91av

WHEN life emerged from the primordial soup, what was it like? One way to find
out is to try to replay Genesis in the laboratory, which is what Harvard
geneticist Jack Szostak and his colleagues want to do.

They are attempting to produce likely candidates for Earth鈥檚 original
self-replicating molecules and ultimately hope to recreate the first cells. The
system by which modern cells divide is far too complex to have emerged directly
from the chemical stew of the early Earth. In this system DNA holds the genetic
blueprint, RNA carries this information to the protein-making factories, and the
proteins act as the cell鈥檚 workhorses.

One of these molecules, RNA, can do all three jobs, and some scientists
suspect it may have been the key to the evolution of life. RNA can store genetic
information and be copied to make new RNA. What鈥檚 more, specialised forms of RNA
called ribozymes, that can break chemical bonds and catalyse reactions, have
been discovered in the past decade, suggesting that RNA may once have been able
to copy itself. RNA, it seems, can do the work of both DNA and proteins.

But is RNA capable of performing all the reactions needed for a simple cell
to live? Compared with proteins, RNA has significant limitations as a catalytic
molecule. Instead of 20 building blocks, RNA has only four, so the number of
different possible RNA molecules of any given length is much smaller. Also, the
building blocks themselves are less varied than those of proteins. 鈥淭he only way
to really see what RNA can do is to make those molecules themselves,鈥� says
Szostak.

So Szostak鈥檚 team synthesises huge pools of random RNA sequences鈥攁round
10¹⁵ different combinations鈥攁nd extracts those best at doing a particular
job. They mutate these slightly and screen them again. Using this kind of
Darwinian selection, they have evolved RNAs that can bind to specific targets
and catalyse various reactions. Some of these reactions involve the kind of
chemistry needed to build proteins and copy RNA. It is at least plausible,
Szostak concludes, that ribozymes carried out the replication functions and
metabolic reactions of the first cells.

At some point in evolution, proteins joined in. Szostak has begun
experimenting with protein evolution in the test tube. This involves translating
a random pool of thousands of RNAs into proteins, and then selecting the 鈥渂est鈥�
protein from the mix鈥攖he one that propels a chemical reaction the fastest,
say. Since proteins do not retain the genetic information that gave rise to
them, Szostak and his colleague Richard Roberts engineered RNA molecules that
would stick to the protein they encoded after translation. That way, after the
most efficient protein was selected, its genetic code could be read directly off
its RNA tag, and the next round of mutation and selection could begin.

Szostak鈥檚 team and others are now thinking about the next step in the lab
simulation of primordial life: artificial cells. Although these are still years
away, Szostak says the essential components will include an RNA genome and
ribozymes capable of making material for cell membranes and producing the
ribonucleotide bases of which RNA is made. 鈥淭hey can鈥檛 really tell us what
actually happened,鈥� Szostak says, 鈥渂ut at least they can tell us what is
辫辞蝉蝉颈产濒别.鈥