Things move fast these days for us science writers.
I’m writing this just after returning home from a radio studio in New Haven, where I was interviewed on NPR’s The Takeaway about my article in today’s New York Times on the gene’s identity crisis. (Speaking of fast, the segment is already archived for your listening pleasure here.)
And now I sit down to find that my next order of business is to respond to a pretty harsh judgment of the article that appeared overnight, from a scientist I respect.
Time for the fourth cup of coffee of the morning!
So: The scientist in question is Michael Eisen, a biologist at the Univiersity of California at Berkeley. Among the many interesting things Eisen studies are transcription factors, proteins that grab onto DNA to switch on and off the production of RNA and proteins. The sites where transcription factors bind to DNA lie outside the conventional boundaries of genes, realms that have sometimes been erroneously referred to with the blanket term of “junk DNA.”
A few weeks ago I took some journalists and press release writers to task about their description of a newly discovered site (known as an enhancer) that may have played a key role in the evolution of the distinctly human hand. They erroneously claimed that the enhancer had been previously considered useless junk DNA, and so the news was that it actually had a function. I pointed out that people have known about enhancers for about 30 years, and other important sites outside conventional genes for 50. I argued that the news about this hand enhancer was interesting enough without the messed-up history.
Eisen liked that post, but he thinks I’ve failed to follow my own advice in my new article, where I describe a lot of features of our DNA that don’t fit the classical concept of the gene, such as alternative splicing, noncoding RNA, and epigenetics:
Well, he needs to apply the same standard to himself. Alternative splicing was discovered in the late 1970’s. Non-coding RNAs in the 1980’s. And epigenetic effects were described over 50 years ago, with molecular mechanisms first worked out over 25 years ago…Science writers play a very important role as honest interpreters of science for the public. But if they don’t present science history accurately, they can’t be taken seriously as authorities on science present.
With all due respect, I would like to propose a hypothesis: Dr. Eisen blogged about my article without actually reading it all the way through.
I base this hypothesis on the fact that in my article I repeatedly pointed out that scientists knew of examples of alternative splicing and the rest long ago. So why write an article now? Because–as I said in the article–what were once considered exceptions to the rule have become the rule. What could be set aside when scientists reflected on the concept of the gene can no longer be ignored. This is a story not about one particular experiment that yielded one particular result, but the story of a large-scale change in the way scientists think. I like the way a scientist named Mark Gerstein put it during an interview:
“The way biology works is different from mathematics,” said Mark Gerstein, a bioinformatician at Yale. “If you find one counterexample in mathematics, you go back and rethink the definitions. Biology is not like that. One or two counterexamples — people are willing to deal with that.”
I am not claiming that I’ve looked at the raw data from recent experiments and am telling scientists something they don’t already know. Just the reverse: I only wrote the story after talking to a number of scientists, some quoted in the article and some not, who all expressed a similar feeling that the weight of evidence today is leading to a remarkable change in how they think about genes. Nor did these scientists keep this feeling secret until I pried it from them–take a look, for example, at a new review in American Scientist co-authored by Gerstein that offers a new and improved definition of the gene, based on the research I describe in my article. I am also not claiming that all scientists feel the same way about the new results–for example, there’s a pretty spirited debate over how much of the noncoding RNA rolling off of our genomes does anything at all. And so I tried to squeeze a few quotes into the story to convey that disagreement.
Am I trying to have my cake and eat it too, as Eisen implies? I don’t think so, for the reasons I’ve laid out here. But maybe I’m wrong. I’m open to further criticism, but I still feel like these kinds of stories are worth writing. Some of the most interesting stories in science are move slowly on many fronts, rather than being one quick hit.
November 11th, 2008 at 12:31 pm
Just to be clear – I wasn’t criticizing the article as a whole. It is a great exploration of the things that are emerging from our newfound power to systematically characterize transcription, epigenetics and other properties of the genome. And I did read the entire article. But it was the hook that bothered me, so it’s what I wrote about.
What frustrates me – and you too I think, judging from your post on junk DNA – is that the need/desire to frame complex and inherently inherently science stories as revolutionary. Your nut graf clearly bases its pitch for people to keep reading on the idea that new research is overturning long-held beliefs about the structure of genes.
