Effects of a Bacillus-based direct-fed microbial on performance, blood parameters, fecal characteristics, rumen morphometrics, and intestinal gene expression in finishing beef bulls

Abstract

We evaluated the effects of supplementing direct-fed microbials (DFM), containing Bacillus licheniformis and Bacillus subtilis, on performance, rumen morphometrics, intestinal gene expression, and blood and fecal parameters in finishing bulls. Nellore × Angus bulls (n = 144; initial BW = 401 ± 45.5 kg) were distributed at random in 36 pens (4 bulls/pen and 18 pens/treatment), following a completely randomized design. A ground corn-based finishing diet was offered for ad libitum intake twice a day for 84 d, containing the following treatments: 1) control (without DFM); 2) DFM (B. licheniformis and B. subtilis) at 6.4 × 10⁹ CFU (2 g) per animal. The data were analyzed using the MIXED procedure of SAS, with a pen representing an experimental unit, the fixed effect of the treatment, and the random effect of pen nested within the treatment. For fecal parameters (two collections made), the collection effect and its interaction with the treatment were included in the model. Bulls that received the DFM had a decreased dry matter intake (P ≤ 0.01), did not differ in average daily gain (2.05 kg; P = 0.39), and had a 6% improvement in gain:feed (P = 0.05). The other performance variables, final BW, hot carcass weight, and hot carcass yield, did not differ (P > 0.10). Plasma urea-N concentration decreased by 6.2% (P = 0.02) in the bulls that received DFM. Glucose, haptoglobin, and lipopolysaccharides were not different between treatments (P > 0.10). Ruminal morphometrics were not affected by the treatment (P > 0.10). The use of DFM tended to reduce fecal starch (P = 0.10). At slaughter, bulls fed DFM had an increased duodenal gene expression of tryptophan hydroxylase-1 (P = 0.02) and of superoxide dismutase-1 (P = 0.03). Overall, supplementation with DFM based on B. licheniformis and B. subtilis to Nellore × Angus bulls in the finishing phase decreased dry matter intake, did not influence ADG, improved gain:feed, and increased the expression of genes important for duodenal function.

Supplementation with direct-fed microbials based on Bacillus licheniformis and Bacillus subtilis offered to Nellore × Angus bulls in the finishing phase decreased dry matter intake, no influence on average daily gain, improved gain:feed, and an increase in expression of genes important for duodenal function.

Introduction

Finishing cattle on high-energy diets allows for earlier slaughter and improved carcass quality (McGregor et al., 2012; Drouillard, 2018). However, finishing cattle can develop metabolic disorders when fed high-energy diets (Nagaraja and Titgemeyer, 2007; Cooke, 2017), such as ruminal acidosis. To mitigate these problems, feed additives can be used to improve the performance and health of feedlot cattle (Nagaraja and Taylor, 1987; Marques and Cooke, 2021).

Direct-fed microbials (DFM) have been adopted as alternatives to traditional antibiotic additives (El Jeni et al., 2024). Among eubiotic approaches, bacteria of the Bacillus genus show promise for application in livestock systems, as they are supplied in the form of spores, ensuring their survival and stability during storage and feeding processes (Bernardeau et al., 2017; Cappellozza et al., 2023a). In addition, Bacillus spp. demonstrate an ability to grow both in the rumen and in the intestine, which may positively impact the health and function of the entire gastrointestinal tract (Green et al., 1999; Fuerniss et al., 2022). Bacillus spp. are characterized as producers of a wide range of enzymes (Qiao et al., 2010; van Dijl and Hecker, 2013), and can increase fiber and starch degradation in vitro (Pan et al., 2022; Cappellozza et al., 2023b). Supplementation with bacterial endospores of the Bacillus genus resulted in increased production of milk, protein, and solids, and improved milk/dry matter intake (DMI) in dairy cows (Sun et al., 2013; Souza et al., 2017). Moreover, for beef cattle, previous studies indicate improved G:F (Calaca et al., 2022; Dias et al., 2022; Lopez et al., 2024).

