Effect of the dose of exogenous fibrolytic enzyme preparations

J. Dairy Sci. 98 :406–417

http://wendang.chazidian.com/ 10.3168/jds.2014-8285 ®© American Dairy Science Association, 2015 .

Effect of the dose of exogenous fibrolytic enzyme preparations

on preingestive fiber hydrolysis, ruminal fermentation, and in vitro digestibility of bermudagrass haylage

2J. J. R omero ,1 M. A. Z arate , and A. T. A desogan

Department of Animal Sciences, Institute of Food and Agricultural Sciences, University of Florida, Gainesville 32608

ABSTRACT

Our objectives were to evaluate the effects of the dose

rates of 5 Trichoderma reesei and Aspergillus oryzae

exogenous fibrolytic enzymes (EFE; 1A, 2A, 11C, 13D,

and 15D) on in vitro digestibility, fermentation charac-

teristics, and preingestive hydrolysis of bermudagrass

haylage and to identify the optimal dose of each EFE

for subsequent in vitro and in vivo studies. In experi-

ment 1, EFE were diluted in citrate-phosphate buffer

(pH 6) and applied in quadruplicate in each of 2 runs

at 0× (control), 0.5× , 1× , 2×, and 3×; where 1× was

the respective manufacturer-recommended dose (2.25,

2.25, 10, 15, and 15 g of EFE/kg of dry matter). The

suspension was incubated for 24 h at 25°C and for a fur-

ther 24 h at 39°C after the addition of ruminal fluid. In

experiment 2, a similar approach to that in experiment

1 was used to evaluate simulated preingestive effects,

except that sodium azide (0.02% wt/vol) was added to

the EFE solution. The suspension was incubated for 24

h at 25°C and then 15 mL of water was added before

filtration to extract water-soluble compounds. For both

experiments, data for each enzyme were analyzed sepa-

rately as a completely randomized block design with a

model that included effects of EFE dose, run, and their

interaction. In experiment 1, increasing the EFE dose

rate nonlinearly increased the DM digestibility of 1A,

2A, 11C, and 13D and the neutral detergent fiber di-

gestibility (NDFD) of 1A, 2A, 11C, and 13D. Optimal

doses of 1A, 2A, 11C, 13D, and 15D, as indicated by the

greatest increases in NDFD at the lowest dose tested,

were 2×, 2×, 1×, 0.5×, and 0.5×, respectively. Increas-

ing the dose rate of 2A, 11C, and 13D nonlinearly in-

creased concentrations of total volatile fatty acids and

propionate (mM), decreased their acetate-to-propionate

ratios and linearly decreased those of samples treated

with 1A and 15D. In experiment 2, increasing the dose

eceived April 24, 2014.R

Accepted September 4, 2014. 1 Current address: Department of Crop Science, North Carolina

