The effect of arabinoxylanase and protease supplementation on nutritional value of diets containing wheat bran or rice bran in growing pig *

One digestibility and one performance experiments were conducted to investigate the effects of a feed enzyme of arabinoxylanase and arabinoxylanase plus protease on digestibility of nutrients, nitrogen (N) fl ow and performance of the growing pigs fed a maize-wheat bran-based diet or a maize-rice bran-based diet. In the digestion experiment, six Landrace × Large White × Chinese Black male pigs of approximately 20 kg initial body weight , fi tted with a simple T-cannula at the terminal ileum, were used to study dietary nutrient digestibility and nitrogen fl ow of the experimental diets. The six experimental diets were prepared in this study: maize-wheat bran diet, consists of maize, wheat bran, soyabean, of rapeseed meal, without feed enzyme supplementation (MWB); MWB supplemented with arabinoxylanase (A); MWB supplemented with arabinoxylanase plus protease (AP); maize-rice bran diet, consists of maize, rice bran, soyabean, rapeseed meal, without feed enzyme supplementation (MRB); MRB supplemented with arabinoxylanase (A); MRB supplemented with arabinoxylanase plus protease (AP). The MWB diet contained: %: soluble nonstrach polysaccharide (NSP) 0.8, insoluble NSP 19, crude protein 22, and 18.6 MJ/kg gross energy (GE), while the MRB diet contained, %, respectively: 0.5, 22, 21, and 18.9 MJ/kg GE. In the performance experiment, lasting 40 days, 60 pigs (30 gilts and 30 castrates) with body weight of 20.5±0.76 kg were randomly allocated to the six dietary treatments with a completely randomized design (5 gilts and 5 castrates per treatment). Feed and water were provided ad libitum. 446 DIETS SUPPLEMENTED WITH ENZYME FOR PIGS The results show there were no signifi cant differences in average daily gain (ADG) (P=0.09), feed gain ratio (F/G; P=0.09), ileal digestibility of nutrients (P=0.09) and N fl ow (P>0.08) between the MWB and MRB diet, although the crude fi bre level in the MRB diet had higher crude fi bre content (9.7 g/kg) than that the MWB diet. The dry matter intake (DMI) of the MWB diet was increased (P<0.05) by the enzyme A supplementation. The ADG and F/G of the pigs fed the MWB diet were improved (P<0.05) by either arabinoxylanase or arabinoxylanase plus protease supplementation. The ileal endogenous N excretion determined with enzymatic hydrolysed casein diet in this experiment was half higher than that calculated from the N-free diet. The ileal digestibility of soluble NSP fractions of the MWB and MRB diets were signifi cantly (P<0.05) increased by either enzyme arabinoxylanase or arabinoxylanase plus protease supplementation. There was no enzyme effect on insoluble NSP fraction digestibility. The results of this study suggest that NSP degrading enzyme do mainly reduce the negative nutritional effect of the soluble NSP in the small intestine of growing pigs.


INTRODUCTION
Rice bran and wheat bran are both widely used in pig diets, especially in Asia, but digestibility and animal performance experiments have shown that inclusion of either of the bran reduces feed intake (Chen, 1981), decreases ileal and overall digestibility (Chen 1979;Graham et al., 1986), and results in poor growth rates (Hanrahan and O'Grady, 1970;Chen, 1981).Recent studies on the cereal shows that plant cell walls contain primarily complex carbohydrates referred to as nonstarch polysaccharides (NSP), which can negatively affect nutrient utilization (Englyst and Hudson, 1996).The NSP constituents of the endosperm cell walls of wheat and barley are mainly arabinoxylans and barley also contains high levels of ß-glucans (Yin et al., 2000a,b,c;2001a,b).Many studies have investigated ways of removing these polymers, which encapsulate the desired nutrients in the endosperm.However, feed processing techniques such as extrusion cooking, gamma irradiation, micronization and fl aking did not signifi cantly improve nutrient utilization (McClean, 1993).
Xylanase enzyme, in combination with ß-glucanase has been reported to improve hulless barley apparent ileal digestibility of nutrients (Yin et al., 2000a;2001a,b) and growth performance of young pigs (Baidoo et al., 1998).However, the effects of feed enzyme on nutrient utilization and nitrogen fl ow through the terminal ileum and excretion in faeces in growing pigs fed rice or wheat bran based diet were seldom reported in the literature.The objects of this study were to determine the effects of carbohydrase and protease supplementation of the maizerice bran and maize-wheat bran-based diets on apparent and true digestibility of nutrients, nitrogen fl ow and performance of growing pigs.YIN Y.-L.ET AL.
After surgery, the pigs were housed individually in stainless steel metabolic crates in a temperature-controlled barn (20 to 22 o C).During a 13-d recovery period, the pigs were fed a commercial starter diet containing 18% of protein.
The experiment was designed and conducted according to a 6 × 6 Latin square design.After recovering from surgery, the pigs were offered one of the 5 diets (Table 1, diet MRB + AP was excluded) at 08:00 and 18:00 daily as a wet mash.The sixth diet was of the enzyme hydrolysed casein (EHC) diet which was used for determination of the ileal endogenous nitrogen (N) and amino acid (AA) fl ow.Chromic oxide was incorporated into all of the diets as an indigestible marker.The EHC diet was formulated with g/kg of: enzyme hydrolysed casein 150, (MW>5000Da, New Zealand Pharmaceuticals Ltd. Palmerston North, New Zealand), rapeseed oil 35, cellulose 50, sucrose 70, maize starch 629 and chromic oxide 2, and vitamin and mineral mixture 64 (for composition see Table 1).Pigs were fed at 2.6 times of their maintenance energy requirement (NRC, 1998) based on their BW at the beginning of each experimental period.Water was added to each meal in a 2:1 ratio.Additional water was available from a nipple drinker.
The experimental period lasted 14 days each.Pigs were allowed to adapt to their respective experimental diets for 7 d and then 3 days of faeces collection.On the 11 th day, ileal digesta samples were collected continuously for 24 h using plastic bags (length, 16 cm; i.d., 10 cm) attached to the cannula.The bag contained approximately 10 ml of a formic acid solution (10%, v/v) to minimize further microbial activity.The bag was removed and replaced as soon as it was partially fi lled with digesta.Digesta were immediately frozen at -20 o C following collection.

