Levels of endogenous β-glucanase activity in barley affect the efficacy of exogenous enzymes used to supplement barley-based diets for poultry

ABSTRACT

To improve the nutritive value of barley-based diet for broilers, 2 experiments using 2 different barley lots were performed to evaluate the capacity of a mesophilic cellulase when fused to a β-glucan specific family 11 carbohydrate-binding module. The data revealed that the recombinant β-glucanase derivatives were not appropriate for feed supplementation because of a lack of stability at acidic pH levels. However, under the same experimental conditions, a commercial enzyme mixture improved the nutritive value of 1 of the cereal lots used. Analysis of the nutritive value of the 2 barleys revealed intrinsic differences in the levels of endogenous β-glucanase activity. These differences were extensively evident when the studies were expanded to a range of 64 barley lots. Thus, to clarify the effect of endogenous cellulases on the efficacy of exogenous β-glucanases used to supplement barley-based diets for poultry, 2 barley lots presenting low and high levels of endogenous plant cell wall-degrading enzymes were selected. These lots were used to prepare 2 barley-based diets, which were supplemented with or without a commercial enzyme product and fed to broiler chicks. The data revealed that the exogenous enzymes were effective when the basal diet presented low levels of endogenous β-glucanases but were unable to improve the nutritive value of the barley lot displaying higher β-glucanase activity. Thus, these studies suggest that levels of endogenous β-glucanases may affect the efficacy of exogenous enzymes used to improve the nutritive value of barley-based diets for broilers. The development of a quick β-glucanase assay that could be applied for cereal-based feeds may help identify those barley-based diets that are more responsive to the action of feed enzymes.

INTRODUCTION

Most cereals contain a significant proportion of soluble nonstarch polysaccharides (NSP), which are known to display a variety of antinutritive properties for monogastric animals, particularly for poultry (Hesselman and Aman, 1986). Incorporation of exogenous β-1,3–1,4-glucanases, in barley, and β-1,4-xylanases, in wheat or rye-based diets, improves the efficiency of feed utilization, enhances growth, and contributes to a better use of low cost feed ingredients (Chesson, 1993; Bedford, 2000). Plant cell wall-degrading enzymes contribute to reduce digesta viscosity that is associated with the intake of soluble indigestible carbohydrates, therefore improving the rate of diffusion of substrates, endogenous digestive enzymes, and nutrients (White et al., 1981; Fengler and Marquardt, 1988; Bedford et al., 1991; Bedford and Classen, 1992). A reduction in digesta viscosity also increases the velocity of feed passage, thus decreasing the proliferation of fermenting microbes in the upper regions of the gastrointestinal (GI) tract (van der Klis et al., 1993) while improving feed intake. However, it is well known that in certain cases the inclusion of exogenous enzymes in diets containing a high percentage of wheat, barley, or rye fails to have any effect in animal performance (Bedford, 2000). Numerous hypotheses have been advanced to explain this observation. Strong evidence exists that the availability of energy from cereal grains in poultry is inversely related to the content of soluble NSP (Villamide et al., 1997). Thus, levels of NSP may vary between cereal lots, resulting in cereals expressing different nutritive values. It is well established that a multitude of factors may affect cereal content in NSP, which include cereal genotype and growing conditions, length of the cereal storage period, grain cultivar, growing season, or soil type, among others (Villamide et al., 1997). Although other factors may explain the unpredictable response to enzyme supplementation, such as the levels of endogenous cellulase and hemicellulase activities within the grain, these remain relatively uncharacterized.

Glycoside hydrolases (EC 3.2.1) that participate in the hydrolysis of plant cell wall carbohydrates display remarkably elaborate molecular architectures comprising both catalytic and noncatalytic carbohydrate-binding modules (CBM). These CBM contribute to establish a close interaction between enzymes and plant polysaccharides allowing the appended catalytic domain to intimately contact its target substrates, thus potentiating catalysis (Boraston et al., 2004). Carbohydrate-binding modules have been classified into sequence-based families in the CAZy database (http://www.cazy.org; Boraston et al., 2004). The role of CBM in the function of feed enzymes has been investigated. A family 6 xylan-binding domain was shown to improve the efficacy of a microbial recombinant xylanase in vivo when the enzyme was used to supplement wheat or rye-based diets for poultry (Fontes et al., 2004). Thus, animals supplemented with a bimodular xylanase containing both catalytic and xylan-binding domains grew significantly faster than animals fed diets containing exclusively the xylanase catalytic domain. More recently, a family 11 CBM that is highly specific for β-1,3–1,4-glucans was shown to improve the efficacy of the associated catalytic domain used to supplement a barley-based diet (Ribeiro et al., 2008). The higher efficacy of the modular enzyme allowed for a significant reduction in enzyme dosage.

The primary objective of this work was to evaluate, in 2 replicated experiments, the capacity of a family 11 CBM that binds specifically to β-1,3–1,4-mixed linked glucans to improve the efficacy of a single-domain cellulase from Cellvibrio mixtus. A different response to enzyme supplementation in these 2 initial experiments motivated the measurement of the levels of β-glucan, β-glucanase activity, and viscosity in the 2 barley lots used. The experiments were extended to a range of more than 60 barleys. Because levels of endogenous plant β-glucanases were found to vary widely in barley, a third experiment was conducted in which the capacity of exogenous enzymes to improve the nutritive value of barley-based diets containing different levels of plant β-glucanases was evaluated.

MATERIALS AND METHODS

Enzyme Preparation

The gene encoding the family 11 CBM of Clostridium thermocellum CtLic26A-Cel5E was amplified through PCR from genomic DNA using the thermostable polymerase NZYPremium (NZYTech Ltd., Lisboa, Portugal) and the following primers: forward, 5′-CTCGTCGACCCAACTCCAAGACCGACC; reverse, 5′-CACCTCGAGAGCACCAATCAGCTTGAT. The PCR product was cloned into pNZY28, sequenced to ensure that no mutations accumulated during PCR, and subsequently subcloned into the XhoI site of pCF1 (Fontes et al., 1997) generating pTR1. Plasmid pCF1 encoded the single-domain β-glucanase Cel5A and pTR1 encoded the unimodular cellulase fused to the clostridial CBM11, which was termed Cel5A-CBM11 (Figure 1). The 2 plasmids were used to transform BL21 Escherichia coli cells. Recombinant E. coli strains were grown on Luria-Bertani media to mid exponential phase (absorbance at 600 nm = 0.5) and recombinant gene expression was induced by adding isopropyl β-d-thiogalactoside to a final concentration of 1 mM. Cells were collected after 5 h of induction at 37°C and protein extracts were prepared by ultrasonication followed by centrifugation. The recombinant proteins were purified by metal-affinity chromatography as described by Fontes et al. (2004). Both recombinant proteins retain considerable catalytic activity at 40°C and are resistant to proteolytic degradation (Fontes et al., 1997).

