Abstract During the degrading of diet protein, protease is the first key enzyme which hydrolyses protein to peptide or amino acid. However there is few genetic information of protease from rumen microbiota due to the limitation of pure-cultured method. The objective of the experiment was to screen, sequence and bioinformatics analyze the protease clone from a Fosmid library of the dairy cow rumen microbiota. Using two different protease selective medium, including 1% skim milk or 1% isolated soybean protein, respectively, fourteen protease activity clones were obtained from rumen microbiota Fosmid library containing 30 000 clones. Different protease decomposition ability of the fourteen clones was measured by Folin-phenol reagent method. The results showed that each clone had its unique ability of protease decomposition. When casein was used as the substrate, the range of enzyme activity was from 0.59 to 2.74 U/mg. While the isolated soybean protein served as substrate, the range of enzyme activity was from 0.70 to 7.19 U/mg. Furthermore, the same clone had different sizes of enzyme activity for different substrates. After end sequences of ten positive cloning(GenBank Accession numbers: JY084410~JY084429), the sequences were blasted by Blastn and Blastx. The results showed that 45% of the genetic sequences could not match with the known genes encoding. The end sequence of pro10F could match to metal peptidase, with the similarity of 54%, which belonged to peptidase family M13.The optimal pH of the clone was 7.0.These results provide the basicdatum for the further study on the clone's enzymatic properties and gene characteristics