Experiment 2. Performance experiment
Sixty Landrace × Large White × Chinese Black pigs (30 gilts and 30 barrows) with body weight of 20.5±0.76 kg were randomly allocated to the six dietary treatments (Table 1) with a completely randomised design (5 gilts and 5 barrows per treatment).The pigs were housed individually in fl oor pens (1.2 m × 1.3 m) in an environmentally controlled room and at an environmental temperature of at least 18ºC.The pens were fi tted with fl ooring made of plastic covered woven iron and stainless steel single-space feeders and nipple drinker; pen partitioning were made of metal bars.Feed and water were provided ad libitum.
Individual pig weights and pen feed intake were determined weekly.If feed was dry (not contaminated) the weight was recorded directly.If contamination had occurred the feed was dried and weighed for correction of feed intake.The experimental period lasted 40 days.

Chemical analysis
After the conclusion of the animal trial, faeces and digesta were freeze-dried, pooled for each within animal and period, ground through a 1.00-mm mesh screen, and mixed before analyses.
For the digesta collected from the EHC diet were thawed and pooled for each pig over the each collection periods.A sub-sample of approximately 100 g was taken from each digesta pool, and centrifuged at 7000 g for 11 min.The supernatant fractions were than ultra fi ltered using Centriprep YM-10 (Amicon-Millipore, Beyerly, MA, USA; Cat.No. 4305, 10000 Da MM cut-off) according to the manufacture's instructions.The precipitate from the centrifugation step was added to the retentate from the ultra fi ltration, and the material freeze-dried and fi nely ground.
Dry matter (DM), gross energy (GE), crude protein (CP, N × 6.25, macro-Kjeldahl), crude fi bre and crude fat contents were determined according to AOAC (1990).Samples were also analysed for chromic oxide (Fenton and Fenton, 1979).Amino acid (AA) contents were determined by ion exchange chromatography following hydrolysis in 6N hydrochloric acid at 110 o C for 24 h.Methionine and cysteine were determined as methionine sulphone and cysteic acid after oxidation with performic acid.Soluble and insoluble NSP sugars and uronic acids in diets were determined by gas chromatography (Pye Unicam 304) according to Englyst and Cummings (1984).

Calculations and statistical analyses
The apparent digestibility of energy, crude protein and AA were calculated as: where D D is the apparent digestibility of a nutrient in the diet; I D is the marker concentration in the diet; A F is the nutrient concentration in faeces or ileal digesta; I F is the marker concentration in faeces or ileal digesta; and A D is the nutrient concentration in the diet.All variables were expressed in percentages.
The ileal true digestibility (TID) of CP and AA were calculated as: where AA E is the gut endogenous AA or N losses expressed as g/kg dry matter intake determined with the EHC or protein free diet reported by CVB (1998) and the AA E was calculated as: The results of the growth performance were analysed as a completely randomised design, with a 2 × 3 factorial treatment, in which animals were randomly assigned to the MWB or MRB diet, with or without enzyme A or enzyme B supplementation.Differences between means were compared using the Student-Neuman-Keuls (SNK) method at a signifi cance level of P<0.05.Data from the digestion trial were fi rst subjected to analysis of variance (ANOVA).Sources of variation were diets (n=6), periods (n=6) and pigs (n=6).Where appropriate, treatment means were compared using the Student-Newman Keuls' multiple range tests.The ANOVA, comparisons of means, and regression analyses were carried out using the General Linear Model Procedure of SAS Institute, Inc. (1988).