Figure 1

Domain organization of Cellvibrio mixtus and its recombinant derivative used in this study. The cellulase (GH5) and β-glucan binding domain (CBM11) are indicated. The gray and the black boxes represent the signal peptide and linker sequences, respectively.

Birds, Diets, and Management

The composition of the barley-based diets used in this study, which were formulated to contain adequate nutrient levels as defined by the NRC (1994), is presented in Table 1. In the first studies (experiments 1 and 2), 2 different barley lots were used to produce basal diets that were supplemented with no enzyme or with 15 U/kg of Cel5A (referred to as treatment Cel5A) or Cel5A-CBM11 (referred to as treatment Cel5A-CBM11). This level of supplementation corresponds to the calculated level of supplementation of the positive control enzyme (see below). In addition, a fourth treatment included a basal diet supplemented with a calculated 15 U/kg of the commercial enzyme cocktail Rovabio Excel AP (Adisseo, Antony, France; referred to as treatment C+), which corresponds to an incorporation ratio of 50 g of enzyme/tonne of feed as recommended by the manufacturer. In experiment 3, 2 basal diets were produced using barley batches that contained high or low levels of endogenous β-glucanases. The diets were supplemented with or without the commercial enzyme product Rovabio Excel AP as described above. Thus, the effect of β-glucanase supplementation was tested in barley lots presenting high and low levels of endogenous β-glucanase. Basal diets were provided in the pelleted form and enzyme preparations were mixed with the feed just before administration to the animals. Animal experiments were conducted in accordance with the principles and specific guidelines presented in European Union (1986), reviewed by the Ethics Committee of CIISA Faculdade de Medicina Veterinária, and approved by the Animal Care Committee of the National Veterinary Authority (Direcção-Geral de Veterinária, Lisboa, Portugal).

Ingredient composition and calculated analysis of the cereal-based feed

Table 1

Ingredient composition and calculated analysis of the cereal-based feed

Item (% unless noted)	Value
Ingredient
Barley	61.48
Soybean meal (47%)	29.40
Soybean oil	5.90
Sodium chloride	0.25
Calcium carbonate	0.80
Dicalcium phosphate (18%)	1.80
dl-Methionine	0.17
Mineral and vitamin premix1	0.20
Calculated nutrient content
Energy (MJ of ME/kg of DM)	12.1
CP	20.8
Ether extract	7.5
Crude cellulose	5.1
Ash	5.7
Methionine	0.47

¹Mineral and vitamin premix provided the following per kilogram of diet: vitamin A, 9,000 IU; vitamin D₃, 2,100 IU; vitamin E, 20 mg; nicotinic acid, 30 mg; vitamin B₁₂, 0.12 mg; calcium pantothenate, 10 mg; vitamin K₃, 2 mg; thiamin, 1 mg; riboflavin, 4.2 mg; vitamin B₆, 1.7 mg; folic acid, 0.5 mg; biotin, 0.5 mg; Fe, 80 mg; Cu, 10 mg; Mn, 100 mg; Zn, 80 mg; Co, 0.2 mg; I, 1.0 mg; Se, 0.3 mg; monensin, 100 mg.

For each experiment, 1-d-old chicks (Ross 308; n = 160) were divided into 40 battery brooders (capacity of 4 animals/pen) exposed to constant light for the duration of the trial. Water and diets were available ad libitum throughout the experiment and were provided from automatic drinking nipples and a hanging feeder, respectively. The brooders were located in an environmentally controlled room, which was adjusted daily to the recommended temperatures according to standard brooding practice. Birds were individually weighed at the beginning of the experiment and were randomly assigned to one of the 4 treatments, with 10 replicates/treatment. Feed consumption and individual BW were recorded weekly. Feed conversion ratios were calculated by dividing the weight gain per pen per week and at the end of the experiment, including the weight gain of any dead birds, by the total feed consumed during the respective period. Bird mortality was recorded daily. At the end of the experiment, on d 28, 1 bird/pen was slaughtered by an intravenous injection of an aqueous isotonic solution of 125 mg of Tiopental Braun (Braun, Barcelona, Spain). The size of the various GI compartments was measured or weighted and digesta samples were collected and stored at −20°C for later analysis. In experiments 1 and 2 the weight of the GI compartments was determined full. In contrast, in experiment 3 the various portions of the GI tract were emptied before weighing. Levels of β-glucanase activity in the GI tract were measured as described below.

Analytical Procedures

To standardize the number of enzyme units used to supplement the basal diets, the catalytic activity of the various exogenous enzymes, including the commercial mixture, was determined under identical experimental conditions. Catalytic activity was determined at 40°C by measuring reducing sugar released, following the method described by Fontes et al. (2000), using barley β-glucan (Megazyme, Bray, UK) as the substrate. One unit of catalytic activity is defined as the amount of enzyme required to release 1 µmol of product/min. The extract containing Rovabio Excel AP enzymes was prepared by ressuspending 75 mg of the enzyme mixture in 10 mL of 50 mM Na-HEPES buffer (pH 7.5) followed by incubation overnight at 4°C with gentle agitation and a centrifugation at 13,000 × g for 5 min. Previous to detection of β-glucanase activity, digesta samples were centrifuged and the supernatant was recovered for analysis. Initially, qualitative analysis of β-glucanase activity in the digesta samples recovered from the various GI compartments was assessed in agar plates, using barley β-glucan (Megazyme) at 0.1% (wt/vol) final concentration, in 10 mM Tris-HCl (pH 7.0). Catalytic activity was detected after 16 h of incubation at 37°C through the Congo red assay plate, as described in Ponte et al. (2004) and Mourão et al. (2006). For measuring the viscosity of small intestine contents, samples were centrifuged for 10 min at 7,500 × g and the viscosity of the supernatant was measured using a Brookfield viscometer (model LVDVCP-II, Brookfield Engineering Laboratories, Middleboro, MA) with a cup maintained at 24°C. Analyses for DM (method 934.01), crude fat (method 920.39), CP (method 954.01), NDF (method 2002.04), and ADF and acid detergent lignin (method 973.18) were performed according to the methods of AOAC (1980).