In non-ruminants, bacteria of the Bacillus genus are widely used to improve intestinal health (Luise et al., 2022). This may be possible in ruminants since the spores may be resistant in passing through the abomasum and able to function in the intestine (Leser et al., 2008; Luise et al., 2022). For beef cattle consuming high-energy diets, this may be a beneficial situation since a large amount of starch can reach the large intestine of cattle which can have negative consequences, such as acidosis and mucosal lesions (Penner et al., 2014; Steele et al., 2016; Aschenbach et al., 2019; Trotta et al., 2021). However, up to now, the intestinal impacts of Bacillus based DFM feeding remain unclear in cattle fed finishing diets.

Given these considerations, the hypothesis of this study was that beef cattle fed an additive containing B. licheniformis and B. subtilis would influence the expression of genes in the cecum (heat shock protein family) and in the duodenum (OCLN, CAT, SOD1, TPH1, TJP1, and CLDN4), and improve growth performance. The aim of the study was to evaluate the effect of a commercial DFM product containing B. licheniformis and B. subtilis on performance, rumen morphometrics, intestinal gene expression, and blood and fecal parameters in finishing beef bulls.

Materials and Methods

This experiment was conducted at a commercial farm in the county/municipality of Extrema, Minas Gerais, Brazil (22°47’52”S, 46°18’02”W, and altitude of 890 m) from May to August 2022. The mean temperature in each respective month over the experimental period was 15.9, 15.4, 17.5, and 16.1 °C, while the total rainfall was 49.2, 16.2, 10.8, and 39 mm. All bulls used in this study were cared for according to acceptable practices and experimental protocols that were reviewed and approved by the Ethics Committee on Animal Use of the University of Lavras (031/22).

Animals, housing, treatments, and diets

One hundred forty-four Nellore × Angus (Bos taurus indicus × Bos taurus taurus) bulls, with adjusted initial BW of 401 ± 45.5 kg (BW × 0.96; assuming 4% fill) were used. After arriving, all bulls were individually identified with ear tags and vaccinated against clostridial diseases (RESGUARD MULTI; Vaxxinova Brasil, Vargem Grande Paulista, Brazil; 5 mL/animal), rabies (RAI-VET; Vaxxinova Brasil, Vargem Grande Paulista, Brazil; 2 mL/animal), and respiratory diseases (CATTLEMASTER GOLD; Zoetis Brasil, São Paulo, Brazil; 5 mL/animal), and dewormed against endo- and ectoparasites (AGEBENDAZOL; Agener União Saúde Animal, Embu-Guaçu, Brazil; 1 mL/44 kg weight; TRUCID; Elanco Animal Health, Greenfield, United States; 1 mL/50 kg weight). To acclimate bulls to the facilities, all bulls were fed a common diet based on corn silage, dry-ground corn, soybean meal, and mineral mix (without any additive) for 5 d prior to experiment initiation (from day −5 to day −1). The pens (4 × 14 m; n = 36) were divided by wooden posts and wire, equipped with individual waterers and feed troughs (4 m). Pen was assigned to one of two treatments: finishing a diet without any additive (CON), or the same diet containing 2 g of DFM/animal per day (DFM). The DFM was composed of B. licheniformis 809 and B. subtilis 810 (3.2 × 10⁹ CFU/g; BOVACILLUS; Novonesis, Hørsholm, Denmark). A batch of a vitamin and mineral supplement with and without the DFM was received every 30 d. The CON supplement contained clay which is an inert substance to replace the direct-fed microbial in the DFM supplement. To administer the product, the supplement with DFM was first mixed with the concentrate and, before feeding, with the other ingredients of the diet. The complete composition and the nutritional profile of the diets are shown in Table 1. Treatment diets were fed for 84 d. The adaptation period was 15 d and consisted of three diets with progressive increases in concentrate (5 d each), varying from a roughage:concentrate proportion (dry basis) of 50:50 to 41:59 and finally 32:68. The final diet had a proportion of 23:77 and was formulated using the NASEM (2016) recommendations to provide an average daily gain (ADG) of 1.9 kg/d. Throughout the experimental period, the diets were provided twice a day (at 0800 and 1500 hours) in equal proportions, as a total feed mixture. The diet was weighed daily with an electronic balance and offered to ensure ad libitum intake, allowing 3% to 5% of orts. To minimize the risk of cross-contamination, each treatment was managed by a different group of people that did not participate in the management of the other. In addition, all equipment used, such as pails, bags, and shovels, were marked with specific colors and designated exclusively for only one of the treatments.

Table 1.

