Multiprotease improves amino acid release in vitro, energy, and nutrient utilization in broilers fed diets varying in crude protein levels

Abstract

Low crude protein (CP) diets can reduce nitrogen (N) excretion and costs by increasing N utilization efficiency. Exogenous proteases may further improve protein digestibility in low CP diets. This study first evaluated in vitro the efficacy of a multiprotease on amino acid (AA) release from feedstuffs and broiler feed. Later, a broiler study evaluated the effect of feeding corn-soybean meal diets containing 3 CP levels (17, 19, and 21% CP) with supplementation on top of 0 or 2,400 U/kg multiprotease on chicken growth performance, total tract CP, and ileal AA digestibilities, and energy utilization. Ross 708 male chickens were placed in 42 cages and assigned to 6 treatments resulting from a 3 × 2 factorial arrangement. Three isocaloric basal diets were formulated to reduce CP, but all diets maintained digestible Lys:CP in 5.47% and the same ideal protein profile. Data were analyzed in a completely randomized design. On average, the multiprotease increased (P < 0.05) in vitro free AA release by 27.81% in most feedstuffs evaluated compared to the control. For broiler feed, 1,200 U/g multiprotease addition improved (P < 0.001) in vitro free AA release by 18.90%. This multiprotease showed interaction effects (P < 0.05) on chicken FCR, energy, and CP digestibility. As expected, BW at 24 d, BW gain, and FCR (8–24 d) worsened (P < 0.001) as dietary CP reduced from 21 to 17%, and multiprotease addition did not improve (P > 0.05) these parameters. BW gain decreased by 12.9% when N intake was reduced from 49.32 to 38.49 g/bird. Multiprotease supplementation improved (P < 0.01) AMEn by 71 kcal/kg, CP digestibility from 59.45 to 63.51%, ileal AA digestibility, and DM digestibility from 67.08 to 73.49%, but only in the 21% CP diet. No differences in ileal AA digestibility due to CP level (P > 0.05) were detected, except for Cys digestibility (P < 0.01). In conclusion, low CP diets reduced growth performance and improved N utilization but negatively affected energy utilization efficiency. Exogenous multiprotease supplementation improved AME, AMEn, protein, ileal AA, and DM digestibility in the 21% CP diet without significantly affecting growth performance.

INTRODUCTION

Low crude protein (CP) diets can mitigate the environmental impact of nitrogen (N) and ammonia emissions, minimize soybean meal dependence, and reduce broiler production costs (Greenhalgh et al., 2020; Amer et al., 2021). Corn-soybean meal diets remain one of the most efficient feeds for poultry and are known to be highly digestible. However, it is possible to further improve the efficiency of nutrient utilization and digestibility by observing details such as accurate feed formulation close to broiler amino acid (AA) needs, improving the balance between protein and starch digestibility dynamics (Moss et al., 2018; Chrystal et al., 2020; Liu et al., 2021), and adding exogenous enzymes that increase AA digestibility (Pan et al., 2016; Borda-Molina et al., 2019; Wealleans et al., 2023). Improvements in apparent ileal AA digestibility in broilers are reported to result in improvements in growth performance and reduced N excretion and may improve overall gut health due to reduced putrefaction in the distal intestine (Cowieson et al., 2017; Cardinal et al., 2019).

Reducing dietary CP levels affect AA digestibilities and can cause AA imbalances (Liu et al., 2021). Overall digestibility of AA in corn-based diets is reduced by 8.36% in the distal ileum when low CP diets are used (Moss et al., 2018). Nevertheless, maize-based diets are more conducive to CP reductions than wheat-based diets (Chrystal et al., 2021). Low-CP corn-soybean meal diets contain high quantities of maize starch that, at the ileal level, may compete with AA for intestinal uptake via Na+ -dependent and -independent transporters with overlapping specificities (Moss et al., 2018; Chrystal et al., 2020). Traditionally, non-bound AA has been added to low-CP diets in the synthetic form to meet the requirements of main limiting AA (Lys, Met, Thr, Val) and other AA (Ile, Leu, Trp, Glu, Gly). However, these synthetic AA sources are more digestible and increase competition with starch.

Consequently, formulating low-CP diets with a better AA balance and minimizing synthetic AA inclusion could be proposed to improve the performance of broilers fed low-CP diets. However, the best feed conversion ratios (FCR) have been reported when dietary starch:protein ratios are between 1.50 and 1.85. Therefore, it is expected that FCR will irremediably worsen in low-CP compared to a control diet with better starch: protein ratios (Liu et al., 2021).

In order to estimate the optimum dietary levels of AAs to use in low-CP diets where essential and nonessential AA will be reduced, the concept of digestible Lys (dig. Lys) to true protein (TP) ratio can be used. Alhotan and Pesti (2016) estimated that dig. Lys requirements for BW gain and FCR can be between 4.92 ± 0.51% and 5.58 ± 0.70% of TP, respectively. To apply the TP concept in feed formulation, Alhotan and Pesti (2016) proposed obtaining each feedstuff's N content and determining TP using the prediction equation TP = 0.873 x CP + 0.435 (R² = 0.99). Since this concept of TP is not widely applied, and optimums for each age are still not completely elucidated, in this experiment, the TP and dig. Lys: The TP ratio was calculated, but a similar ratio was expressed as a dig. Lys: CP, as others have used (Ng'ambi et al., 2009) and using one ideal AA ratio. However, ideal AA ratios are probably different in low-CP diets (Chrystal et al., 2020).

Exogenous protease enhances protein digestibility in broiler nutrition when evaluating feedstuffs (Adebiyi and Olukosi, 2015; Stefanello et al., 2016) or complete diets (Angel et al., 2011; Cowieson and Roos, 2016). Proteases can shift the site of protein digestion to more proximal intestinal segments (Cowieson and Roos, 2016), which can benefit low CP diets. Exogenous proteases have been reported to reduce the antinutritional effects of various protein raw materials and to increase intestinal resilience by reducing inflammatory responses, improving tight junction and mucin integrity (Cowieson and Roos, 2016; Cowieson et al., 2017), reducing populations of intestinal Clostridium (Kamel et al., 2015), and improving intestinal mucosa traits (Xu et al., 2017). The simultaneous supplementation of multiple proteases with different pH optima and substrate specificity may improve the efficacy and consistency of protease supplementation. It is important to test the activities of these multiproteases under different pH conditions and for diverse substrates to determine their efficacy.

Enzymes, including mono-component protease, have been proposed to improve the utilization of reduced CP diets with considerable synthetic AA supplementation (Vieira et al., 2016; Chrystal et al., 2020; Jabbar et al., 2021). Broilers fed a protein-deficient diet (lower CP and AA) supplemented with multiprotease were found to improve growth performance with increased ileal AA and CP digestibility (Cho et al. 2020).

The present study evaluated first the effects of a multiprotease supplementation in vitro across a range of feed ingredients and commercial broiler feed to test its efficacy in releasing free AA during in vitro digestibility. The second objective was to evaluate in a broiler study the in vivo efficacy of the multiprotease when applied to common corn-soybean meal diets but varying in CP level to observe variability in feed ingredient inclusion. Grower broiler diets with 3 CP levels (17, 19, and 21%) maintaining digestible Lys in 5.47% of CP, and the same ideal protein profile was used for the study, and the effects of protease supplementation on performance energy utilization, CP, and AA ileal digestibility was evaluated.

