EVALUATION OF CELLULOLYTIC ENZYMES FROM A NEWLY ISOLATED BREVIBACILLUS SP. JXL; AND OPTIMIZATION OF COSLIF PRETREATMENT VARIABLES OF SWEET SORGHUM BAGASSE USING A RESPONSE SURFACE METHOD

Author

Jemil N. Yesuf, Southern Illinois University CarbondaleFollow

Date of Award

5-1-2012

Degree Name

Doctor of Philosophy

Department

Engineering Science

First Advisor

Liang, Yanna

Abstract

The first part of the dissertation presented a potentially novel aerobic, thermophilic, and cellulolytic bacterium identified as Brevibacillus sp. Strain JXL which was isolated from swine waste. Strain JXL can utilize a broad range of carbohydrates including: cellulose, carboxymethylcellulose (CMC), xylan, cellobiose, glucose, and xylose. In two different media supplemented with crystalline cellulose and CMC at 57°C under aeration, strain JXL produced a basal level of cellulases as FPU of 0.02 IU/ml in the crude culture supernatant. When glucose or cellobiose was used besides cellulose, cellulase activities were enhanced ten times during the first 24 h, but with no significant difference between the effects caused by these two simple sugars. After the end of the 24 hour period, however, culture with glucose demonstrated higher cellulase activities compared with that from cellobiose. Similar trend and effect on cellulase activities were also observed when glucose or cellobiose served as a single substrate. The optimal doses of cellobiose and glucose for cellulase induction were 0.5 and 1%. These inducing effects were further confirmed by scanning electron microscopy (SEM) images, which indicated the presence of extracellular protuberant structures. These cellulosome-resembling structures were most abundant in culture with glucose, followed by cellobiose and without sugar addition. With respect to cellulase activity assay, crude cellulases had an optimal temperature of 50°C and optimal pH range of 6-8. These cellulases also had high thermotolerance as demonstrated by retaining more than 50% activity after 1 h at 100°C. In summary, this is the first study to show that the genus Brevibacillus may have strains that can degrade cellulose. In the second part of the dissertation, the effect of Cellulose- and Organic-Solvent based Lignocellulose Fractionation (COSLIF) (Zhang, Y.-H. P.; Ding, S.-Y.; Mielenz, J. R.; Elander, R.; Laser, M.; Himmel, M.; McMillan, J. D.; Lynd, L. R. Biotechnol. Bioeng.2007, 97 (2), 214−223) pretreatment conditions on sweet sorghum bagasse (SSB) feedstock was studied using Response Surface Methodology (RSM). Batch experimental matrix was set up based on response surface method's central composite design in two factors to determine the effects of reaction time and temperature on the yield of simple sugars after a sequential pretreatment-enzyme hydrolysis process. Accordingly, changes in delignification, total reducing sugar (TRS) yield, glucan retention, digestibility and overall sugar yields resulting from various combinations of reaction times and temperatures were determined. The results suggested that both pretreatment temperature and reaction time were significant factors, although temperature was more so than reaction time. COSLIF pretreatment conditions of 50°C and 40 min were found to be the optimum pretreatment conditions for the saccharification of SSB. At the end of pretreatment and enzymatic hydrolysis, maximum values of 51.4% delignification, 85% overall glucose yield, and 44% overall xylose yield at an ACCELERASE®1500 loading of 0.25 mL/g sweet sorghum bagasse were achieved. Optimum ACCELERASE®1500 dosage of 0.1 mL/g of sweet sorghum bagasse was identified which resulted in an overall glucose yield of 82.2%±1.05. An effort has also been made to prescribe predictive models which represented the correlation between independent variables (reaction time and temperature), and dependent variables (delignification, and overall glucose yield) using RSM. The significance of the correlations and adequacy of these models were statistically tested for the selected objective functions. The outcomes suggested very competent and statistically adequate regression models which provided quantitative information both for delignification and overall glucose yield for the batch experiments studied.