Making molecules that copy themselves.

Making molecules that copy themselves

Imagine building a machine whose physical structure provides instructions for its own construction and whose parts do double duty as the construction tools. Such a machine might even manage to make a copy of itself. Chemists at MIT have now assembled a rudimentary molecular machine of this sort, and they say it could serve as a model for probing the origins of the self-replicating biochemical systems inside cells.

"At best this can be regarded as a primitive sign of life," suggest Julius Rebek Jr. and his coauthors in the Jan. 31 JOURNAL OF THE AMERICAN CHEMICAL SOCIETY. At the very least, they say, their self-copying molecule should help scientists address a profound question of biochemistry: How did nucleic acids such as DNA and RNA come to embody a blueprint for making proteins -- the stuff of hair, muscles and key biological molecules such as enzymes?

The researchers assembled one of the system's two reactants, pentafluorophenyl ester, by covalently bonding it to one arm of a U-shaped molecular frame (Fig. 1 at right). With weaker hydrogen bonds, the second reactant -- called amino adenosine (Fig. 2) -- temporarily sticks to the frame's other arm. Amino adenosine closely resembles adenosine, a building block of nucleic acids. The specific patterns of hydrogen bonds that form between nucleic acid components endow the nucleic acids with their ability to carry and duplicate genetic information.

Thus positioned on the U-shaped frame, the reactants readily link to form an amide bond, the same kind of bond that links amino acids into proteins. The newly formed molecule, at first joined to both arms of the frame, jackknifes upward as the amino adenosine severs its hydrogen bonds with the frame and snaps in place over the ester. This yields an upright assembly (Fig. 3) whose chemical features form a template that attracts, positions and helps bond another pair of reactants (Fig. 4).

"My work shows how you can take information that exists in base pairs [as in the hydrogen-bonded pairs in DNA] and use it to drive chemical reactions that make amide and peptide bonds [as in proteins]," Rebek says. Still, he cautions, the system's limitations make it a stepping stone on the path to more sophisticated self-replicating reactions rather than an end unto itself. For instance, once the second pair of reactants links up, the resulting molecule tends to stay on the frame instead of vacating it so that new reactants can come in.

Chemist Jonathan Sessler of the University of Texas in Austin ranks the ongoing quest to duplicate the protein-making and information-carrying feats of biochemical systems as one of the grandest in science. "To succeed would represent a landmark achievement in the history of chemistry," he says. Although two other groups in the mid-1980s described laboratory-created replicating systems based on small nucleic acid molecules, Sessler says Rebek's work takes a step beyond them by using molecules with properties of both nucleic acids and proteins.