MATERIALS AND METHODS

The North Carolina State University Institutional Animal Care and Use Committee approved all procedures involving broiler chicken used in the present experiment (Protocol No. 21-145).

In Vitro Substrate Degradation by a Multiprotease

To test the efficacy of the multiprotease to release free AA, an in vitro 2-step method with sequential incubations at acidic and neutral conditions was designed to mimic the gastrointestinal conditions. In the acidic incubation, each treatment was made up to 50 mL with 0.1 M phosphate buffer (pH 2.5) and 10% (w/v) of pepsin and incubated for 1 h at 37°C under 90 rpm of continuous stirring. The samples were subsequently neutralized by 50 mL of 0.1 M phosphate buffer (pH 12.0) to pH 6.8 and incubated at 37 °C for 2 h under 150 rpm of continuous stirring. 10.0 g of individual protein raw material was weighed into a 250 mL conical flask for each treatment. Each type of protein raw material was treated by blank control or an exogenous multiprotease (KEMZYME^TMProtease, Kemin Industries, Des Moines, IA) supplemented at an equivalent dosage to 1,200 U/kg feed. Furthermore, maize, sorghum, rice, wheat, and barley were supplemented with the multiprotease at a dose equivalent to 2,400 U/kg.

The nine types of protein and cereal raw materials and 3 complete diets used in this study, as summarized in Table 1, were commercial products bought on the open feed market. Pepsin from porcine stomach mucosa were obtained from Sino pharm Group Co., Ltd., China. All other chemicals were obtained from Guangzhou Chemical Reagent Co, China.

Table 1.

Feed ingredients used in the in vitro study.

Material	Commercial origin	Crude protein, %
Protein-sources
Soybean meal 1	Guangdong province, China	47.00
Soybean meal 2	Jiangxi province, China	43.36
Dehulled soybean meal	Jiangxi province, China	46.00
Fermented soybean meal	Jiangxi province, China	50.00
Cereals
Corn	Guangdong province, China	8.24
Sorghum	Guangdong province, China	9.36
Rice	Guangdong province, China	10.40
Wheat	Guangdong province, China	13.90
Barley	Guangdong province, China	11.33
Complete broiler feeds
Commercial feed 1	Guangdong province, China	17.62
Commercial feed 2	Guangdong province, China	15.65
Commercial feed 3	Guangdong province, China	17.03

The Ninhydrin reaction has been widely used to analyze AA, peptides, and proteins in biomedical studies (Friedman, 2004) and recently for chicken protein digestibility in feedstuffs (Bryan et al., 2018; 2019; Bryan and Classen, 2020). The method was used to analyze the number of amino groups, which was used to indicate the degree of hydrolysis of protein materials. In this assay, 1 mL of digestion supernatant was mixed with 1 mL of ninhydrin reagent solution (1% solution of ninhydrin, Sigma, in ethylene glycol monomethyl ether) and 1 mL of sodium acetate-acetic acid buffer (0.2 M, pH 5.4, Guangshi, China). The mixture was shaken to homogeneity by a mixture (G560E, Scientific Industries, INC) and then boiled for 15 min in a 100°C water bath (HWS-24, Yiheng Instruments, China). The solution was cooled to room temperature and diluted with 3 mL of 60% ethanol. The absorbance against deionized water at 570 nm was measured by ultraviolet and visible spectrophotometer (2,100, Unico). The absorbance value of the supernatant needed to be within the range of 0.2 to 0.8. Otherwise, dilution was needed till the absorbance reached the valid range.

A standard curve was plotted using different dilutions of lysine (Sigma, 0.06–0.34 mmol/L) as the y-axis and the corresponding absorbance as the x-axis. The assay was performed in duplicate for each sample, and the results were calculated based on the standard curve. The level of improvement was calculated based on the amount of free AA in the protease-supplemented group vs. the control group.

In Vivo Evaluation of Multiprotease Efficacy

Chicken Husbandry

A total of 294 Ross-708 d-old male chicks were placed in 42 Alternative Design (Siloam Springs, AR) Super Brooder battery cages (96.5 × 78.7 × 38.1 cm) in an environmentally controlled room (3.50 × 12 m). Each cage had one stainless steel linear feeder and 2 nipple drinkers with cups. Room temperature was recorded and adjusted daily to guarantee thermoneutral temperatures at each age. The room temperature decreased from 32 to 28°C the first wk, and was maintained at 27 and 23°C during the second and third wk, respectively. A 23 h light and one h dark cycle was used the first week, and 18 h light and 6 h darkness the following wk. Mortality was recorded daily. Feed and water were offered ad libitum to the chicks during the experimental period.

Treatments and Experimental Diets

Six treatments resulting from a 3 × 2 factorial arrangement of 3 dietary CP levels (17, 19, and 21%) and 2 levels of exogenous multiprotease (KEMZYME^TMProtease, Kemin Industries, Des Moines, IA) supplementation (0 and 2,400 U/kg) were evaluated in a completely randomized design with seven replicates per treatment. The 2,400 U/kg equates to 300 g/t feed and is in line with the use of this protease in other studies (Chandrasekar et al., 2017; De Leon et al., 2021; Wealleans et al., 2023). The multiprotease includes a combination of acid (pepsin type protease), neutral (metallo endo-peptidase), and alkaline (serine endopeptidase), proteases, produced by Aspergillus niger, Bacillus subtilis, and Bacillus licheniformis, respectively (Lim et al., 2014). All protease activities are measured in a standardized assay: one unit of protease activity is defined as the amount of enzyme that solubilizes one microgram of Azo-casein per minute at pH 7.0 and 37°C (Pan et al., 2016).

Until 7 d of age, chicks received a standard corn-soybean starter diet as crumbles, formulated to meet Ross 708 requirements (Aviagen, 2019). The experimental grower diets were produced from 3 basal diets (Table 2) with their respective CP level. The nutrient composition for these experimental diets is presented in Table 2. These 3 basal diets were isocaloric (3,100 kcal/kg ME) formulated to reduce CP, but all diets maintained dig. Lys in 5.47 % of CP and the same ideal protein profile. This dig. Lys:CP ratio is estimated to be 6.08 ± 0.04% dig. Lys:TP ratio, which is within the optimum levels suggested by Alhotan and Pesti (2016). The ideal protein profile chosen corresponds to Ross Broiler (Aviagen, 2019) nutrient recommendations, with dig. total sulfur AA (TSAA), Thr, and Val maintained 76, 67, and 78% of dig. Lys, respectively. Synthetic Lys, Met, and Thr were used to obtain desired AA levels of balanced protein but not to keep similar AA concentrations across the 3 CP diets. The Lys:energy ratios were 0.30, 0.34, and 0.37%/Mcal ME. All diets contained titanium dioxide (TiO₂) as an indigestible marker. Two batches of each basal diet were used to add either sand or the multiprotease on top at a rate of 2,400 units of protease. Phytase and protease activity in samples of experimental crumbled diets were analyzed at Kemin Industries Labs.

Table 2.

Ingredient and nutrient composition of the experimental diets.

