Human Genome

Is Junk DNA Really Junky?

The delicious, religious debate over what most of our genome is good for.

The silk dancers' interpretation of DNA loops.
Dancers’ interpretation of DNA loops

Courtesy of Tom Whipps/Nature

In the 12 years since the human genome was sequenced, so many critters have had their DNA deciphered—oysters, bees, eels, camels, clawed frogs, elephant sharks—that it’s hard to suppress a yawn sometimes. But every so often, a genome cuts through the indifference and makes geneticists’ eyes goggle out. Take the humped bladderwort, a humble aquatic plant whose DNA was sequenced this past May.

The humped bladderwort has yellow, snapdragon-like flowers, and it’s actually carnivorous, capable of trapping and eating not just insects but even tadpoles and tiny fish. But this combination of beauty and death isn’t what makes the bladderwort special. Most organisms have loads of junk DNA—less pejoratively, noncoding DNA—cluttering their cells. The bladderwort doesn’t: 97 percent of its DNA is classic, hardworking, protein-building DNA. And that lean, mean bladderwort DNA challenges some trendy notions about how all DNA works, including (if not especially) in human beings.

First, a primer on junk DNA, one of the most reviled terms in science. Anyone who took Bio 101 remembers (if only vaguely) that DNA gets turned into RNA, which in turn gets turned into proteins. The protein-producing stretches of DNA are called genes, and genes reside on much longer molecules called chromosomes.

A century ago, as biologists came to grips with the vast number of different proteins needed to build and maintain the body, they decided that genes must be packed very tightly together on chromosomes, since tight packing would be more efficient. They couldn’t have been more wrong. In humans, a typical species in this regard, less than 2 percent of our 3 billion letters of DNA actually builds proteins. Chromosomes were more like vast Saharan wastelands, broken up only sporadically by oases of genes.

So what does that extra 98 percent do? Here’s where things get contentious. Some of the excess—the pseudogenes, the transposons, the tedious stretches where Mother Nature held her fingers down on the keyboard (ACACACACACA … )—does look like garbage. Heck, 8 percent of our genome is nothing but old, broken-down virus DNA, the genetic equivalent of a Pontiac Firebird on cinderblocks. The name junk DNA emerged in the early 1970s as a catchall term for this cruft.

Even at the time, though, some scientists objected to the term as too dismissive. Molecular biologists had already discovered bits of junk that, far from being irrelevant, actually managed genes: They turned genes on or off and regulated when and where genes were active. As more and more examples of this type of control emerged in the 1980s, the term junk DNA seemed less and less appropriate.

The protests grew especially loud after the Human Genome Project decided (somewhat arbitrarily) to declare the genome fully sequenced in 2003. Before, most geneticists argued that, based on our size and smarts, human beings must have around 100,000 genes. The Human Genome Project turned up just 23,000. (That’s fewer than the bladderwort’s 28,500, incidentally.) Biologists now faced a doozy of a dilemma: How can a species as complex as Homo sapiens get by with so few genes?

One good bet was noncoding DNA. Again, noncoding DNA can switch genes on or off or make them produce proteins faster or slower. It also helps splice and remix genetic material, allowing different types of cells (neurons especially) to customize their RNA and proteins. In other words, noncoding DNA allows us to use one gene in many ways, multiplying the effective number. Perhaps, then, it wasn’t genes alone that made human beings special; it was how we used genes that counted.

Noncoding DNA also offered new leads on curing diseases. Frankly, the sequencing of the human genome hasn’t lived up to its hype here: Almost no new treatments have emerged, and there aren’t many in the pipeline, either. Things look especially bleak for common killers such as diabetes and heart disease. Those ailments clearly have a genetic component. But when scientists survey genes looking for which mutations patients have in common, they come up empty. In other words, far from curing these diseases using genetics, scientists can’t even find the right DNA to target. Geneticists are still hashing out the details of why this approach whiffed, but part of the problem could be a failure to understand how noncoding DNA contributes to diseases.

These high expectations for noncoding DNA peaked last fall thanks to an ambitious project known as ENCODE. It set out to catalogue every last bit of noncoding DNA in the human genome. It cost $288 million and produced a supernova burst of 30 scientific papers last September, including an overview article in Nature with 442 co-authors. Given its size and scope, ENCODE covered a lot. But its leaders trumpeted one main finding above all: that 80 percent of noncoding DNA had some sort of biological function. This was like Columbus discovering five new continents at once—whole new worlds of unexpected genetic activity and potential therapeutic targets to exploit. If the Human Genome Project deflated hopes about genetic medicine, ENCODE pumped them right back up.