A large range of barley lots from different varieties was selected in September and October 2009, and levels of viscosity and contents in β-glucan and β-glucanase were measured. All barleys were harvested in 2009 and thus could be considered as young barleys because the assays were performed in the autumn of 2009. Barley content in β-glucan and levels of cereal β-glucanase activities were determined with a β-glucan assay kit and a β-glucanase assay kit (Megazyme), respectively. The method for determining β-glucanase activity followed the manufacturer’s protocol with a modification on the incubation period that was extended from 10 min to 3 h. The method used for measuring the levels of viscosity followed the steps described above, although the barley samples were milled at 0.5 mm and mixed with 15 mL of phosphate citrate buffer (pH 6.5) by vigorous shaking for 5 min before the start of the experiment.

Statistical Analysis

Statistical analysis of data related to bird performance was conducted by ANOVA using the GLM procedure of SAS (SAS Inst. Inc., Cary, NC). Means with a significant F-ratio were separated by the least significant difference test. The experimental unit was a cage of 4 birds. Unless otherwise stated, differences were considered significant at P < 0.05. Regression analyses were conducted to test for linearity between level of β-glucans and viscosity, between level of β-glucans and β-glucanase activity, and between viscosity and β-glucanase activity in barley lots.

RESULTS AND DISCUSSION

Recombinant Derivatives of Cel5A from C. mixtus Were Unable to Improve the Nutritive Value of Barley-Based Diets for Poultry

In contrast to the majority of plant cell wall-degrading enzymes, Cel5A from C. mixtus is a unimodular enzyme which, despite being produced by a mesophilic bacterium, presents resistance to proteolytic inactivation and significant thermostability (Fontes et al., 1997). Previously, it was shown that a family 11 CBM has the capacity to increase the catalytic activity of the associated β-glucanase and cellulase catalytic domains in vivo by improving the efficacy of the catalytic modules to enhance the nutritive value of barley-based diets for poultry (Guerreiro et al., 2008; Ribeiro et al., 2008). Here, we investigated the capacity of a recombinant form of Cel5A from C. mixtus to decrease the levels of β-glucans found in barley-based diets for poultry. In addition, the capacity of the family 11 CBM to enhance the efficacy of Cel5A was evaluated by engineering a recombinant fusion enzyme combining the 2 modules (Figure 1).

In experiment 1, BW of birds fed barley-based diets supplemented with or without the exogenous polysaccharidases was not significantly different on d 7, 14, 21, or 28 (final BW; bird mortality was 3.1%; Table 2). No differences were found in weight gain among the different groups during the entire trial. In addition, feed intake and feed conversion ratios did not differ among the groups. Taken together, the results suggest that the exogenous enzymes of microbial (CelA and CelA-CBM11) or fungal (Rovabio Excel AP; treatment C+) origin were unable to improve the nutritive value of the barley-based diet used in experiment 1, which contained more than 60% (wt/wt) of barley.

Performance of broilers fed a barley-based diet supplemented with different β-glucanase preparations studied in experiment 1

Table 2

Performance of broilers fed a barley-based diet supplemented with different β-glucanase preparations studied in experiment 1

Item	Treatment1				SEM	p(F)2
	C−	C+	CelA	CelA-CBM11
BW (g)
0 d	42.9	43	42.9	42.9	0.150	0.945
7 d	135.9	137.3	130.4	131.4	2.570	0.182
14 d	310.5	331.4	310.2	307.4	7.330	0.095
21 d	594.5	627.8	591.3	589.1	11.590	0.077
28 d	1,028.7	1,068.5	1,015.1	1,040.3	22.350	0.391
Weight gain (g)
0–7 d	93	94.3	87.5	88.5	2.57	0.186
7–14 d	174.6	194.1	179.8	176.1	5.56	0.070
14–21 d	284.1	296.5	281.1	281.7	6.16	0.268
21–28 d	434.2	440.7	423.9	451.3	15.59	0.655
0–28 d	985.8	1,025.5	972.2	997.4	22.32	0.391
Feed intake (g)
0–7 d	122.0	127.1	123.2	120.6	3.66	0.636
7–14 d	261.4	288.6	274.3	272.7	6.64	0.052
14–21 d	438.6	455.8	444.4	432.3	14.65	0.709
21–28 d	695	698.7	716.7	714.4	16.58	0.729
0–28 d	1,516.9	1,570.2	1,558.6	1,540.0	32.81	0.683
Feed conversion ratio
0–7 d	1.320	1.360	1.410	1.360	0.044	0.545
7–14 d	1.500	1.490	1.540	1.580	0.044	0.520
14–21 d	1.550	1.540	1.590	1.560	0.033	0.708
21–28 d	1.650	1.590	1.700	1.630	0.042	0.381
0–28 d	1.570	1.540	1.600	1.630	0.036	0.257

¹C−: birds not supplemented with an exogenous enzyme; C+: birds supplemented with an enzymatic commercial mixture; CelA: birds supplemented with Cellvibrio mixtus; CelA-CBM11: birds supplemented with the protein CelA-CBM11.

²p(F) < 0.05 is statistically significant.

The exact same experiment was repeated (experiment 2) using a basal diet including the same proportion of ingredients (as described in Table 1) but preparing it using different barley, soybean meal, and soybean oil; all other remaining components of the diet had the same origin as in experiment 1. Bird BW, weight gain, feed intake, and feed conversion ratio of experiment 2 are summarized in Table 3 (bird mortality was 1.9%). The data revealed that final BW of birds fed the basal diet supplemented with the commercial enzyme cocktail (Rovabio Excel AP; treatment C+) was significantly higher than birds of the other groups. Differences in BW were visible as soon as d 7 and remained significant for the duration of the experiment. However, differences in weight gain were particularly acute on d 7, revealing that the exogenous enzymes were particularly effective in earlier stages of growth. In addition, no differences in feed intake were found between treatments, although feed conversion ratio in animals receiving the fungal enzymes was smaller when compared with the other groups (0–28 d). Thus, these data suggest that an improvement in BW in animals receiving the commercial feed enzymes resulted from a better use of the feed ingredients rather from an increase in feed intake. Taken together the results revealed that, in contrast to the commercial enzyme mixture, the recombinant enzymes were unable to improve the nutritive value of the barley-based diet of experiment 2. In addition, broilers fed the basal diets of experiments 1 and 2 responded differently to the addition of the fungal exogenous enzymes that were effective only in experiment 2.