Items	Starter 0-7d	Grower basal diets (8 - 24 d)
		21%	19%	17%
Ingredients	————————- % ———————
Yellow dent-corn	51.70	55.77	61.73	68.67
Soybean meal solvent-extracted	29.68	24.88	19.67	19.40
Full-fat soybean meal	10.00	10.00	10.00	4.47
Poultry fat	2.19	3.18	2.42	2.36
DDGS	2.00	2.00	2.00	0.90
Limestone fine	1.32	1.00	1.01	1.01
Dicalcium phosphate	1.17	0.99	1.03	1.09
DL-Methionine	0.37	0.33	0.28	0.24
Sodium bicarbonate	0.34	0.26	0.27	0.28
Salt	0.30	0.30	0.29	0.30
L-Lysine	0.24	0.22	0.24	0.24
Trace mineral premix1	0.20	0.20	0.20	0.20
Choline chloride	0.18	0.18	0.18	0.18
L-Threonine	0.15	0.12	0.11	0.10
Vitamin premix2	0.10	0.10	0.10	0.10
Coban	0.05	0.05	0.05	0.05
Phytase3	0.03	0.03	0.03	0.03
Titanium dioxide4	0.00	0.40	0.40	0.40
Energy and nutrient content
Metabolizable energy, kcal/kg	3,000	3,100	3,100	3,100
Crude protein, %	23.00	21.00	19.00	17.00
Fat, %	7.03	8.05	7.40	6.50
Crude fiber, %	2.64	2.53	2.42	2.11
Calcium, %	0.96	0.87	0.87	0.87
Total phosphorus, %	0.58	0.52	0.51	0.50
Ash, %	5.56	5.22	4.98	4.74
Available phosphorus, %4	0.48	0.43	0.43	0.43
Phytate phosphorus, %	0.22	0.22	0.21	0.20
Dig Lys:CP, %	5.57	5.47	5.47	5.47
Dig Lysine, %	1.28	1.15	1.04	0.93
Dig. TSAA, %	0.95	0.87	0.79	0.70
Dig. Thr, %	0.86	0.77	0.70	0.62
Dig Trp, %	0.24	0.22	0.19	0.17
Dig. Ile, %	0.88	0.80	0.71	0.63
Dig. Leu, %	1.83	1.73	1.63	1.53
Dig. Val, %	0.97	0.89	0.81	0.73
Dig. Arg, %	1.40	1.26	1.11	0.98
Dig. Lys:ME, % per Mcal	0.427	0.371	0.335	0.300
Dietary Electrolyte Balance, mEq/kg	300	268	246	225

¹Trace minerals provided per kg of premix: manganese (Mn SO₄), 60 g; zinc (ZnSO₄), 60 g; iron (FeSO4), 40 g; copper (CuSO₄), 5 g; iodine (Ca(IO₃)₂),1.25 g.

²Vitamins provided per kg of premix: vitamin A, 13,227,513 IU; vitamin D3, 3,968,253 IU; vitamin E, 66,137 IU; vitamin B12, 39.6 mg; riboflavin, 13,227 mg; niacin, 110,229 mg; d-pantothenic acid, 22,045 mg; menadione, 3,968 mg; folic acid, 2,204 mg; vitamin B6, 7,936 mg; thiamine, 3,968 mg; biotin, 253.5 mg.

³Phytase (1,000 FTU/kg) was expected to release at least 0.12% available phosphorus and 0.06% Ca (Kemin Industries, Des Moines, Iowa 50317, USA).

⁴Undigestible marker Titanium dioxide (Venator, Hombitan AFDC101, CAS 13463-67-7, Duisburg, Germany).

Diet Manufacturing

Corn was ground in hammermill tip speed 3,335 m/min and #24 screen and obtained a particle size of 927 µm of geometric mean diameter (d_gw) and 3.64 geometric standard deviation (S_gw). The soybean had 980 ± 2.71 µm d_gw, and full-fat soybean was ground to obtain a similar particle size of 953 ± 3.26 µm d_gw. Diets were mixed in a twin-shaft counterpoise ribbon mixer (Model TRDB126060, Hayes & Stolz, Fort Worth, TX) for 180 s. All diets were pelletized and crumbled. Mixed mash feeds were conveyed to a single pass conditioner (model C18LL4/F6, California Pellet Mill, Crawfordsville, IN) and conditioned at 85°C for 30 s. Diets were pelleted using a 30 HP CPM pellet mill (model PM1112− 2, California Pellet Mill, Crawfordsville, IN) equipped with a 4.4 × 35.2 mm die with 548 cm2 working surface area at a production rate of 16.34 kg/min (980 kg/h). The steam pressure was 207 kPa. The pellet mill die was warmed with feed before pelleting the experimental batches. After pelleting, pellets were cooled in a counterflow cooler (Model VK09 × 09KL, Geelen Counterflow USA, Inc, Orlando, FL) and crumbled.

Experimental Procedure

Group BW and feed intake were obtained at 7 and 24 d of age, and BW gain and FCR (adjusted for mortality) were calculated. On d 23 and 24, excreta without feathers or down were collected in 2 shifts, immediately mixed, pooled by cage, and stored in a freezer at -15°C. Ileal digesta contents were obtained from all chicks at 25 d after euthanizing them by CO₂ inhalation, and digesta samples were collected from the half-distal portion of the ileum. The ileum was defined as that portion of the small intestine extending from the Meckel's diverticulum to 40 mm proximal to the ileo-caecal junction. Ileal digesta contents were gently flushed with de-ionized water into plastic containers, pooled by cage, and immediately stored in a freezer at -15°C. The excreta samples were dried at 55°C in a forced-air oven, and ileal digesta samples were lyophilized using FreeZone 6 (NC State University Phytotron). Subsequently, samples were ground in a Retsch mill (Retsch USA Verder Scientific, Inc. Newtown, PA) to pass through a 0.5-mm screen in a grinder.

Chemical Analysis and Parameter Calculations

Diets, excreta, and ileal samples were analyzed for dry matter (DM), CP, and titanium content. Excreta samples were further analyzed for gross energy (GE) and nitrogen content. Feed and ileal digesta samples were analyzed for AA content with HPLC (AOAC, 2006 method 982.30 Ea,b,c). Dry matter was determined according to the method, and CP (N x 6.25) was determined by the combustion method (LECO, AOAC International, 2006; method 990.03). The GE analysis was determined using an oxygen bomb calorimeter (IKA C5003; IKA Labortechnik), and AME was corrected to zero N retention (AMEn) using a factor of 8.22 kcal/g (Hill and Anderson, 1958). Titanium, CP, DM, and AA concentrations were measured in triplicate for dietary feeds and duplicates for fecal and ileal digesta samples. Titanium was measured on a UV spectrophotometer following the method described by Myers et al. (2004). Apparent ileal digestibility, total tract utilization, and AMEn were calculated using the following equations (Kong and Adeola, 2014):

$$Digestibility(\%)=\left[1-\left(M_i/M_o\right)\times \left(E_o/E_i\right)\right]\times 100$$

$$AMEn\left(kcal/kg\right)=G{E}_i-\left[G{E}_o\times \left(M_i/M_o\right)\right]-8.22\times \left\{N_i-\left[N_o\times M_i/M_o\right]\right\}$$

Where M_i represents the concentration of inert maker titanium dioxide in the diet in g/kg DM; Mo represents the concentration of titanium dioxide in the excreta and ileal digesta in g/kg DM output; E_i represents the concentration of DM, CP, energy, or AA in the diet in mg/kg of DM; and E_o represents the concentration of DM, CP, AA and energy in the excreta and ileal digesta, in mg/kg DM. The GE_i is gross energy (kcal/kg) in the diet; GE_o is the GE (kcal/kg) in the excreta; N_i represents nitrogen concentration in the diet, and N_o represents nitrogen concentration in the excreta in g/kg DM.