RESULTS
The proximate analysis of a number of chemical constituents of the wheat bran and rice bran diets, including AA is shown in Table 2.The CP and most AA were slightly higher for the wheat bran diet.Conversely, the crude fi bre and insoluble NSP were higher for rice bran diet, although the soluble NSP were lower for the rice bran diet.
The effect of dietary treatments on the pig performance is shown in Table 3. Pigs fed the MWB diet supplemented with either the arabinoxylanase (MWB+A) had a higher feed intake (P<0.05),higher daily gains (P<0.05) and better feed utilization (P<0.05)than pigs fed unsupplemented diet.Additional supplementation of protease (diet MWB+AP) did not improve daily gains or feed utilization, but lowered the feed intake to the MWB diet level.Enzyme treatment did not improved (P>0.05) the performance of pigs fed the MRB diets and there were no differences (P>0.05) in performance between pigs fed the MWB diet and MRB without or enzyme supplemented diets.
The effects of dietary treatments on the total fresh faeces excretion, total and ileal endogenous N excretion is shown in Table 4. Enzyme supplementation reduced (P<0.05) the fresh faeces and total N fl ow (P<0.05).The endogenous N fl ow, calculated according the Centraal Veevoeder Bureau (1998) data based on N-free diet, was less than half of endogenous N determined in our experiment using EHC diet.
The digestibility of DM, GE, CP, NSP and AA are given in Tables 5 and 6.Pigs fed the MWB diet supplemented with either arabinoxylanase or arabinoxylanase plus protease had a higher (P<0.05)apparent ileal digestibility of DM, GE, CP and apparent and true ileal digestibility of arginine, histidine, isoleucine, lysine and threonine than the pigs fed the other diets.Enzyme supplementation increased (P<0.05) the ileal digestibility of soluble NSP fractions, however, there were no enzyme effects (P>0.05) on digestibility of insoluble NSP.The digestibility of DM, GE, CP and AA were not improved (P<0.05) by addition of arabinoxylanase to the MRB diet.There were no signifi cant differences in total digestibility between the unsupplemented an supplemented MWB and MRB diets.
Value of DE at the ileal (IDE) and total tract (TDE) levels and true ileal digestible CP and essential AA contents are presented in Table 7. Pigs fed the MWB diet supplemented with either the enzymes A or AP had a higher (P<0.05)IDE, CP, arginine, histidine, isoleucine, lysine and threonine than the pigs fed other diets.There were no difference (P<0.05) in DE and true ileal digestible CP and AA contents among the other diets.The effect of enzyme supplementation on hindgut fermentation of nutrients is shown in Table 8.The hindgut fermentation of CP and soluble NSP, expressed either by subtracting overall from ileal digestibility or as a proportion of undigested material reaching the hindgut, were reduced (P<0.05)by enzyme supplementation, but the hindgut fermentation of insoluble NSP was not affected by enzyme treatment and there was no difference (P<0.05) in hindgut fermentation between the MWB diet and the MRB diet.

