search for


Probiotic potentials of non-starch polysaccharide hydrolyzing Bacillus strains newly isolated from guts of the termite Termes propinquus: a source of probiotic bacteria§
Korean J. Microbiol. 2024;60(2):68-82
Published online June 30, 2024
© 2024 The Microbiological Society of Korea.

Putsawee Tomtong and Pinsurang Deevong*

Department of Microbiology, Faculty of Science, Kasetsart University, Bangkok 10900, Thailand
Correspondence to: *E-mail:;
Tel.: +66-2-562-5555; Fax: +66-2-579-2081

§Supplemental material for this article may be found at
Received March 13, 2024; Revised May 14, 2024; Accepted May 16, 2024.
Termite guts are extreme environment harboring diverse microorganisms with ability to degrade a wide range of plant polysaccharides. Beneficial effects of the gut microorganisms are relevant to the various potential properties of individual strains, and relate to the original sources of the strains. Non- starch polysaccharides (NSPs) are main biopolymers in seeds and grains. These antinutritional factors are found abundantly in plant-based feed and able to reduce nutrient digestibility in animals. The present study aimed to investigate termite gut bacteria capable of NSP hydrolysis and potential probiotic properties. Among 172 isolates obtained from the soil-wood interface-feeding termite Termes propinquus, seven isolates were selected based on the presence of three hydrolytic enzymes (cellulase, xylanase, and pectinase). Molecular identification showed that the different isolated strains were closely related to the species in Bacillus subtilis group including Bacillus velezensis, Bacillus amyloliquefaciens, Bacillus siamensis, and Bacillus subtilis. All strains were considered safe based on hemolytic activity and antibiotic susceptibility; six strains survived in both 0.3% bile salts and pH 2.5 conditions for 3 h; four strains inhibited 12 strains of foodborne and gastrointestinal pathogenic bacteria. The strains were evaluated for NaCl and phenol tolerances, antioxidant activity, cell-surface properties including hydrophobicity, auto- and co-aggregations, and adhesion ability to intestinal epithelial cells, Caco-2 and HT-29. All Bacillus strains, except TpNSP_11, showed high survival in simulated gastrointestinal tract conditions and the NSP degradation ability was detected. According to this study, the guts of T. propinquus are indicated as a source of potential probiotic bacteria useful in feed biotechnology.
Keywords : Bacillus, Termes propinquus, non-starch polysaccharides, probiotic, termite gut

Non-starch polysaccharides (NSPs) are abundant natural polymers found in plants, animals, and microorganisms. NSP polymers that mainly occur in plants are cellulose (complex polysaccharide), hemicellulose (such as arabinoxylan, β-glucan, and mannan), and pectin (D-galacturonic acid units) (Deng et al., 2021). Plant products such as seed and grains are employed as the main and most common ingredients of animal feed, to decrease the cost of animal production. After digestion, the dietary NSPs remain undigested constituents with a high water-holding capacity and swell into a gel when combined with various nutrients, impeding nutrient absorption and the interaction ability among epithelium, mucous, and microflora in the gut system (Bedford and Partridge, 2010). One approach to solve this issue is NSP hydrolysis by microbial enzymes. After microbial hydrolysis, NSPs are decreased, and availability of feed nutrients can be improved. In addition, some digested compounds act as prebiotics, increasing growth and activity of beneficial gut bacteria (Sargautiene et al., 2018).

Termites (Isoptera) play an important role in decomposition of plant materials and organic matters in terrestrial ecosystems. Although they are mainly known for causing damage affecting humans, as edible insects, they show potential as an alternative protein source and at least 43 termite species are widely consumed as traditional human food, livestock feed, and therapeutic resource in 29 countries across Africa, America, and Asia (De Figueirêdo et al., 2015). Termites are ordinarily divided into higher and lower termites, and the higher termites show variety in their feeding behavior and have a more complicated gut structure than lower termites. The higher termite Termes propinquus (family Termitidae, subfamily Termitinae) used as bacterial source in the present study feed on wood and humus. Its gut comprises compartments which maintain various physicochemical conditions, such as oxygen and hydrogen levels, pH, and redox potential (Thongaram et al., 2003). Generally, termite guts harbor diverse microorganisms, which are involved in the digestion of plant-cell wall components and numerous studies have previously investigated NSP-hydrolyzing bacteria from termite guts (Tsegaye et al., 2019). However, the beneficial effects of the termite gut bacteria are relevant to the various potential properties of individual strains, and relate to the termite species that is original source of the bacterial strains. The majority of gut bacteria are specific symbionts that have coevolved with termites and their community structure is basically consistent within a genus of termites (Hongoh et al., 2005).