Additionally, the coefficients of apparent DM fecal and ileal digestibility were calculated. The N intake was calculated using N dietary content and feed intake. Using calorie and nitrogen intake, the conversion ratios to live BW were estimated.

Statistical Analysis

All data were evaluated in completely randomized designs. The in vitro data was analyzed in a one-way ANOVA to test the effect of multiprotease with mean separation by student's t or Tukey's HSD test. In the chicken experiment, a 3 × 2 factorial arrangement of treatments with 3 CP levels and 2 multiprotease supplementation levels was evaluated. Each treatment had seven replicates distributed equally in 42 cages. Data was analyzed in 2-way ANOVA using JMP 16 (SAS Institute. Inc., Cary, NC). Means were separated by the LS means procedure using Tukey's HSD at a significance level of alpha 0.05. Correlation coefficients between formulated and analyzed CP and AA content were calculated. Finally, energy, CP, and Lys intakes were calculated, and correlation analysis with chicken growth performance was conducted.

RESULTS AND DISCUSSION

In Vitro Protein Degradation by a Multiprotease

As an initial investigative step, the ability of the multiprotease to release AA was tested in vitro in nine raw feed materials and 3 broiler commercial feeds using a methodology of CP digestibility developed and validated by Bryan et al. (2018, 2019) and positively correlated with in vivo AA digestibility (Bryan and Classen, 2020). The results are shown in Table 3. This data was intended to preliminarily explore the substrate-specificity of the multiprotease as a foundational step before animal trials. The addition of exogenous multiprotease improved (P < 0.05) AA release on top of the action of pepsin in all but one sample. The improvement ratio following supplementation with 1,200 U/g in the feed ingredients ranged from 13.54 to 60.22 %, and in the complete broiler feed, from 15.93 to 24.69 %, averaging 18.90%. Uplifts in free AA release from cereals were generally higher (P < 0.05) with supplementation of 2,400 U/g, with improvement ratios ranging from 28.35 to 87.20% compared to the control, averaging 57.81 %, but not significantly different from 1,200 U/g in sorghum, rice, and barley.

Table 3.

Release of free amino acids from plant-based protein sources and cereals, and complete commercial broiler feeds following supplementation with 1,200 or 2,400 U/g of an exogenous multiprotease.

	Free Amino Acid Content (mmol/L)			P-value
	Control	1,200 U/g	2,400 U/g
Protein-sources
Soybean meal 1	11.83 ± 1.3	13.9 ± 0.95	–	NS
Soybean meal 2	11.47 ± 0.03b	14.35 ± 0.42a	–	0.001
Dehulled soybean meal	15.33 ± 0.63b	18.98 ± 0.76a	–	0.010
Fermented soybean meal	21.2 ± 0.41b	24.07 ± 0.25a	–	0.001
Cereals
Corn	3.95 ± 0.02c	4.56 ± 0.01b	5.07 ± 0.09a	0.001
Sorghum	3.26 ± 0.18b	4.15 ± 0.21a	4.94 ± 0.23a	0.050
Rice	3.67 ± 0.23b	5.88 ± 0.20a	6.65 ± 0.06a	0.001
Wheat	5.69 ± 0.10c	7.96 ± 0.19b	10.21 ± 0.02a	0.001
Barley	4.08 ± 0.13b	5.2 ± 0.19a	6.06 ± 0.20a	0.010
Complete broiler feeds
Commercial feed 1	7.21 ± 0.09b	8.37 ± 0.07a	–	0.001
Commercial feed 2	9.23 ± 0.17b	10.7 ± 0.01a	–	0.001
Commercial feed 3	7.21 ± 0.12b	8.99 ± 0.05a	–	0.001

^a-cMeans in rows followed by different superscript letters are statistically different (P < 0.05) by t-test or Tukey's test. n = 2 per treatment and each ingredient. NS = non-significant.

Across the nine feedstuffs tested, the average uplift in free AA release was 27.81% compared to the control. No significant uplifts were seen for one of the soybean meal sources. This difference in efficacy may affect the response to the multiprotease in vivo. This is in accordance with the reported ileal digestibility values of such alternate protein sources in broilers and provides an avenue for future research into the correlation between in vitro and in vivo efficacy of multiprotease.

Though the 2 assessed soybean meal samples had similar basal digestibility in the presence of pepsin (11.83 and 11.47 mmol/L AA release), the relative uplift following multiprotease supplementation was different between samples (17.5 and 25.1 %). Significant digestibility improvement was observed in soybean meal 2 (P < 0.001) and dehulled soybean meal (P < 0.01) but not in soybean meal 1 (P > 0.05). The presence of protease inhibitors in soybeans is well known, and the concentration varies between soybean varieties, processing conditions, and growing microclimate (Sakkas et al., 2019; Yang et al., 2022). The difference in AA release as a percentage of the control may be due to the undetermined presence of such anti-nutritional factors. Aderibigbe et al. (2021) suggested that the effect of protease on AA release and bird performance was independent of the levels of trypsin inhibitor present in the diet. However, Wedekind et al. (2020) suggested that exogenous protease was more effective in diets containing high trypsin inhibitor levels. Another possible reason could be the differences in protein solubility between the 2 soybean sources (Zhao et al., 2022). The uplift of AAs observed in fermented soybean meal (P < 0.001) and other types of soybean meal could be due to the destruction of macromolecular proteins, the deactivation of anti-nutritional factors, or a combination of both factors during the fermentation process (Bao et al., 2014).

In summary, the in vitro evaluation indicated that the multiprotease applied in this test acts over various feedstuffs used in poultry diets and complete broiler feed. However, the degree of efficacy may vary according to the feed ingredient evaluated. The following step is to test the efficacy of the multiprotease in vivo when applied to chicken corn-soybean diets varying in nutrient composition and feedstuff proportions. Modifying the CP levels will cause the desired dietary variability. The multiprotease will be added to chicken feed at 2,400 U/kg since corn free-AA release was increased (P < 0.001) by 11.18 % using this level instead of 1,200 U/g (Table 3). The specific objectives were to test the effects of the multiprotease on CP, AA, and energy utilization by chickens fed these diets.

In Vivo Evaluation of a Multiprotease in Broiler Diets Varying in CP Content

Growth Performance

Results of laboratory analysis of samples of experimental diets (Table 4) indicated good agreement (P < 0.001) between formulated and analyzed values for CP (r = 0.991) and AA (r = 0.997). The analyzed activity of phytase for all diets and protease for supplemented diets is presented in Table 5. The values observed in pelletized and crumbled diets matched the expected activity of phytase supplemented in the basal diets and the inclusion of protease according to the treatment. No protease activity was detected in control diets.

Table 4.

Formulated and analyzed¹ values for crude protein (CP) and amino acids (AA) of experimental diets per CP treatment.

