Martin Barfoed

BIOCHEMICAL PROPERTIES OF CLONED LIPASES FROM THE PSEUDOMONAS FAMILY

Three Pseudomonas lipases, representing three subfamilies, were analysed for pH optima, destabilization by EGTA and surfactants, phospholipase and cholesterolesterase side activities. All the Pseudomonas lipases tested showed alkaline pH optima. The Pseudomonas cepacia and the P. pseudoalcaligenes lipases were totally inhibited by EGTA at pH 9, and the latter was also fully inhibited at pH 7. The lipase from P. mendocina was not inhibited by EGTA at any of the pH values tested. These findings indicate that a calcium binding site exists in some of the Pseudomonas lipases. The P. pseudoalcaligenes, P. cepacia and P. mendocina lipases were inhibited by the anionic surfactant SDS at concentrations between 0.01-0.5 mg/ml. The P. pseudoalcaligenes and P. cepacia lipases were not inhibited by the nonionic surfactant Brij35 in concentration up to 1 mg/ml, whereas the lipase from P. mendocina was inhibited at 0.1 mg/ml. The P. pseudoalcaligenes and P. cepacia lipases were found to possess high cholesterol esterase activity. P. pseudoalcaligenes lipase was further found to have high phospholipase activity. Ten Pseudomonas lipase sequences were compared by automatic sequence alignment. On the basis of sequence identity we have classified Pseudomonas lipases into five subfamilies.

Application of a neural network in the optimization of an enzymatic synthesis

Abstract A neural network has been trained to predict the outcome of a synthesis controlled by several parameters. Furthermore, this ability was used in the search for optimized reaction conditions. Finally, an excellent correlation between predicted yields and experimental results was obtained.