When NASA scientist Andrew Pohorille first began to research the origin of life, he thought the topic seemed more a question for philosophers than scientists. "We called this type of research project a ‘Café Discipline,’ which is something you talk about in a café rather than do in a laboratory," says the Polish-born Pohorille. Charles Darwin himself, when asked about the origin of life, scoffed, "You may as well question the origin of matter."

Today, scientists like Pohorille have come a long way in understanding how a random collection of atoms and molecules could have slowly, over billions of years, combined into the complex living systems that we know today. They’ve pieced together much of the puzzle. They know how the first cell membrane formed. They know how the simplest living creature, the bacterium, came into being. They’ve traced the evolution of bacteria, plus or minus a few gaps in the fossil record, to the formation of higher animals and eventually to humans.

But one piece of the puzzle is missing: What was the first living molecule? Before bacteria, before cellular components, what simple living structure emerged from the so-called primordial soup?

Scientists have few clues to go on. Fossil records from this era are incomplete, and researchers are still debating whether the stuff of life zoomed to Earth on meteorites or bubbled up from deep-sea thermal vents. Wherever life came from, scientists believe it appeared on Earth around 3.8 billion years ago. They think the key ingredients were nonliving molecules like carbon dioxide and methane. Early attempts to re-create life by combining these ingredients in a chemistry beaker, however, failed miserably. Life has a highly ordered structure. It is more like a lasagna, in which ingredients are combined and layered in a particular order, than like soup, in which ingredients are simply tossed into a pot.

Instead, scientists have to work backwards, like chefs who have tasted a delicious dish and are trying to re-create the recipe. Since they can’t travel back in time, scientists are trying to create the first living molecule in chemistry labs and on computers.

To be considered "living," this primordial molecule must fulfill certain requirements. It must have been able to be reproduce. It must have had some sort of purpose, or carried some key bit of information in its structure. It must have been capable of evolving into more complex structures. And finally, it must have been able to kick-start other chemical reactions around it, acting as a catalyst to spur reactions. Without these boosters, known as enzymes, the chemical reactions of life would take eons, and life as we know it would not exist.

Over the years, researchers have narrowed the search for the first living molecule down to two main suspects: gene-like material and protein-like material. Some scientists think RNA, a type of genetic material, kicked off the development of all other living entities. Others, like NASA’s Andrew Pohorille, think amino acids, the building blocks of proteins, did the job. The answer might lie somewhere in between.

The RNA World

Sound evidence points to the theory that RNA is the first living molecule. RNA is good at copying itself and it carries a lot of information in its structure. In today’s cell, RNA acts as a sort of office messenger. If DNA is the master architect, RNA is like an assistant who takes a blueprint from DNA and plugs it into cellular machines. (These machines, called ribosomes, are actually made up of RNA as well.) These ribosomes read the blueprint like a set of instructions that tell which amino acids to hook together.

Billions of years ago, however, RNA may have been the main molecule in charge, acting as the architect, messenger, and cellular machine. Not only is RNA a good copier, it can play the part of a dynamic manager that convinces employees to work faster and more efficiently. In other words, RNA can act as an enzyme to galvanize other chemical reactions within the cell. For the discovery of RNA enzymes, or ribozymes, Thomas Cech’s team at the University of Colorado and Sidney Altman’s group at Yale University were awarded the 1989 Nobel Prize in chemistry.

Since RNA is good at reproducing itself, carrying information, evolving, and catalyzing, it made sense to Cech and others that it could be the first living entity. If RNA was first, they reasoned, it should be possible to build an RNA molecule out of the ingredients lying about in the primordial kitchen. But here chemists ran into problems.

When chemists tried to brew RNA in their chemistry beakers, they succeeded only in creating a tangled mess. Researchers began to realize that rather than being the very first living entity, the "RNA world" was already pretty advanced. There had to be a predecessor.

The Amino Acid World

The amino acid is the other strong contender for the job of first living molecule. Amino acids are very simple in structure, and very necessary for life. They are the basic units of proteins, which do all sorts of useful work inside the cell. One thing proteins do particularly well is act as enzymes, catalyzing other chemical reactions.

As a candidate for the first life form, amino acids have one thing going for them: they are easy to make with materials known to have existed on early Earth. The first team to do it was Stanley Miller and Harold Urey at the University of Chicago in 1953.

But if making amino acids was easy, making proteins proved almost impossible. To make a protein, the amino acids must first be hooked onto each other in a very specific order in a long chain called a peptide. Then these peptides must be folded very precisely into some very odd-looking shapes, and only then do we call them proteins. In today’s cell, the cellular machine called the ribosome does the hooking. But in the early days of life, ribosomes didn’t yet exist.

