Underneath life’s variety, from mites to mastodons, there’s a profound sameness. Every cell of every organism, with the exception of a few viruses, encodes genetic information in DNA; DNA dispatches a single strand of RNA to make proteins; and the proteins do all the cellular grunt work. This system, so universal and uniform, poses a puzzle: where did it all come from? Complex as it is, it is hard to see how it could have sprung full-blown from the primordial soup. Might the first organisms instead have used some kind of pre-NA?
Some fascinating hints are now emerging, many of them from the lab of chemist Stanley Miller at the University of California at San Diego. Miller is famous for a 1953 experiment in which he mixed up atmospheric gases and water in a flask and put a spark to it, thus simulating lightning ripping through primeval skies. The energy caused the molecules to combine into many simple organic compounds. Researchers believed that these building blocks might have formed into the first genes.
Despite Miller’s work and other work that flowed from it, a messy paradox remained: the division of labor in the cell was too clean. DNA and RNA were only information carriers, and proteins could only do chemical chores. Each needed the other to exist, and so, like chicken and egg, neither could come first. But in the early 1980s, researchers discovered that RNA could catalyze some chemical reactions--acting like a protein as well as an information carrier. That immediately made it a strong candidate for the first biomolecule. The picture of the primordial soup as an RNA world has now become conventional wisdom; only later, in this view, did organisms evolve DNA. Less fragile than RNA and thus more secure as a storehouse of genetic information, DNA ultimately grabbed control of life, demoting RNA to the rank of errand boy.
But could RNA itself have formed spontaneously and abundantly in the primordial soup? From a genetic point of view the key components of both RNA and DNA are the four bases that make up the genetic alphabet. In the 1960s researchers were able to synthesize two of RNA’s bases--adenine and guanine--from precursor molecules that are likely to have been present on the early Earth. This spring Miller and one of his graduate students, Michael Robertson, finally finished the job: they discovered reactions that could efficiently produce copious amounts of the other two bases, cytosine and uracil. The secret, they found, was to load up their flasks with urea, a carbon-nitrogen compound likely to have formed in the early ocean. Then it works like a charm, says Miller. If you had urea in the ocean and seawater evaporated in a lagoon, you’d get a very high concentration.
That lagoon, then, might have given rise to all four RNA bases. Case closed? Far from it. DNA and RNA require not only bases but a backbone on which to hang them. The backbone is made of two repeating building blocks. Every base is attached to a sugar molecule called ribose (or deoxyribose, in the case of DNA). The sugars alternate with phosphate molecules, each of which consists of a phosphorus atom surrounded by four oxygen atoms; two of those oxygens bind to sugar carbons on either side, thus linking two sugars together.
The backbone, as recent experiments have helped show, is the Achilles’ heel of RNA. Rosa Larralde, another student of Miller’s (now at Harvard), took a close look at ribose. On the early Earth, ribose probably formed out of formaldehyde, which is a single carbon atom bonded to a water molecule; it has been found not only in Miller’s spark chambers but also in comets, which in Earth’s youth were constantly pelting it. Chains of formaldehyde molecules spontaneously curl into rings, forming various sugars. A fraction of 1 percent of those rings are ribose.
That’s not an encouraging percentage, but Larralde has discovered something far more disheartening. She put ribose in a variety of solutions designed to span the gamut of possible primordial soups, heated them up, and measured how much ribose remained as time passed. At the boiling point, ribose had a half-life of only 73 minutes. Larralde calculates that its half-life in a room-temperature primordial soup would have been less than a year--which wouldn’t have allowed it much time to bind with bases and phosphates and form RNA.
Even with plenty of time, ribose would have trouble hooking up with phosphate. Organisms today have enzymes to grab phosphate out of their environment and link it with sugars, but those enzymes would not have been available in the primordial soup--and phosphate on its own, it turns out, is not very reactive. Tony Keefe, a postdoc with Miller, tried dozens of experiments and couldn’t make an RNA backbone with phosphate. Even if such a backbone could have formed on the early Earth, Miller argues, it wouldn’t have lasted long enough to begin acting like life: ribose and phosphate are linked by the same weak carbon-oxygen bond that makes ribose itself so fragile. Everybody who’s looking at the phosphate-ribose backbone is just barking up the wrong tree, Miller concludes.
There’s hope, though, in the possibility that life began with a different backbone altogether. In 1991 Danish researchers led by Peter Nielsen of the Panum Institute in Copenhagen created an artificial molecule that could bind to a specified sequence of DNA bases. Instead of using ribose and phosphate, they created a backbone of carbon and nitrogen arranged in repeating subunits, with one of the four DNA bases dangling off each subunit. The carbon-nitrogen bond that holds this backbone together-- it’s called a peptide bond, and it’s the same one found in all proteins--is much stronger than the ribose-phosphate bond. And it wasn’t just the backbone of Nielsen’s molecule that turned out to be stunningly strong: when two strands of it were woven into a double helix, they withstood heating and clasped each other tightly long after a double helix of DNA had fallen apart. Nielsen calls his molecule PNA (the p is for peptide).
No one has yet shown that PNA could have formed spontaneously in the primordial ocean. Nevertheless, Nielsen believes that the sturdiness of its backbone recommends it (or a similar sugarless polymer) as a plausible first molecule of life. So far, though, PNA has shown no signs of having RNA’s catalytic ability. Thus PNA organisms that managed to create RNA would have had an evolutionary advantage. Nielsen has recently found that a string of bases on a PNA backbone can indeed guide the formation of a single complementary strand of RNA, much as DNA does.
In the long run, life had a strong incentive to switch over completely to RNA and DNA; the PNA double helix is so hard to pry apart that it’s not very good at surrendering its genetic information. Taken together, then, all this recent work suggests there may have been three distinct ages of life: the age of PNA, when the first biomolecules hauled themselves out of the soup; the age of RNA, a sort of transitional period; and finally the age of DNA, when life as we know it began to carpet the planet. We haven’t evolved out of that age yet.