Treatment	21		19		17
Nutrient	Formulated	Analyzed	Formulated	Analyzed	Formulated	Analyzed
	—————————————————————————- % —————————————————————————
CP	21.00	21.36	19.00	18.96	17.00	17.08
Total AA
Lys	1.26	1.38	1.14	1.24	1.01	1.10
TSAA	0.96	0.98	0.87	0.90	0.77	0.82
Thr	0.90	0.89	0.81	0.81	0.72	0.72
Trp	0.25	0.24	0.22	0.22	0.19	0.20
Ile	0.88	0.96	0.78	0.84	0.69	0.74
Leu	1.77	1.83	1.65	1.64	1.52	1.50
Val	0.97	1.08	0.88	0.88	0.78	0.81
Arg	1.38	1.36	1.22	1.18	1.07	1.04
Glu	3.65	3.78	3.30	3.31	2.96	2.95
Asp	2.08	2.13	1.83	1.82	1.60	1.60
Phe	1.02	1.08	0.92	0.95	0.82	0.85
Tyr	0.60	0.74	0.51	0.65	0.41	0.57
Ala	1.03	1.06	0.96	0.96	0.88	0.89
Gly	0.85	0.88	0.76	0.78	0.68	0.70
Ser	1.01	0.96	0.91	0.84	0.81	0.76
His	0.54	0.55	0.49	0.49	0.44	0.44
Pro	1.21	1.22	1.13	1.14	1.04	1.06

¹Analyzed values are the means of six samples, 3 replicates for each experimental diet.

Table 5.

Analyzed activity of enzymes in the pelletized and crumbled experimental diets¹

Diet	Phytase (FTU/kg)	Multiprotease (U/kg)2
Starter (0–7 d)	1,214	-
Grower basal diet (21%)	1,137	-
Grower basal diet (19%)	1,180	-
Grower basal diet (17%)	1,042	-
Grower basal diet (21%)+ Protease	1,136	2,160
Grower basal diet (19%)+ Protease	1,184	2,280
Grower basal diet (17%)+ Protease	1,045	2,230

¹Values represent the average of 3 replicates samples per diet. Enzyme recovery tests conducted by Kemin Industries (USA).

²One unit of protease activity is defined as the amount of enzyme that solubilizes one microgram of Azo-casein per minute at pH 7.0 and 37°C.

The results of the pre-experimental period up to 7 d of age indicated the experiment started with a uniform group of chickens (Table 6) for all treatments with an average BW of 180 ± 2 g, which is 23 g below the current standard BW objective (203 g) for this genetic line (Aviagen, 2022). No differences (P > 0.05) in BW (Table 6), feed intake (132 ± 2 g), and FCR (0.976 ± 0.011 g:g) between groups of chickens assigned to the different treatments, but fed a common starter diet were observed up to 7 d of age.

Table 6.

Effect of exogenous multiprotease on growth performance of broilers at 24 d, fed grower diets (8 to 24 d) with 3 levels of CP (17, 19, and 21%).

CP level %	Multiprotease U/kg	BW 7 d	BW 24 d	BWG 8-24 d	FI 8-24 d
		———————————————–G—————————————————————
17		183	1,246c	1,067c	1,498
19		179	1,338b	1,159b	1,515
21		180	1,405a	1,225a	1,508
SEM		1.7	11.6	11.4	16.9
	0	181	1,324	1,145	1,498
	2,400	179	1,334	1,155	1,516
	SEM	1.4	9.4	9.3	13.8
CV, %1		3.49	3.25	3.70	4.20
Source of variation		————————————————–P-values———————————————————
CP level		0.342	<0.001	<0.001	0.772
Multiprotease		0.264	0.428	0.443	0.356
CP level*Multiprotease		0.792	0.104	0.159	0.107

^a-cMeans in columns followed by different superscript letters are statistically different (P<0.05) by Tukey's test. n = 7.

¹Coefficient of variation or normalized root-mean square deviation. Ratio of the standard deviation to the mean (%).

Performance results from 8 to 24 d of age are presented in Table 6. Interaction effects (P < 0.05) were observed on FCR (Figure 1). The main effects of CP level were observed (P < 0.001) in BW at 24 d and BW gain (8-24d). As expected, BW at 24 d (1,405, 1,338, and 1,246 g), BW gain, and FCR (8–24 d) worsened as dietary CP reduced from 21 to 17%, but multiprotease supplementation did not improve (P > 0.05) these parameters. The BW gain decreased by 12.9 % when N intake reduced from 48.96 to 38.49 g/bird. Feed intake was not affected (P > 0.05) by treatments.

Figure 1.

Effect of exogenous multiprotease (0 and 2,400 U/kg) on feed conversion ratio of broilers at 24 d, fed grower diets (8 to 24 d) with 3 levels of CP (17, 19, and 21%). ^a-c Means in bars with different superscript letters are statistically different (P < 0.001) by Tukey's test.

These results partially agree with Yang et al. (2015), who reported increments in feed intake and reduction in growth and feed efficiency in Arbor Acres chickens when CP and AAs were diluted 10, 20 and 30% (21.32 to 15.40 % CP) in starter diets fed during similar dietary phase (8–14 d). Several authors (Bregendahl et al., 2002; Yadav and Sah, 2005; Angel et al., 2011; Freitas et al., 2011) have reported that reduced-CP diets produced significant negative responses in growth performance. Even with smaller CP reductions, several authors reported worse chicken growth performance when essential AA were not maintained at similar levels using synthetic sources. Cowieson et al. (2017) reported that Ross 308 broiler chickens fed diets containing 20.5% CP and 1.08% dig. Lys had lower BW gain and higher FCR than those fed diets with 21.0% CP and 1.13% dig. Lys. Rehman et al. (2018) concluded that Hubbard chickens receiving starter diets containing 19.3% CP were 21 g heavier at d 14 of age than those receiving diets that contained 18.8 % CP with 7% less dig. AA (1.00 vs. 0.93% total lysine). Ndazigaruye et al. (2019) also observed 11 g lighter Ross 308 male chickens at 21 d of age (962 vs. 973 g) and worse FCR (1.364 vs. 1.399) by reducing CP from 20.58 or 19.72% CP (1.20 vs. 1.10% total Lys) in grower diets fed after 8 d of age.

Nevertheless, the present study evaluated a more significant CP reduction from 21 to 17% CP and dig. Lys from 1.15% to 0.93. In the current study, chickens fed 21% and 19% CP grower diets were heavier than the current standard for Ross-708 male broilers (Aviagen, 2022), while chickens fed the 17% CP diets had similar BW (1,246 g) to this standard (1,250 g) at d 24. The dietary balance (5.47 % dig. Lys:CP) and AA profile used in this experiment supported adequate growth. Optimum dig. Lys:CP ratios have been previously evaluated (Ng'ambi et al., 2009), and results indicated that in each feeding phase, one ratio could optimize growth rate, FCR, and breast meat yield; however, the ratios can be different for each parameter and dig. Lys:TP ratio could be a better parameter to formulate these low-CP diets (Alhotan and Pesti, 2016). However, FCR worsened significantly as CP decreased from 21 % CP (1.230 g:g) to 19% (1.273) and 17% CP (1.355 g:g). The interaction effect (P > 0.001) indicated FCR response was dependent on multiprotease supplementation; however, no significant improvement due to multiprotease was detected at any CP level (Figure 1). Higher starch content in low-CP diets could partially explain the adverse effects of increasing FCR (Chrystal et al., 2020; Liu et al., 2021).

