Isolation of bacterial consortia with probiotic potential from the rumen of tropical calves

Abstract

Probiotics are live microorganisms that confer health benefits to their animal host by balancing the composition of its gastrointestinal microbiota and modulating its immune response. In this work, we studied bacterial consortia isolated from the rumen of 28‐ and 42‐day‐old calves to select those showing probiotic capacity. Consortia were characterized and their growth dynamics were determined in several growth media. The number of viable bacteria was larger in the Man, Rogosa and Sharpe broth (MRS) than in nutritive medium A (MNA) and the largest was for A3D42. Antibiotic susceptibility of bacterial consortia in MRS was higher than in MNA and the most susceptible samples were A1D28 and A3D42. In turn, A3D42 showed the highest tolerance to bile salts in MRS and MNA. Moreover, all bacterial consortia showed optimal growth at pH 5, 5.5, 6 and 7 in both media, while their temperature tolerance was higher in MRS. The antagonistic activity of bacterial consortia in MNA was higher than in MRS with A2D42 showing the best antagonistic activity for Pseudomona aureginosa (ATCC 9027) and Staphylococcus aureus (ATCC 6538) in MNA. Additionally, A1D42 and A2D42 in MRS and A3D42 in MNA had significant adhesion to mucins, and A1D42 in MRS had the highest. Regarding their species composition, all bacterial consortia in MRS belonged to the phylum Firmicutes, and the class Bacilli and bacterial consortia in MNA belonged to three phyla; Proteobacteria, Firmicutes, and Bacteroidetes. Lactobacillus casei, Lactobacillus rhamnosus, Lactobacillus fermentum, and Lactobacillus johnsonii were identified in all bacterial consortia in MRS broth. Based on these results, A1D42 and A3D42 grown in MRS showed the best potential as probiotics for calves, which could result in health benefits and improve their production.

1. INTRODUCTION

Probiotics are microorganisms that confer health benefits to their host animal when administered in adequate doses owing to their nutritional, immunological, bacteriostatic and bactericide effects (Food and Agriculture Organization of the United Nations/World Health Organization FAO/WHO, 2002; Guarner et al., 2005; Pineiro & Stanton, 2007). Microorganisms with probiotic capacity are an alternative for the prevention and treatment of some calf pathologies, as well as to improve their productivity (Uyeno et al., 2015). Calf weaning and changes in feeding cause stress, which inhibit the colonization of the gastrointestinal tract (GIT) by beneficial microbiota and increase the risk of its colonization by pathogens. This results in common problems in cattle production such as disease (Signorini et al., 2012; Ventura et al., 2013) and the subsequent economic losses due to the cost of treatment, calf mortality and calf morbidity (Ok et al., 2009).

There is evidence that GIT microbiota stimulates the host immune function (Hooper et al., 2012; Olszak et al., 2012; Rosmini et al., 2004). Moreover, probiotics may also prevent the colonization of the digestive tract of young animals by pathogens that cause gastrointestinal affections (Rosmini et al., 2004). The mechanism used by these beneficial bacteria is colonization resistance or barrier effect (Isolauri et al., 2001), which prevents colonization pathogens intestinal sites. Some lactic acid bacteria (LAB) have been reported to have immunomodulatory and protective effects and may prevent infectious pathologies (Hooper et al., 2012; Isolauri et al., 2001).

Hence, the ideal microbial characteristics for probiotic candidates are being a normal inhabitant of the target species, non‐toxic, non‐pathogenic, resistant to the GIT pH, temperature, and bile salts, antagonist to pathogenic bacteria, adherent to the intestinal mucosa, with the potential of modulating the immune response, and genetically stable (Hernández‐Peñaranda, 2003). Therefore, when searching for wild‐type probiotic candidates, the best strategy is to obtain the bacterial strains or consortia (BC) from the familiar environment of the microorganism in the GIT of the host. Those used in the food industry are LAB, mainly Streptococcus termophilus and Lactobacillus, Bifidobacterium sp., Propionobacterium sp., Enterococcus faecium, Lactococcus lactis, and Escherichia coli strain Nissle (Adams & Marteau, 1995; Pineiro & Stanton, 2007; Weimer, 2015).

In this work, we isolated and characterized bacterial communities from the rumen of 28‐ and 42‐day‐old calves from the Mexican tropic that show probiotic properties, which could be used to develop calf‐specific probiotics and improve their health conditions increasing their productive performance.

2. MATERIALS AND METHODS

2.1. Isolation of bacterial consortia and culture conditions

A total of six ruminal fluid samples were collected through an orogastric probe from clinically healthy F1 Holstein × Zebu calves raised at the Centro de Enseñanza, Investigación y Extensión en Ganadería Tropical (CEIEGT, FMVZ‐UNAM) of 28 (A1D28, A2D28, A3D28) and 42 days of age (A1D42, A2D42, A3D42) and stored in sterile tubes. Experimental protocols were approved by the internal committee for the care and use of experimental animals (CICUAE.DC‐2019/4‐2, UNAM) according to the NORMA Oficial Mexicana NOM‐062‐ZOO‐1999 (2001). At birth, the calves were fed colostrum and were with their mothers until weaning. They began to eat forage after the first week of age. All the animals were sacrificed at CEIEGT and rumen fluid samples were frozen and stored at −70°C until further analysis (Liang et al., 2014, 2016). All rumen fluid samples were cultured in Man, Rogosa and Sharpe broth (MRS, DIFCO™, USA; Sigma‐Aldrich), thioglycolate broth (TG, Becton, Dickinson, and Company Sparks), and nutritive medium A (MNA) (Aguilar‐González & Mora‐Izaguirre, 2016) at 37°C under anaerobic conditions until they reached 0.5−0.6 optical density at 600 nm (OD_600 nm) measured with an HP/Agilent 8453UV‐Vis spectrophotometer (SpectraLab Scientific). From each culture, a 1 ml sample was frozen in MRS, TG, and MNA medium containing 80% of glycerol and stored at −70°C to evaluate probiotic properties, and Propionibacterium acidipropionici (ATCC 4965) was used as positive control.

