2007 WINNER

By Ed Yong
Cancer Research UK

Winner of the 20-28 category

Grammar, a weapon against disease

Grammar is not just the province of pedants and proof-readers, it's our newest weapon against lethal diseases like MRSA and anthrax. The grammatical rules that govern our languages have now inspired scientists to create a new powerful line of synthetic proteins that could give us an edge over deadly drug-resistant bacteria.

At first glance, languages and proteins could not be more different. The former is the stuff of human minds, and the latter are responsible for building those minds in the first place. But, on closer inspection, they share uncanny similarities. Genes and proteins have their own parallels of letters, words, sentences and even grammar.

Proteins are built from chains of molecules called amino acids, in the same way that sentences are chains of words. But random strings of words are pointless. To form sentences and carry meaning, they must be united in specific ways defined by a set of rules - a grammar. A protein's amino acids must be connected using similar "grammatical" rules. Their order determines

the protein's shape and function, and haphazard combinations would result in molecular gibberish.

For years, popular science writers have used this linguistic metaphor to help people appreciate and understand the genetic world. But Christopher Loose and colleagues from the Massachusetts Institute of Technology, near Boston, have used it in a more practical way - to design new medicines.

Loose focused on a group of small proteins called anti-microbial peptides or AmPs. AmPs are normally deployed by the immune system to destroy bacteria and other infectious invaders and they have tremendous potential as cures for bacterial diseases. Unlike conventional antibiotics such as penicillin, they don't seem to trigger resistance in their targets, and they conveniently give the rest of the host's immune system a boost.

As a group, AmPs carry out similar functions and individual members have many amino acid sequences in common. For example, nine in every 10 insect AmPs, known as cecropins, contain the sequence QxEAGxLxKxxK, where each capital letter represents a specific amino acid and each small x can be filled by any of them. Loose likened these shared sequences to common phrases in a language.

He identified over 700 such "phrases" that commonly occur in over 500 well-known AmPs, Each one describes a sequence of 10 amino acids, and together these rules make up the grammar of an "AmP language". In a spark of creativity, Loose realised that he could use this language to design all-new AmPs that might outperform their natural counterparts. He designed a set of synthetic "gramatically-correct" AmPs, each just 20 amino acids long. In each of these designs, every set of 10 consecutive amino acids matched at least one of the 700 phrases. He excluded any designs that were too similar to a naturally occurring AmP, leaving behind 40 promising candidates. These were synthesised in the lab and tested against living bacteria.

Amazingly, they worked. Eighteen of the synthetic AmPs blocked the growth of common species of bacteria and Loose picked the two best ones, snappily named D28 and D51, for further development. Both successfully killed a bacterium called Bacillus cereus, at the same sorts of concentrations as natural AmPs.

Bacillus cereus causes mild food poisoning, but belongs to a bacterial group which includes Bacillus anthracis, the bug that causes anthrax, and Staphylococcus aureus, the basis of the drug-resistant MRSA strain. Low concentrations of D28 and D51 stopped both of bugs from growing.

Deploying naturally occurring AmPs against bacteria is risky; if the bugs then developed resistance, it could put public health in serious jeopardy. But synthetic proteins only bear passing resemblance to our natural arsenal. If their targets developed resistance, it wouldn't compromise our own defences.

The study proves that language-based models can help scientists to design new medicines. This method could give us a multitude of new weapons against bacterial infections, greatly reducing the chances of resistance. With an enemy that evolves so quickly, this is an invaluable advantage.

Ed Yong
Ed Yong works at Cancer Research UK, London and has a blog at http://notexactlyrocketscience.wordpress.com