Probiotics are live microorganisms that confer a health benefit on the host; they therefore play an important role in maintaining and improving animal and human health. Their functions involve the digestion of nutrients to increase the nutritional value, the neutralization of pathogenic microorganisms in the intestinal tract, and the stimulation of immunity and physiological processes. Probiotic properties are found in a variety of microorganisms such as those of the genera Bacillus, Lactobacillus, Lactococcus, Streptococcus, Bifidobacterium, Enterococcus, Pediococcus, and various yeast species. Different microbial strains belonging to similar or same species have different probiotic properties and effects, depending on the host species and digestive systems (Bernardeau and Vernoux, 2013). Bacillus spp. are endospore-forming bacteria with high heat and acid resistance, suitable for cultivation and storage. Widely used probiotic species include Bacillus subtilis, Bacillus polyfermenticus, Bacillus coagulans, Bacillus licheniformis, Bacillus pumilus, and some strains of Bacillus cereus (Cutting, 2011). The probiotic Bacillus spp. can tolerate stressful conditions, inhibit pathogens, and have antioxidant and immunity adjustment functions. In addition, Bacillus spp. have functional properties promoting health through the production of extracellular enzymes; these enzymes have been used in human food and animal feed to aid digestion and improve nutrient bioavailability (Nualkul et al., 2022).

Bacteria with probiotic capability have been previously investigated from guts of many insects, such as honey bees (Chandran and Keerthi, 2018; Elzeini et al., 2021). However, less information is known about probiotic potentials of bacteria obtained from termite guts. The present study aimed to isolate Bacillus species that can use NSPs and have probiotic potential from intestines of the soil-wood interface-feeding higher termite, T. propinquus. The isolated strains showing NSP-hydrolytic enzyme production were investigated in vitro probiotic properties. Bacterial survival and enzyme activity in simulated gastrointestinal tract conditions were investigated. Here, we reported the intensive study of potential probiotic bacteria with NSP hydrolytic capability from the intestinal microbiota of termite. This study provides probiotic agents and new knowledge in feed microbiology and biotechnology.

Materials and Methods

Bacterial isolation from termite guts

Termites (Termes propinquus) were collected from the Sakaerat Environmental Research Station, Nakhon Ratchasima, Thailand. On an ice-cold plate, 50 entire guts of worker termites were collected and the content was squeezed, using a sterile pestle, in a microcentrifuge tube containing normal saline. To select endospore-forming bacteria, the homogenate was heated at 80°C for 30 min and subsequently cooled down on ice. The suspension was then diluted and spread on nutrient agar (NA; pH 7), followed by incubation at 37°C for 24–72 h. Bacterial colonies with different morphological features were selected and purified using the streak plate method. Gram-positive endospore-forming bacilli were selected for the next step.

Screening of extracellular enzyme-producing bacteria

NSP-degrading enzyme and other hydrolytic enzyme activities were evaluated using the drop assay. For inoculum preparation, bacterial isolates were cultivated in nutrient broth (NB; pH 7) at 37°C for 16–18 h, then diluted to an OD600 value of 0.2. Bacterial suspension (2 μl) was dropped on NA supplemented with 1% different substrates: NSPs including carboxymethyl cellulose (CMC; Sigma-Aldrich), beech-wood xylan (Megazyme) and citrus pectin (Sigma-Aldrich), and other substrates including skimmed milk (Himedia), and soluble starch (Sigma-Aldrich), tested for cellulase, xylanase, pectinase, protease, and amylase, respectively. Phytase screening medium containing 0.5% Na-phytate from rice (Sigma-Aldrich) and Tween 80 agar containing 0.1% Tween 80 (Ajax Finechem) were used for phytase and lipase assays, respectively. After incubation at 37°C for 48 h, capacity of enzyme production was evaluated by the ratio of the hydrolysis zone diameter (mm) to the bacterial colony diameter (mm). The isolates with broad-spectrum hydrolytic activities were selected for subsequent experiments.

Assay for NSP-hydrolytic enzyme activity

Bacterial inoculum (OD600 = 0.5; 1% v/v) was inoculated to liquid medium, with the composition (g/L) modified from Sreena and Sebastian (2018): yeast extract, 5 g; (NH4)2SO4, 4.5 g; CaCl2 × 2H2O, 0.1 g; MgSO4 × 7H2O, 0.1 g; NaCl, 0.1 g; KH2PO4, 0.7 g; MnSO4 × 4H2O, 0.01 g; FeSO4 × 7H2O, 0.01 g, supplemented with different substrates (CMC, xylan, and pectin). During incubation at 37°C, samples were collected every 24 h for 7 days. After centrifugation, the supernatant was collected and used as crude enzyme, which was incubated with 0.5% each substrate in 50 mM phosphate buffer (pH 7.0) at 37°C for 30 min. The enzyme activity was evaluated by the 3,5-dinitrosalicylic acid (DNS) assay. Absorbance was measured at 540 nm, using glucose, xylose, and galacturonic acid as standards. One unit of enzyme activity was defined as the amount of enzyme that released 1 μmol of reducing sugar per minute under the above assay conditions.