The growth rate of all BC cultured in MRS, TG, and MNA broths was determined by measuring their OD_600 nm every 2 h until an OD of 0.5−0.6 was reached. The data were expressed as a time and OD growth rate curve.

2.2. Quantification of viable bacteria

BC were grown in MRS, TG, and MNA broths to an OD_600 nm of 0.5−0.6 and diluted to a ratio of 1:1 × 10⁹ (v/v) by serial 10‐fold dilutions in sterile water, seeded on MRS and MNA agar plates, and incubated anaerobically at 37°C for 24 h. Colonies were counted and CFU/ml were determined. MNA was used to maintain the stability of bacterial populations as a non‐specific broth.

2.3. Probiotic tests

2.3.1. Susceptibility to antibiotics

One hundred microliters of BC culture (OD_600 nm = 0.5−0.6) was seeded on MRS and MNA agar plates. Multidisc with antibiotics Gram (−)‐specific: AK, amikacin (30 µg); AM, ampicillin (10 µg); CB, carbenicillin (100 µg); CL, chloramphenicol (30 µg); NET, netilmicin (30 µg); NF, nitrofurantoin (300 µg); NOF, norfloxacin (10 µg). Gram (+)‐specific: DC, dicloxacillin (1 µg); CLM, clindamicin (30 µg); E, erythromycin (15 µg); PE, penicillin (10 U); VA, vancomycin (30 µg); TE, tetracycline (30 µg). Gram (−) and (+)‐specific: AM, ampicillin (30 and 10 µg respectively); CF, cephalothin (30 µg); CFX, cefotaxime (30 µg); CPF, ciprofloxacin (5 µg); GE, gentamicin (10 µg); SXT, sulfamethoxazole/trimethoprim (25 µg) (MULTIBAC‐ID, Investigación Diagnostica) were placed on the inoculated plates and incubated anaerobically at 37°C for 24 h. The diameter of the inhibition zone (mm) was measured and the results were expressed as resistant (R), intermediate (I) and susceptible (S) according to the manufacturer's instructions. Propionibacterium acidipropionici (ATCC 4965) was grown on MRS agar plates.

2.3.2. Resistance to bile salts

Based on the antibiotic susceptibility results, the resistance to bile salts was determined only for samples from 42‐day‐old calves.

One hundred microliters of BC culture (OD_600 nm = 0.5−0.6) was inoculated in MRS and MNA broths at pH 7.5 with or without 0.3% ox‐bile salt and grown anaerobically at 37°C. The growth curve of each BC culture was monitored by measuring OD_600 nm every 2 h up to 6 h. The cultures were then diluted by 10‐fold serial dilutions to 1:1 × 10⁸ and 1:1 × 10⁹ (v/v, ratio), plated on MRS and MNA agar, and incubated anaerobically at 37°C for 24 h. The survival rate was expressed as percentage of CFU/ml.

2.3.3. Resistance to pH

One hundred microliters of BC culture (OD_600 nm = 0.5−0.6) was inoculated in MRS and MNA broths at pHs 2, 3, 4, 5, 5.5, 6, 7 and control (pH = 6.5) and incubated anaerobically at 37°C for 24 and 48 h. The growth curve of each BC culture was monitored by measuring OD_600 nm and six 10‐fold serial dilutions (v/v, ratio) starting at 1:1 × 10¹ (v/v, ratio) were plated on MRS and MNA agar and grown anaerobically at 37°C for 24 h. Results were expressed as survival in percentage of CFU/ml.

2.3.4. Tolerance to temperature

One hundred microliters of BC culture (OD_600 nm = 0.5−0.6) was inoculated in MRS and MNA broths and incubated anaerobically at 30°C, 37°C, and 45°C, with OD_600 nm measured at 24 and 48 h. The cultures were diluted to 1:1 × 10⁶, 1:1 × 10⁷, 1:1 × 10⁸, and 1:1 × 10⁹ (v/v, ratio) and plated on MRS and MNA agar at 37°C for 24 h. Survival results were expressed as percentage of CFU/ml.

2.3.5. Antagonistic activity

Bacterial antagonism was evaluated by disc diffusion method (Bhunia et al., 1988) with modifications, employing two pathogenic strains (Salmonella typhimurium ATCC 19028 and E. coli ATCC 11229, donated by the Facultad de Medicina, Universidad Autónoma de Quéretaro). The BC cultures were incubated anaerobically in MRS and MNA broths, and the pathogenic strains in LB broth at 37°C overnight. One hundred microliters of pathogens was spread on Müller−Hinton agar plates. BC cultures were centrifugated at 1006 g for 5 min; after that, BC cultures supernatants were added to sterile paper discs and placed on the Müller−Hinton agar plates. After incubation at 37°C for 24 h, the growth inhibition zone (mm) was measured. The antagonistic activity was considered positive when the average growth inhibition zone diameter was ≥10 mm.

2.3.6. Adhesion to mucus

According to all the previous results, it was decided to continue working with that BC that showed better results in MRS. Sheep small intestinal mucin was prepared according to the methods described by Vélez et al. (2010) and used to assess the adhesion of BC. Briefly, BC were labelled by incubation with 250 μl of 4,6‐dichlorotriazinyl aminofluorescein in 500 μl of PBS at 60°C for 2 h and washed three times by resuspension in PBS and collected by centrifugation at 10,000g 10 min at 4°C. Ten microliters of labelled BC was mixed with 200 μl of ovine mucin and incubated at 37°C for 2 h, washed three times with PBS and diluted 1:1000 and 1:100,000, and fixed with paraformaldehyde (4%) on microscope slides. Adherence was expressed as CFU/ml. Viable bacteria that adhered to the mucus were observed under a microscope. ImageJ was used to quantitate viable bacteria. For microscopic analysis, the MosaiX module for the APOTOME system with the ×40/1.30 DIC (UV) VIS‐IR M27 Plan‐Apochromat oil immersion objective was used to obtain a full mosaic image (1 mm²). Three individual image stacks were collected and assembled by the MosaiX system (Carl Zeiss) for each histological slide. ImageJ64 and FIJI were used for the analysis of the 2D images. Quantitation of particles was performed using Yen's image thresholding method (Yen et al., 1995).