Performance of broilers fed a barley-based diet supplemented with different β-glucanases studied in experiment 2

Table 3

Performance of broilers fed a barley-based diet supplemented with different β-glucanases studied in experiment 2

Item	Treatment1				SEM	p(F)2
	C−	C+	CelA	CelA-CBM11
BW (g)
0 d	45.3	44.7	45.2	45.3	0.450	0.759
7 d	141.2b	157.3a	132.3b	133.6b	4.730	0.002
14 d	365.0b	397.3a	344.9b	351.5b	9.020	0.001
21 d	699.2b	750.8a	664.3b	677.1b	16.100	0.003
28 d	1,171.9b	1,230.4a	1,112.6b	1,154.0b	23.210	0.009
Weight gain (g)
0–7 d	96.0b	112.6a	87.1b	88.3b	4.67	0.002
7–14 d	223.8	240.1	212.6	218.0	8.60	0.146
14–21 d	334.3ac	353.5a	319.4bc	325.6bc	8.10	0.029
21–28 d	472.6	479.8	458.4	477.0	11.06	0.538
0–28 d	1,126.6ac	1,185.9a	1,077.5bc	1,108.7bc	23.90	0.022
Feed intake (g)
0–7 d	140.8	148.4	136.0	142.0	6.07	0.551
7–14 d	327.1	338.5	322.7	335.2	9.65	0.640
14–21 d	591.9	595.6	587.5	612.4	16.17	0.716
21–28 d	917.7	877.5	910.1	917.6	21.69	0.513
0–28 d	1,977.5	1,960.0	1,956.2	2,007.5	38.98	0.783
Feed conversion ratio
0–7 d	1.460	1.370	1.560	1.630	0.079	0.117
7–14 d	1.480	1.440	1.520	1.550	0.058	0.535
14–21 d	1.780	1.690	1.840	1.890	0.058	0.091
21–28 d	1.940	1.840	2.000	1.930	0.049	0.158
0–28 d	1.760ab	1.660b	1.820a	1.820a	0.038	0.014

^a–cMeans within the same row with different superscripts are significantly different (P < 0.05).

¹C−: birds not supplemented with an exogenous enzyme; C+: birds supplemented with an enzymatic commercial mixture; CelA: birds supplemented with Cellvibrio mixtus; CelA-CBM11: birds supplemented with the protein CelA-CBM11.

²p(F) < 0.05 is statistically significant.

It is well known that diets presenting high levels of soluble NSP induce considerable enlargement of some portions of the GI tract (Brenes et al., 1993) and stimulate an increase in protein turnover rates (Dänicke et al., 2000). Enzyme supplementation decreases digesta viscosity and therefore improves the feed passage rate and nutrient absorption. In addition, the relative weight of the digestive tract decreases, leading to an increase in carcass yield (Petterson and Aman, 1989). Fuente et al. (1998) found an equation relating the empty weight of the digestive tract to digesta viscosity. Therefore, the effects of dietary treatments in the relative length and weight of GI tract compartments of broiler chickens of experiments 1 and 2 were evaluated. Enzyme supplementation had no effect on crop, gizzard, and liver relative weights, on digesta viscosity, or on the duodenum, jejunum, and cecum relative lengths in experiment 1 (data not shown). In contrast, in experiment 2 jejunum and ileum relative lengths were significantly reduced (P < 0.05) in birds receiving the commercial enzyme mixture when compared with animals of other treatments (Table 4). Thus, exogenous enzymes in treatment C+ contributed to reduce the levels of digestive antinutritive β-glucans, which through an increase in viscosity contribute to increase the size of the digestive compartments. Although it significantly contributed to improve broiler performance, the commercial β-glucanase (treatment C+) had no effect on the viscosity of the intestinal contents (Table 4). Digesta viscosity was determined exclusively at the end of the experimental period, on d 28. At later stages of animal growth, the endogenous microflora might contribute to considerably reduce the chain length of soluble glucans, thus reducing digesta viscosity.

Relative weight and length of the gastrointestinal tract and viscosity of digesta samples of broilers fed a barley-based feed supplemented with different exogenous β-glucanases (data from experiment 2)

Table 4

Relative weight and length of the gastrointestinal tract and viscosity of digesta samples of broilers fed a barley-based feed supplemented with different exogenous β-glucanases (data from experiment 2)

Item	Treatment1				SEM	p(F)2
	C−	C+	CelA	CelA-CBM11
Relative weight (g/100 g of BW)
Crop	3.790a	2.940b	3.130b	3.460ab	0.221	0.051
Gizzard	13.710	12.600	13.420	13.260	0.888	0.840
Relative length (cm/kg of BW)
Duodenum	22.06	21.82	23.37	21.54	0.686	0.261
Jejunum	60.28b	53.35a	63.83b	60.10b	1.813	0.002
Ileum	64.04ab	59.39a	68.67b	61.42ab	2.196	0.030
Cecum	14.70	13.32	15.07	14.34	0.542	0.140
Content viscosity (cP)
Duodenum + jejunum	6.753	5.983	6.632	6.937	0.483	0.919
Ileum	12.529	9.553	12.485	12.128	0.974	0.670

^a,bMeans within the same row with different superscripts are significantly different (P < 0.05).

¹C−: birds not supplemented with an exogenous enzyme; C+: birds supplemented with an enzymatic commercial mixture; CelA: birds supplemented with Cellvibrio mixtus; CelA-CBM11: birds supplemented with the protein CelA-CBM11.

²p(F) < 0.05 is statistically significant.