Molecular identification of bacteria

The selected bacteria were identified based on nucleotide sequences of 16S rRNA (~1.5 kb), DNA gyrase A and B (gyrA, ~1.0 kb and gyrB, ~1.2 kb), and RNA polymerase beta (rpoB, ~0.5 kb) genes. Genomic DNA was extracted using the Genomic DNA Purification Kit (Thermo Scientific) according to the manufacturer’s instruction. Polymerase chain reaction (PCR) was carried out using Taq DNA polymerase with the primer pairs 616V/1492R, gyrA-F/gyrA-R, UP-1-F/UP-2r-R, and rpoB-F/rpoB-R, respectively. The primer sequences and PCR conditions were previously described by Nualkul et al. (2022). DNA sequencing was performed by Macrogen, Inc., and the obtained nucleotide sequences were taxonomically identified using BLAST ( A maximum-likelihood tree with 1,000 bootstrap replicates was constructed using MEGA X.

Evaluation of safety profile based on hemolytic activity and antibiotic susceptibility

The bacterial culture was streaked on 5% (v/v) sheep blood agar. After incubation at 37°C for 48 h, the culture plates were determined for hemolytic activity. According to the methods described by Somashekaraiah et al. (2019), antibiotic susceptibility of the isolates was assayed on NA plates using the disc diffusion test. The susceptibility pattern was assessed using ampicillin (10 μg), vancomycin (30 μg), ciprofloxacin (5 μg), streptomycin (10 μg), erythromycin (15 μg), penicillin (6 μg), and tetracycline (30 μg). The diameter (mm) of inhibition zone was measured, and the antibiotic susceptibility was evaluated based on the interpretive standards of the Clinical and Laboratory Standards Institute (CLSI, 2020).

Evaluation of acid and bile salt tolerance

Bacterial inoculum (OD600 = 0.2; 1% v/v) was subjected to NB with pH 2.5 and NB supplemented with 0.3% (w/v) dried bile bovine (Sigma-Aldrich). After incubation at 37°C, the culture was collected at the initial time point (0 h) and 3 h, further diluted in sterile phosphate-buffered saline (PBS; pH 7.4), and spread on NA plates, followed by incubation at 37°C for 24 h. Bacterial survival (%) = (number of colonies at 3 h/number of colonies at the initial time) × 100.

Tolerance to NaCl and phenol

Bacterial isolate was cultivated in NB containing different percentages of NaCl (0.0%, 2.0%, 3.5%, 5.0%, and 10.0%) and phenol (0.0%, 0.1%, 0.2%, 0.3%, and 0.4%). After incubation at 37°C for 24 h, the cultures were serially diluted and spread on NA plates. Cell viability was evaluated using the plate count method.

Assay of antioxidant activity

The assays of hydroxyl radical- and DPPH radical-scavenging activity were modified from Kadaikunnan et al. (2015). For the assay of hydroxyl radical-scavenging activity, the reaction mixture contained 2 ml of phosphate buffer (pH 7.4), and 1 ml each of solutions of 1,10-phenanthroline (0.75 mM), FeSO4 (0.75 mM), and bacterial suspension (109 CFU/ml). Then, 1.0 ml of 0.01% v/v hydrogen peroxide (H2O2) was added to initiate the reaction, and incubated at 37°C for 90 min. Absorbance of the collected supernatant was measured at 536 nm. Ascorbic acid solution was used as a standard and positive control. Hydroxyl radical-scavenging activity (%) = [(As−Ac)/(Ab−Ac)]×100, where As is the absorbance of the test sample; Ac is the absorbance of the control including 1,10-phenanthroline, FeSO4, and H2O2; and Ab is the absorbance of the blank including 1,10-phenanthroline and FeSO4.

For the assay of DPPH radical-scavenging activity, the reaction mixture contained 1 ml of DPPH methanol solution (0.05 mM) and 1 ml of bacterial suspension (109 CFU/ml). The mixture was incubated in the dark at 37°C for 60 min, and absorbance of the collected supernatant was measured at 517 nm. Trolox solution was used as a standard and positive control. DPPH radical-scavenging activity (%) = [(Ac−As)/Ac]×100, where As is the absorbance of the test sample; Ac is the absorbance of the control including DPPH methanol solution and distilled water.