Ingredients, chemical composition, and particle size distribution of the diets used

Item	Diet
	Step 1 (0 to 5 d)	Step 2 (6 to 10 d)	Step 3 (11 to 15 d)	Final diet (16 to 84 d)
Ingredients, % DM
Corn silage	50.0	41.0	32.0	23.0
Ground corn	38.0	47.0	56.0	65.0
Soybean meal	8.00	8.00	8.00	8.00
Urea	1.00	1.00	1.00	1.00
Mineral/vitamin supplement1	3.00	3.00	3.00	3.00
Chemical composition
Dry matter	48.3	55.0	60.1	62.9
Ash	4.74	4.80	4.71	5.31
Crude protein	13.1	13.1	13.1	13.4
Neutral detergent insoluble fiber	34.5	27.1	25.4	22.9
apNDF2	32.3	26.4	24.7	22.2
peNDF3	17.2	13.6	12.2	10.8
Ether extract	3.04	3.60	3.40	3.67
Starch	37.6	43.5	49.7	50.4
Non-fiber carbohydrates	44.6	51.4	53.4	54.7
Non-starch non-fiber carbohydrates	6.96	7.92	3.69	4.33
Particle size distribution of silage		% of the total
>8 mm	31.5	26.0	21.9	18.1
<8 to >4 mm	28.0	25.6	20.2	17.9
<4 mm	40.5	48.4	57.9	63.9
Particle size distribution of ground corn		% of the total
>3.35 mm	0.00
<3.35 and >2.36 mm	4.00
<2.36 and >1.7 mm	16.0
<1.7 and >1.18 mm	20.0
<1.18 and >0.6 mm	30.0
<0.6 mm	30.0
Mean particle size of corn4, mm	1.03

¹Composition: calcium 190 g/kg, phosphorus 30 g/kg, sulfur 30 g/kg, magnesium 30 g/kg, sodium 75 g/kg, butylated hydroxytoluene (BHT) 1,000 mg/kg, cobalt 30 mg/kg, copper 750 mg/kg, fluorine 300 mg/kg, iodine 40 mg/kg, manganese 1,500 mg/kg, selenium 12 mg/kg, zinc 3,000 mg/kg, vitamin A 1,50,000 IU/kg, vitamin D3 15,000 IU/kg, vitamin E 500 IU/kg.

²Fiber insoluble in neutral detergent corrected for ash and protein.

³Percent peNDF was estimated by multiplying the percentage of sample larger than 8 mm in particle size (top 2 sieves) by the percent NDF of those particle sizes.

⁴Based on Yu et al. (1998).

Chemical analyses

Diet formulation and composition were based on weekly dry matter (DM) analyses, actual nutrient values, and corresponding feed batching records. Weekly ingredient samples were stored in a freezer at −20 °C until nutrient analyses were completed. After weekly DM determination, weekly samples from each ingredient were composited by month and analyzed. All samples were dried in a forced-air circulation oven at 55 °C for 72 h and then ground in a Wiley mill (Wiley TE-680, Philadelphia, United States). After milling, the absolute DM concentration was obtained by drying in a laboratory oven at 105 °C (method 934.01; AOAC, 1990), crude protein (CP) was obtained by the Kjeldahl method calculated as total N × 6.25 (method 984.13), and mineral matter (MM) concentration was determined by complete combustion in a muffle furnace at 600 °C for 5 h (method 942.05). The ether extract (EE) was quantified using the Soxhlet method (method 963.15) according to the AOAC (1990). Neutral detergent fiber (NDF) was analyzed with thermostable amylase, free of sodium sulfite, and corrected for ash and protein (Mertens, 2002; method F-012/1, Detmann et al., 2021). The starch (ST) plus free glucose was determined using the enzymes α-amylase and amyloglucosidase and obtained by colorimetry for glucose, as described in Fernandes et al. (2022), adapted from Hall et al. (2015). The data from dietary nutrient analyses were used to calculate the concentrations of NFC (100–CP–NDF–EE–MM) and NFC without starch (NFCWS; 100–CP–NDF–EE–MM–ST). Particle size distribution of the corn silage was evaluated with a Penn State particle separator, containing screens of 19, 8, and 4 mm and the bottom pan (Heinrichs and Jones, 2013). Percent physically effective fiber (peNDF) was estimated by multiplying the percentage of sample larger than 8 mm (top 2 sieves) by the percent NDF (as a decimal) of those particle sizes. Weekly samples of the ground corn used in preparing the diets were characterized for particle size distribution by using an orbital sieve shaker with intermittent tapping motion—rotap (MA 750, Marconi, Piracicaba, Brazil), and the sieves had mesh openings of 6.70, 4.75, 3.35, 2.36, 1.70, 1.18, 0.60 mm, and the bottom pan. Approximately 100 g of dry sample was shaken for 10 min. Three rubber sieve balls were used to assist in detaching the finer particles in the sieves of 1.70, 1.18, and 0.60 mm, as described in Kalivoda et al. (2017), whereas the mean particle size was obtained according to Yu et al. (1998).