In contrast with previous reports (Philip et al., 1995; Guerreiro et al., 2008; Ribeiro et al., 2008) that confirmed the capacity of recombinant cellulases Cel5E (Philip et al., 1995) and CtLic26A-Cel5E (Guerreiro et al., 2008; Ribeiro et al., 2008) from C. thermocellum to improve the nutritive value of barley-based diets for poultry, data presented here suggest that Cel5A from C. mixtus is unable to reduce the antinutritive effects of soluble glucans present in barley-based diets. One of the major actions of feed β-glucanases is to decrease the degree of polymerization of β-1,3–1,4-glucans present in barley through the random cleavage of glycosydic bonds of the polysaccharide backbone. The reduction in carbohydrate chain length contributes to decrease the levels of digesta viscosity (Fengler and Marquardt, 1988; Bedford and Morgan, 1996; Apajalahti and Bedford, 1999). Data presented here suggest that the exogenous recombinant CelA derivatives were not catalytically active. Thus, to evaluate the stability of exogenous glycoside hydrolases during passage through the GI tract, β-glucanase activity was qualitatively determined in digesta samples collected in the various digestive compartments of 10 animals/treatment. The data, presented in Table 5, revealed that the majority of digesta samples collected from birds receiving the commercial enzyme mixture expressed β-glucanase activity. However, the levels of enzyme activity present in the GI tract of birds supplemented with the recombinant enzymes were very small or below the assay detection limit, particularly after the crop. The lower activity of the recombinant enzymes in the crop when compared with animals supplemented with the fungal β-glucanases is difficult to explain considering that the crop presents a milder environment in terms of pH and proteases. It is possible that some degree of enzyme inhibition may have occurred as a result of the presence of specific CelA inhibitors in the diet, although this possibility was unexplored in the current work. As was anticipated, only birds not receiving exogenous enzymes expressed significant levels of β-glucanase activity in the cecum. Thus, the data suggest that although a similar level of exogenous enzymes was added to the basal diets of treatments C+, Cel5A, and Cel5A-CBM11, the activity of the recombinant enzymes was significantly reduced after the passage through the proventriculus. This suggests that CelA and its recombinant derivatives are particularly sensitive to denaturation under acidic conditions. Assays performed in vitro confirmed this hypothesis; the 2 recombinant derivatives were shown to retain only approximately 10% of their initial activity after incubation for 10 min at pH 3.

Qualitative detection¹ of cellulase activity in digesta collected from the gastrointestinal compartments of 40 broilers fed a barley-based feed supplemented with recombinant cellulases displaying different molecular architectures² (data from experiment 2)

Table 5

Qualitative detection¹ of cellulase activity in digesta collected from the gastrointestinal compartments of 40 broilers fed a barley-based feed supplemented with recombinant cellulases displaying different molecular architectures² (data from experiment 2)

Item	C−	C+	CelA	CelA-CBM11
Crop	+/–/–/+/–/–/–/–/–/–	+/+/+/+/+/+/+/+/+/+	+/+/+/–/–/+/+/–/–/–	–/–/+/–/+/–/+/–/+/+
Gizzard	–/–/–/–/+/–/–/–/–/+	+/+/+/+/–/+/+/+/–/+	–/–/–/–/–/–/–/–/–/–	–/–/–/–/–/–/–/–/+/–
Duodenum	–/–/–/+/–/–/–/–/–/–	+/+/–/+/–/–/+/+/+/–	–/–/–/–/–/–/–/–/–/–	–/–/–/–/–/–/–/–/–/–
Jejunum	–/–/–/–/–/–/–/–/–/–	+/+/–/–/–/–/+/–/+/+	–/–/–/–/–/–/–/–/–/–	–/–/–/–/–/–/–/–/+/–
Ileum	–/–/–/–/+/+/–/–/–/–	+/–/–/–/+/–/–/–/+/+	–/–/–/–/–/–/–/–/+/–	–/–/+/–/–/–/–/–/–/–
Cecum	+/+/+/+/+/+/+/+/+/+	+/+/+/+/+/+/+/+/+/+	+/+/+/+/+/+/+/+/+/+	+/+/+/+/+/+/+/+/+/+

¹Symbols refer to no (–) or detectable (+) β-glucanase activity.

²C−: birds not supplemented with an exogenous enzyme; C+: birds supplemented with an enzymatic commercial mixture; CelA: birds supplemented with Cellvibrio mixtus; CelA-CBM11: birds supplemented with the protein CelA-CBM11.

Taken together the data presented here revealed that the nutritive value of the barley lots used in the 2 experiments differed because the diets reacted differently to the supplementation with exogenous cellulases. In addition, the recombinant enzymes derived from C. mixtus CelA are not appropriate for feed supplementation because of their limited stability at lower pH levels.

Different Barley Lots Expressed Different Levels of Endogenous β-Glucanase Activity

Data revealed that the exogenous enzymes of Rovabio Excel AP were unable to improve the nutritive value of the basal diet of experiment 1. Because diets of experiments 1 and 2 contained more than 60% of barley but presented the same formulation, it is suggested that the observed differential response to enzyme supplementation resulted from the different composition of the 2 barley lots used in experiments 1 and 2. To test this possibility, viscosity, β-glucan, and β-glucanase contents of the 2 barley lots used in the above-described experiments were determined. The data revealed that levels of β-glucan and viscosity were higher in the barley lot used in experiment 2 (barley of experiment 2 contained 3.44% of β-glucan and 1.36 cP viscosity, whereas the barley lot of experiment 1 presented levels of β-glucan and viscosity of 2.26% and 1.25 cP). In contrast, the barley lot used in experiment 1 presented almost 5 times more β-glucanase activity than the barley lot of experiment 2 (589 and 134 U/kg of β-glucanase activity in barleys of experiments 1 and 2, respectively). Thus, the data suggest that a variation exists in the levels of β-glucan and digesta viscosity but particularly of endogenous β-glucanase activity in the barleys used in experiments 1 and 2. These observed variations may explain the differential response to enzyme supplementation, as will be explored below.

Levels of viscosity, β-glucan, and β-glucanase activity were determined in a large range of barley lots to assess the variation in barley composition in relation to those factors (see Materials and Methods section). The data, depicted in Table 6, revealed that considerable variation exists in the levels of β-glucanase activity expressed by the different barley lots, which ranged from >1,300 U/kg to <60 U/kg (in Table 6, barleys B9 and B11 correspond to the cereals used in experiments 1 and 2, respectively). A regression analysis to test the existence of linear effects between the level of β-glucan and viscosity, between the level of β-glucan and β-glucanase activity, and between viscosity and β-glucanase activity was performed. However, the coefficients of determination were 0.019, 0.097, and 0.331, respectively, which indicates a low correlation between the variables. However, β-glucanase activity varies much more widely than the levels of viscosity and β-glucans. Taken together, these observations suggest that endogenous levels of β-glucanase activity and not exclusively the content in β-glucans may affect the nutritive value of barley for poultry.