DISCUSSION
Although the crude fi bre content for the MRB diet was 24% higher than for the MWB diet, ADG, average daily feed intake (ADFI), F/G, DE and ileal true digestible AA for the MRB diet were similar (P=0.08-0.09) to those for the MWB diet.This response may be due to an inadequate number of replicates being conducted.Chen (1981) found that there were no differences in feed to gain or DE to gain ratio between rice bran-and wheat bran diets, when both were added at a rate of 20% into a maizesoyabean meal-based diet for growing-fi nishing pigs.Yin et al. (l993) reported that the nutritional value of rice bran mainly depends on the rice milling procedure.The rice bran used in this trial comes from the rice fi ne milling process.In this procedure, the outside ring of rice hull was at fi rst dehulled and taken away and the bran come from the second rice milling.The rice bran from the fi ne milling showed a higher ileal digestibility of crude protein (73%) and total amino acids (74%) than rice bran from rough rice milling (56 and 64%, respectively).However, both rice bran and wheat bran increase gut fi ll and also infl uence other physiological actions such as transit time, nutrient absorption and digestive secretions (Chen, 1979;Graham et al., 1986;Knudsen and Hansen, 1991).Kass et al. (1980) reported that a level of 400 g/kg lucerne meal, resulting in a concentration of 311 g/kg NDF in the diet for growing pigs, signifi cantly affected the daily gain and feed conversion effi ciency compared with the control diet (228 g/kg of NDF).The average daily gain was reduced from 700 to 520 g and the feed conversion effi ciency increased from 3.23 to 4.91.
In view of the results reported in the literature (Taverner and Campbell, 1988;Van Lunen and Schulze, 1996;Baidoo et al., 1998;Yin et al., 2000a,b), it was expected that inclusion of enzymes would increase the DE and available AA for the MWB-based diet or MRB-based diet.Since the mechanism of such improvement seems to involve breakdown of the cell wall non-starch polysaccharides (NSP) this would have possible benefi ts in terms of reducing viscosity, as seen in poultry (Bedford and Classen, 1992;Campbell and Bedford, 1992) and to a lesser extent in pigs (Inborr et al., 1993).Van Lunen and Schulze (1996) found a positive effect of xylanase addition on growth rate and F/G, with an improvement of 9.2 and 5.3%, respectively, regardless of the level of wheat and maize inclusion in diets for weaned pigs.In the present study, enzyme supplementation only improved the performance for the pigs fed the wheat bran, which had higher soluble NSP fractions than the rice bran.Similar to the performance data, the digestibility of nutrients was improved by the enzyme supplementation only with the wheat bran diet.However, enzyme supplementation improved the soluble NSP fraction digestibility for wheat and rice bran diets, although there were no effects on digestibility of insoluble NSP fractions.These results demonstrated that the enzymes used in this study can only breakdown the soluble NSP fractions of the bran as reported by Marquardt et al. (1994).The reason for the no enzyme effect on the performance and digestion in pigs fed the MRB diet may be that there is less scope for enzyme action, since the total soluble NSP in the MRB diet was much lower than that of the MWB diet (4.38 vs 6.76 g/ kg).This indicated that the response of enzyme mixture supplementation to wheat bran diet appears to result largely from the functions of arabinoxylanase, rather than protease.However, it is clear from this study that arabinoxylanase addition has reduced faeces and N excretion in faeces and shifted part of digestion from the hindgut to the small intestine, which results in less environmental pollutions and less nutrients available for bacterial fermentation in the hindgut.Undigested soluble fi bre may exacerbate such diseases as non-specifi c colitis (Taylor, 1999) and, more recently, Pluske et al. (1996) have shown that soluble fi bre may provoke Serpulina hyodysenteriae.Hence, controlling hindgut microbial activity, particularly through nutrimental means, may be of increasing relevance as the use of in-feed antibiotics is more and more constrained.
Recent studies suggest that endogenous gut N losses are higher than previously estimated and that the differences in apparent digestibility between feedstuffs are attributed to difference in endogenous gut N losses rather than to differences in true digestibility (Moughan and Rutherfurd, 1990;Marty et al., 1994).Traditionally, the endogenous loss of N and AA has been determined following protein free alimentation.This method can be criticised, however, due to the physiologically abnormal nature of the protein free state.There are at least two major potential problems with this approach.First, the rate of whole body protein metabolism will decrease when animals are fed a protein free diet (Moughan and Rutherfurd, 1990) and this may lead to a lowered secretion of proteinaceous material into the gut.In addition to this direct effect of the protein-deplete metabolic state, the absence of stimulation from dietary peptides or protein, per se, may lead to a reduced secretion of protein into the gut.The present study with the EHC method confi rmed these conclusion and therefore it would be useful to determine ileal endogenous AA follow with the EHC method for correcting the apparent ileal digestibility of AA to true ones.
In summary, it can be concluded that the rice and wheat bran used in this study have the similar feeding values for growing pigs.The soluble NSP in the wheat and rice bran diet can be partly hydrolysed in the gastro-intestinal tract of pigs by arabinoxylanase supplementation of feeds.

TABLE 2
Proximate analysis of maize wheat bran and maize rice bran diets and contents of essential amino acids and non-starch polysaccharide (NSP) contents, g/kg DM

TABLE 5
Ileal apparent and true digestibility (%) of the essential amino acids in pigs fed the maize-wheat branbased diet (MWB), that diet supplemented with arabinoxylanase (MWB+A) or with arabinoxylanase + protease (MWB +AP) and maize-rice bran diet (MRB), that diet supplemented with arabinoxylanase (MWB+A) and arabinoxylanase + protease (MWB +AP) calculated from the enzyme hydrolysed casein based diet method a,b means in the same row with difference superscripts differ at P<0.05 YIN Y.-L.ET AL.