2.4. Genomic characterization of the rumen microbiome of 42‐day‐old calves

Genomic DNA of the BC rumen microbiota was used to determine the bacterial diversity. Rumen DNA extraction of BC A1D42, A1D42 and A3D42 in MRS was performed according to the RBB + C method (Yu & Morrison, 2004). DNA concentration was measured with a Qubit 3.0 fluorometer (Life Technologies), and DNA integrity was verified by agarose gel electrophoresis (1%). The purified DNA was subject to partial 16S DNA gene sequencing to confirm bacteria identity. Sequencing of the rumen microbiome was performed by Research and Testing Laboratory (RTL) Genomics with Illumina Myseq sequencer. Illumina produces FASTQ files with a prhred offset of +33. FASTQ files contain all the raw sequence data generated by the sequencer; they may contain information regarding the primer for amplicon sequencing. To allow for more run flexibility, RTL genomics used a two‐step PCR process for Illumina sequencing that uses universal adapters and sequences the forward and reverse primers. FASTQ files generated by Illumina sequencer come in two forms depending on the method used: paired or single end; in this case, we use pair‐end sequence.

Results were expressed as relative abundance (%). Data quality control and analyses were mostly performed using the USEARCH V.11 pipeline. The FASTQ forward and reverse files were merged into a single FASTQ file per sample; quality control and processing included removing adapters and cutting the sequences to length‐based filtering of 400 bp (reads smaller than 200 bp were excluded from the analysis). The resulting read files were then aligned to RDP V.16 to define operational taxonomic units (OTUs) for taxonomy assignment; the OTU table was generated at 97% identity. The Uclust method was used to cluster the reads into OTUs (Edgar, 2010).

2.5. Relative quantification of Lactobacilli and Enterobacteria

Rumen DNA extraction was performed as described above. A total of 50 ng of DNA measured in a Nanodrop 1000 spectrophotometer was analysed in a StepOne Real‐Time PCR System. The primers used were: Universal primers: BACT1369 F: CGGTGAATACGTTCYCGG, PROK1492 R: GGWTACCTTGTTACGACTT (Suzuki et al., 2000); Lactobacillus F: AGCAGTAGGGAATCTTCCA, Lactobacillus R: CACCGCTACACATGGAG (Rinttila et al., 2004) and Enterobacteria F: ATGTTACAACCAAAGCGTACA, Enterobacteria R: TTACCYTGACGCTTAACTGC (Takahashi et al., 2017). Conditions of PCR: 40 cycles; 95°C 10 min, 95°C 10 s, 56°C 15 s, 72°C 15 s. Melting curve: 56°C 1 min, increasing temperature 0.3°/s up to 95°C. Samples were analysed in duplicate. Relative abundance of the DNA target was calculated by the De Gregoris et al. (2011) method.

2.6. Identification of Lactobacilli

The molecular identification of Lactobacilli in BC cultured in MRS broth was made only in BC A1D42 and A3D42, which showed better results and it was performed by PCR. The Lactobacillus tested were Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus rhamnosus, Lactobacillus fermentum, Lactobacillus ruminis, and Lactobacillus johnsonii, using primers designed in this study (Table 1). PCR conditions were 95°C for 10 min, 35 cycles at 57°C 30 s for primer annealing and 72°C for 60 s for extension, and a final extension at 72°C for 10 min. PCR products were analysed by gel electrophoresis (1%) and visualized under UV light using ImageMaster (Pharmacia Biotech). A 1 kb ladder (Invitrogen) was used to verify PCR size. Illumina MiSeq sequencing of the DNA samples was performed using primers covering the V4 region (357wF:3'‐CCTACGGGNGGCWGCAG‐5'; 785R :3'‐GACTACHVGGGTATCTAATCC‐5') obtaining 250 nucleotides paired reads. The obtained sequence data were compared with the Genbank database using the Basic Local Alignment Search Tool (BLAST) for the final identification of the Lactobacillus spp.

Table 1.

PCR primers used to detect Lactobacillus

Target	Sequence (5'−3')	Annealing temp. (°C)	Amplicon size (bp)
Bacterial primersa	357wF CCTACGGGNGGCWGCAG	—	—
	785R  GACTACHVGGGTATCTAATCC
L. acidophilus	LacidopF CAACCAGAAAGTCAGGCTA	57.5	302
	LacidopR CTAATCCTGTTCGCTACCAT
L. casei	LcaseiF CGCAGCTAACGCATTAAGCA	57.9	462
	LcaseiR TTGCAGCCTACAGTCCGAA
L. Rammusus	LrammuF CAACGCGAAGAACCTTACCAG	57	352
	LramnuR TTGCAGCCTACAGTCCGAAC
L. fermentum	LfermenF CGGTGTGCCTAATACATGCAA	56.5	300
	LfermenR TCAGTCTCTCAACTCGGCTA
L. ruminis	LruminisF TTTGTAACACCCCAAAGTCG	52.7	414
	LruminisR ATATCGCAGTTAGTCACGTC
L. johnsonii	LjohnsonF CACGGCCCAAACTCCTACGG	58	455
	LjohnsonR TTCGCTACCCATGCTTTCGAG

^a Primers used by RTL genomics.