Viscosity and levels of β-glucan and β-glucanase activity in different barley lots^1,2

Table 6

Viscosity and levels of β-glucan and β-glucanase activity in different barley lots^1,2

Barley no.		Variety	β-Glucan (%)	Viscosity (cP)	β-Glucanase activity (U/kg)
B1		Scarlett	2.96	1.97	554
B2		Braeman	2.83	1.23	930
B3		Braeman	3.52	1.86	569
B4		Braeman	3.14	1.71	1,209
B5		Unk	3.43	1.41	1,010
B6		Unk	2.62	2.12	541
B7		Unk	3.01	1.97	221
B8		Unk	2.89	1.53	339
B9		Scarlett	2.26	1.25	589
B10		Unk	2.35	1.23	672
B11		Unk	3.44	1.36	134
B12		Unk	2.50	1.23	1,381
B13		Unk	2.80	1.46	1,047
B14		Unk	3.02	1.41	1,125
B15		Unk	2.50	1.46	1,304
B16		Unk	3.24	2.05	174
B17		Unk	2.40	1.20	1,202
B18		Unk	3.40	1.61	254
B19		Unk	3.20	1.43	1,024
B20		Unk	3.24	1.46	843
B21		Unk	3.08	1.20	1,064
B22		Scarlett	2.44	1.43	113
B23		Unk	2.58	1.49	287
B24		Braeman	2.68	1.23	407
W1		Esterel	1.65	2.02	108
W2		Regina	2.75	1.69	97
W3		Wintmalt	3.94	1.46	315
W4		Cartel	3.47	1.69	126
W5		Malice	3.41	1.77	95
W6		Metaxa	4.12	1.71	128
W7		Malicorne	3.57	1.84	123
W8		Manureva	3.87	1.89	113
W9		Melodica	3.62	1.64	60
W10		Violetta	3.91	2.00	66
W11		Vanessa	2.95	2.12	111
CS1		Scarlett	3.73	2.04	61
CS2		Prestige	2.99	2.10	78
CS3		Vivaldi	3.11	2.71	234
CS4		Maltasia	3.27	2.25	150
CS5		Kangoo	3.79	2.27	241
CS6		Jolika	3.06	2.12	1,038
CS7		Belgravia	3.34	2.28	157
CS8		Thorgal	3.22	1.99	98
CS9		Cropton	3.07	1.99	296
CS10		Jennifer	2.84	1.82	436
CS11		Signora	2.98	2.22	177
CS12		Calcule	4.38	1.56	188
CS13		Concerto	3.08	2.07	173
CS14		Azalea	3.10	1.64	84
CS15		Primadonna	2.93	1.64	93
CS16		Pewter	3.40	2.04	100
CS17		Margret	3.97	2.33	57
IB1		Quench	3.65	1.84	52
IB2		Clairion	3.60	1.74	267
IB3		Concerto	3.78	2.15	77
IB4		Erlina	3.62	2.27	67
IB5		NFC106–119	3.28	1.87	58
IB6		180/02 5B	3.78	1.77	115
IB7		302/02-G	3.75	2.05	203
IB8		IN0616	3.86	2.68	67
IB9		SYN407–143	1.12	2.50	117
IB10		Charmay	2.39	2.25	133
IB11		Chogun	3.26	1.66	151
IB12		Parigus	3.67	1.38	111

¹Barley B9 was used in experiment 1; barley B11 was used in experiment 2; barleys B13 and B23 were used in experiment 3.

²The Unk varieties are mixtures of several barley varieties. Barley identification numbers with the letter B refer to Portuguese barleys; barley identification numbers with the letter W refer to European winter barleys; barley identification numbers with the letters CS refer to European spring barleys; and barley identification numbers with the letters IB refer to Iberian spring barleys.

Levels of Endogenous β-Glucanase Activity Affect the Nutritive Value of Barley-Based Diets for Poultry

Data presented above suggest that levels of endogenous plant β-glucanases in barley may affect the efficacy of exogenous cellulases used to supplement barley-based diets for poultry. To test this possibility 2 barley lots were selected for a comparative study aiming to evaluate the capacity of an exogenous cellulase mixture to improve the nutritive value of barley-based diets with different levels of endogenous plant enzymes for broilers. The 2 barley lots selected were B13 (presenting high β-glucanase activity; HA) and B23 (presenting low β-glucanase activity; LA), with barley B13 displaying approximately 4 times more β-glucanase activity than barley B23. In contrast, the viscosity and levels of β-glucan were similar in barleys B13 and B23 (Table 6). The chemical composition of the 2 selected barley lots was similar (Table 7), although levels of NDF were higher in barley HA (Table 7). Thus, the formula of Table 1 was used to produce 2 different barley-based diets using either barley B13 or barley B23. The 2 basal diets, alone or supplemented with the commercial enzyme mixture Rovabio Excel AP, were used to feed broiler chicks until d 28 as described in the Materials and Methods section (experiment 3). The results of experiment 3, expressed as final BW, weight gain, feed intake, and feed conversion ratios are summarized in Table 8 (bird mortality was 3.6%). Final BW of birds fed the HA basal diet were significantly higher than those of animals fed the LA diet. In contrast with what was observed in the 2 preceding experiments, final BW of birds fed HA diets was similar to what is expected in standard commercial conditions, confirming that barley B13 displayed a higher nutritive value when compared with the other barleys used in this study. The addition of exogenous β-glucanases had no effect on the weight gain of animals receiving the HA diet. In contrast, addition of exogenous enzymes to diet LA significantly improved bird final BW, although birds did not reach the final weight of birds receiving the HA diets. No variations were found in feed intake among the 4 groups, suggesting that barley source and exogenous enzymes lead to different efficiencies of nutrient utilization rather to an increase or decrease of feed intake. This was confirmed by analyzing the data on feed conversion ratios, which were substantially smaller for the groups receiving the HA diet and the group receiving the LA diet supplemented with the exogenous enzymes when compared with the animals receiving the LA basal diet not supplemented with microbial enzymes. Taken together, the data suggest that levels of endogenous β-glucanase activity may have a major effect on the nutritive value of barley-based diets. Hence, addition of exogenous β-glucanases may be effective only in the case of barley-based diets containing lower levels of endogenous plant cell wall-degrading activity, such as barley B23. Notwithstanding the suggested implications of endogenous plant enzymes in the effectiveness of exogenous enzymes used to supplement barley diets, it is clear that diets used in experiment 3 may have presented unaccounted variations that may have affected the nutritive value of the 2 barley lots. Clear proof of the concept would be possible by comparing the nutritive value of a barley batch similar to B13 before and after specifically inactivating the endogenous enzymes (for example, through heating). However, although it can be anticipated that endogenous plant β-glucanases are thermolabile, it is clear that a relative thermostability exists because diets used in this study were subjected to pelleting, where temperatures can reach temperatures of 80°C. However, it is also expected that a considerable degree of enzyme substrate protection exists, which contributes to enhance the stability of the endogenous enzyme (Fontes et al., 1995).