But, with all due respect to Mark Gerstein and other scientists who make similar points, I don’t think science viewed alternative splicing, non-coding transcription or epigenetics as oddities until these new studies came along (trans-splicing is another story). Rather, they were accepted as general phenomena whose scale and role in biology were poorly understood. What the new genomic studies have done is give us a detailed picture of precisely which alternative splice forms are made, which regions of non-coding DNA are transcribed, and what types of epigenetic marks are found in different regions of the genome. For all of our acceptance that these existed, we had little idea about what role they play in biology, or how they are generated and regulated.
Or, said more simply. We already knew that the genome was a dizzyingly complex place. The new data gives us the means to actually understand this complexity, and I don’t see why your story couldn’t have been framed around that idea.
This is more than just a minor pet peeve. The portrayal of science in the popular press is always around the idea that new experiments are showing that everything we’ve believed for decades was wrong, rather than showing it as a continuous process of discovery and elucidation. Yes, our initial models of genes were overly simplistic. But they weren’t really wrong either. They were useful abstractions of reality that allowed us to uncover the true complexity that exists in biology.
And let me just end by placing blame where blame is due. The stories on junk DNA that both you and I have criticized originate primarily with the authors of the studies in question and their press offices – who write press releases that pitch every new observation as novel and every new ideas as revolutionary. They do it because they think it’s the only way to get reporters to bite at their stories. We – scientists and science journalists – should work together to make sure this is unnecessary.
November 11th, 2008 at 3:24 pm
Good points all. “Paradigm shift” is quickly turning into nails on a chalkboard. I thought the story was well balanced — you even included the original definition of “gene” and how we’re moving back to the more generalized use that it initially had (I gave the same answer once at a talk). I am very glad that you tempered the “It’s all functional” claim with some “It’s potentially noise” with the transcription bits. If I had to be picky, it would only be that not all TEs are endogenous viruses. I also am not sure epigenetic impacts on the phenotype are so relevant evolutionarily — 1) do they accumulate, 2) are they reversible, and 3) are they actually encoded elsewhere? Seems more like plasticity than evolution when this occurs, but intriguing.
Carl: ‘Sup with no Genomicron on the blogroll?
November 13th, 2008 at 12:49 pm
Worth mentioning – From Thomas Gingeras’ website:
(http://gradschool.cshl.edu/staff/gingeras_.html)
“Genomes from yeast to man are almost totally transcribed. More than half of this transcription is composed of non-protein coding (nc)RNA. The organization of transcribed regions found in each cell is highly individualized and consists of overlapping RNAs in which the same genomic sequences encoded in both protein coding and non-coding transcripts. We have developed genome-wide transcription maps in several cell types belonging to multiple organisms. Our biochemical characterization of the mapped transcripts has revealed several common features found in eukaryotic transcriptomes. These include: 1) the number of RNA isoforms for protein coding genes is significantly underestimated relative to what most databases indicate, 2) more than 70% of all protein coding genes have at least one antisense transcript overlapping the coding transcript, 3) the 5’ and 3’ boundaries of most protein coding genes can be positioned hundreds of thousands to millions of base pairs away from the annotated ends of a gene, 4) chimeric RNAs are made by eukaryotic cells encoding portions of a genome mapping at great distances from each chimeric section, 5) the fate of many nc-transcripts is to be processed into short RNAs and 6) the 5’ and 3’ ends of protein coding genes are sites at which collections or short RNAs can be found. Many questions have emerged concerning this pervasive transcription. These questions focus on the nature of the regulatory regions and enzymatic processing pathways that control the expression and maturation of non-protein coding RNAs, respectively as well as the functions of these long and short ncRNAs. Answers to these questions will provide a new view of what a gene is and how genomes are organized and regulated.”
November 13th, 2008 at 12:56 pm
Reading the information above and other cutting edge research, inquiring minds are hard-pressed to explain how these phenomena can be attributed to random, accidental or non-directed events.
Clearly, there is a massive intelligence at work.
November 14th, 2008 at 12:43 am
Charlie Wagner, I thouht you wrote in jest.
Then I checked your site link.
Then I realized…