2.7. Statistical analysis

Data were measured in triplicate (antibiotic susceptibility, bile salts resistance, resistance to pH, temperature tolerance, and antagonistic activity) or duplicate (in vitro adhesion assay), and expressed as means ± standard deviation (SD); differences between samples were analysed by one‐way analysis of variance (ANOVA) and Tukey−Kramer tests, and the differences with p < 0.05 were considered significant. Data were analysed using the GraphPad Prism version 5.0 for Windows (GraphPad Software) and the JMP version 8.0 software (SAS Institute). Taxonomic assignment was performed using the RDP V.16 reference database with 97% identity. Data were analysed in the R platform (3.6) with the Phyloseq library.

3. RESULTS

3.1. Growth rate

There was no difference in the growth rate of the BC in MRS and TG broths. Nevertheless, sample A1D42 had the highest rate in both media (Figure 1a,b). A2D28 and A2D42 in MNA broth had the lowest growth, whereas A3D42 had the highest rate in MRS, MNA and TG, although it was not significant (p ˃ 0.05) (Figure 1c).

Figure 1.

Growth rate (OD/incubation time (h)) of BC (A1D28, A2D28, A3D28, A1D42, A2D42 and A3D42) inoculated in (a) MRS, (b) TG and (c) MNA broth. Colony formation units (CFU)/ml of BC inoculated in (d) MRS, TG and MNA broth, and (e) MRS and MNA. Data are mean ± SD. Values with different letters are different between media; Tukey−Kramer (p < 0.05), n = 3. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth, thioglycolate broth [Color figure can be viewed at wileyonlinelibrary.com]

3.2. Quantitation of viable bacteria

Consortia in TG had the highest growth performance in terms of viable bacteria (CFU/ml) at both ages (28 and 42 days, p < 0.05), except for A3D28, while A2D42 had the largest count in TG broth (Figure 1d). When only MRS and MNA were compared, larger counts were observed in the former (Figure 1e, p < 0.05) with A3D42 showing the largest count in MRS.

3.3. Probiotic tests

3.3.1. Susceptibility to antibiotics

The susceptibility profile for antibiotics specific for Gram (−) and Gram (+) bacteria of BC isolates of 28‐ and 42‐day‐old calves and P. acidipropinici are shown in Tables 2 and 3 respectively. BC grown in MRS broth for both ages was more susceptible than in MNA broth for both antibiotic types. For Gram (−) antibiotics, BC from A1D28 and A3D42 were the most susceptible. For Gram (+), BC isolates from 42 days in MRS were more susceptible than those from Day 28. Propionibacterium acidipropionici was susceptible to all Gram (−) antibiotics tested, but was resistant to the Gram (+) antibiotics, gentamycin and tetracycline (Table 3).

Table 2.

Susceptibility profile to Gram (−) and (+) antibiotics of potential probiotic BC grown on MRS and MNA broths

Gram (−)						Gram (+)
A1D28a	MRSb	MNAb	A1D42a	MRSb	MNAb	A1D28a	MRSb	MNAb	A1D42a	MRSb	MNAb
AK	R	R	AK	R	R	AM	R	R	AM	S	R
AM	R	R	AM	R	R	CF	S	R	CF	S	R
CB	I	R	CB	I	R	CFX	S	S	CFX	S	R
CF	S	I	CF	R	S	DC	R	R	DC	S	R
CFX	S	I	CFX	S	I	CPF	R	R	CPF	R	I
CPF	R	I	CPF	R	I	GE	S	R	GE	R	R
CL	S	R	CL	S	I	CLM	S	R	CLM	S	R
GE	S	I	GE	R	R	E	S	R	E	S	R
NET	R	R	NET	R	R	SXT	R	R	SXT	R	R
NF	S	S	NF	S	S	PE	S	R	PE	S	R
NOF	R	R	NOF	R	I	VA	S	R	VA	S	R
SXT	R	R	SXT	R	R	TE	R	R	TE	R	R

A2D28 a			A2D42 a			A2D28 a			A2D42 a
AK	R	R	AK	R	R	AM	S	R	AM	S	R
AM	R	R	AM	R	R	CF	R	R	CF	S	R
CB	R	R	CB	S	R	CFX	I	R	CFX	S	R
CF	I	S	CF	S	I	DC	R	R	DC	S	R
CFX	I	R	CFX	S	I	CPF	R	S	CPF	I	S
CPF	R	I	CPF	R	R	GE	R	R	GE	R	R
CL	I	R	CL	S	R	CLM	S	R	CLM	S	R
GE	R	R	GE	R	R	E	I	R	E	S	R
NET	R	R	NET	R	R	SXT	R	R	SXT	I	R
NF	S	S	NF	R	S	PE	S	R	PE	S	R
NOF	R	S	NOF	R	R	VA	R	R	VA	S	R
SXT	R	R	SXT	R	R	TE	R	R	TE	R	R

A3D28 a			A3D42 a			A3D28 a			A3D42 a
AK	R	R	AK	R	R	AM	R	R	AM	S	R
AM	R	R	AM	R	R	CF	I	R	CF	S	R
CB	R	R	CB	S	R	CFX	I	R	CFX	S	S
CF	I	I	CF	S	R	DC	R	R	DC	S	R
CFX	I	I	CFX	S	R	CPF	R	S	CPF	R	S
CPF	R	I	CPF	R	R	GE	R	R	GE	S	R
CL	S	I	CL	S	R	CLM	R	R	CLM	I	R
GE	R	R	GE	R	R	E	S	R	E	R	R
NET	R	R	NET	R	R	SXT	R	R	SXT	S	R
NF	I	S	NF	S	S	PE	S	R	PE	S	R
NOF	R	R	NOF	R	R	VA	R	R	VA	R	R
SXT	R	R	SXT	R	R	TE	R	R	TE	S	R

Abbreviations: MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth.

