Introduction
Soybean meal (SBM) is the most important and preferred source of quality protein in animal feeds (Banaszkiewicz, 2011). Among the oil seeds, SBM contains the highest crude protein (CP) and has the best AA profile, with only 6% crude fibre (Dei, 2011). The AAs in SBM are highly digestible by poultry (Newkirk, 2010). Based on the CP content, SBM can generally be classified as high protein (47% - 50%), or low protein (44% - 46%). Popescu & Criste (2003) suggested that full-fat SBM could completely replace conventional commercial SBM and supplemental oils in diets for broiler chickens, but these full-fat products require heat treatment. Untreated soybeans contain several anti-nutritional factors (ANFs), including trypsin inhibitors, lectins and phytate (Liu, 1998; Newkirk, 2010). Trypsin inhibitors and lectins inhibit protease and other enzymes in the digestive tract, particularly the activity of trypsin and chymotrypsin, and contribute to reducing protein digestion in non-ruminant animals and young ruminants (Leiner, 1994).
Birds that are fed on diets that contain raw soybean experience low FI, slow body growth and increased weight of pancreas (Mogridge et al., 1996; Perez-Maldonado et al., 2003; Newkirk, 2010). In recent studies, the authors examined the potential of improving the quality of diets containing RFSB for broiler chickens through supplementation with high-potency protease and phytase products. The enzymes reduced the negative impact of RFSB, although productivity was still poorer than that observed with a control diet that was devoid of RFSB (Erdaw et al., 2017a; 2017b). In previous studies, protease alone, and especially in combination with phytase, improved the in vitro digestion of protein (Erdaw et al., 2016). The phytate content of RFSB was also reduced. The mechanisms behind these in vitro and in vivo tests were not fully understood, particularly in the absence of a control diet that contained RFSB without microbial enzyme supplementation. The direct effect on soy protein of protease or in combination with phytase was not assessed either. Therefore, the objectives of this study were to evaluate the attributes of individual and combined (synergistic) effects on protein digestibility and growth of broiler chickens of microbial protease and phytase in diets containing RFSB. The authors hypothesized that there would be a reduction in the negative effects of ANFs in soya, trypsin inhibitors (TI) on birds through supplementation with microbial protease and/or phytase enzymes.
Materials and Methods
The seeds of RFSB were purchased from a local supplier in northern New South Wales, Australia, cleaned and hammer-milled to pass through a 2-mm sieve. This milled ingredient was then incorporated in diets, as described below. The samples and diets were further milled and then analysed for CP content (AOAC, 2006a), amino acid (AA) profiles (AOAC, 2006b) and available lysine contents (AOAC, 2006b), urease activities (UA) (AOCS, 2011a), TI concentrations (AOCS, 2011b) and protein solubility (KOH) (Araba & Dale, 1990).
A sub-sample of the RFSB was milled to pass through a 0.2-mm screen. About 1.0 g of this sample was incubated with 0.2 mg microbial protease, 0.2 mg phytase or a combination (0.2 mg each) of the two enzymes to determine the in vitro digestibility of protein in a two-step enzymatic procedure (Boisen & Fernández, 1997; Park et al., 2012). Additionally, the material was incubated in an enzyme-free medium.
The RFSB that was milled to 2-mm particle size was used to replace commercial SBM at 25% (7.5% of diet) that was fed as such (NC) or supplemented with microbial protease (0 or 15000 PROT/kg) (NC-Pro), phytase (0 or 2000 FYT/kg) (NC-Phy). A positive control (PC) diet was prepared without RFSB or enzyme. The enzyme recovery rates are shown in Table 1.
Table 1 Recovery rates of test enzymes in diets (containing 75 g of raw, full-fat soybean per kg of diet (grower diet 10 - 24 days) or raw, full-fat soybean used as a sole source of crude protein and amino acids (special diet 25 - 31 days)
In general, the recovery rate of protease was slightly lower than calculated, with a coefficient of variation (CV) of between 3% and 11%. Phytase recovery was slightly higher than calculated (CV, 9 - 13%).