Assay of antimicrobial activity

Antibacterial activity was evaluated against the following foodborne and enteric pathogenic bacteria: Bacillus cereus ATCC 10876, Staphylococcus aureus ATCC 25923, Escherichia coli ATCC 25922, Aeromonas hydrophila ATCC 7966, Aeromonas schubertii ATCC 43700, Salmonella Typhimurium ATCC 13311, Pseudomonas aeruginosa ATCC 9027, Listeria innocua ATCC 33090, Listeria monocytogenes ATCC 19111, Vibrio cholerae ATCC 14033, Vibrio parahaemolyticus ATCC 17802, and Proteus mirabilis ATCC 25933. Using the method modified from Lee et al. (2017), bacterial inoculum (2 μl; 108 CFU/ml) was dropped on an appropriated agar medium which was densely streaked with each pathogen (108 CFU/ml). After incubation at 37°C for 24 h, capacity of pathogen inhibition was evaluated by the ratio of the inhibition zone diameter (mm) to the bacterial colony diameter (mm).

Investigation of cell-surface properties based on hydrophobicity, auto- and co-aggregations

The method for hydrophobicity assay was modified from Somashekaraiah et al. (2019). Bacterial cells were washed and resuspended with PBS to an OD600 value of 0.2 (A0) to standardize the bacterial number. The suspension was mixed with equal volume of each following solvent: xylene, chloroform, and ethyl acetate. The mixture was incubated at 37°C without shaking for 1 h to separate the aqueous and organic solvent phases. After removing the aqueous solution, an absorbance of the remaining solution was measured at 600 nm (A1). Hydrophobicity (%) = (1−A1/A0) × 100.

Auto-aggregation was assayed according to the method modified from Nwagu et al. (2020). Bacterial suspension was prepared as above and vortexed for 30 sec, subsequently, absorbance at 600 nm was measured at the initial time point (A0), and 1, 2, 3, and 4 h (At). Auto-aggregation (%) = 1−(At/A0) × 100. Co-aggregation with foodborne pathogenic bacteria was assayed using L. monocytogenes ATCC 19111, S. Typhimurium ATCC 13311, and B. cereus ATCC 10876 according to the method modified from Liu et al. (2021). Cell suspensions of the isolates and pathogenic bacteria were mixed and vortexed for 10 sec, followed by incubation at 37°C without shaking. The absorbance was measured at 2 h. Co-aggregation (%) = [(Ax + Ay)−2(Amix)/(Ax + Ay)] × 100, where, Ax and Ay indicate the absorbance of test strains and pathogens in the control tubes and Amix is the absorbance of the combination of two bacteria.

Cell adhesion assay

Caco-2 and HT-29 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Himedia) with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco) in 6-well tissue culture plates. The method was modified from Daneshazari et al. (2023). The Caco-2 and HT-29 cells in a monolayer were washed three times with PBS. Bacterial cultures (108 CFU/ml) suspended in 1 ml DMEM without serum and antibiotics were inoculated to each well. After incubation at 37°C for 3 h, the monolayer was washed three times with PBS to remove non-adherent bacteria. To enumerate the viable adhered bacteria, the cells from monolayer were detached by 0.1% (v/v) Triton X-100. The suspension of lysed cells and bacteria was serially diluted with saline solution and plated on NA agar. Cell adhesion (%) = (Bf/Bi) × 100, where Bi and Bf are the initial and final count of bacteria (CFU/ml).

Evaluation of bacterial viability and NSP-hydrolytic activity in simulated gastrointestinal tract conditions

Bacterial survival under three stressful conditions of gastrointestinal tract was determined using in vitro simulation. Gastric phase (GP) and enteric phase-I (EP-I) were modified from Lo Curto et al. (2011). However, enteric phase-II (EP-II), a newly invented treatment, was used as final intestinal phase in this study. The GP contained pepsin (3 mg/L) and lipase (0.9 mg/L), adjusted to pH 2.5. The EP-I contained bile salts (10 g/L) and pancreatin solutions (1 g/L), adjusted to pH 5.5. The EP-II (pH 7.4) was the EP-I solution spiked with three NSP substrates (CMC, xylan, and pectin) for determination of the activities of cellulase, xylanase, and pectinase, produced by the bacterial isolates. At the beginning, bacterial cells were washed with PBS, followed by suspension with 5 ml of GP solution to an OD600 value of 0.2. After incubation, cell viability in the sample was evaluated using the spread plate method on NA plates, and subsequently, the total bacterial cells were harvested from the suspension and continually resuspended with the treatment solution of the next phase (EP-I). The experiment was performed using the same feature for all phases. Each treatment was incubated at 37°C, and the bacterial survival was evaluated in the 0- and 2-h samples. Additionally, after 2-h incubation in the final phase (EP-II), enzyme activity was detected by the presence of reducing sugars in Benedict's solution. After another 24-h incubation, enzyme activity in the EP-II treatment was determined using agar well diffusion on NA supplemented with each substrate. Survival rate (%) = [T2/T0]×100, where T0 and T2 are the number of colonies at 0 and 2 h, respectively.