State University, Raleigh, 1104A Williams Hall, Raleigh, NC 27695-

7620. 2 Corresponding author: a

desogan@ufl.edu rate of each EFE nonlinearly decreased concentrations of netural detergent fiber; also, increasing the dose rate of 1A, 2A, 11C, and 1D nonlinearly increased concen-trations of water-soluble carbohydrates and free ferulic acid (μg/g). Application of increasing doses of the EFE increased NDF hydrolysis, NDFD, and ruminal fluid fermentation of bermudagrass haylage, but the optimal dose varied with the EFE. Key words: fibrolytic enzyme , dairy cattle , bermu-dagrass , in vitro digestibility , dose INTRODUCTION Warm-season grasses are extensively used for cattle production in the southeast United States. Bermudag-rass is the most important of such grasses that is used for cattle production (Newman, 2007), but as with other warm-season grasses, the quality of bermudagrass is low (Hanna and Sollenberger, 2007). Exogenous fibrolytic en-zyme (EFE) treatment has been proposed as a method to improve forage quality and animal performance, but results of published studies have been equivocal (Ad-esogan et al., 2014). Various enzyme, animal, feed, and management factors influence the efficacy of fibrolytic EFE (Beauchemin et al., 2003; Adesogan et al., 2014), many of which are challenging to control. One factor that is easily controlled is the dose of the EFE. To our knowledge, only 2 studies (Dean et al., 2005; Krueger et al., 2008) have been conducted on effects of the dose of EFE on the nutritive value of bermudagrass. Dean et al. (2005) reported that 48-h in vitro NDF digestibility (NDFD) increased quadratically with increasing doses of 1 of 3 cellulase-xylanase EFE applied at the point of ensiling to a 5-wk regrowth of Tifton 85 bermudagrass. Krueger et al. (2008) reported that applying increasing doses of an EFE with high esterase activity to Coastal or Tifton 85 bermudagrass hay had no effect on 6-, 24-, and 48-h in vitro NDFD, except for a linear increase in 6-h NDFD of the Tifton 85 cultivar. More studies are needed to examine effects of EFE dose rates on the quality of bermudagrass hay, silage, and haylage due to the important role of these forages in the diets of dairy and beef cattle in the southeast United States. This is

406

EFFECTS OF EXOGENOUS ENZYME DOSE RATES407

because EFE can be ineffective if applied in insufficient or excessive amounts (Sanchez at al., 1996; Beauchemin et al., 2004). Low doses do not fully exploit the hydro-lytic potential of EFE, especially during short incuba-tion times. In contrast, excessively high doses decrease availability of substrates for catalysis or accessibility of substrates to these sites by crowding the substrate sur-face, which reduces the enzymatic hydrolysis rate (Bom-marius et al., 2008). In the rumen, competition between excessively high doses of EFE and ruminal endogenous cellulolytic bacterial enzymes for substrates can decrease vided by 3 manufacturers were examined at 4 doses (0×, 0.5×, 1×, and 2×, where 1× was the manufac-turer’s recommended dose). Table 1 lists the enzymatic activities and protein concentrations, form, doses, and biological sources of the EFE preparations. Endogluca-nase (EC 3.2.1.4), exoglucanase (EC 3.2.1.91), xylanase (EC 3.2.1.8), and β-glucosidase (EC 3.2.1.21) activities were quantified using carboxymethyl cellulose, avicel, oat-spelt xylan, and cellobiose as artificial substrates, respectively (Wood and Bhat, 1988). Enzymes were incubated with the respective substrates for 15, 120, 5, fiber digestibility (Nsereko et al., 2002) and consequently reduce animal performance (Kung et al., 2000). There-fore, optimization of the EFE dose is critical for using EFE to improve the digestibility of forages.

The objective of our study was to determine the opti-mum dose of 5 EFE (1A, 2A, 11C, 13D, and 15D) that were selected as the most promising of 12 candidates from 3 companies at improving the NDFD of bermu-dagrass haylage (BH; Romero et al., in press). The hypothesis was that increasing the dose of each EFE would increase the NDFD of BH in a quadratic manner.

MATERIALS AND METHODS

Bermudagrass Substrate

An established stand of bermudagrass (Cynodon dac-tylon cultivar Tifton 85) in Alachua, Alachua County, Florida, was staged in June 2010, by mowing to a 4-cm stubble and removing the residue. The field was fertil-ized subsequently with N (95 kg/ha) and the grass was allowed to regrow for 4-wk such that the harvest day was July 7, 2010. On harvest day, the grass was mowed within 1 d to a 4-cm stubble with a Claas 3500 mower conditioner (Claas North America, Omaha, NE). The grass was wilted for 2.5 h in the windrow and then rolled into round 280-kg round bales without inoculant addition. Bales were wrapped with 7 layers of 6-mm plastic and ensiled for 53 d. Ensiled bermudagrass was chosen over hay because it is more typically used in this form by dairy producers due to the high humid-ity and frequent summer rainfall in Florida (Staples, 2003). Representative haylage samples were collected as substrate for our study, dried at 60°C for 48 h, and ground to pass the 1-mm screen of a Wiley mill (Arthur H. Thomas, Philadelphia, PA). The haylage had 49.4% of DM and 93.5, 68.1, 34.2, 3.7, and 18.7% of OM, NDF, ADF, ADL, and CP, respectively (DM basis).Enzymes