All the while, however, a few scientists were grumbling about ENCODE, and in a slew of papers from earlier this year, they argued that ENCODE was vastly overselling itself. In particular, they disputed the claim that 80 percent of the genome was somehow “active” or “functional.” For instance, cells sometimes transcribe DNA into RNA only to turn around and destroy that RNA moments later. It’s Sisyphean work that doesn’t benefit the organism at all—but ENCODE nevertheless counted that DNA as functional. (One ENCODE leader even admitted that, in trying to determine what percentage of DNA they should count as “functional,” his team had played around with different definitions, resulting in a range between 20 and 80 percent. They ultimately decided to push the 80 percent because “the bigger number … brings home the impact of this work to a much wider audience.”)

In other words, critics complained, ENCODE had defined “functional DNA” so broadly that the term lost all meaning. No biologist really disputes ENCODE’s goal: Some junk DNA does have a function, and we need to understand it. But in saving the baby, ENCODE may have saved a lot of scummy bathwater, too.

The scientific backlash was harsh enough, but critics really worked themselves into a lather over ENCODE’s marketing campaign, which included scores of interviews, an iPad app, and a promotional cartoon. One critic equated the media push with “sleight-of-hand.” Another announced that he was “ready to drink [him]self into a stupor.” One especially vitriolic paper chided the ENCODE scientists’ “absurd conclusions” and “self-serving” behavior; even the acknowledgments section contains snide comments. The paper concluded, less than magnanimously, that “the ENCODE results were predicted by one of its authors to necessitate the rewriting of textbooks. We agree, many textbooks dealing with marketing, mass-media hype, and public relations may well have to be rewritten.”

If all this wasn’t enough, the intelligent design crowd soon jumped into the fray, making the debate even more acrimonious. In short, proponents of intelligent design—a branch of creationism that uses scientific language but not scientific ideas or standards of evidence—strongly backed ENCODE. Why? Because the sheer bulk of junk DNA inside us doesn’t reflect well on God’s engineering skills. Why not eliminate the waste, Big Guy? But if most noncoding DNA does have an essential purpose, then perhaps God crafted us molecule by molecule after all. And if that’s the case, well, then evolution is bunk and Jesus Christ died for our sins (or something like that—their reasoning’s a bit fuzzy). To be clear, no one claims that ENCODE scientists support intelligent design. But critics worried that, in addition to producing flawed science, ENCODE also provided ammunition for the enemy.

Now, finally, the poor humped bladderwort has been dragged into the contretemps. Millions upon millions of years ago, the bladderwort had a normal complement of junk DNA. (We know this because it shares an ancestor with grapes and tomatoes, both of which are pretty bloated with junk.) For whatever reason, though, it began to shed that extraneous DNA generation by generation, until it arrived at the svelte genome of today. At the same time—and here’s the key—the bladderwort has been evolving a lot, even picking up new traits. Indeed, the water traps it evolved to catch bugs and tadpoles are one of the marvels of the plant kingdom, capable of snapping shut in less than a millisecond, hundreds of time faster than you can blink your eye. In other words, even while the bladderwort’s genome got vastly simpler, its body got more complex—which undermines the idea that 80 percent of junk DNA does something vital. And sure enough, the scientists who decoded the bladderwort genome took a swipe at ENCODE in their paper.

Of course, you could now argue that it’s the ENCODE critics who are stretching things. After all, how much can the genome of the bladderwort, a plant, really tell us about the genome of Homo sapiens, an animal? But even within the animal kingdom there’s basically no correlation between genome size and sophistication. The lungfish, for instance, has an enormous genome, some 133 billion letters long, 40 times bigger than the human genome. Meanwhile certain species of Takifugu puffer fish—(in)famous as an occasionally poisonous delicacy in Japan—have a wee genome, just 365 million letters. And neither fish is obviously more complex or well-adapted. Some creatures load up on junk, others slim down, but both can thrive.

In an unsettled area of science, it’s easy to whip between two extremes. Junk DNA is completely necessary! No, no, wait. It’s completely unnecessary! Take a breath: The answer almost certainly lies somewhere in between. But no one knows which end of the spectrum is closer to the truth. Will the epithet junk DNA prove more accurate than we realize, or will it go down as the greatest misnomer in science? Part of the answer lies within, and it will become clearer as we sequence more and more human beings. But we also need to keep sequencing sponges and worms and elephant sharks, to give us a wider perspective. Those organisms may seem a little obscure or dull, it’s true. But then again, so too did the humped bladderwort.