Chemical composition of the 2 barleys used in experiment 3¹

Table 7

Chemical composition of the 2 barleys used in experiment 3¹

Chemical composition (% unless noted)	Barley HA (B13)	Barley LA (B23)
DM	11.65	13.66
Crude energy (MJ/kg of DM)	16.30	16.05
CP	9.15	9.00
Ash	2.32	1.90
NDF	25.73	20.35
ADF	4.99	5.70
Acid detergent lignin	1.56	1.57

¹Barley HA = barley with high β-glucanase activity. Barley LA = barley with low β-glucanase activity. See Table 6 for explanation of B13 and B23.

Performance of broilers fed 2 barley-based diets supplemented or not with a commercial β-glucanase preparation (data from experiment 3)

Table 8

Performance of broilers fed 2 barley-based diets supplemented or not with a commercial β-glucanase preparation (data from experiment 3)

Item	Treatment1				SEM	p(F)2
	HA−	HA+	LA−	LA+
BW (g)
0 d	48.06	47.72	47.71	48.19	0.248	0.42266
7 d	154.46ab	159.71a	131.19c	144.21b	4.934	0.00130
14 d	404.16a	421.30a	293.93c	361.25b	11.457	<0.0001
21 d	784.67a	811.52a	576.74c	693.33b	21.923	<0.0001
28 d	1,323.00a	1,304.57a	992.65c	1,163.93b	30.847	<0.0001
Weight gain (g)
0–7 d	106.38ab	111.93b	83.95c	96.14ac	5.006	0.0019
7–14 d	249.68a	261.16a	162.29b	217.27c	7.735	<0.0001
14–21 d	378.88c	390.24c	282.85a	332.08b	12.826	<0.0001
21–28 d	533.58c	493.08c	415.92a	470.62b	14.134	<0.0001
0–28 d	1,274.86c	1,256.85c	944.95a	1,115.86b	30.922	<0.0001
Feed intake (g)
0–7 d	141.74	144.29	152.95	132.9	10.202	0.5854
7–14 d	412.14	394.43	348.51	373.27	24.524	0.3051
14–21 d	721.13b	681.82ab	596.38a	662.38ab	37.420	0.1406
21–28 d	986.53	925.78	854.71	867.86	55.615	0.3319
0–28 d	1,957.88	1,888.00	1,952.54	1,875.27	79.271	0.8307
Feed conversion ratio
0–7 d	1.27a	1.25a	1.83b	1.40a	0.063	<0.0001
7–14 d	1.50a	1.47a	2.23b	1.72a	0.095	<0.0001
14–21 d	1.75a	1.70a	2.15b	2.01b	0.086	0.0016
21–28 d	1.69a	1.82a	2.08b	1.86ab	0.090	0.0333
0–28 d	1.67a	1.66a	2.09b	1.84a	0.068	0.0002

^a–cMeans within the same row with different superscripts are significantly different (P < 0.05).

¹HA−: barley-based diet displaying high endogenous β-glucanase activity not supplemented with an enzymatic commercial mixture of cellulases; HA+: barley-based diet displaying high endogenous β-glucanase activity supplemented with an enzymatic commercial mixture of cellulases; LA−: barley-based diet displaying low endogenous β-glucanase activity not supplemented with an enzymatic commercial mixture of cellulases; LA+: barley-based diet displaying low endogenous β-glucanase activity supplemented with an enzymatic commercial mixture of cellulases.

²p(F) < 0.05 is statistically significant.

The effects of the different dietary treatments on the relative length or weight of different organs and GI tract compartments of broiler chickens of experiment 3 were evaluated; the data are presented in Table 9. The data revealed lower relative lengths and weights for the duodenum, jejunum, and ileum of birds receiving the HA diet when compared with birds receiving the LA treatment. In addition, supplementation of the LA diet with β-glucanase activity reduced organ sizes. As discussed above, these data are in agreement with results reported by several other authors who have shown that exogenous enzymes decrease digestive viscosity affecting the digestive tract weight, or length, or both, when expressed as a percentage of live weight (Brenes et al., 1993; Peterson et al., 1993; Viveros et al., 1994). Digesta viscosity in the hindgut and foregut of birds fed the different dietary treatments was measured; the data are presented in Table 9. Duodenum and jejunum viscosities were higher in digesta samples from birds receiving the LA diet not supplemented with exogenous enzymes when compared with the 3 other groups. Therefore, the data suggest that endogenous and exogenous enzymes have a similar capacity to reduce the degree of polymerization of soluble glucans, thus contributing to a considerable reduction in digesta viscosity. Because only 1 animal/pen was slaughtered for sample collection, as a result of logistic limitations, these values and those reported for experiments 1 and 2 should be viewed with some caution considering that variation among animals for the measured variables can be significantly high.