^a BC, bacterial consortium, Gram (−)‐specific: AK, Amikacin (30 µg); AM, Ampicillin (10 µg); CB, Carbenicillin (100 µg); CL, Chloramphenicol (30 µg); NET, Netilmicin (30 µg); NF, Nitrofurantoin (300 µg); NOF, Norfloxacin (10 µg). Gram (+)‐specific: DC, Dicloxacillin (1 µg); CLM, Clindamicin (30 µg); E, Erythromycin (15 µg); PE, penicillin (10 U); VA, Vancomycin (30 µg); TE, Tetracycline (30 µg). Gram (−) and (+)‐specific: AM, Ampicillin (30 and 10 µg respectively); CF, Cephalothin (30 µg); CFX, Cefotaxime (30 µg); CPF, Ciprofloxacin (5 µg); GE, Gentamicin (10 µg); SXT, Sulfamethoxazole/Trimethoprim (25 µg).

^b R, resistant; I, intermediate; S, susceptible, n = 3.

Table 3.

Gram (−) and (+) antibiotic susceptibility of P. acidipropionici

Gram (−)a	MNAb	Gram (+)a	MNAb
AK	S	AM	S
AM	S	CF	R
CB	R	CFX	I
CF	S	DC	R
CFX	I	CPF	I
CPF	I	GE	R
CL	S	CLM	R
GE	S	E	R
NET	S	SXT	I
NF	S	PE	R
NOF	R	VA	S
SXT	S	TE	R

Abbreviation: MNA, nutritive medium A.

^a AK, Amikacin (30 µg); AM, Ampicillin (10 µg); CF, Cephalothin (30 µg); CL, Chloramphenicol (30 µg); CPF, Ciprofloxacin (5 µg); GE, Gentamicin (10 µg); NET, Netilmicin (30 µg); NF, Nitrofurantoin (300 µg); SXT, Sulfamethoxazole/Trimethoprim (25 µg); VA, Vancomycin (30 µg); TE, Tetracycline (30 µg).

^b R, resistant; I, intermediate; S, susceptible, n = 3.

Since the 42‐day‐old animals were more susceptible to antibiotics than 28‐day‐old animals, the following probiotic tests were carried out only on the samples from 42‐day‐old animals.

3.3.2. Resistance to bile salts

All BC isolates showed similar tolerance to 0.3% ox‐bile salts in the broths tested (Figure 2), and A3D42 showed the highest survival (53.73% for MRS and 57.34% for MNA broth, p < 0.05, Figure 2).

Figure 2.

Bile salts survival of BC (A1D42, A2D42 and A3D42) inoculated in MRS and MNA broth. Data are mean ± SD. Values with different letters are different between bacterial consortia; Tukey−Kramer (p < 0.05), n = 3. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth

3.3.3. Tolerance to pH

Survival (%) of BC improved with the increasing pH in MNA broth. At 24 h (Figure 3a), there were differences (p < 0.05) at pH 4 and 7, in which A1D42 and A3D42 had the highest survival. In contrast, at 48 h (Figure 4a), there were differences (p < 0.05) between them, at pH 5.5, and A3D42 had the highest survival rate. Regarding the survival of BC in MRS at 24 h, the lowest rates were at pH 2, 3 and 4, which improved from pH 5 upwards (Figure 3b), and A3D42 had the highest rate. Similar survival results were obtained at 48 h (Figure 4b), which were slightly higher than at 24 h. Moreover, there was a difference at pH 2, in which A2D42 had the highest rate. At pH 3 and 5, A1D42 had the highest survival and P. acidipropionici (ATCC 4965) also showed high survival.

Figure 3.

pH survival of BC (A1D42, A2D42 and A3D42) inoculated in (a) MNA and (b) MRS broth for 24 h. Data are mean ± SD. Values with different letters are different between bacterial consortia; Tukey−Kramer (p < 0.05), n = 3. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth

Figure 4.

pH survival of potential BC (A1D42, A2D42 and A3D42) inoculated in (a) MNA and (b) MRS broth for 48 h. Data are mean ± SD. Values with different letters are different between bacterial consortia; Tukey−Kramer (p < 0.05), n = 3. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth

3.3.4. Tolerance to temperature

All BC isolates in MNA and MRS broths at 24 and 48 h had large numbers of viable bacteria (CFU/ml) at 37°C, and there were no differences between them (Table 4). In general, BC grown in MRS had larger counts than in MNA and were larger at 30°C than at 45°C. There were no differences between the BC in MRS at 24 h, except A1D42 at 45°C, which was lower than the others BC. The significant differences in BC in MNA were: A2D42 had the largest count at 30°C for 24 h, while A3D42 had the largest at 45°C in MNA. The CFU/ml was slightly lower in the BC at 48 h than 24 h.

Table 4.

Temperature tolerance of BC (A1D42, A2D42 and A3D42) grown on MRS and MNA broths ^†

	MRS 24 h			MNA 24 h
BC	30°C	37°C	45°C	30°C	37°C	45°C
A1D42	1.92E+11 ± 2.52E+10a	2.10E+11 ± 6.24E+10a	1.30E+11 ± 1.00E+10b	1.67E+08 ± 5.77E+07b	5.40E+10 ± 1.27E+10a	1.00E+09 ± 0.00E+00b
A2D42	1.67E+11 ± 6.00E+09a	2.07E+11 ± 2.52E+10a	1.63E+11 ± 5.77E+09a	1.03E+10 ± 2.52E+09a	3.57E+10 ± 7.02E+09a	1.33E+07 ± 5.77E+06b
A3D42	1.82E+11 ± 9.07E+09a	2.30E+11 ± 2.00E+10a	1.60E+11 ± 1.00E+10a	4.00E+08 ± 1.00E+08b	4.47E+10 ± 8.33E+09a	1.33E+10 ± 5.77E+09a

	MRS 48 h			MNA 48 h
BC	30°C	37°C	45°C	30°C	37°C	45°C
A1D42	1.20E+11 ± 2.92E+10a	1.53E+11 ± 2.69E+10a	8.17E+10 ± 2.65E+10a	6.33E+09 ± 1.53E+09a	4.00E+10 ± 1.00E+10a	1.33E+07 ± 5.77E+06b
A2D42	1.17E+11 ± 3.16E+10a	1.77E+11 ± 2.56E+10a	5.67E+10 ± 3.31E+10a	5.67E+09 ± 1.53E+09a	5.00E+10 ± 1.00E+10a	4.00E+08 ± 1.00E+08a
A3D42	1.23E+11 ± 4.73E+10a	1.80E+11 ± 2.64E+10a	1.00E+11 ± 2E+10a	4.67E+09 ± 5.77E+08a	4.33E+10 ± 5.77E+09a	1.00E+08 ± 0.00E+00b

Abbreviations: BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth.