Five previously selected (Romero et al., in press) commercial and experimental EFE preparations pro-and 30 min as suggested by Colombatto and Beauche-min (2003). Glucose was used as the standard for mea-suring endoglucanase, exoglucanase, and β-glucosidase activity, whereas xylose was used as that for measuring xylanase activity. Ferulic acid esterase (EC 3.1.1.73) activity was measured using ethyl ferulate as the sub-strate with an incubation period of 5 min with the en-zymes (Lai et al., 2009). All activities were measured at 39°C and a pH of 6 to mimic conditions in the rumen as previously recommended for enzyme studies for lactat-ing dairy cows (Colombatto and Beauchemin, 2003). Activities measured at 20°C and pH 6 were included as a reference for the simulated preingestive hydrolysis assay that was conducted at 25°C. Protein concentra-tion was measured using the Bio-Rad protein assay (Bradford, 1976) with BSA as the standard (Bio-Rad Laboratories, Hercules, CA).In Vitro Ruminal Digestibility (Experiment 1)All EFE were evaluated with a 24-h in vitro ruminal digestibility assay (Goering and Van Soest, 1970) us-ing BH as the substrate. As described by Krueger and Adesogan (2008), EFE were diluted in 0.1 M citrate-phosphate buffer (pH 6) and 2 mL of the solution con-taining the requisite EFE dose was applied to 0.5 g of substrate. The 0× (control) treatment consisted only of the citrate-phosphate buffer and the substrate. Treat-ments were applied in quadruplicate to the substrate in 100-mL polypropylene tubes capped with a rubber stopper fitted with a one-way gas-release valve. Two blank tubes per treatment, containing no substrate, were used to correct for the substrate contribution from the ruminal inoculum. Tubes were tapped gently to en-sure proper mixing of EFE solution with the substrate and the suspensions were subsequently incubated at 25°C for 24 h before addition of buffered ruminal fluid. The ruminal fluid was representatively collected by aspiration 3 h after feeding (0800 h) from 2 nonlactat-ing, nonpregnant, ruminally cannulated Holstein cows consuming a ration consisting of coastal bermudagrass hay ad libitum supplemented with corn (0.45 kg), cot-tonseed hulls (0.46 kg), soybean meal (0.90 kg), and a

Journal of Dairy Science Vol. 98 No. 1, 2015

408

ROMERO ET AL.

vitamin-mineral mix (35.8 g; DM basis). The ruminal fluid collection protocol was approved by the University of Florida, Institute of Food and Agricultural Sciences, Animal Research Committee. The ruminal fluid col-lected was filtered through 4 layers of cheesecloth and mixed with prewarmed artificial saliva (Goering and Van Soest, 1970). Buffered ruminal fluid (52 mL) was dispensed into prewarmed tubes. Tubes were incubated at 39°C for 24 h. Fermentation was terminated by placing the tubes on ice. Tube contents were filtered through previously dried (60°C for 48 h) and weighed 125-mm Whatman No. 451 paper (Fisher Scientific, Pittsburgh, PA). Filtrate and residues were collected for further analysis. Residues were dried at 60°C for 48 h, weighed, and analyzed for NDF (Van Soest et al., 1991) and ADF (AOAC International, 2000) sequen-tially using an Ankom 200 Fiber Analyzer (Ankom, Macedon, NY). No sodium sulfite was used for the NDF analysis. Hemicellulose (HEM) was calculated as the difference between NDF and ADF. Residue weights and their fiber concentrations were used to calculate true digestibility of DM, NDF, HEM, and ADF (DMD, HEMD, and ADFD; Mertens, 2007). Filtrate samples were analyzed for pH using an Accumet XL25 pH me-ter with an automatic temperature correction feature (Fisher Scientific), acidified with 50% H2SO4 (1% vol/vol) and centrifuged at 8,000 × g for 20 min at 4°C. The supernatant was frozen (?20°C) and subsequently analyzed for concentrations of VFA (Muck and Dicker-son, 1988) using a Merck Hitachi Elite LaChrome High Performance Liquid Chromatograph system (Hitachi