Performance, intake, and efficiency measurements

The individual live weight of the bulls was collected on days 0 and 84 of the study and was used to calculate adjusted weight gain (BW × 0.96; final weight minus initial weight) and ADG (adjusted weight gain/84 d) during the experiment. Samples of the diet offered, and orts were collected daily, weighed, and dried. Thus, DMI was calculated as the total kilogram of DM delivered to each pen after subtracting the weight of dry orts. At the end of the experiment, the ADG and the DMI were used to calculate G:F.

Fecal parameters

Feces were collected twice in periods of three consecutive days (from the 33rd to the 35th day and from the 63rd to the 65th day of the experimental period). Fecal samples from each animal (50 g/animal) were collected from the floor of each pen immediately after defecation. To ensure representativeness, the collections were carried out at different times throughout the day: from 1700 to 1800 hours on the first day, 1200 to 1300 hours. On the second day, and 0600 to 0700 hours on the third day. At the end of each daily collection and at the end of the 3 d of each collection, a composite sample was formed per pen (200 g/animal) and stored in plastic bags at −20 °C.

To determine fecal pH, 15 g of feces were added to test tubes containing 10 mL of distilled water and homogenized using a vortex mixer (VORTEX MIXER, Kasvi, São José dos Pinhais, Brazil). The pH was determined using a multiparameter pH meter (Medidor edge HI2020-02, Hanna Instruments, Tamboré Barueri, Brazil). Fecal samples were dried, ground, and analyzed regarding nitrogen and starch concentration, as already described.

Blood parameters

Blood samples (10 mL) from two bulls per pen (n = 72) were collected on the last day of the experimental period after the morning feeding (approximately at 0900 hours). Samples were taken from the jugular vein using vacutainer tubes with sodium fluoride (for glucose) or heparin (for urea, haptoglobin, and lipopolysaccharides) to prevent blood clotting. After collection, the samples were placed on ice and centrifuged at 2,700 × g for 20 min, and the plasma was harvested and frozen until analysis.

Glucose concentrations were quantified using a colorimetric kit (K082; GLICOSE MONOREAGENTE, Bioclin, Belo Horizonte, Brazil). Urea was determined with a commercial colorimetric kit (UREA UV; Bioclin, Belo Horizonte, Brazil), whereas haptoglobin was determined with the Cat:ELK7354 kit (KIT ELISA HAPTOGLOBIN CATTLE, ELK Biotechnology, Denver, USA). Lastly, lipopolysaccharides were quantified using the EU3126 kit (LPS ELISA KIT, Fine Test, Wuhan, China), according to manufacturer’s recommendations. Results were obtained using a Multiskan GO spectrophotometer (Thermo Scientific; Waltham, USA).

Slaughter and sample collection

At the end of the experimental period, bulls were slaughtered in a commercial slaughterhouse (Frigorífico Angelelli Ltda., Piracicaba, SP, Brazil). The carcasses were weighed to determine hot carcass weight (HCW). Carcass yield was determined considering the ratio of HCW to adjusted final weight. Samples of the ventral sac of the rumen (atrium ruminis), duodenum, and cecum (same bulls selected for blood collection; n = 72) were collected. Rumen tissue samples were fixed on polyethylene slides and stored in universal sample containers containing buffered formalin. The duodenum and cecum samples were placed in 5-mL cryogenic tubes (E-lab, Monte Alto, Brazil) and then in liquid nitrogen for transport, and subsequently stored at −80 °C until the beginning of gene expression analyses. Liver abscess score was evaluated according to its occurrence and seriousness, as described by Brink et al. (1990): 0 (without abscess), 1 (1 or 2 abscesses < 2.5 cm or abscess scars), 2 (3 to 4 abscesses < 2.5 cm), and 3 (1 abscess > 2.5 cm or several small abscesses).