Simply heating up the amino acids doesn’t make them hook together properly--it is tricky to get the bond to form just right. Instead of forming peptide chains, the amino acids just melt together. But in the late 1980s, Bernd Rode, of the University of Innsbruck in Austria, found a solution to this problem. He mixed up some amino acids, salt, and copper atoms in a solution. When he evaporated the liquid, he found that bridges formed between the amino acids. Then he rinsed the molecules with fresh water to wash away the copper, leaving the amino acids linked together. Even more exciting, this cycle of dry periods followed by watery flushes is similar to what happened on early Earth, when dry spells alternated with rainy periods.

With a plausible mechanism for joining amino acids together, scientists began to explore the idea that these molecules could be the first living biological structures.

Andrew Pohorille is one of these scientists. As director of the Center for Computational Astrobiology at NASA’s Ames Research Center in Mountain View, California, Pohorille is using a computer to model how the hodgepodge of Earth’s early molecules might have assembled into peptides that can catalyze reactions. Pohorille and his colleague Michael New designed a computer model that reenacts the early steps in life. On the computer, they create imaginary peptide chains of various lengths. By running the computer program, they simulate how peptide chains randomly encounter each other and form longer, more complex strands.

In Pohorille’s model, the short peptides interact with each other for a prescribed amount of time, hooking up and breaking up. Once in a while, purely by random meetings, the peptides form a catalyst, able to incite nearby chemical reactions.

Pohorille has found that over time his peptides evolve into longer chains that become better and better at their task of catalyzing the formation of other nearby peptides. Since the process is random, however, there is always a chance that instead of forming a matchmaker, the system will produce the opposite: a catalyst that breaks up peptides. But this is not all bad, says Pohorille. He thinks these peptide-breakers could break up dysfunctional partnerships, thus freeing them to find better-suited partners.

Although some critics are skeptical of Pohorille’s work because they think the chances are slim that peptides can spontaneously form catalysts, others think he is on the right track.

Jean Chmielewski, a protein chemist at Purdue University in Lafayette, Indiana, has created peptides that can replicate in test tubes. She and post-doctoral researcher Shao Yao built a 35-amino-acid-long chain capable of copying itself, a task peptides normally cannot do.

When conditions are just right, the kinked chain acts as a sticky template to which free-floating amino acids will gravitate. Once these newcomers are lined up opposite their identical mates, they can bond to each other and form a new chain. The new chain can then separate and float away, resulting in an identical copy of the original peptide chain.

Chmielewski and Yao can control this replication by changing the salt concentration in the mixture. Adding salt causes the amino acids to alter their shape slightly, making the beaded chain more compact and easier to copy. "We can use these conditions as an on/off switch for peptide self-replication," says Chmielewski.

Chmielewski’s system is promising. It could indicate how peptides were the first living entities. But there is a hitch. Her system does not necessarily mimic what happened during the early days of the Earth. She artificially rigged the peptides to favor self-replication. But Chmielewski says her system provides a useful model of how peptide replication could have worked. "Sure, we’ve stacked the deck, but that doesn’t mean that self-replication couldn’t happen on its own," says Chmielewski.

Another scientist has built a more lifelike system. Reza Ghadiri, a chemist at the Scripps Research Institute in La Jolla, California, chemically tweaked an existing peptide chain and coaxed it to replicate.

Not content with just showing that self-replicating peptides can exist, Ghadiri is now trying to prove that they can evolve into more complex forms. In other words, he is trying to prove that what Pohorille saw in his computer model can exist in real life. Ghadiri is building entire networks of peptides in a test tube. He is creating what he calls a "molecular ecosystem," where various "species," consisting of peptides of different lengths, roam around in a big tube, interacting with each other and "evolving" into larger, more powerful catalysts. "We have shown that minimal systems are capable of doing a lot," says Ghadiri. "Now we want to create more complex systems with capabilities that are far greater than the sum of the molecular parts."

To look at how peptides could have evolved, Tony Keefe and Jack Szostak at Harvard Medical School have fashioned a peculiar system. Keefe took a newly synthesized peptide (made the modern way, using a ribosome) and fused it on one end to a corresponding RNA tag. Since RNA is very good at copying itself, Keefe can copy the RNA-peptide as much as he wants. He uses the peptide part of the molecule to study how peptides might have evolved from simple amino-acid strands into complex, powerful proteins.

First, Keefe creates a number of random RNA-peptides. Then he tests them for catalytic function. If he finds any, he copies only the catalyzing peptides. With each copying step, random mutations spring up that can improve catalytic function. After each round of copying, Keefe tests the batch to see whether the catalytic function has improved. In a sense Keefe is reproducing evolution in a test tube. "It is an artificial system, so it doesn’t really tell us about how life evolved," says Keefe. "But the questions we can answer are relevant to the origins of life."

The work of Pohorille, Chmielewski, Ghadiri, and Keefe shows that peptides can replicate and they can even evolve into more complex peptides. But, as with the RNA hypothesis, not all is smooth sailing for the amino acids.