Journal of Dairy Science Vol. 98 No. 1, 2015

L2400, Tokyo, Japan) fitted with a Bio-Rad Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA). Ammonia-N was determined with a Technicon Auto Analyzer (Technicon, Tarrytown, NY) and an ad-aptation of the Noel and Hambleton (1976) procedure that involved colorimetric N quantification.Preingestive Fiber Hydrolysis (Experiment 2)

The same EFE doses examined in experiment 1 were tested in this experiment to ascertain their ef-fects on preingestive hydrolysis of BH. In experiment 2, a similar approach as that in experiment 1 was used to evaluate simulated preingestive effects, except that 50-mL centrifuge tubes were used and sodium azide (0.02% wt/vol) was added to the EFE solution to prevent microbial degradation of substrate (Krueger and Adesogan, 2008).Two blank tubes per treatment, containing no substrate, were included as EFE blanks. After the incubation at 25°C, 15 mL of double-distilled water were added and tubes were shaken for 1 h at 260 oscillations/min with an Eberbach Reciprocating Shaker Model 6000 (Eberbach Corporation, Ann Ar-bor, MI) to extract water-soluble compounds. Tubes were filtered through previously dried (60°C for 48 h) and weighed 125-mm Whatman No. 451 filter paper (Fisher Scientific, Pittsburgh) and filtrate and residue samples were collected. Residues were dried at 60°C for 48 h, weighed, and analyzed for NDF and ADF as described previously. Residue and sample dry weights and DM concentrations were used to calculate DM

EFFECTS OF EXOGENOUS ENZYME DOSE RATES409

losses. Residue weights and their fiber concentrations were used to calculate fiber fraction concentrations. Filtrate samples were frozen (?20°C) and subsequently analyzed for water-soluble carbohydrates (WSC; Du-Bois et al., 1956) and ferulic (FER) and p-coumaric acids (COU; Bio-Rad, 2011) using the HPLC system trol, the 2× dose was the lowest dose that resulted in the greatest increases in NDFD and HEMD (7.4 and 11.5%, respectively). This dose also resulted in the greatest increase in NDFD per unit of EFE 1A used (0.56% NDFD per gram of EFE). Applying EFE 1A did not affect DMD, but the 2× and 3× doses resulted and the HPX-87H column described previously.Statistical Analyses

Data for each EFE were analyzed separately. A ran-domized complete block design with 4 replicates per treatment and 2 runs was used to determine the effects of EFE preparations on digestibility and fermentation measures (experiment 1) and substrate DM and fiber disappearance and release of sugars and phenolic acids (experiment 2). Run was the blocking factor.

The model used to analyze digestibility, fermentation and preingestive hydrolysis data was

Yijk = μ + Di + Rj + DRij + Eijk,

where μ = general mean; Di = effect of dose i; Rj = effect of run j; DRij = effect of the dose i × run j interaction; and Eijk = experimental error.

The GLM procedure of SAS v.9.1 (SAS Institute Inc., Cary, NC) was used to analyze each EFE separately, because comparing dose rate effects among EFE was not of interest. Polynomial contrasts were used to de-termine dose effects and the Fisher’s F-protected least significance difference test was used to determine the optimal dose. Both of these mean characterization and separation tests were considered necessary to properly interpret the results because they depict the polyno-mial trend and the optimal dose, respectively. The final decision on the optimal dose of the EFE for future in vitro and animal trials was defined as the least dose that resulted in a greater increase in NDFD than lower doses and a similar or greater response relative to higher doses. Neutral detergent fiber digestibility was chosen as the response of choice for selecting the optimal dose because of its correlation with DMI and milk produc-tion (Oba and Allen, 1999). Significance was declared at P < 0.05 and tendencies at P > 0.05 < 0.10.