Rumen morphometrics

Samples collected during slaughter were fixed in buffered formalin for at least 48 h before further processing. Tissues were gradually dehydrated with ethanol and clarified with xylene. Samples were placed in paraffin and then cut into 5-µm-thick sections and mounted on a glass microscope slide. At least five histological sections were obtained from each sample. Hematoxylin and eosin were used to stain the slides. Histological images of each section were obtained using an optical microscope (BX43, Olympus LS, Tokyo, Japan) in a 20X lens linked to a camera to capture images with 300 DPI resolution. Height, width, and thickness of the keratinized layer and the papillary area were determined using the ImageJ software (version 1.52A; National Institutes of Health, Bethesda, MD, USA).

Gene expression

According to the manufacturer’s recommendations, the total ribonucleic acid (RNA) of the duodenum and cecum samples was extracted by the SV Total RNA Isolation System (Promega, Madison, WI, USA). Subsequently, 1.0% (w/v) agarose gel electrophoresis was performed on the RNA of all the tissues, and they were stained with GelRed Nucleic Acid Gel Stain (Biotium, Hayward, CA, USA). Gel was visualized in an E-Gel Imager Camera Hood (Life Technologies, Neve Yamin, Israel). Samples were quantified (260/280 nm and 260/230 nm) using a nano-spectrophotometer (DeNovix DS-11, Wilmington, DE, USA). Complementary DNA (cDNA) was synthesized using the GoScript Reverse Transcription System (Promega, Madison, WI, USA).

All the primers used for the target genes and reference genes are described in Table 2. Actin beta (ACTB) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) reference genes were used. Primers were designed using the Primer3Plus web interface program and were based on the sequences indexed and published in the GenBank public database, NCBI (National Center for Biotechnology Information) platform. Primers were synthesized by Invitrogen (Carlsbad, CA, USA).

Table 2.

Sequences of primers used in the RT-qPCR trial and tissue in which gene expression was determined

Gene	Sequence (5' to 3')	Accession number	Tissue analyzed
HSPB1	F: CTGAAACACCGCCTGCTAAA	NM_001025569.1	Cecum
	R: CGAGTGAAGCAACGGGAAAT
HSPA4	F: AGAGTAGAGCCACCACTTCG	NM_001114192.2	Cecum
	R: GGCCAGCTGAACTCTTGAAG
CLDN4	F: TACTCCGCTAAGTACTCCGC	NM_001014391.2	Duodenum
	R: ACCAGTTTGTAGCACCTCCA
TJP1	F: CCTGCCCAACTCAACTCATG	XM_024982012.1	Duodenum
	R: TGCTTTATTGTGTGGAGGCG
TPH1	F: AACTCTCTCCACTGCTAGCC	XM_002693039.5	Duodenum
	R: GCCTCCAGAGTTACCCGTTA
SOD1	F: CGGCGTCGTTTTCTCTACTT	NM_174615.2	Duodenum
	R: ATTACACCACAGGCCAAACG
CAT	F: CATTGCAGTTCGCTTCTCCA	NM_001035386.2	Duodenum
	R: CAAGCCATGATGGTGCTGAA
OCLN	F: CCCTTTCTGCTTCTTCAGGC	NM_001082433.2	Duodenum
	R: CTCCAAGTTACCACTGCTGC
ACTB	F: GTCCACCTTCCAGCAGATGT	NM_173979.3	Reference gene
	R: CAGTCCGCCTAGAAGCATTT
GAPDH	F: CATTGCCCTCAACGACCACTT	NM_001034034.1	Reference gene
	R: TCCACCACCCTGTTGCTGTA

HSPB1 = heat shock protein family B, member 1; HSPA4 = heat shock protein family A, member 4; CLDN4 = claudin-4; TJP1 = tight junction protein-1; TPH1 = tryptophan hydroxylase-1; SOD1 = superoxide dismutase-1; CAT = catalase; OCLN = occludin; ACTB = actin beta; GAPDH = glyceraldehyde 3-phosphate dehydrogenase.