Statistical analysis

All tests were carried out in triplicate, and data are presented as mean ± SD. The data were analyzed using one-way ANOVA, followed by the Duncan post-hoc test. P-value < 0.05 was considered statistically significant.


Bacterial isolates and NSP-hydrolytic activity

A total of 172 bacterial isolates were obtained from guts of Termes propinquus (Tp). Then, the isolates were evaluated NSP-degrading enzyme and other hydrolytic enzyme activities using the drop assay. The results showed that 67 isolates could produce at least one of the NSP-hydrolytic enzymes (cellulase, xylanase, and pectinase) (Supplementary data Table S1). Among them, seven isolated strains (TpNSP_1, TpNSP_11, TpNSP_22, TpNSP_33, TpNSP_72, TpNSP_84, and TpNSP_122) showed the ability to produce all three NSP-hydrolytic enzymes and additionally, they also produced another four hydrolytic enzymes (protease, amylase, phytase, and lipase) (Table 1). All seven strains harboring broad-spectrum hydrolytic activities were selected for the subsequent experiments, with the perspective of their use in feed and biotechnological applications. NSP-hydrolytic enzyme activity was determined every 24 h for 1 week using the DNS assay and all strains exhibited NSP-hydrolytic activities for all three enzymes (Fig. 1). The optimum incubation period to obtain the highest activity value of cellulase and xylanase was varieties, whereas for the highest pectinase activity, the optimum incubation period was 48 h (Table 1). Among them, TpNSP_122 showed significant highest cellulase activity, with 0.49 U/ml at 24 h, and the highest pectinase activity, with 0.87 U/ml at 48 h. The xylanase activity of TpNSP_1 was highest, with 1.18 U/ml at 48 h. To observe morphological characteristics, the bacteria were cultured on NA at 37°C for 12 h. The seven isolates were Gram-positive endospore-forming bacilli, and their large colonies were creamy-white, translucent, undulate, and grew rapidly on NA (Supplementary data Fig. S1). Over time, the shape of circular colonies became irregular. The isolate TpNSP_11 had convex and smooth colonies, differing from the other six isolates, which displayed raised and rough colonies.

Ratio of extracellular enzyme production and maximum NSP-hydrolyzing enzyme activity at optimal incubation period of the selected strains


TpNSP_1 TpNSP_11 TpNSP_22 TpNSP_33 TpNSP_72 TpNSP_84 TpNSP_122
Enzyme production ratio
Cellulase 1.94 ± 0.06 2.28 ± 0.03 3.17 ± 0.76 1.98 ± 0.71 1.84 ± 0.58 1.31 ± 0.10 3.84 ± 0.37
Xylanase 1.52 ± 0.21 3.61 ± 0.56 1.90 ± 0.14 2.10 ± 0.14 4.75 ± 0.08 2.82 ± 0.25 1.08 ± 0.10
Pectinase 3.61 ± 0.08 2.21 ± 0.13 4.00 ± 0.00 2.02 ± 0.25 3.89 ± 0.02 3.18 ± 0.25 2.95 ± 0.06
Protease 4.80 ± 1.13 4.67 ± 1.18 3.06 ± 0.23 2.45 ± 0.49 3.07 ± 0.96 2.36 ± 0.20 2.32 ± 0.18
Amylase 3.33 ± 0.94 4.33 ± 0.94 3.50 ± 1.18 2.45 ± 0.07 3.37 ± 0.18 4.50 ± 0.23 3.00 ± 0.00
Phytase 2.25 ± 0.18 1.33 ± 0.23 2.00 ± 0.47 2.33 ± 0.32 1.67 ± 0.19 1.33 ± 0.12 2.00 ± 0.35
Lipase 1.21 ± 0.06 1.07 ± 0.01 1.40 ± 0.18 1.73 ± 0.16 2.10 ± 0.19 2.90 ± 0.31 4.69 ± 1.00

