The diversity of nature may be enormous, but for Michael Hecht it is just a starting point.
Hecht, a Princeton professor of chemistry, has invented a technique
for making protein molecules from scratch, a long-sought advance that
will allow scientists to design the most basic building blocks of all
living things with a variety of shapes and compositions far greater
than those available in nature.
The technique, which Hecht developed over the last 10 years and
validated in experiments to be published in November, could prove
useful in a wide range of fields. Custom-designed proteins, for
example, could become a source of new drugs or could form the basis of
new materials that mimic the strength and resilience of natural
substances.
The range of proteins present in nature, while great, has evolved
only as far as the needs of biological organisms, said Hecht. "Why
should we be limited by a mere few million proteins?" he said. "We can
now not only ask what already exists in the biological world, but go
beyond that and ask what might be possible."
Hecht and colleagues from Princeton and Rutgers University reported
the advance in a paper to be published Nov. 11 in the Proceedings of
the National Academy of Sciences. Co-authors of the article are former
Princeton graduate student Yinan Wei and Rutgers chemistry professor
Jean Baum and her colleagues Seho Kim and David Fela.
Nearly all the internal workings of living things are built from
proteins. While genes are the "blueprints" for organisms, proteins are
the products built from those instructions. The molecules that transmit
signals in the brain, carry oxygen in the blood and turn genes on and
off are all proteins.
Scientists have long wanted to design their own proteins, but doing
so has proved a major challenge. Proteins are strings of chemical units
called amino acids and are often more than 100 amino acids long. When
cells make them, these long chains fold spontaneously into complex
three-dimensional shapes that fit like puzzle pieces with other
molecules and give proteins their unique abilities. There are 20
different amino acids, so the number of possible combinations is
enormous. However, the vast majority of these combinations are useless
because they cannot fold into protein-like structures.
The advance reported by Hecht and colleagues involves a simple
system for designing amino acid sequences that fold like natural
proteins. First publishing the idea in 1993, Hecht realized that some
amino acids were strongly "water-loving" while others were
"oil-loving." The two types naturally separate from each other, with
the oil-loving ones clustering in the protein core and water-loving
ones forming the perimeter. He also saw that natural proteins with good
structures tend to have certain repeating patterns of oil-loving and
water-loving amino acids. For example, taking a string of water-loving
units — no matter which ones — and inserting any oil-loving unit
every three or four positions typically creates proteins that fold into
bundles of helices.
Hecht used this method to create a "library" of genes encoding
millions of novel proteins, each designed to fold into a bundle of four
helices. Initial tests of the library in the early 1990s showed that
most of the proteins folded into compact arrangements, but these were
"mushy," fluctuating shapes instead of well-ordered, rigid structures.
Hecht suspected that these proteins were simply too short to achieve
a good structure and added amino acids to each sequence, making them 40
percent longer. In their latest findings, the researchers found that
this new library contained well-folded proteins. They subjected one to
a painstaking test — a type of MRI for molecules — and verified its
three-dimensional structure. The experimentally determined structure
closely matches that expected from the design.
The results are "quite important work," according to Jane
Richardson, a biochemist at Duke University, and are a "direct
demonstration of the importance of one simple and central factor in
protein folding, which has not in the past been much emphasized in
either the design or the folding fields."
Previously, the only ways for scientists to invent new proteins have
been to churn out random sequences and screen them for well-folded
proteins or to calculate, atom-by-atom, combinations that will fold
into a desired shape. The first is difficult because there are too many
combinations to try them all, said Hecht. Making every possible
sequence of 100 amino acids would require more than all the atoms in
universe, he said. The second, the calculation method, yields only one
protein at a time.
Hecht’s method offers a middle ground because it limits the number
of possible sequences to those that fit the correct
oil-loving/water-loving patterns.
Having a rich variety of custom proteins may allow scientists to
consider using them for tasks that do not exist in nature, such as
catalyzing industrial chemical reactions, Hecht said. "Critters in
nature haven’t been challenged to solve the technological problems
we’re faced with today," he said. "If we are limited by what nature has
given us, we are not going to tackle those problems."
Princeton University. October 2003.