Positive effects of serine proteases on growth performance parameters have been reported (Cowieson et al., 2017; Rehman et al., 2018; Park et al., 2020; Saleh et al., 2020; Jabbar et al., 2021), while others did not show a response over the same chicken growing period from 7 to 24 d of age (Ding et al., 2016, Law et al., 2018; Cardinal et al., 2019). Proteases have been tested to improve the performance of chickens fed low-CP diets (Ding et al., 2016; Cho et al., 2020; Jabbar et al., 2021). Ding et al. (2016) reported that a reduced CP diet (starter diets 21, 20, and 19% CP) negatively affected performance. The supplementation with an exogenous protease produced by fermentation using Bacillus licheniformis at 150 and 300 mg/kg had no significant effects on growth performance. Those results are similar to the data obtained in the present study. In contrast, Mohammadigheisar and Kim (2018) showed significant improvements in BW gain due to the addition of a serine protease, also expressed in Bacillus licheniformis to low CP diets (19.4% CP in starter and 17.5% CP in grower diets) and a positive trend was observed in improving FCR. In a recent study by Cho et al. (2020), formulating diets with the same concept used in the current study and using the same multiprotease observed similar responses in growth performance with a reduction in dietary CP (21.5–18.5% CP). However, multiprotease supplementation could restore the production parameters.

It is evident that the dietary reduction of CP affects growth performance. However, the response to protease supplementation has been inconsistent across different studies, presumably due to the differences in feed ingredients or the research methodologies used. In the present study, it should be emphasized that all diets maintained dig. Lys:CP at 5.47 % and the same ideal protein profile and only synthetic Lys, Met, and Thr were used to obtain the desired AA levels of balanced protein but not to keep similar essential AA concentrations across the three CP diets.

Energy and Total Tract Digestibility

Energy utilization and total tract digestibility results are presented in Table 7. Interaction effects (P < 0.05) were observed for AME, AMEn, CP, and total and apparent ileal DM digestibility. Chickens fed diets with 17 and 19 % CP had better AME and AMEn than chickens fed 21% CP diets. The highest total tract CP digestibility was observed in chickens fed diets with 17 % CP and the worst in chickens fed a diet with 21% CP without the multiprotease. Multiprotease supplementation improved (P < 0.01) AMEn by 71 kcal/kg, CP digestibility from 59.45 to 63.51 %, and ileal DM digestibility from 67.08 to 73.49%, but only in diets containing 21 % CP. Ileal DM and total tract CP digestibility only improved with multiprotease supplementation in 21% CP diets. However, the total tract digestibility of DM decreased (P < 0.05) as dietary CP increased, and the addition of multiprotease did not produce any significant effect. However, calorie conversion worsened from 3.66 to 4.25 kcal/g BW gain as the CP level decreased while N conversion improved from 40.0 to 36.1 g/kg BWG (Table 8). These results indicated better N utilization but decreased energy utilization with reduced dietary CP content. Research has shown better N utilization with dietary CP reduction both without and with performance depression. Belloir et al. (2017) reported increased N retention efficiency with reducing CP content (3.5 points/CP percentage point) while observing a reduced performance. Reduction in CP levels (20.5–17.4% CP) increased the N-utilization from 62 % to 73 %, with no performance drop in the study by Ulrich et al. (2018).

Table 7.

Effect of exogenous multiprotease (2,400 U/kg) on apparent metabolizable energy (AME), nitrogen-corrected AME (AMEn), total tract apparent digestibility of CP and dry matter, and the apparent ileal dry matter digestibility of broilers at 24 d fed diets with 3 levels of crude protein (17, 19, and 21%).

CP level, %	Multiprotease, U/kg	AME	AMEn	Total tract	Ileal
				Crude protein digestibility	Dry matter digestibility	Dry matter digestibility
		—————- Kcal/kg —————		—————————- % ——————–———
17	0	3,417a	3,389a	65.02a	72.32a	72.93ab
19		3,388a	3,351ab	60.77bc	69.97ab	73.80ab
21		3,286b	3,234c	59.45b	66.89c	67.08c
17	2,400	3,366ab	3,337ab	61.83abc	70.64ab	72.78b
19		3,366ab	3,328ab	61.56abc	70.32ab	75.14a
21		3,351ab	3,305b	63.51a	68.96bc	73.49ab
SEM		29	29	0.90	0.61	0.46
CV, %1		2.17	2.15	3.81	2.30	1.58
Source of variation		————————————————— P-values ——————————————————
CP level		0.009	0.001	0.465	<0.001	<0.001
Multiprotease		0.907	0.945	0.047	0.629	<0.001
CP level*Multiprotease		0.048	0.038	0.002	0.0172	<0.001

^a-cMeans in columns followed by different superscript letters are statistically different (P <0.05) by Tukey's test. n = 7.

¹Coefficient of variation or normalized root-mean square deviation. Ratio of the standard deviation to the mean (%).

Table 8.

Effect of exogenous multiprotease on nitrogen intake, calorie, and nitrogen conversion of broilers at 24 d, fed grower diets (8 to 24 d) with 3 crude protein levels (17, 19, and 21%).

CP level, %	Multiprotease, U/kg	Nitrogen intake	Calorie conversion	Nitrogen conversion
		– g/bird –	– cal/g BWG –	– g/kg BWG –
17		38.49c	4.25a	36.09c
19		43.79b	3.99b	37.77b
21		49.32a	3.66c	40.11a
SEM		0.50	0.03	0.22
	0	43.64	3.96	38.00
	2,400	44.10	3.97	37.97
	SEM	0.41	0.02	0.18
CV, %1		4.23	2.77	2.12
Source of variation		———————————————– P-values ————————————————
CP level		0.001	0.001	0.001
Multiprotease		0.437	0.651	0.908
CP level*Multiprotease		0.196	0.969	0.409

^a-cMeans in columns followed by different superscript letters are statistically different (P < 0.05) by Tukey's test. n = 7.

¹Coefficient of variation or normalized root-mean square deviation. Ratio of the standard deviation to the mean (%)

Some studies have reported increased N retention and improvements in growth performance or carcass traits by supplementing exogenous proteases in broiler diets (Xu et al., 2017; Rehman et al., 2018; Jabbar et al., 2021), differing from the present study. Xu et al. (2017) used the same multiprotease evaluated in the present experiment in corn and sorghum-soybean diets with corn gluten meal (21.8 and 21.3% CP). Rehman et al. (2018) compared keratinase and serine proteases in diets (19.3 and 18.8% CP) containing maize, rice polishing, soybean, rapeseed, sunflower, cottonseed, and fish meals. Jabbar et al. (2021) evaluated a neutral protease in diets with a reduction in CP (17, 19, and 21% CP) containing corn, soybean, canola, sunflower, gluten, and bone meals, and poultry by-products, balanced with several synthetic AA (Lys, Met, Val, Thr, Ile, Arg). Proteases can have more beneficial effects in more complex plant-based diets.

Nevertheless, Yu et al. (2007) reported that the supplementation of a cocktail of a serine protease in combination with carbohydrases in low-CP corn-soybean diets (20.39 and 18.93% CP in starter and grower diets) resulted in lower dietary-protein waste, N excretion, and emissions into the environment without a reduction in the performance of broiler chickens. In the same way, Freitas et al. (2011) observed improvements in FCR, fat, and CP digestibilities using low-CP corn-soybean meal-based diets (21.5 % CP in starter and 19.0% CP in grower diets) with meat and bone meal (∼6%) supplemented with increased levels of protease (7,500 to 120,000 PROT units/kg) in 42-day-old broilers. Likewise, Angel et al. (2011) reported that CP digestibility improved due to serine protease supplementation compared to control (22.5% CP) or low-CP (20.25% CP) diets. However, they did not observe differences among various concentrations of protease supplementation (7,500, 15,000, 30,000, and 60,000 PROT units/kg) in 22-day-old chickens. Similarly, Fru-Nji et al. (2011) concluded that exogenous serine protease (15,000 PROT/kg feed) enhanced the protein and energy digestibility of broilers fed corn-soybean meal diets (21.1 and 20.0 % CP in starter and 20.0 and 19 % CP in grower diets). In more recent work, Kamel et al. (2015) indicated that the same serine protease on top (15,000 PROT/g) of corn-soybean meal diets with corn gluten meal (21.5 % CP in starter diets) improved growth performance, carcass dressing percentage, CP digestibility in broiler chickens and reduced ileal clostridia species counts. Proteases may also affect endogenous enzyme activity. Ding et al. (2016) reported an improvement of 16 % in trypsin activity in the pancreas when a serine protease was added to the diet (21 % CP in starter and 19 % in grower diets) at the dose of 300 mg/kg, which could help improve protein digestibility.

