The protein folding revolution

The protein folding revolution


Proteins carry out the labor in our cells
— so we really need to know what they do and how they work. The key to proteins is their shape because
that dictates their function. They can take many different shapes. Scientists call these folds. Unfolded, a protein is a long string of amino acids. There are 20 different amino acids and each
one of them has its own chemical behaviours. When a protein folds up, you get a long tangled
piece of spaghetti with all these different chemical functionalities on it It’s not exactly like spaghetti – because
its 3D shape evolved over billions of years to do very specific jobs.
If you can understand the minute details of the structure of proteins– not only do you
get insights into their function — you might be able to change that function. So researchers have been trying for many years
to solve the protein folding problem: Can we just look at the sequence of amino acids
and predict how a protein is going to fold? You could take the amino acid sequence, plug
it into a computer, and see if your algorithms are good enough to make sense of how it might fold. You can use X-ray crystallography or
other techniques to image a protein structure but that hasn’t been done for very many kinds of proteins. A couple of decades ago folks asked a separate
question –could all the genome sequence data- the three billion letters in our genome and
the billions in all the other genomes out there – could they scan that code, which is
separate from the amino-acid code of proteins, and learn anything about how proteins might fold? The DNA in our genes codes for RNA, which
is translated into proteins. So there’s a relationship between the 4-letter DNA code
and the 20- letter amino acid sequences of proteins. Because a protein wraps around in many different
twists and turns–the 6th amino acid in that chain might end up next to the 18th amino acid. If they end up next to each other, researchers
realized there might be an interaction between the pair that is critical to the shape of
the protein and therefore its function. If that’s true then a mutation in the DNA
that changes one of the amino acids must be accompanied by another mutation to the other
member of the pair to preserve the interaction. In essence, they co-evolve. Well, if you can log maybe a hundred or more
of those cases of close-by neighbors in 3D space, based on looking at many genome sequences, then you plug that into your folding program.
Now that it has all these tight constraints it gives a much better chance of getting a
really accurate structure. And it works! The upshot is scientists can fold lots of
proteins that they never could before. That’s important because it will give new
insights into how those proteins work. Beyond that-researchers have been steadily
improving the ability of computers to model the shape of proteins, and this now enables
them to design their own proteins—making things never seen before in nature. The most obvious application is medicine. They can target very specific parts of the
flu virus with a special built protein– enabling a vaccine that works across flu strains They’ve designed proteins that naturally assemble
into tiny cages that can deliver different molecules in the body Or, new materials,like engineered surfaces
that self assemble could be used in solar cells and electronic devices. you can go in a thousand different directions.