Real-time quantitative polymerase chain reaction (RT-qPCR) analysis was carried out using the SYBR Green detection system (Applied Biosystems, USA) in the Eppendorf Realplex thermal cycler (Eppendorf, Hamburg, Germany). The entire RT-qPCR assay for each gene was conducted based on cDNA obtained from 36 biological replicates per treatment and 2 technical replicates for each treatment. Relative expression levels were calculated using the Delta Delta Ct method, according to Livak and Schmittgen (2001).

Statistical analysis

Results were analyzed in a completely randomized design using the MIXED procedure of SAS (SAS Inst., Inc., Cary, NC, USA). Pen was considered the experimental unit. The statistical model considered the fixed effects of the treatment (CON vs. DFM) and pen nested within the treatment as a random effect. When significant (P < 0.05), the adjusted initial weight was incorporated as a covariate in statistical analysis of the corresponding parameters. If the adjusted initial weight was not significant (P > 0.05), it was removed from the model. For the fecal parameters, the collection effect (1st and 2nd) and its interaction with the treatment effects were added to the model. For all the variables and their interactions, the data are shown as treatment means ± standard error of the mean. The covariance structure adopted was defined by the lowest value for the Akaike information criterion. Statistical differences were considered significant at P < 0.05. Trends were noted if the P-value was greater than 0.05 but less than or equal to 0.10.

Results

At the beginning of the experiment, the adjusted initial BW did not differ among the treatments (P = 0.85; Table 3). In line with this response, no differences were observed for adjusted final weight (P = 0.39), HCW (P = 0.30), and dressing (P = 0.70). Although ADG (P = 0.39) did not differ between treatments, bulls consuming the DFM had 3.1% lower DMI (0.4 kg) and, consequently, had 6% greater G:F. At slaughter, there was no incidence of liver abscesses (data not shown).

Table 3.

Performance and carcass characteristics of Angus × Nellore bulls with and without intake of DFM composed of Bacillus

Item1	Treatment		SEM3	P-value
	Control	DFM2
Adjusted initial weight, kg	402.6	399.8	10.651	0.85
Adjusted final weight, kg	570.0	575.4	4.363	0.39
ADG, kg/d	2.01	2.08	0.052	0.39
DMI, kg/d	12.9	12.5	0.085	0.01
G:F	0.157	0.167	0.004	0.05
HCW, kg	307.1	311.2	2.763	0.30
Dressing, %	54.9	55.1	0.258	0.70

¹Adjusted weight (final and initial) = BW × 0.96, assuming 4% fill; ADG = average daily weight gain; DMI = dry matter intake; G:F = gain:feed; HCW = hot carcass weight.

²Direct-fed microbials, composed of B. licheniformis and B. subtilis supplied at a dose of 2 g per animal/d.

³Standard error of the mean.

Regarding blood metabolites, treatment did not influence plasma concentration of glucose, haptoglobin, and LPS (P ≥ 0.05; Table 4). Bulls that consumed DFM had lower plasma urea concentrations than CON bulls (P = 0.017). No difference was found for height, width, and thickness of the keratinized layer and rumen papillary area (P ≥ 0.05; Table 5).

Table 4.

Blood parameters of Angus × Nellore bulls with and without intake of DFM composed of Bacillus

Item	Treatment		SEM2	P-value
	Control	DFM1
Glucose (mg/dL)	79.0	76.7	2.122	0.36
Urea (mg/dL)	29.2	27.4	0.968	0.02
Haptoglobin (ng/mL)	7.73	9.19	0.719	0.26
LPS (µg/mL)3	2.31	2.86	0.556	0.15

¹Direct-fed microbials, composed of B. licheniformis and B. subtilis supplied at a dose of 2 g per animal/d.

²Standard error of the mean.

³Lipopolysaccharides.

Table 5.

Ruminal morphometrics of Angus × Nellore bulls with or without intake of DFM composed of Bacillus

Item	Treatment		SEM2	P-value
	Control	DFM1
Papillae height, mm	4.36	4.73	0.208	0.22
Papillae width, mm	0.37	0.38	0.010	0.34
Keratin depth, μm	11.9	11.2	0.475	0.47
Area, mm2	1.79	1.74	0.031	0.25

¹Direct-fed microbials, composed of B. licheniformis and B. subtilis supplied at a dose of 2 g per animal/d.