RESULTS AND DISCUSSION

Experiment 1: EFE Dose Effects on Measures of in Vitro Digestion and Fermentation

EFE Dose Effects on Digestibility Measures. Increasing the dose of EFE 1A linearly increased (P < 0.01) HEMD and had a cubic effect (P < 0.05) on DMD and NDFD (Table 2). Compared with the con-in greater DMD than the 0.5× dose. Applying increas-ing doses of EFE 1A resulted in a cubic effect (P < 0.05) on ADFD partly because its value decreased to a nadir when the 0.5× and 1× doses were applied (?8.4 and ?10.9%, respectively; P < 0.05). The decrease in ADFD with increasing doses of EFE 1A is probably attributable to the declining reactivity of residual cel-lulose during enzymatic hydrolysis due to the decrease in surface area and number of accessible chain ends or adsorption of inactive cellulase on the surface of cel-lulose (Zhang and Lynd, 2004). Yet, if the EFE dose is too low, the supply of auxiliary enzymes and proteins, such as swollenin, may be insufficient to remove barri-ers preventing increases in ADFD.Applying increasing doses of EFE 2A increased (P < 0.05, quadratic) DMD, NDFD, and HEMD (Table 2). The 2× dose was the lowest dose that resulted in the greatest (P < 0.05) increases in NDFD and HEMD (10.8 and 16.2%, respectively), but the dose that gave the greatest increase in NDFD per unit of EFE 2A was 0.5× (1.97% NDFD units per gram of EFE). The choice of which dose to use depends on the desired objective. If the intent is to maximize NDFD, the 2× dose should be selected, whereas the 0.5× dose should be chosen from an efficiency standpoint, provided it is subsequently shown to provide an economic increase in animal performance. Increasing the dose of EFE 2A did not increase ADFD. Therefore, the EFE exerted its hy-drolytic effect on HEM rather than ADF. Hemicellulose typically represents about half of the fiber concentra-tion in grasses (Van Soest, 1994), and it was the fiber fraction most effectively hydrolyzed by adding EFE to BH (Romero et al., 2013).Increasing the dose of EFE 11C increased ADFD (P < 0.01, linear), DMD (P < 0.01, quadratic), and NDFD and HEMD (P < 0.01, cubic; Table 2). The 1× dose was the lowest dose that resulted in the greatest increases in NDFD, HEMD, and ADFD (16.2, 22.6, and 6.7%, respectively; P < 0.05). The dose resulting in the greatest increase in NDFD per unit of EFE 11C was 0.5× (0.37% NDFD per gram of EFE). As was the case for EFE 1A and 2A, EFE 11C had its greatest effects on the HEM fiber fraction. However, unlike EFE 1A and 2A, 11C also increased ADFD when applied at 1× or 3×. This may be because EFE 11C supplied more exoglucanases on an applied basis than the other EFE. Exoglucanase I is usually the most secreted enzyme protein in Trichoderma reesei (60%) commercial prepa-

Journal of Dairy Science Vol. 98 No. 1, 2015

410

ROMERO ET AL.

rations and is arguably the most relevant enzyme in cellulose hydrolysis (Selig et al., 2008). Exoglucanases progressively cleave cellulose chains at the reducing and nonreducing ends to release cellobiose or glucose after

Journal of Dairy Science Vol. 98 No. 1, 2015

the hydrolytic and oxidative cleavage of internal parts of the chain by endoglucanases (Zhang et al., 2006) and novel polysaccharide monooxygenases, respectively (Dimarogona et al., 2013). This enzymatic depolymer-