^† Data are presented as CFU/ml mean ± standard deviation. Values with different letters are different between BC at the same temperature and time; Tukey−Kramer (p < 0.05), n = 3.

3.3.5. Antagonistic activity

All the BC isolates showed positive inhibition against S. typhimurium and E. coli because they had a zone of inhibition bigger than 10 mm, except for A3D42 on MRS, which had no inhibition for E. coli, since it had an inhibition zone of 9.33 mm. Although no significant difference was shown, A1D42 in MRS had a bigger zone of inhibition for both pathogens. While A3D42 in MNA had a larger zone of inhibition for both pathogens as well (Table 5).

Table 5.

Antagonistic activity of BC (A1D42, A2D42, and A3D42) grown on MRS and MNA broths ^†

	Zone of inhibition (mm) ‡
	MRS		MNA
BC	Salmonella T. ATCC	E. coli ATCC	Salmonella T. ATCC	E. coli ATCC
A1D42	13.33 ± 0.57a	13.0 ± 1.0a	12.0 ± 0a	10.67 ± 2.0a
A2D42	11.66 ± 1.15a	11.0 ± 1.0ab	11.67 ± 1.15a	11.33 ± 1.15a
A3D42	12.0 ± 1.0a	9.33 ± 0.57b	13.0 ± 1a	13.67 ± 1.52ª

Abbreviations: BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth.

^†Data are presented as mean ± standard deviation, values with different letters are different between BC; Tukey−Kramer (p < 0.05).

^‡The antagonistic activity was considered positive when the average growth inhibition zone diameter was ≥10 mm.

3.3.6. Adhesion to mucus

Adhesion of BC to mucus was higher in MRS than in MNA (Figure 5). A1D42 and A2D42 in MRS had the highest adhesion in MRS (p < 0.05) and A3D42 in MNA. A1D42 in MRS had the highest adhesion, which was even higher than the positive control (P. acidipropionici).

Figure 5.

Adhesion capacity of (a) BC (A1D42, A2D42 and A3D42) inoculated in MNA and MRS broth to intestinal mucus. Images of bacteria bound to intestinal mucus (b) negative control, (c) positive control, and (d) BC A1D42 inoculated in MRS. Each bar represents the mean of the viable count of six data sections of bound bacteria to intestinal mucus and, error bars are ±standard deviation. Strain Propionibacterium acidipropionici was used as a positive control for adhesion. Values with different letters are different between bacterial consortia; Tukey−Kramer (p < 0.05), n = 6. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth [Color figure can be viewed at wileyonlinelibrary.com]

3.4. Characterization of rumen microbiome of 42‐day‐old calves

Sequencing of the microbiome of the BC grown in MNA showed that they consist of species belonging to three phyla: Proteobacteria with the highest relative abundance (60%−95%), followed by Firmicutes (1%−31%), and Bacteroidetes with the lowest relative abundance (0%−8%). It is noteworthy that one sample (A2D42) had no Bacteroidetes, had a low relative abundance of Firmicutes (0% of Clostridia and 5% of Bacilli), and hence, had the highest relative abundance of Proteobacteria (95%). The phylum Firmicutes was represented by the Classes Clostridia and Bacilli, Bacteroidetes had only Bacteroidia, while Proteobacteria had only Grammaproteobacteria. All BC had higher relative abundances of Clostridia (31%−24%) than Bacilli (1%, 5%, 2%), except A2D42 which had no Clostridia and had the highest relative abundance of Bacilli (5%). This BC also lacked Bacteroidia (Bacteroidetes), but most OTUs present belonged to the Class grammaproteobacteria (Proteobacteria) (95%). Moreover, the microbiomes of A1D42 and A3D42 were similar to each other (data are not shown). Regarding the order, A1D42 contained Clostridiales, Enterobacteriales and Bacteroidales with the highest relative abundances (37.6%, 36.2%, and 22.5% respectively) (Figure 6a). Also, the sample A2D42 contained 92.7% of Enterobacteriales (Figure 6a) while A3D42 had 57.1% of Enterobacteriales, 22.9% of Clostridiales and 11.4% of Bacteroidales (Figure 6a). Remarkably, the microbiomes of all samples grown in MRS were entirely composed of the class Bacilli of the phylum Firmicutes (data not shown), all of which belonged to the Order Lactobaciliales (Figure 6a).

Figure 6.

(a) Orders of rumen microbiome sequencing of BC (% relative abundance) inoculated in MRS and MNA broth. Agarose gel electrophoresis of PCR products after amplification of genomic DNA. (b) 1—1 kb Marker, 2—Negative control, 3—Lactobacillus acidophilus A1D42, 4—L. acidophilus A2D42, 5—L. acidophilus A3D42, 6—Lactobacillus casei A1D42, 7—L. casei A2D42, 8—L. casei A3D42, 9—Lactobacillus rhamnosus A1D42, 10—L. rhamnosus A2D42, 11—L. rhamnosus A3D42, 12—Lactobacillus fermentum A1D42 and 13—L. fermentum A2D42. (c) 1—1 kb Marker, 2—Negative control (without DNA), 3—L. fermentum A3D42, 4—Lactobacillus ruminis A1D42, 5—L. ruminis A2D42, 6—L. ruminis A3D42, 7—Lactobacillus johnsonii A1D42, 8—L. johnsonii A2D42, 9—L. johnsonii A3D42, n = 3. BC, bacterial consortium; MNA, nutritive medium A; MRS, Man, Rogosa and Sharpe broth [Color figure can be viewed at wileyonlinelibrary.com]

3.5. Relative quantification of Lactobacilli and Enterobacteria

The relative abundance of Lactobacilli in A3D42 in MRS was 5.64%, which was higher than in A1D42 in MRS (0.13%), whereas the relative abundance of Enterobacteriaceae was: 0.032% in A3D42 in MRS and 0.018% in A1D42 in MRS. The Enterobacteriaceae species found were Klebsiella oxytoca (64%), Klebsiella sp (30%), and E. coli (5%) (data not shown).