A positive control (PC) diet without RFSB or enzyme was also prepared (Tables 2 and 3). The diets were offered to birds in crumbled form for starters (0 - 10 days), in pelleted form for growers (11 - 25 days) and as mash (26 - 31 days). Each diet was replicated six times, with nine birds per replicate. A completely randomized design with a 2 x 2 +1 arrangement was used in both phases. Three sets of tests were conducted in this study:
1. In vitro nutrient digestibility test on RFSB
2. A feeding trial between zero and 25 days (Phase 1)
3. A feeding trail between 26 and 31 days (Phase 2)
A total of 270 one-day-old Ross 308 male broiler chicks (initial bodyweight 40.54 ± 0.18 g) were obtained from a local commercial hatchery (Baiada Poultry Pty., Ltd., Tamworth, Australia). Groups of nine chicks were allocated to each of 30 cages (60 x 42 x 23 cm). The birds were raised in a climate-controlled room and were offered corn-soybean-based diets (starter and grower) formulated to Aviagen standards for Ross 308 broilers (Aviagen, 2014).
Table 2 Ingredient composition (g/kg) of starter, grower and special diets of broiler chickens
Table 3 Composition (g/kg) of starter (0 - 10 days), grower (10 - 26 days), and special (26 - 31 days) diets for broiler chickens
Each cage was equipped with a medium-sized feeder (78 x 12 x 8 cm) and two nipple drinkers. The temperature of the room was set at 33 ºC for the first two days, with a relative humidity of between 49% and 60%. This temperature was gradually reduced to 24 °C by day 19, and maintained at this level for the remaining period. For the first two days, 24 hours of light (20 lux) were provided. This was then reduced to 23 hours for the next six consecutive days, followed by 20 hours of light (10 lux) for the remaining days. Feed was provided ad libitum in the form of crumble for starters and pellets for growers and special diets in phase 2, and the birds had free access to water.
One and two birds were randomly selected on days 10 and 25, respectively, and electrically stunned and killed by cervical dislocation to assess the weights of the pancreas and duodenum. Samples of ileal digesta were also collected on day 25 to assess protein and amino acid digestibility.
In the second phase, birds (6 per replicate) were supplied with special diets, which contained commercial SBM or RFSB as the sole source of protein and amino acids (Tables 2 and 3). These diets were fed continuously for five days, that is, between 26 and 31 days old. As in the first phase, there were five diets, namely NC, NC-Pro, NC-Phy, NC-Enz (combined) and PC. The main objective of this phase was to assess the digestibility of RFSB protein in the presence of the enzymes. Titanium dioxide (TiO2) was added to grower and phase 2 diets as a marker to enable assessment of ileal nutrient digestibility. The pancreas and duodenum were collected and weighed at 25 and 31 days old. The concentration of TiO2 in the ileal digesta and diets was determined using the method described by Short et al. (1996).
At sampling on days 10 and 25, blood was collected into EDTA-coated tubes to assess blood composition and liver function. The blood was collected from the right jugular vein, shaken gently, and placed on ice. The sample was centrifuged at 3000 x g for 10 min to obtain the serum. Serum samples were analysed at Laverty Pathology (Armidale, NSW, Australia) for aspartate aminotransferase (AST) (EC 2.6.1.1) and alanine aminotransferase (ALT) (EC 2.6.1.2). The methods were based on those recommended by the International Federation for Clinical Chemistry (IFCC), which were developed by Shumann et al. (2002).
Colorimetric methods were used to evaluate the recovery rate of exogenous protease enzymes in the mixed diets (Keltti, 2014). Evaluation of the recovery rate of exogenous phytase enzyme activities in the mixed diets was carried out according to the procedures described by Yasar & Desen (2014).
Descriptive statistics were used to present the nutrient and anti-nutrient composition of the ingredients and diets. The general linear model (GLM) of Minitab software version 17 (Minitab, 2013) was used to analyse the data from the feeding trial. Differences between mean values were separated by Duncan’s multiple range test and were considered significant at P < 0.05. This study was approved by the University of New England Animal Ethics Committee (Authority No AEC16-062).
Results
The analysed concentration of TI in the RFSB as a feed ingredient was 16564 TIU/g. The concentrations of crude fibre, CP and ether extract were 62, 405 and 201 g/kg, respectively. The concentration of TI in the PC diet was 1940.0 TIU/kg (Table 4). When the commercial SBM was replaced with RFSB at 25% (75 g/kg diet), the concentration of TI was increased by about fivefold (NC). Available lysine content was slightly reduced while protein solubility was increased.
Table 4 Effects of partial replacement of soybean meal with raw full-fat soybean on diet quality of broiler diets
The in vitro digestibility of CP in RFSB was improved (P < 0.001) significantly owing to supplementation of microbial protease or a combination of protease and phytase (Figure 1). However, phytase on its own did not improve the digestibility of CP in this study.