Maximum NSP-hydrolyzing enzyme activity (U/ml) at optimal incubation period (h)
Cellulase 0.39 ± 0.06c (96 h) 0.27 ± 0.03a (96 h) 0.32 ± 0.05abc (72 h) 0.29 ± 0.01ab (72 h) 0.37 ± 0.00bc (24 h) 0.38 ± 0.03bc (24 h) 0.49 ± 0.05d (24 h)
Xylanase 1.18 ± 0.11c (48 h) 0.90 ± 0.05a (24 h) 1.01 ± 0.01ab (24 h) 1.07 ± 0.05bc (48 h) 1.07 ± 0.02bc (24 h) 1.05 ± 0.01abc (24 h) 0.90 ± 0.05a (24 h)
Pectinase 0.46 ± 0.08a (48 h) 0.54 ± 0.05a (48 h) 0.78 ± 0.05bc (48 h) 0.74 ± 0.01b (48 h) 0.84 ± 0.04bc (48 h) 0.84 ± 0.00bc (48 h) 0.87 ± 0.02c (48 h)

Data are the mean ± SD (n = 3); different superscripts indicate significantly difference (P < 0.05) in different strains; -, not detectable; enzyme production ratio = ratio of hydrolysis zone diameter (mm) to bacterial colony diameter (mm); Time course of NSP-hydrolytic enzyme activity was shown in Fig. 1.

Fig. 1. Time course of non-starch polysaccharide (NSP) hydrolytic enzyme activity (U/ml; bar graph) of cellulase (black), xylanase (gray), and pectinase (white), and estimated bacterial density (OD600 value; line graph) of seven selected isolates TpNSP_1 (A), TpNSP_11 (B), TpNSP_22 (C), TpNSP_33 (D), TpNSP_72 (E), TpNSP_84 (F), and TpNSP_122 (G) in a liquid medium containing different substrates (cellulose, xylan, and pectin). Data are the mean ± SD (n = 3).

Molecular identification of bacteria

All different isolated strains showed a high similarity to species in the Bacillus subtilis group, based on the nucleotide sequences of 16S rRNA gene (99.20–100% identity) and the nucleotide and amino acid sequences of the housekeeping genes gyrA, gyrB, and rpoB (98.80–100% and 99.30–100% identity, respectively). Figure 2 shows the phylogenetic trees based on nucleotide sequences. Based on the similarity of 16S rRNA gene sequences, TpNSP_1 and TpNSP_33 were related to Bacillus velezensis B268, TpNSP_11 and TpNSP_22 to Bacillus siamensis cqsM9 and B. velezensis Y17W, TpNSP_72 and TpNSP_84 to B. velezensis Bac57, B. siamensis SB1001, and Bacillus amyloliquefaciens S2S95, and TpNSP_122 to Bacillus subtilis JCL16. Both phylogenetic trees based on concatenated nucleotide and amino acid sequences of the three housekeeping genes (Fig. 2 and Supplementary data Fig. S2) indicated that TpNSP_1, TpNSP_11, TpNSP_22, and TpNSP_33 were closely related to strains of B. amyloliquefaciens and B. velezensis, TpNSP_72 and TpNSP_84 to B. siamensis strains, and TpNSP_122 to Bacillus subtilis strains. Moreover, phylogenetic analysis of the concatenated nucleotide sequences from these three genes could present a slight dissimilarity among the closely related isolates. The sequences were deposited at GenBank under the accession numbers OQ776779–OQ776785 for 16S rRNA genes, OR224548–OR224554 for gyrA, OR224555–OR224561 for gyrB, and OR224562–OR224568 for rpoB.

Fig. 2. Phylogenetic trees based on nucleotide sequences of 16S rRNA gene (A), and concatenated nucleotide sequences of three housekeeping genes gyrA, gyrB, and rpoB (B) grouping within Bacillus spp. The tree was constructed using MEGA X. Numbers at the nodes indicate the bootstrap value (%) based on a maximum-likelihood analysis of 1,000 resampled datasets and only the values ≥ 50% are presented. The scale bar corresponds to a number of nucleotide substitutions per site.

Safety profile of bacteria

Safety of the bacterial strains was evaluated based on hemolysis and antibiotic susceptibility. In Table 2, all selected strains exhibited alpha-hemolysis, except TpNSP_11 and TpNSP_33, which exhibited gamma-hemolysis. They were sensitive or intermediate to six antibiotics (ampicillin, ciprofloxacin, erythromycin, tetracycline, vancomycin, and streptomycin), except TpNSP_22, TpNSP_72, and TpNSP_122, which were resistant to erythromycin. However, six strains (without TpNSP_1) resisted penicillin. Ampicillin was the most potent antibiotic, whereas penicillin had a weak effect on the isolated strains.