3.6. Identification of Lactobacilli

The presence of several species of Lactobacilli was assessed by PCR in BC cultured in MRS broth. All samples revealed the presence of L. casei, L. rhamnosus, L. fermentum, and L. johnsonii (Figure 6b,c).

4. DISCUSSION

The development of probiotics for calves as an alternative to antibiotics to improve their health and productivity has recently taken on great importance. A viable option is to isolate probiotic BC from their natural host due to their adaptation to proliferate in such environment, which increases their chances of survival and of exerting beneficial effects upon their administration to a similar host (Dowarah et al., 2018).

As in vitro analysis is an efficient means to select suitable BC with the highest potential and best probiotic properties for ruminants (Ridwan et al., 2018), we analysed cultured rumen liquid consortia from 28‐ and 42‐day‐old calves in different media and characterized them regarding various parameters. We observed robust growth of all samples in MRS broth which is highly selective for LAB bacterial isolates of the phylum Firmicutes, such as Lactobacilli, common inhabitants of the rumen, and widely used as probiotics in livestock to improve the microbial balance of feed digestibility and animal's health (Food and Agriculture Organization FAO, 2016; Uyeno et al., 2015).

Antibiotic susceptibility is another important property for selecting a potential BC. Probiotics must not contribute to the spread of antibiotic resistance and should not carry transferable antibiotic resistance (Hernández‐Peñaranda, 2003; Venkatasatyanarayana et al., 2017). The resistance of these organisms, however, may be intrinsic due to the absence of targets, low affinity to targets, low permeability, or to their efflux mechanisms. In some cases, however, this resistance may be a desirable trait, as probiotics may help restore the host's intestinal homoeostasis (Berebon et al., 2018; Ouwehand et al., 2016; Wong et al., 2015). We observed that the majority of BC in MNA showed more resistance than in MRS. LAB grows in MRS broth, suggesting that the LAB present in these BC in MRS may be more susceptible to antibiotics than the bacteria present in BC in MNA. LAB are reported to be susceptible to ampicillin, penicillin, chloramphenicol, ceftriaxone and novobiocin (Anandharaj & Sivasankari, 2014; Puniya et al., 2016; Rine et al., 2019) and resistant to oxacillin, vancomycin, ciprofloxacin, streptomycin, tetracycline and gentamicin (Jose et al., 2015; Rine et al., 2019).

A3D42 in MNA was resistant to the entire Gram (−) antibiotic, except to nitrofurantoin, and in MRS was susceptible to carbenicillin, cephalothin, cefotaxime, chloramphenicol and nitrofurantoin.

While A3D42 in MNA also was resistant to all the Gram (+) antibiotics, except to cefotaxime ciprofloxacin, and in A3D42 in MRS was susceptible to ampicillin, cephalothin, cefotaxime, dicloxacillin, gentamycin, sulfamethoxazole/trimethoprim, penicillin and tetracycline. This resistance pattern indicates that they can also be used for the treatment and control of intestinal infections when administered with antibiotics (Kim et al., 2011). These findings also show that antibiotic resistance varies from study to study due to differences in experimental conditions like broth used. The lack of standard methods to test antibiotic susceptibility of candidate probiotic bacteria is the reason why some LAB strains reported to harbour tetracycline‐ or erythromycin‐resistant patterns, as in our results, later turned out to be negative for the resistance tet (M) or erm (B) genes (Berebon et al., 2018; Bertel et al., 2019). This could occur because the microbial community of BC underwent different ecological succession processes in the tests carried out. This is the problem when working with BC.

The ability of the potential probiotic BC to survive intestinal bile salts and grow in the GIT is another important requirement for probiotic selection (Ehrmann et al., 2002). Exposure of probiotic bacteria to bile salts triggers alterations in cellular homoeostasis leading to dissociation of the lipid bilayer and integral proteins of their cell membranes, resulting in leakage of bacterial contents and ultimately cell death. Strompfová and Lauková (2004) proposed that bile salt resistance is the second most important criterion for the colonization and metabolic activity of probiotic bacteria in the host small intestine. In the present study, all BC evaluated tolerated bile salts at 0.3%, and A3D42 had the highest survival, which could be due to its high relative abundance of LAB, that are reported to have good tolerance to bile salts at 0.3% (Yaneisy et al., 2016). Hence, these results suggest that when passing through the intestinal transit, there will be a selection of the most resistant BC strains that can survive the bile salts in the calf intestine, a desirable trait as stated above. These surviving strains will play their beneficial roles in the host if they are beneficial.

Potential probiotic BC must also tolerate the pH of the ruminant GIT environment (reticulum, rumen, omasum, abomasum, small intestine and colon), which ranges from pH 2 to 7.2, albeit it also varies with the diet (Gueimonde & Salminen, 2006; Kern et al., 1974). Our results show that all LAB in MNA survived at pH 2; although the percentage of survival was low, A3D42 had higher survival (%) at 24 h than the other LAB, but it was not significant (p ˃ 0.05). LAB has the property to tolerate acidic pH; a study evaluating the potential of probiotic LAB used in animals survived at acidic pH (Vizoso et al., 2006; Yaneisy et al., 2016). This could be the reason why our BCs survived at acidic pH. Although the percentage of survival of BC in MRS at pH 2, 3 and 4 was low, it increased and had the optimal development between pH 5 and pH 7.

