Isolation, screening, partial purification and characterization of halophilic protease from different samples

In this research the protease producing bacteria were screened from Indonesian traditional fermented food and saline soil sample from Indonesia. The purpose of this research was to partially purify and characterize microbial protease from halophilic bacteria. During the study, 4 halophilic protease producers were isolated from tauco and terasi. Among these isolates, halophilic bacteria isolate TANN 4 was recorded as the best protease producer. The isolate TANN 4 was grown in 18 % MGM (Modified Growth Medium) agar and broth containing 1% skim milk. Extracellular protease from isolate TANN 4 was partially purified using amonium sulfate precipitation and dialysis. The protease was partially purified with final yield of 72.87 % and also with 25.41 fold purity. This moderate thermoactive and alkaliphilic protease showed a pH optimum of 8.0 and temperature optimum was 50 °C. The enzyme also was active at salt concentrations ranging from 1 to 15 % (w/v), with optimum activity at 1 % NaCl (w/v). Ethylenediaminetetraacetic acid (EDTA) completely inhibited the enzyme activity suggesting that it was a metalloprotease. Among metal ions, the Ca2+, K+ and Mg2+ ions enhanced the activity of enzyme. The KM and Vmax values exhibited by partially purified protease were 0.0649 mM and 216.45 U mg−1 using casein as substrate. The molecular weight was estimated to be 19.8 kDa on SDS PAGE. The enzyme also fairly stable in Triton X-100 (1 and 5 %), SDS (0.1 and 0.5 %), 1 % commercial detergents (OMO and Ariel) and 25 % methanol. In addition, this enzyme was capable of hydrolyzing casein, hemoglobin and bovine serum albumin (BSA). Automated ribotyping analysis revealed that 3 isolate (TANN 4, TR 2 and TR 4) resembled Halobacillus trueperi that exhibited 71, 68 and 69 % similarity respectively, and isolate (TR 1) resembled Virgibacillus pantothenticus with 64 % similarity. To improve this study, further protein purification, bacterial identification and media optimization are needed.