²Standard error of the mean.

Bulls consuming DFM tended to have lower fecal starch (P = 0.096; Table 6). The fecal starch concentration was greater (P < 0.0001) during the 2nd feces collection (9.88%; from the 63rd to the 65th day of the experimental period) compared to the 1st feces collection (6.05%; from the 33rd to the 35th day of the experimental period). The opposite was found for fecal pH (5.86 vs. 5.60; P < 0.0001), which did not differ between treatments (P = 0.33).

Table 6.

Fecal starch, pH, and nitrogen of Angus × Nellore bulls with or without intake of DFM composed of Bacillus in two collection periods

Item	Treatment	Collection		Mean	SEM	P-value
		First	Second			Treatment	Collection	Treatment*Collection
Starch, % DM	Control	6.56	10.3	8.45	0.542	0.10	<0.0001	0.86
	DFM	5.54	9.42	7.48
Mean		6.05	9.88
pH	Control	5.82	5.61	5.71	0.034	0.33	<0.0001	0.14
	DFM	5.90	5.59	5.75
Mean		5.86	5.60
Nitrogen, % DM	Control	2.29	2.22	2.25	0.044	0.12	0.31	0.32
	DFM	2.18	2.20	2.19
Mean		2.23	2.21

¹Direct fed microbials, composed of B. licheniformis and B. subtilis supplied at a dose of 2 g per animal/d.

²Standard error of the mean.

In the duodenum (Figure 1), the expression of tryptophan hydroxylase-1 (TPH1; P = 0.01) and superoxide dismutase-1 (SOD1; P = 0.02) was greater in bulls that consumed DFM. Nevertheless, no differences were observed in expression of claudin-4 (CLDN4; P = 0.22), tight junction protein-1 (TJP1; P = 0.60), catalase (CAT; P = 0.55), and occludin (OCLN; P = 0.57). In the cecum (Figure 2), in turn, no differences were observed among the treatments for either heat shock protein family B member 1 (HSPB1; P = 0.66) or heat shock protein family A member 4 (HSPA4; P = 0.53).

Figure 1.

Relative expression of genes in the duodenum of Angus × Nellore bulls with or without intake of DFM composed of Bacillus species. *P < 0.05. Standard error of the mean = 0.340 (CLDN4 = claudin-4); 0.220 (TJP1 = tight junction protein-1); 0.160 (TPH1 = tryptophan hydroxylase-1); 0.108 (SOD1 = superoxide dismutase-1); 0.396 (CAT = catalase); 0.171 (OCLN = occludin).

Figure 2.

Relative expression of genes in the cecum of Angus × Nellore bulls with or without intake of DFM composed of Bacillus species. Standard error of the mean = 0.215 (HSPB1 = heat shock protein family B, member 1); 0.327 (HSPA4 = heat shock protein family A, member 4).

Discussion

Provision of DFM composed of bacteria of the Bacillus genus has the potential to improve beef cattle performance; however, the responses in the literature are inconsistent and scarce. In this study, we have examined the performance and the physiological characteristics of Nellore × Angus bulls consuming a DFM composed of B. licheniformis and B. subtilis bacteria. Bacteria with sporulation capacity, such as Bacillus spp., have greater thermal stability, viability during storage, and, above all, tolerance to the adverse conditions of the gastrointestinal tract than other feed additives (Leser et al., 2008; Luise et al., 2022; Cappellozza et al., 2023a). Beef cattle receiving high-energy diets can benefit from supplementation with Bacillus spp. through its effects in regulating the microbiome, preventing and combating colonization of potentially harmful bacteria, and improving the barrier function and integrity of the gastrointestinal tract (McAllister et al., 2011; Lambo et al., 2021; Boll et al., 2024; El Jeni et al., 2024).