In particular, the physiological pH range in the rumen has been reported to be between 5.5 and 6.9 (Choudhury et al., 2015). In keeping with this, all the BC we evaluated in MNA broth had the optimum development at a similar pH range of 6.5 and 7 at 24 and 48 h. This same pattern was also found in other works in which the highest bacterial growth was shown to occur at pH close to 7 with a significant decrease at acidic pH (Ávila et al., 2010; Bertel et al., 2019; Castillo et al., 2018 and Landa‐Salgado et al., 2019). Moreover, Marrero et al. (2010) evaluated different strains of bacilli and reported high growth of cellulolytic bacteria at pH at the rumen level of 6.6.

The change of temperature to 30°C and 45°C significantly affected the growth of BC in MNA broth, which could be because the rumen temperature is 38−41°C and the ideal temperature for the development of these bacteria is 37°C (Choudhury et al., 2015). On the other hand, the CFU/ml was not affected in MRS broth by the change in temperature, the optimum development was also at 37 °C, followed by 30 °C, and was lower at 45 °C. Then, it could be that all BC grown in MRS withstood GIT temperature better than in MNA. However, Bertel et al. (2019) showed that the highest growth for their rumen liquid samples occurred at 40°C. The work of Lara and Pineda (2013) showed that temperature affects growth depending on the strain, although the optimal growth temperatures were also found to lie between 37°C and 40°C. These results were expected since these microorganisms were isolated from the rumen, where the temperature is normally 40°C due to the enormous number of metabolic processes taking place (Ávila et al., 2010). Regarding the antagonistic activity of BC against Salmonella T. ATCC 19028 and E. coli ATCC 11229, A1D42 in MRS had a higher inhibition zone, and A3D42 in MNA had a bigger inhibition zone for both pathogens. This could be due to the different types of bacteria that grow in MRS and MNA present in the BC, which could be more aggressive to pathogens. E. coli has been reported to be sensitive to L. casei, Lactobacillus plantarum and Lactobacillus helveticus (Çadirci & Çitak, 2005). Lara and Pineda (2013) and Gaitán and Pérez (2007) also reported that Bacilli strains inhibit the growth of E. coli, and Zamudio and Zavaleta (2003) reported large inhibition halos of E. coli and L. monocytogenes. Moreover, Sánchez et al. (2015) showed that LAB bacteria inhibited the growth of Staphylococcus aureus, E. coli, Salmonella sp. and L. monocytogenes. The same pattern was reflected in the study of Gámez et al. (2009) in which L. plantarum inhibited the growth of E. coli, S. typhimurium, Clostridium perfringens and S. aureus. The nature of the predominant inhibitory action was also related to the presence of organic acids, mostly lactic acid, which is the main product of carbohydrate catabolism and contributes to lower pH, creating a hostile environment for pathogenic microorganisms (Gámez et al., 2009; Sánchez et al., 2015), and the presence of other antimicrobial metabolites like acetic acid, hydrogen peroxide, bacteriocins and bacteriocin‐like substances (Juven et al., 1992).

Another important characteristic of a probiotic is its ability to promote bacterial colonization of the intestinal mucosa (Kolter & Greenberg, 2006). In general, we observed good adhesion of all BC which was higher in MRS than in MNA. A study with Lactobacilli spp. strains in dairy calves that reported high adhesion to mucus (Fernández et al., 2018) is consistent with our results. Moreover, Cueto‐Vigil et al. (2010) showed LAB high adhesion to mucus while López and Espinoza (2017) found that L. plantarum can adhere and colonize intestinal cells in vitro. The ability of probiotics to reduce pathogenic microorganisms in the digestive tract may be due to the production of bacteriocin and exclusion due to competition by adherence to the intestinal epithelium. Indeed, some strains of Lactobacilli and Bifidobacterium have hydrophobic surface proteins that promote non‐specific adhesion to animal cells, cover receptor binding sites and prevent pathogenic microorganisms from binding to intestinal epithelium (López & Espinoza, 2017; Molina, 2019).

Finally, the relative species abundance based on DNA sequences of potential probiotic BC showed that all of those cultured in MRS belong to the phylum Firmicutes and all of them to the class Bacilli, whereas the BC cultured in MNA broth belong mainly to the phylum Proteobacteria (class Grammaproteobacteria) (data not shown). Castillo et al. (2018) and Bertel et al. (2019), as in our study, identified Enterococcus and Bacillus in samples with the best results in the probiotic tests. Zhou et al. (2015) found that in ruminal populations the most abundant phyla always fell on Firmicutes and Bacteroidetes. Similarly, Tapio et al. (2017) reported that the main phylum is represented by Firmicutes (55.9%−86.8%) followed by Bacteroidetes (8%−24.4%) and Proteobacteria (0.9%−13.4%). Other rumen sequences, as in our study, report Firmicutes and Bacteroidetes as the phyla with the largest relative abundance in the rumen (Kim et al., 2011). The results of this study also confirmed that the species of BC grown in MRS all belong to the order Lactobacillales. Lactobacilli was more abundant in A3D42 than in A1D42, although both of them have Enterobacteria. Lactobacilli isolates from animal rumen contents have been reported to have potential use as probiotics (Neethu et al., 2015).

These results revealed that all BC tested have functional characteristics of a probiotic although some of them have higher potential.

5. CONCLUSIONS

According to the results described herein, BC A1D42 and A3D42 cultured in MRS broth are good candidates to be used as probiotics in calves, which could improve their health and production. The potential found in these BC is related to their high concentration of Lactobacilli spp.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.

Rodríguez‐González, S. , González‐Dávalos, L. , Robles‐Rodríguez, C. , Lozano‐Flores, C. , Varela‐Echavarría, A. , Shimada, A. , & Mora‐Izaguirre, O. (2023). Isolation of bacterial consortia with probiotic potential from the rumen of tropical calves. Journal of Animal Physiology and Animal Nutrition, 107, 62–76. 10.1111/jpn.13699