Relative weight and length of the gastrointestinal tract and viscosity of digesta samples of broilers fed 2 barley-based diets supplemented or not with exogenous β-glucanases (data from experiment 3)

Table 9

Relative weight and length of the gastrointestinal tract and viscosity of digesta samples of broilers fed 2 barley-based diets supplemented or not with exogenous β-glucanases (data from experiment 3)

Item	Treatment1				SEM	p(F)2
	HA−	HA+	LA−	LA+
Relative weight (g/100 g of BW)
Crop	3.51	3.07	3.91	3.53	0.231	0.1062
Gizzard	15.98ab	17.09ac	14.71bcd	13.03d	0.866	0.0141
Relative length (cm/kg of BW)
Duodenum	20.85ab	19.91a	27.16c	22.07b	0.749	<0.0001
Jejunum	55.04a	54.50a	68.82b	56.15a	2.302	0.0002
Ileum	55.34a	54.63a	75.28b	57.68a	2.266	<0.0001
Cecum	12.72a	12.62a	16.01b	13.90a	0.530	0.0002
Content viscosity (cP)
Duodenum + jejunum	3.463b	2.813b	8.612a	3.596b	0.437	<0.0001
Ileum	6.509b	7.836b	4.998b	12.678a	0.853	0.0019

^a–cMeans within the same row with different superscripts are significantly different (P < 0.05).

¹HA−: barley-based diet displaying high endogenous β-glucanase activity not supplemented with an enzymatic commercial mixture of cellulases; HA+: barley-based diet displaying high endogenous β-glucanase activity supplemented with an enzymatic commercial mixture of cellulases; LA−: barley-based diet displaying low endogenous β-glucanase activity not supplemented with an enzymatic commercial mixture of cellulases; LA+: barley-based diet displaying low endogenous β-glucanase activity supplemented with an enzymatic commercial mixture of cellulases.

²p(F) < 0.05 is statistically significant.

Various authors have suggested that a range of factors such as animal age, microbial challenge, cereal genotype, cereal growing conditions, or cereal storage length (Fuente et al., 1998; Svihus and Gullord, 2002) may contribute to the reduction of effectiveness of exogenous β-1,3–1,4-glucanases used to supplement poultry diets. However, the molecular mechanisms underlying this lack of response, in a variety of circumstances, remain unknown. It is clear that if levels of β-glucans are smaller the resulting digesta viscosity will be lower and consequently enzymes have no margin to function (Campbell et al., 1989; Choct, 2006). In addition, it is known that cereal genotype, growing conditions, and storage length affect the levels of barley β-glucan. Data presented here revealed that barley may also express significant but variable levels of endogenous β-glucanases. The plant enzymes seem to be active under the conditions of the GI tract and thus contribute to a significant depolymerization of the antinutritive β-glucans. For example, plant β-glucanases of barley B13 were sufficient to decrease duodenal viscosity of nonsupplemented birds to the levels observed in animals supplemented with the exogenous enzyme. Thus, endogenous enzymes may contribute to affect microbial enzyme effectiveness by reducing substrate availability to the exogenous enzymes. The role of endogenous β-glucanases in plant metabolism remains relatively unexplored. During germination of barley grains, the cell walls of the starchy endosperm are degraded by β-1,3–1,4-glucanases. This allows the amylases and proteases secreted from the surrounding aleurone or scutellar tissues to reach their starch and storage protein substrates inside the endosperm cells (Fincher et al., 1986). The activation of β-glucanase activity most possibly contributes to improve the nutritive value of soaked or germinated barley (Svihus et al., 1997). In these examples, the activity of the endogenous enzymes reduces the degree of polymerization of the antinutritive β-glucans, resulting in an improvement of the diet nutritive value. However, data presented here suggest that the basal levels of expression of barley cellulases and β-1,3–1,4-glucanases before the germination process starts may vary and can reach significant values in a range of barley lots.

The molecular structure of barley cellulases remains to be identified. In addition, the barley genome sequence, which could reveal the complete set of plant cell wall hydrolase genes encoded by this cereal, is presently unknown. However, work developed in the last 20 yr led to the cloning of 2 β-1,3–1,4-glucanase encoding genes (Fincher et al., 1986). In addition, 12 expressed sequence tags, which contain plant cell wall hydrolase genes, were identified in various plant tissues (data obtained from GenBank; http://www.ncbi.nlm.nih.gov/genbank/). Cellulase and β-1,3–1,4-glucanase activities are of utmost importance during the germination of cereal grains to allow the degradation of cell walls of the nonliving starchy endosperm (Fincher et al., 1986). However, levels of expression of genes encoding plant cell wall hydrolases in barley grains are presently unknown because most studies have been performed in germinating seeds. Therefore, more work is required to characterize the expression profiles of genes encoding plant cell wall-degrading enzymes in grains, particularly of β-1,3–1,4-glucanases and cellulases. Differences in the expression pattern of the different genes should be evaluated with cereal lots from different origins. The collected data may contribute to the development of novel targets for barley breed improvement. Clearly the development of barley varieties expressing higher levels of endogenous cellulase activity would contribute to improve the nutritive value of the corresponding barley-based diets for monogastric animals. The development of novel barley varieties with increased β-1,3–1,4-glucanase activity could eventually contribute to preclude the requirement for feed supplementation of barley-based diets in monogastric nutrition.

Conclusions

Data presented here revealed that endogenous β-glucanases may be present in moderate levels in barley seeds and may remain active throughout the GI tract, thus contributing to reduce digesta viscosity and consequently the effectiveness of exogenous cellulases used to supplement poultry feed. In addition, the results revealed that levels of endogenous β-glucanases vary largely between different barley lots, a variation that is more pronounced than the levels of β-glucans. Therefore, a variable response to enzyme supplementation may result, among other factors, from the variable levels of β-glucanase activity present in barley grains. The implementation of a quick assay to detect the levels of both barley β-glucan and β-glucanase may help rationalize the supplementation of poultry feed with exogenous enzymes.

ACKNOWLEDGMENTS

We thank Sociedade Agrícola da Quinta da Freiria SA (Bombarral, Portugal) for supplying the 1-d-old broilers used in these experiments and Central de Cervejas (Vialonga, Portugal) for kindly donating the barley used in experiment 3. This work was supported by Fundação para a Ciência e a Tecnologia (Lisboa, Portugal; grants PTDC/CVT/69329/2006 and PTDC/CVT/103942/2008). Patrícia I. P. Ponte and Teresa Ribeiro were supported by Fundação para a Ciência e a Tecnologia through individual fellowships SFRH/BD/17969/2004 and SFRH/BD/32321/2006, respectively.