Problems with Peptides

When peptides copy themselves, they are sloppy about it. They don’t copy exactly. The newly minted peptides are often incomplete, missing one or more amino acids, or containing a substituted amino acid. Compared to DNA or RNA, peptide replication is quite shoddy work.

But Pohorille thinks sloppy copying could be an advantage. Maybe careless copying is needed to kick-start evolution. It’s like having a machine that turns out thousands of error-ridden copies, one of which might yield a useful function. "Once you find something good, you want to remember it," says Pohorille. "And that is where some sort of RNA genome has to come in. But in the very beginning you want sloppiness so you can explore more possibilities."

For this idea to work, however, living systems would have had to switch from sloppy peptide copying to precise genomic copying. And herein lies the problem: No one yet knows how that could have occurred. Andrew Ellington, a biochemist at the University of Texas at Austin, is skeptical that such a progression is even possible. "If there is so little fidelity in copying from peptide to peptide, how did we suddenly gain fidelity going from peptides to nucleic acids?" he asks. To Ellington, this information gap is proof that a peptide could not possibly have been the first molecule.

Another knock against the amino acid theory is that self-replicating peptides do not carry as much information as RNA. Thomas Cech, the Nobelist who first explored the RNA world, says self-catalyzing peptides do not fit the criteria for a living molecule. "The simplest definition of life is replication (reproduction) plus mutation," says Cech. "Self-assembling peptides could have preceded the RNA world, but since they are not an information-rich molecule with significant coding capacity, capable of being reproduced, then this does not constitute the ‘origin of life,’ but rather just a plausible chemical event," says Cech.

Another stumbling block is a concept called chirality. Some molecules are perfectly symmetrical; that is, if one cuts a line down the center of the molecule, the left side is a mirror image of the right side. Amino acids are not symmetrical, but chiral. Just as humans possess left and right hands that are equal in shape but face in opposite directions, so molecules are said to be either right-handed or left-handed.

It turns out all naturally occurring amino acids are left-handed, says Jeffrey Bada, a geochemist at the University of California, San Diego. "We cannot imagine how evolution could separate out left- and right-handed molecules," says Bada. In other words, scientists can find no evolutionary mechanism by which left-handed amino acids would be selected over right-handed ones. For Bada, these facts point to the conclusion that a symmetric molecule developed first, and eventually twisted itself into a left-handed configuration.

Amino acids and their resulting peptide chains don’t carry enough information, and on top of that they have this chirality problem. In fact, RNA has the same chirality problem. So what is the answer?

Scientists are starting to consider other options. Leslie Orgel, a chemist at the Salk Institute for Biological Studies in La Jolla, thinks the possibilities are vast. "Presumably if there were catalytic peptides in the early days, they could have been completely different from modern proteins," says Orgel, "You have to ask what amino acids might have been available, and all the different ways they could have joined together."

Hybrid Molecules

The answer could be a combination molecule, a hybrid of peptides and RNA that existed in the early years of Earth’s formation but eventually died out once it lost its usefulness. One candidate is a peptide-nucleic acid hybrid, known as PNA. This molecule has a peptide backbone with RNA-like units sticking out of it. Bada and others, including amino-acid pioneer Stanley Miller, now at the University of California, San Diego, believe PNA was the first living entity. "The beauty of PNA is that its backbone is easy to make under early Earth conditions," says Bada.

Recently, Miller showed that PNA can arise spontaneously from the ingredients available on early Earth. Unlike RNA, PNA is easy to make. Unlike amino acids, PNA carries lots of information. And unlike both of them, PNA is symmetrical so it doesn’t have the chirality problem.

Which molecule is indeed the first living entity, whether RNA, amino acids, or a hybrid molecule such as PNA, is still a hot subject of debate. Using chemistry, computers, and innovative hybrid molecules, scientists are piecing together this mystery. Despite their differences over which ingredient came first, they all agree on one thing: Life arose from chemical and biological principles that can be understood, meaning that someday, we may develop the capacity to create life anew in the laboratory.

Pohorille hopes someday to take his work on the origin of life one step further, to show that life can exist elsewhere in the universe. He says such life will probably not look that different from our own, at least on the biochemical level. "There may be differences in the molecules used in the genetic code," says Pohorille, "but there will be an underlying similarity that will be recognizable." To Pohorille, the hunt for the origin of life is no longer just a "Cafe Discipline," but a serious line of scientific inquiry that can be followed and--eventually--resolved.





WRITER Catherine Zandonella
B.S. University of California at Santa Barbara (Pharmacology); M.P.H. University of California,
Berkeley (Environmental Health).
Internships: New Scientist Magazine; Santa Cruz County Sentinel; Stanford University News Service;
Newsday (Summer 2000).
ILLUSTRATOR Kimberlee Heldt
B.A, University of California, Berkeley; M.S., University of Hawaii. Manoa; G.C .Sci. Ill, University of California, Santa Cruz
Internship: Harcourt

Text © 2000 Catherine Zandonella
Illustrations © 2000 Kimberlee Heldt