Cho et al. (2020), who worked with the same multiprotease used in the present study, demonstrated that supplementation of multiprotease to 21.5 to 18.5% CP diets improved the ileal digestibility of DM and CP of chickens on d 21 and 35 compared to those fed diets without a multiprotease. In another broiler study conducted by De Leon et al. (2021), the same multiprotease improved the growth performance of low-CP (0.5% reduction) and AA (2% reduction) diets when supplemented at 300g/ton (2,400 U/kg). De Leon et al. (2021) observed improvements in DM, CP, and energy ileal digestibility coefficient. On the other hand, Saleh et al. (2020) studied the effect of serine-protease, and they found that the digestibility of DM and CP was enhanced by adding both 200 and 300 mg/kg of protease compared to the control without supplementation in starter diets containing 21 % CP.

Likewise, Jabbar et al. (2021) concluded that a neutral protease expressed in Bacillus subtilis (30,000 PROT units/kg feed) improved BW gain and FCR in diets with 17 and 19% CP during a trial period of 15 to 28 d. Protease supplementation only improved N retention (62.71 vs. 64.00 %) and reduced abdominal fat in 17% CP diets. However, no significant protease improvements in growth performance or N retention, or changes in abdominal fat were observed during that study in chickens fed a 21 % CP diet. In the study described by Jabbar et al. (2021), all diets maintain similar Lys (1.15 %) and TSAA (0.90%) levels by adding variable levels of synthetic AA (L-Lys HCl, DL-Met, Val, Thr, Iso, Arg), and did not contain phytase, or it was not reported. This neutral protease improved the response of this low-CP but had no significant benefits in the 21% CP diet. This is opposite to the effects observed with the multiprotease evaluated in the experiment reported here, where phytase was added to all diets.

Several exogenous enzymes are frequently utilized in poultry production. Velásquez-De Lucio et al. (2021) mentioned that the enzyme most commonly used in poultry diets worldwide is phytase, followed by xylanase and amylase. All diets used in the present experiment contained phytase, which has been suggested to improve the utilization of low-CP diets (Chrystal et al., 2020). Protease is currently less well adopted in commercial poultry diets, and the results of studies conducted with mono-component proteases are frequently inconsistent. In this context, the effects of exogenous protease on broilers are unclear due to differences in the types of proteases tested and dissimilar methodology, differences in the composition of the negative control diet, and the addition of other enzymes like phytase and carbohydrases.

Inconsistency in the effect of exogenous proteases on growth performance or AA digestibility in poultry diets has been attributed to the inherent digestibility of AA in the diets (Cowieson and Roos, 2016), which vary in low-CP diets (Liu et al., 2021). The neutral proteases supplemented in most of the experiments in the literature are categorized as serine proteases, the same family of proteases as trypsin and chymotrypsin, and have activity over a broad range of substrates (Odetallah et al., 2005). Additionally, Freitas et al. (2011) suggested that proteases supplemented in combination with other enzymes might contribute to the conflicting reports surrounding protease supplementation in poultry diets. In the present study, all diets contained at least 1,000 FTU/kg phytase. The interactive effects between enzymes could be partially responsible for some of the results observed herein, as Erdaw et al. (2017) had discussed, but differentiating these effects was not the scope of this project.

Amino Acid Ileal Digestibility

Interaction effects (P < 0.05) were observed for almost all ileal AA digestibility values except for Cys and Trp (Table 9). The multiprotease improved the ileal digestibility of all AA in diets with 21% CP, except for Cys and Trp. The multiprotease improved the digestibility of Cys and Trp (P < 0.05) independently of the CP level. No differences in ileal AA digestibility due to dietary CP level (P > 0.05) were detected, except for Cys digestibility (P < 0.01), which was better in 19 % diets than in 21 % CP, while 17% was intermediate. A similar balance of AA across all diets may cause similar ileal AA digestibility. The multiprotease had positive effects when AA levels exceeded chicken requirements in 21% CP diets.

Table 9.

Effect of exogenous multiprotease on apparent ileal amino acid digestibility coefficients at 24 d of age for grower broiler diets with 3 crude protein levels (17, 19, and 21%), fed from 8 to 24 d of age. ^a,1

CP level %	Multiprotease U/kg	Lys	Met	Cys	Thr	Val	Trp	Leu	Ile	Ala	Arg	Glu	His	Asp	Phe	Ser	Gly	CdCP
		———————————————————————————————- % ————————————————————————————————————
17		87.78	93.38	75.68ab	80.49	82.64	89.10	85.81	83.66	84.56	90.00	88.76	85.64	82.89	85.55	84.40	78.79	84.99
19		88.29	93.53	76.79a	81.58	83.52	88.78	86.39	84.85	85.00	90.81	89.42	86.75	83.92	86.45	84.51	80.09	85.69
21		88.41	92.97	73.90b	80.96	83.34	87.99	85.73	84.63	84.11	90.89	89.51	86.41	84.12	86.27	84.39	80.08	85.57
SEM		0.37	0.26	0.61	0.62	0.52	0.38	0.41	0.47	0.47	0.32	0.35	0.46	0.51	0.45	0.47	0.56	0.45
	0	87.82	92.96b	74.93b	80.35	82.60	88.04b	85.46b	83.67b	84.00b	90.23	88.86	85.77	83.24	85.62	83.86b	79.06	84.96
	2,400	88.48	93.61a	76.21a	81.67	83.70	89.21a	86.48a	84.98a	85.10a	90.88	89.57	86.77	84.04	86.56	85.01a	80.21	85.87
	SEM	0.30	0.22	0.50	0.50	0.43	0.31	0.34	0.38	0.39	0.26	0.29	0.37	0.42	0.37	0.39	0.46	0.37
17	0	88.34ab	93.51ab	75.80	80.97ab	83.02ab	88.93	86.01ab	83.96ab	84.96a	90.25ab	89.01ab	85.90ab	83.44ab	85.79ab	84.84ab	79.32ab	85.30ab
19		88.12ab	93.31ab	76.21	81.16ab	83.40ab	88.19	86.30ab	84.40ab	84.76ab	90.64ab	89.30ab	86.37ab	83.76ab	86.16ab	83.96ab	79.80ab	85.49ab
21		86.93b	92.00b	71.93	78.92b	81.15b	87.00	83.81b	82.64b	82.01b	89.73b	88.16b	85.01b	82.53b	84.88b	82.78b	78.06b	84.07b
17	2,400	87.29b	93.28ab	75.52	80.02ab	82.25ab	89.29	85.60ab	83.34b	84.15ab	89.75b	88.51b	85.37ab	82.34b	85.30ab	83.95ab	78.15b	84.67ab
19		88.45ab	93.75a	77.36	82.00ab	83.63ab	89.37	86.47a	85.28ab	85.24a	90.99ab	89.53ab	87.11ab	84.07ab	86.74ab	85.07ab	80.36ab	85.88ab
21		89.67a	93.80a	75.54	82.99a	85.20a	88.98	87.36a	86.32a	85.91a	91.89a	90.65a	87.80a	85.70a	87.64a	86.01a	82.09a	87.06a
SEM		0.52	0.37	0.86	0.87	0.74	0.53	0.58	0.66	0.67	0.46	0.50	0.64	0.73	0.64	0.67	0.79	0.64
CV, %1		1.52	1.06	2.86	2.86	2.32	1.57	1.76	2.05	2.07	1.32	1.47	1.97	2.30	1.96	2.09	2.62	1.98
Source of variation		————————————————————————————— P-values ——————————————————————————————————
CP level		0.579	0.415	0.005	0.464	0.480	0.111	0.374	0.189	0.318	0.134	0.331	0.228	0.209	0.338	0.979	0.160	0.508
Multi-protease		0.121	0.027	0.042	0.073	0.061	0.010	0.025	0.020	0.036	0.078	0.078	0.065	0.188	0.078	0.041	0.084	0.088
CP level*Multi-protease	0.004		0.023	0.106	0.022	0.007	0.334	0.003	0.010	0.004	0.021	0.016	0.045	0.018	0.046	0.014	0.007	0.021