In the present study, bulls consuming the DFM had 3.1% lower DMI than the animals in the control group. Several factors affect DMI in beef cattle, such as body fat content, sex, physiological status, heat stress, and the addition of feed additives (NASEM, 2016). The effect of Bacillus spp. on DMI of cattle has varied among studies (Smock et al., 2020; Dias et al., 2022; Goetz et al., 2023; Lopez et al., 2024; Silva et al., 2024). Under suitable conditions, DMI decreases as a result of increased digestibility of nutrients in the gastrointestinal tract (Tyrrell and Moe, 1975), which allows the animal to meet its energy demands with a smaller amount of feed (Cantalapiedra-Hijar et al., 2018; Jacovaci et al., 2021). In this respect, the ability that B. licheniformis and B. subtilis have in producing a wide range and number of enzymes (Elshaghabee et al., 2017; Luise et al., 2022; Cappellozza et al., 2023b) may have helped increase the digestibility of the nutrients, especially of starch. Using the same bacterial species as the present study, Pan et al. (2022) observed 5.75% improvement in in vitro starch digestibility for grain corn. Related to that, evidence indicates that one of the species used, B. licheniformis, specifically hydrolyzes starch in environments with greater availability of this nutrient (de Boer et al., 1994; Deng et al., 2018; Lopez et al., 2024; Silva et al., 2024). Furthermore, bulls fed DFM tended (P = 0.10) to have lower starch concentration in feces. A potential explanation for this finding is that the DFM may have increased starch degradation, as discussed above, and led to lower fecal excretion, once B. licheniformis and B. subtilis are capable of degrading starch (Pan et al., 2022). However, our study did not perform variables that could confirm this theory. Thus, future studies are warranted in evaluating the effects of DFM on carbohydrates degradation. The lower fecal concentration in DFM bulls may also be related to the lower DMI in those animals, once several studies (Moharrery et al., 2014; Owens et al., 2016) suggested that total tract starch digestibility and fecal starch concentration in cattle closely depend on dietary starch intake.

As the ADG was not affected by the treatment (mean of 2.05 kg), along with the lowered aforementioned DMI, bulls that received the DFM exhibited 6% greater G:F. In agreement with our findings, Lopez et al. (2024) reported that upon feeding crossbred Angus cattle with DFM composed of Lactobacillus animalis, Propionibacterium freudenreichii, B. subtilis, and B. licheniformis in the respective ratio of 1:1:1:3, G:F was improved by 5.7%. Dias et al. (2022) also found a trend of improvement of 4.5% in ADG and 3.4% in G:F in Nellore (Bos indicus) cattle in contrast between different DFM and a control group. Divergences in the effects found in the literature may be attributed to the differences of breeds, microbial species, strains, concentrations, and types of diets. Supplementation of the current DFM seems not to have negative for cattle. Conversely, the potential for improving performance is frequently reported (Sun et al., 2013; Calaca et al., 2022; Dias et al., 2022; Lopez et al., 2024). Furthermore, the fact that the gene expression of a key antioxidant enzyme (SOD1) was greater in the duodenum of DFM-fed bulls may be associated with a greater intestinal nutrient efficiency, but this needs to be addressed and confirmed in future research efforts. Lastly and supporting our results, others have reported that Bacillus spp. stimulates the expression, synthesis, and release of antioxidant enzymes in the lower GIT of livestock animals (Wang et al., 2017).

Plasma urea-N was 6.2% lower for bulls consuming DFM. This response may be attributed to the lower DMI in those bulls, once intake and plasma urea-N concentrations are highly correlated. Another potential explanation for that is the presence of B. licheniformis as a feed additive (Qiao et al., 2010). Those authors reported that the addition of B. licheniformis in the diet decreased ammonia concentration; and consequently, stimulated microbial growth in the rumen of Holstein cows. They suggested that the reduction in ruminal ammonia resulted from greater assimilation by the bacteria. However, it should be highlighted that this explanation should be considered cautiously since we did not evaluate rumen parameters, nitrogen balance, production of microbial protein, and other aspects.

Conclusions

Supplementation with DFM based on B. licheniformis and B. subtilis to Nellore × Angus bulls in the finishing phase decreased DMI, did not influence ADG, improved G:F, and increased the expression of genes important for duodenal function.

Glossary

Abbreviations

ADG

average daily gain

AS

ash

CON

control

CP

crude protein

DFM

direct-fed microbials

DM

dry matter

DMI

dry matter intake

EE

ether extract

G:F

gain:feed ratio

HCW

hot carcass weight

HCY

hot carcass yield

LPS

lipopolysaccharide

NDF

neutral detergent fiber

NFC

non-fiber carbohydrates

peNDF

physically effective neutral detergent fiber

ST

starch

Conflict of interest statement

The authors declare no conflicts of interest.