Screening exogenous fibrolytic enzyme preparations for improved in vitro digestibility of bermudagrass haylage.

Our objectives were to evaluate the effects of 12 exogenous fibrolytic enzyme products (EFE) on ruminal in vitro neutral detergent fiber digestibility (NDFD) and preingestive hydrolysis of a 4-wk regrowth of bermudagrass haylage (BH), to examine the accuracy of predicting NDFD with EFE activity measures, and to examine the protein composition of the most and least effective EFE at increasing NDFD. In experiment 1, effects of 12 EFE on NDFD of BH were tested. Enzymes were applied in quadruplicate to culture tubes containing ground BH. The suspension was incubated for 24 h at 25 °C before addition of rumen fluid media and further incubation for 24 h at 39 °C. The experiment was repeated twice. In addition, regression relationships between EFE activity measures and NDFD were examined. Compared with the values for the control, 9 EFE-treated substrates had greater NDFD (37.8 to 40.4 vs. 35.6%), 6 had greater total VFA concentration (59.1 to 61.2 vs. 55.4 mM), and 4 had lower acetate-to-propionate ratios (3.03 to 3.16 vs. 3.24). In experiment 2, EFE effects on preingestive fiber hydrolysis were evaluated by incubating enzyme-treated and untreated bermudagrass suspensions in quadruplicate for 24 h at 25 °C and examining fiber hydrolysis measures. Compared with values for the control, 3 EFE reduced neutral detergent fiber concentration (62.8 to 63.7 vs. 67.3%), 10 increased release of water-soluble carbohydrates (26.8 to 58.5 vs. 22.8 mg/g), and 8 increased release of ferulic acid (210 to 391 vs. 198 μg/g). Regression analyses revealed that enzyme activities accurately [coefficient of determination (R(2)) = 0.98] predicted preingestive hydrolysis measures (water-soluble carbohydrates, ferulic acid), moderately (R(2) = 0.47) predicted neutral detergent fiber hydrolysis, but poorly (R(2) ≤ 0.1) predicted dry matter and NDFD. In experiment 3, proteomic tools were used to examine the protein composition of the most and least effective EFE at improving NDFD. Relative to the least effective, the most effective EFE at increasing NDFD contained 10 times more endoglucanase III, 17 times more acetylxylan esterase with a cellulose-binding domain 1, 33 times more xylanase III, 25 times more β-xylosidase, and 7.7 times more polysaccharide monooxygenase with cellulose-binding domain 1 and 3 times more swollenin. The most effective EFE had a much greater quantity of fibrolytic enzymes and key proteins necessary for hemicellulose and lignocellulase deconstruction. This study identified several EFE that increased the NDFD and in vitro fermentation of 4-wk BH and revealed why some EFE are more effective than others.