Probiotic properties based on hemolytic activity, antibiotic susceptibility, tolerance to stressful conditions, and antioxidant activity of the selected strains


TpNSP_1 TpNSP_11 TpNSP_22 TpNSP_33 TpNSP_72 TpNSP_84 TpNSP_122
Hemolytic activity Alpha Gamma Alpha Gamma Alpha Alpha Alpha

Antibiotic susceptibility (inhibition zone diameter, mm)
Ampicillin S (50) S (50) S (43) S (52) S (62) S (52) S (45)
Ciprofloxacin S (40) S (37) S (33) S (39) S (40) S (36) S (33)
Erythromycin I (18) I (18) R (15) I (18) R (13) I (19) R (15)
Tetracycline S (36) S (36) S (34) S (34) S (39) S (34) S (38)
Vancomycin S (32) S (33) S (31) S (32) S (37) S (32) S (30)
Streptomycin S (32) S (34) S (28) S (31) S (35) S (30) S (28)
Penicillin I (16) R (15) R (9) R (0) R (11) R (9) R (9)

Survival (%) in stressful conditions
0.3% Bile salts 65.97 ± 1.12a 89.45 ± 1.63d 74.07 ± 3.96b 79.26 ± 4.53bc 90.50 ± 1.72d 76.38 ± 1.83b 83.90 ± 3.52c
pH 2.5 72.38 ± 10.34bc - 94.69 ± 5.71d 83.96 ± 21.56cd 64.00 ± 5.30b 83.45 ± 5.88cd 91.47 ± 6.93d

Salt tolerance (%)
2.0% NaCl 78.00 ± 1.32 97.41 ± 6.35 131.14 ± 5.41 100.00 ± 6.67 122.01 ± 6.35 114.93 ± 8.14 104.78 ± 1.32
3.5% NaCl 84.47 ± 3.24 100.65 ± 2.47 122.10 ± 14.33 103.61 ± 15.23 137.86 ± 3.22 132.35 ± 15.23 133.47 ± 12.41
5.0% NaCl 101.94 ± 1.55 47.92 ± 1.27 73.28 ± 1.19 52.51 ± 1.25 149.31 ± 8.65 75.95 ± 2.21 89.48 ± 5.71
10.0% NaCl 47.34 ± 1.52 47.40 ± 1.55 - 52.51 ± 4.63 64.43 ± 2.45 60.35 ± 1.46 -

Phenol tolerance (%)
0.1% phenol 98.79 ± 4.21 103.21 ± 6.35 96.57 ± 4.21 98.79 ± 1.19 88.84 ± 2.69 87.85 ± 2.21 94.48 ± 2.21
0.2% phenol 93.94 ± 6.68 83.97 ± 1.55 88.58 ± 1.22 93.94 ± 1.22 87.72 ± 1.65 85.65 ± 1.55 91.17 ± 1.65
0.3% phenol 93.94 ± 1.65 82.90 ± 1.22 88.58 ± 1.55 93.94 ± 5.45 86.60 ± 4.21 85.65 ± 1.23 86.75 ± 1.55
0.4% phenol 93.94 ± 0.92 82.90 ± 8.49 88.58 ± 2.21 93.94 ± 5.71 86.60 ± 2.21 85.65 ± 1.19 85.65 ± 1.55

Relative scavenging activity (%)
Hydroxyl radical 61.50 ± 4.59a 55.00 ± 12.73a 52.50 ± 6.36a 61.00 ± 18.38a 52.00 ± 7.07a 58.00 ± 7.07a 65.00 ± 11.31a
DPPH radical 57.72 ± 0.49b 51.96 ± 7.98ab 54.03 ± 4.40ab 58.41 ± 1.14b 45.16 ± 1.63a 53.11 ± 4.07ab 57.83 ± 0.00b

Data are the mean ± SD (n = 3); different superscripts indicate significantly difference (P < 0.05) in different strains; -, not detectable; antibiotic susceptibility was evaluated based on the interpretive standards of Clinical and Laboratory Standards Institute (CLSI): S = sensitive (≥ 21 mm diameter), I = intermediate (16–20 mm diameter), and R = resistant (≤ 15 mm diameter).

Acid and bile salt tolerance

Acid and bile salt tolerances were accessed for each strain by monitoring bacterial survival. All seven strains showed 66.97–90.50% survival in 0.3% bile salts, and TpNSP_72 showed highest bile tolerance (Table 2). All of them, except TpNSP_11, showed 64.00%–94.69% survival at pH 2.5, with TpNSP_22 showing highest acidity tolerance.

