One of the most exciting recent discoveries in medical cell biology was the finding of Japanese researcher Shinya Yamanaka and his coworkers in 2006-7 that fully differentiated adult cells can be returned to a stem-cell-like pluripotent state by introducing into them certain proteins. These seem to reactivate switched-off genes. For this discovery Yamanaka received the 2012 Nobel prize in physiology or medicine. To what extent these “induced pluripotent stem cells” (iPS cells) are really like the embryonic versions remains, however, a matter of research.
Here, then, is one good reason to want to understand how genes are switched on and off. There are plenty of others. Yet Mukherjee’s New Yorker piece could very easily leave the impression that it is all to do with epigenetics. This entails the chemical alteration of genes by enzymes, whereby they acquire little molecular tags that can silence them. This kind of gene switching can also be achieved by chemical modification of the so-called histone proteins on which DNA is wound, like thread on a bobbin, in chromosomes. These epigenetic marks are well established, and indeed one of the questions about iPS cells is to what extent the marks become erased in the pluripotent state—whether or not the “resetting” is complete.
“It seems plain that some of the DNA previously thought to be “junk” accumulated through millions of years of evolution in fact has a biological role”
The complainants, however, accuse Mukherjee of focusing solely on epigenetics in gene regulation, overlooking a far more important issue: the role of so-called transcription factors. It has been recognized since the 1960s, thanks in particular to the work of 1965 Nobel laureates François Jacob and Jacques Monod, that the genome is not simply a passive set of instructions for making proteins, but that there is a complex network of interactions between its genes. Some genes encode proteins called transcription factors that can bind to other genes and either enhance or inhibit the activity of those genes (the rate at which they are transcribed to produce a protein). Transcription factors can generally recognize particular sequences (strings of the four chemical building blocks of DNA), making them able to target specific genes. In this way, the genome is constantly regulating its own “programme.” It was by using transcription factors that Yamanaka was able to create iPS cells.
Epigenetic marks can modify the behaviour of transcription factors—for example preventing such a protein from binding and exerting its effect on gene activity. But they are, Mukherjee’s detractors rightly point out, in general a secondary effect after the gene-regulating role of transcription factors. As Steve Henikoff of the Howard Hughes Medical Institute put it, epigenetic markings like histone modifications “at most act as cogs in the machinery that enforces the often complex programs specified by the binding of transcription factors.”
While plenty is now understood about gene regulation, there is still plenty that is not. The discovery, in the Human Genome Project to sequence our entire complement of DNA, that we have far fewer genes than was previously thought—perhaps just 20,000 or so—suggests that much of the enormous complexity of human cell biology comes not from our genes per se (onions and bananas have more genes than us) but from the network of gene interactions that modifies their collective activity. It also seems plain that some of the DNA previously thought to be “junk” accumulated through millions of years of evolution in fact has a biological role, perhaps encoding RNA molecules (RNA is the intermediary molecule between a stretch of DNA and the protein it encodes) that themselves regulate other genes. It also seems that the three-dimensional shape in which DNA is compacted into chromosomes, folded up in combination with proteins like histones in a material called chromatin, also influences gene activity.
“Some argue that markings on the genome may be induced by an organism’s environment, and so in some respects experience can leave “scars” on our genome”
Epigenetics is a part of this story. But many researchers feel that it has been vastly overhyped, as if it somehow rewrites the basics of molecular biology and genetics. It doesn’t. Mukherjee’s article tapped into that deep discontent. He got some things wrong, for example misrepresenting the message of Yamanaka’s work—but mostly his sins were of omission.
To compound matters, Mukherjee flirted with the popular notion that epigenetics also rewrites Darwinian evolution by suggesting that a form of Lamarckian evolution can also occur. Jean-Baptiste Lamarck’s pre-Darwinian view was that characteristics acquired during an organism’s lifetime can then be passed on to its progeny—in the classic example, by stretching to reach high leaves, giraffes acquire long necks which they bequeath to their offspring. Epigenetic markings on the genome may be induced by an organism’s environment, and so in some respects experience can leave “scars” on our genome. It is generally thought that these marks are wiped clean in the germ cells—eggs and sperm—that give rise to children. But some studies have suggested that epigenetic markings can survive that process, so that children or even grandchildren can exhibit traits acquired by parents. Such claims are controversial—but even if they prove to be true, it seems highly unlikely that the effect will persist for many generations of will have long-term consequences for human evolution.
Yet this suggestion hints at the reasons for the ire. Mukherjee deserved to be taken to task for giving little hint that there’s a broader story beyond epigenetics, but it’s significant that some of the harshest criticism came from evolutionary biologist Jerry Coyne of the University of Chicago. Coyne has long been a staunch defender of Darwinian evolution in the face of US creationists and advocates of its pseudo-scientific version, intelligent design, and he often gives the impression that talk that seems to challenge the Darwinian picture looks like unconscionable recidivism. “Until there is evidence for this kind of evolutionary transformation—ANY evidence”, he growled, “people should stop yammering about this kind of ‘Lamarckian’ evolution.” Coyne also gathered together criticisms of Mukherjee’s piece from the kind of stellar array of Nobel laureates and other biologists that would make even a Pulitzer-prize-winning professor blanch. One called the piece “so wildly wrong that it defies rational analysis,” while another said “It is unfortunate to inflict this article, without proper scientific review, on the audience of The New Yorker.”
We don’t know what all genes do—but does it matter?
What happens when we can’t test scientific theories
Is the notion of a cure for cancer too simplistic?
Mukherjee has admitted that he made “an error” in focusing on epigenetics at the expense of gene regulation by transcription factors. At face value the mistake is puzzling. Mukherjee clearly knows the full story—it is given a more balanced treatment in his new book The Gene (which I praised in Prospect), on which the New Yorker piece is largely based. He says that some of that material was in the original draft article, but it was removed during the editorial process in order to tell a “simpler story.” He and the editors considered that decision “very carefully”, he says. That won’t quite wash, however. All science writers are familiar with the simplifications and compromises demanded when space is tight—but Mukherjee enjoyed the rare luxury of a 6,000-word article, in which one might have expected to see some acknowledgement of the issues that Coyne and others identify.
All the same, Mukherjee has surely been hit by the epigenetics backlash. For this topic has indeed acquired a faddish appeal—a “dangerously fashionable” one, in the words of American criminologists Brian Boutwell and JC Barnes—among the social sciences community. Take a recent, much-reported paper in the journal Molecular Psychiatry, for example, which argued that epigenetics can be linked to depression. It presented evidence that the effects of poverty and low socioeconomic status may impose an imprint on the genome that manifests as depression-related brain function in adolescents. The uncritical reception this claim received belied the statistics on which it rested, which look destined to join the growing ranks of unreproducible claims. You don’t need to be an expert in statistics to wonder how much reliance to put on this “trend” extracted from the scattershot data.
This prompts the question: why, then, have epigenetic explanations of traits become so popular? Boutwell and Barnes feel there is a clear agenda: “Many social scientists felt vindicated by the findings, assuming it represented a triumph of the ‘social’ over the ‘biological.’” And indeed I suspect the appeal has a lot to do with the antidote epigenetics apparently offers to genetic determinism. If the environment can effectively alter our genes, we are no longer slaves to the genomes we inherit. Epigenetics seems to rescue some self-determination and free will in the face of our implacable DNA.
If so, then that perceived need shows how misleading the narrative of genetics has become in the first place. Some aspects of our genetics undoubtedly constrain and define us, not least in the important context of susceptibility to disease. But some of Mukherjee’s detractors might want to consider if they have unwittingly played a part in creating an unhealthy perception that we are no more than what our genomes dictate.