A lathe may be a critical machine shop tool for a manufacturing plants, but what tools go into an industrial biological machine shop? Biological tools are in common use in many manufacturing processes, especially nature's best chemists - enzymes that perform highly specific reactions better than even the most ingenious chemical synthesis reactions. Natural enzymes, however, are sensitive creatures, and they are often unable to withstand the harsh environment necessary for an efficient industrial process. To overcome this deficiency, biological engineers often resort to one of the most effective tools in the biologist's machine shop: evolution. A group of scientists at the University of Toulouse, writing in Protein Science, have used directed evolution to engineer a hardier polymer-making enzyme, demonstrating how we can use evolution to engineer biological tools for manufacturing processes. Amylose is a chain of sugar molecules found in starch. But it's not just useful for its food properties - amylose polymers are used to make films, paper and thickeners, among other things. But to be used efficiently in these industrial applications, amylose synthesis has to be tightly controlled - to have an efficient and economical industrial process, the raw materials have to be used at a high concentration, and high temperatures are needed to control the solubility of your materials and to prevent the growth of contaminating microbes. Temperatures as high as 50 degrees Celsius are often needed, but natural amylose synthesizing enzymes prefer a much cooler 30 degrees. The French research group set out to make an enzyme that could withstand the high heat of the industrial process. To do this they used an evolutionary strategy: they made tens of thousands of random mutations in this enzyme, and sought to pick out a better enzyme from this pool of mutants. The first step was to find enzymes that had merely survived the round of mutation - those that could still make amylose. Roughly 7,000 clones survived this easy step, and were next tested to see how well they survived at high temperatures. Out of the 7,000 mutants, only 3 were found to be effective at higher temperatures - but ultimately one better enzyme was all that the group needed, so although directed evolution burns through a lot of bad solutions, it is effective at finding good ones. The researchers had their improved enzyme, but curiosity took them further: could they look at the engineered enzyme and see why it was better? The changes were simple in all three heat-stable enzymes, just one or two amino acids had been altered. Looking at the structure of this enzyme, the mutations in these three enzymes were in different places - in other words, their directed evolution procedure did not come up with the same solution three separate times. There were some possible physical explanations for why these mutations made enzymes more stable, but not enough clues to make it obvious how to rationally design a heat-stable enzyme from scratch, unfortunately. Which means that directed evolution will still be one of the most valuable tools in the biological engineer's machine shop.