Figure 1 Effects of a two-step enzymatic in vitro test on crude protein digestibility of raw soybean samples abMeans with different superscript letters within a column are significantly different. Each test was replicated five times
The results of FI, BWG and FCR of birds fed on the test diets containing RFSB (75 g/kg of diet) and supplemented with microbial enzymes are shown in Table 5. Between hatch and 10 days, chicks in the PC group consumed less (P < 0.01) feed and gained less weight than those in the NC and NC-enzyme groups. However, over the entire starter and grower phases (1 - 25 days), the PC group consumed more feed (P < 0.05), grew faster (P < 0.001) and were slightly more efficient (not significantly) than birds fed the diets containing RFSB, with or without the enzyme supplements. Supplementation with protease or protease and phytase increased (P < 0.05) FI, but only the combination of enzymes increased (P < 0.001) BWG between hatch and 25 days, compared with the NC.
Table 5 Effects of supplementing enzymes and raw, full-fat soybean on feed intake, body-weight gains and feed conversion ratio for broiler diets
The gross responses of birds to diets in which SBM or RFSB was the sole source of CP and AAs and to microbial enzyme supplements are shown in Table 6. Regardless of enzyme supplementation, birds allocated to PC consumed more feed (P < 0.001), grew faster (P < 0.001) and were more efficient (P < 0.01) than birds on the NC diets. Although enzyme supplementation of the NC diet tended to improve FI, BWG and FCR, this was not significant (P > 0.05).
The relative weight of the pancreas at days 10, 25 and 31 was lower (P < 0.05) in birds fed the PC diet than in those offered NC or NC enzyme-supplemented diets (Table 7). Enzyme supplementation to the NC diet tended to reduce the relative weight of the pancreas. This effect was significantly (P < 0.001) higher for the combined enzymes at days 10 and 31.
Consumption of diets that contained RFSB and those supplemented with protease and/or phytase had no effect on liver AST and ALT or blood PCV (Table 8). Blood pH and electrolyte values were also unaffected (Table 9)
Regardless of the protein source, there were no interacting effects between microbial protease and phytase supplementation across the two trial feeding phases. Birds fed the PC diet had significantly (P < 0.002) higher apparent ileal digestibility (AID) of CP and most AA at day 25 than birds fed the NC diets (Table 10). Supplementation of the NC diet with protease, phytase or both enzymes improved (P < 0.002) the AID of the CP, but only marginally improved the digestibility of some of the AA, compared with the NC group.
The trend in ileal digestibility of protein and AA, which was measured at day 31, when the sole source of protein and AAs was SBM or RFSB, was similar to that at day 25, although the values were much lower in the second phase (Table 11). The AID of CP and most AAs was significantly higher (P < 0.01) in birds on the PC diet than on the NC diets. The AID of CP and a few AAs was improved when the NC diet was supplemented with protease or a combination of protease and phytase, but rarely with phytase only.
Table 6 Response of broilers to enzyme supplements when the sole source of dietary crude proteins and amino-acids was raw, full-fat soybean or commercial soybean meal (25 - 31 days)
Table 7 Influence of supplementing enzymes on the relative weight (g/100 g of bodyweight) of pancreas and duodenum when commercial soybean meal was partially replaced with raw full-fat soybean at days 10 and 25 or when raw, full-fat soybean was used as a sole source of crude protein and amino-acids (at days 31)
Table 8 Activities of liver enzymes, Alanine transaminase; Aspartate transaminase, and packed cell volume in birds fed diets containing raw, full-fat soybean with or without microbial enzyme supplements
Table 9 Effects of raw full-fat soybean and protease supplementations on values of acidity and electrolytes in blood plasma of broilers at days 10 and 25
Discussion
Although the concentration of TI in the test sample of RFSB was 16564 TIU/g, results of an in vitro digestibility of CP test were improved significantly because of the test enzymes, particularly combined enzyme supplementation. Similarly, in the feeding trial (0 - 25 days), the FI and BWG of birds on the PC and NC diets, which were supplemented with combined enzymes, were greatly improved, compared with the same measurements on other diets. These results in general are in line with previous reports (Malathi & Devegowda, 2001; Srinath et al., 2012; Erdaw et al., 2017a; 2018) that better response were found in in vitro and in vivo experiments when combined microbial enzymes were used rather than individual application. These results suggest synergistic and combined effects of the two enzymes on ANF.