^a-cMeans in columns followed by different superscript letters are statistically different (P < 0.05) by Tukey's test. n = 7.

¹Coefficient of variation or normalized root-mean square deviation. Ratio of the standard deviation to the mean (%).

These results differ from those reported by Rada et al. (2016). They concluded that mono-component serine protease did not significantly affect apparent ileal AA digestibility or protease activity in jejunum when feeding diets containing 19.9 and 20.7% CP and different inclusion levels of rapeseed meal. On the other hand, Walk et al. (2019) observed improved AA digestibility with the supplementation of serine and acid proteases in diets containing 19.00 or 21.50% CP but with no improvement in broiler performance parameters, similar to the present study.

Several authors have reported results similar to the present experiment in AA digestibility. For example, Angel et al. (2011) observed that exogenous protease supplementation improved the apparent digestibility (0.01−0.08 %) of Arg, Thr, Ile, Asp, Lys, His, Ser, and Cys in broiler chickens fed control (22.5% CP) or low-CP (20.25% CP) diets. Other authors have reported significant improvements in feed ingredients' apparent ileal AA digestibility (Adebiyi and Olukosi, 2015; Stefanello et al., 2016). Park et al. (2020), researching the effects of protease (0.03, 0.06, and 0.09%) on growth performance and nutrient digestibility, concluded that the digestibility of a variety of AA, like Lys, Met, Cys, Thr, Ile, Leu, His, and Try was enhanced significantly as exogenous protease supplementation increased in broiler starter (21% CP) and grower (19%) diets.

Plant proteins like soybean meal are rich in anti-nutritional factors, particularly protease inhibitors that can suppress the activity of the proteolytic enzymes trypsin and chymotrypsin (Erdaw et al., 2017). A reduction in the activity of these digestive enzymes may decrease the protein and AA digestibility of plant-based diets. Therefore, including protease in livestock feed could effectively attenuate the considerable adverse effects of unbalanced AA and anti-nutritional factors contained in feed materials that primarily consist of plant proteins (Erdaw et al., 2017; Park et al., 2020; Jabbar et al., 2021). In the present experiment, the protease supplementation improved energy utilization and digestibility in the 21% CP diet. This diet had up to 34.88% soybean products (solvent-extracted and full-fat) and more plant protein substrate than the other diets. In this diet type, multiproteases will probably have more pronounced effects.

The in vivo improvements in ileal AA digestibility observed with multiprotease supplementation are consistent with the in vitro evaluation results (Table 3). Although the in vivo study was not designed to evaluate the correlation between in vitro and in vivo efficacy of protease, our results indicated the possibility of deriving a correlation between the 2. Previous attempts to develop a correlation between in vitro and in vivo protease efficacy were unsuccessful (Zheng et al., 2023). Despite the improvement in in vitro crude protein digestibility (IVCPD), the addition of 4 proteases alone and in combination significantly reduced the apparent ileal digestibility of CP in broilers at 31 d and the apparent ileal digestibility of specific AA (Asp, Thr, Ser, Glu, Gly, Ala, Cys, Met, and Phe). This necessitates more judiciously designed in vitro and in vivo experiments to develop such correlation between the in vitro and in vivo effect of multiprotease.

Energy and Nutrient Intake

The energy utilization, CP, and AA digestibility values varied across treatments of CP levels. Dietary treatments did not affect (P > 0.05) feed intake (Table 6). Still, considering the variability in feed intake, energy utilization, and digestibility values per diet, it is possible to calculate the total energy and nutrient intake or retention (Table 10) during the experimental period. These results explain the chicken growth performance results better.

Table 10.

Effect of exogenous multiprotease on total intake of Apparent Metabolizable Energy (AME) and Apparent Metabolizable Energy corrected by nitrogen (AMEn), crude protein (CP), and Lysine of broilers at 24 d, fed grower diets (8 to 24 d) with 3 crude protein levels (17, 19 and 21%).¹

CP level, %	Multiprotease, U/kg	Energy and nutrient intake
		AME	AMEn	CP	Lys
		—– Kcal/broiler —-		——– g/broiler ——-
17	0	4,517	4,468	157.86	14.43
19		4,633	4,576	163.08	15.23
21		4,549	4,498	160.64	14.70
17	2,400	4,665	4,611	160.72	14.97
19		4,482	4,430	155.40	14.33
21		4,508	4,454	161.48	15.01
SEM		83	82	7.88	0.70
CV, %2		4.81	4.84	13.02	12.49
Source of variation		————————————————————- P-values ———————————————————————————
CP level		0.754	0.745	0.966	0.977
Multiprotease		0.835	0.817	0.838	0.978
CP level*Multiprotease		0.203	0.217	0.779	0.546

¹n = 7.

²Coefficient of variation or normalized root-mean square deviation. Ratio of the standard deviation to the mean (%).

No effects (P > 0.05) of CP level or protease supplementation were observed in total energy intake, CP, and Lys intake across treatments (Table 10). However, more protein and AA consumption was related to better BW gain and FCR. Lysine and CP intake had a highly positive (P < 0.001) correlation with BW gain (r = 0.92, 0.85) and a negative correlation with FCR (r = -0.82; -0.77), respectively. Energy consumption (AME and AMEn) was only moderately correlated (r = 0.50) (P < 0.001) with BW gain but not correlated (P > 0.05) with other performance parameters.

CONCLUSIONS

In line with the in vitro results, the multiprotease supplementation improved the chicken AA and CP total tract digestibilities in 21% CP diets. The AMEn was mainly affected by dietary CP levels, and multiprotease supplementation improved the AMEn by 71kcal/kg. Broiler performance was influenced primarily by dietary CP levels, and adding multiprotease did not improve growth performance at any CP level. Dietary CP level or multiprotease supplementation did not affect total energy, CP, and lysine intake. The similarity in energy, CP, and AA intake partially explains the lack of differences in BW gain and FCR between treatments with or without multiprotease supplementation. However, higher CP and AA intake were correlated with better performance. Further evaluation of the effect of multiprotease in poultry diets with lower CP levels and different feedstuffs is recommended to better understand protease application to broiler diets.