The relative weight of the pancreas at days 10, 25 and 31 was lower in birds fed the PC diet than in those offered the NC or NC enzyme-supplemented diets. These results are in line with a previous report (Erdaw et al., 2017c; 2018). Enzyme supplementation of the NC diet tended to reduce the relative weight of the pancreas. This effect was significant for the combined enzymes at days 10 and 31. The weight of the duodenum was reduced significantly due to supplementation of phytase in broiler diets when the sole source of CP and AAs was RFSB. These findings disagree partially with those of other researchers (Mogridge et al., 1996; Mayorga et al., 2011), who reported that birds fed diets that contained RFSB had heavier pancreas and duodenum, although there was no indication that the diets were supplemented with microbial enzymes. The reduction in weight of duodenum in the current study might be due to the positive contribution of microbial enzymes, particularly phytase, against anti-nutritional factors (ANF) in broiler diets.
Consumption of diets containing RFSB or supplemented with protease and/or phytase had no effect on liver AST and ALT or blood PCV. These results confirm the safety of the RFSB at the tested levels.
Birds fed on the diets with RFSB as the sole source of CP and AAs and supplemented with combined enzymes had lighter weight of pancreas. Similarly, the BWG and feed efficiency of these birds were improved by 11.7% and 24.8% when supplemented with protease and phytase, respectively, compared with birds in the NC group (26 - 31 days). Not only were the combined microbial enzymes effective in the latter phase of this study, but supplementation of microbial protease alone into the NC diet contributed to approximately 5.3% and 22.5%, or 11.4% increase in FI, BWG and feed efficiency, respectively, compared with birds on NC between 26 and 31 days. These results in general agree with previous reports (Ao, 2011; Barletta, 2011; Erdaw et al., 2017a; 2017b; 2017c) that ANF, for example proteinaceous anti-nutrients and stored proteins, can be broken down by supplementing the diets or ingredients with microbial protease or phytase.
The AID of CP at day 25 was greatly improved when the diet was supplemented with combined enzymes in NC followed by microbial protease. Similarly, Erdaw et al. (2017d) reported that when broiler diets with ANF were supplemented with microbial protease, the loss of undigested and unabsorbed ileal CP (endogenous losses) was reduced, leading to an increase in the AID and standardized ileal digestibility (SID) of CP.
Table 10 Influence of raw, full-fat soybean and enzyme supplementations on coefficient values of apparent digestibility of CP and AAs at days 25 in diets for broilers
Table 11 Influence of raw, full-fat soybean and enzyme supplementations on the coefficient values of apparent digestibility of crude protein and amino acids at days 31 on broilers
The current findings agree partially with those of previous researchers (Guggenbuhl et al., 2012; Romero & Plumstead 2013; Cowieson et al., 2016), who observed an increase in AID of AAs in poultry and piglets and health benefits in response to inclusion of microbial protease.
When microbial protease, phytase or combined enzymes were supplemented to the NC diet, approximately 8.7%, 2.2% and 6.7% increases of AID were recorded for the total AAs, respectively, at day 31, but these changes were not statistically significant. These results agree with those of other researchers (Ao, 2011; Barletta, 2011; Pettersson & Pontoppidan, 2013; Erdaw et al., 2017a; 2017b; 2017c), who reported that the adverse impact of ANF, particularly TI, was reduced owing to supplementation of microbial enzymes. The improvement in digestibility of CP and AAs in the current study, in which diets contained a high concentration of ANF, including TI, might be because of the supplemented microbial protease and/or phytase. This might be also due to the contribution of the tested microbial feed enzymes, which tend to be potent against some of the ANF, including TI. The test enzymes appear to be more effective as a cocktail.
Conclusion
The results of the current study confirmed the negative impact of RFSB on growth and efficiency of broiler chickens. The test enzymes, particularly protease and when combined with phytase, improved in vitro and in vivo digestibility of protein. The BWG and FCR of birds up to 25 days old were also improved. However, the in vivo digestibility of most of the AA was not significantly improved by enzyme supplementation. The health of the birds was not compromised by RFSB at the level that was tested. Further tests are required to establish the exact mechanisms of the test enzymes. There is also a need to re-develop the test protease to sustain its in vitro activity when it is applied in vivo.
Acknowledgements
The financial support by DSM, Singapore, and the use of the facilities of the University of New England, Australia for the study are gratefully acknowledged.
Authors’ Contributions
MME and PAI were in charge of the experimental design, implementation and writing the manuscript. RAP participated in interpreting and reviewing the results of the study. Only the named authors contributed.
Conflict of Interest Declaration
There are no conflicts of interest
This article was originally published in South African Journal of Animal Science 2019, 49 (No. 3). http://dx.doi.org/10.4314/sajas.v49i3.6. This is an Open Access article under the terms of the Creative Commons Attribution 4.0 South African License (http://creativecommons.org/licenses/by/4.0/za). 