Tolerance to NaCl and phenol stress

Salt and phenol tolerance ability of the strains was evaluated. In Table 2, all strains showed tolerance to 5.0% NaCl, whereas TpNSP_1, TpNSP_11, TpNSP_33, TpNSP_72, and TpNSP_84 tolerated up to 10.0% NaCl. At 5.0% salinity, TpNSP_1 and TpNSP_72 showed growth potential, with ~2 and ~49% increase, respectively. All strains tolerated 0.4% phenol, and only TpNSP_11 could grow in 0.1% phenol, with ~3% increase.

Antioxidant activity

Hydroxyl radical- and DPPH radical-scavenging activity of the strains were assayed. All strains had hydroxyl and DPPH radical-scavenging rates ranging from 52.00% to 65.00% and 45.16% to 58.41%, respectively (Table 2). TpNSP_122 showed the highest value of relative hydroxyl radical-scavenging activity. Besides, strains TpNSP_1, TpNSP_33, and TpNSP_122 showed the highest DPPH radical-scavenging activity, with a relative value of approximately 58%.

Antimicrobial activity

As shown in Table 3, all seven strains showed broad-spectrum antimicrobial activity against all 12 pathogenic bacteria, except E. coli ATCC 25922 tested with TpNSP_1, TpNSP_11, and TpNSP_72. Among all strains, TpNSP_84 exhibited the highest antibacterial activity against the pathogens B. cereus ATCC 10876, S. aureus ATCC 25923, E. coli ATCC 25922, A. hydrophila ATCC 7966, L. innocua ATCC 33090, L. monocytogenes ATCC 19111, and P. mirabilis ATCC 25933. Strains TpNSP_72 and TpNSP_84 showed the highest inhibition value against S. Typhimurium ATCC 13311; TpNSP_22 showed the highest inhibition value against P. aeruginosa ATCC 9027 and V. cholerae ATCC 14033; and TpNSP_11 and TpNSP_72 exhibited the highest inhibition values against V. parahaemolyticus ATCC 17802 and A. schubertii ATCC 43700, respectively.

Antimicrobial activity of the selected strains against 12 foodborne and enteric pathogenic bacteria

Ratio of pathogen inhibition Strains

TpNSP_1 TpNSP_11 TpNSP_22 TpNSP_33 TpNSP_72 TpNSP_84 TpNSP_122
Bacillus cereus ATCC 10876 1.32 ± 0.20 1.43 ± 0.39 1.63 ± 0.01 1.43 ± 0.08 1.48 ± 0.04 1.69 ± 0.18 1.42 ± 0.04
Staphylococcus aureus ATCC 25923 1.28 ± 0.06 1.50 ± 0.12 1.62 ± 0.27 1.62 ± 0.16 1.57 ± 0.00 2.02 ± 0.70 1.56 ± 0.00
Escherichia coli ATCC 25922 - - 1.44 ± 0.10 1.57 ± 0.12 - 1.65 ± 0.19 1.46 ± 0.05
Aeromonas hydrophila ATCC 7966 1.62 ± 0.45 1.70 ± 0.23 1.62 ± 0.14 1.56 ± 0.04 1.52 ± 0.16 1.88 ± 0.22 1.76 ± 0.16
Aeromonas schubertii ATCC 43700 1.24 ± 0.12 1.56 ± 0.31 1.56 ± 0.08 1.64 ± 0.07 2.02 ± 0.36 1.59 ± 0.16 1.70 ± 0.01
Salmonella Typhimurium ATCC 13311 1.69 ± 0.44 1.87 ± 0.05 2.04 ± 0.05 1.74 ± 0.13 2.11 ± 0.16 2.11 ± 0.41 1.83 ± 0.01
Pseudomonas aeruginosa ATCC 9027 1.55 ± 0.17 1.83 ± 0.04 2.00 ± 0.00 1.62 ± 0.11 1.71 ± 0.40 1.83 ± 0.01 1.83 ± 0.24
Listeria innocua ATCC 33090 1.90 ± 0.14 1.89 ± 0.16 1.51 ± 0.34 1.85 ± 1.02 1.29 ± 0.01 2.07 ± 0.10 1.43 ± 0.04
Listeria monocytogenes ATCC 19111 1.49 ± 0.14 1.33 ± 0.00 1.70 ± 0.10 1.26 ± 0.11 1.39 ± 0.28 2.63 ± 1.76 2.03 ± 0.21
Vibrio cholerae ATCC 14033 2.00 ± 0.14 1.64 ± 0.02 2.23 ± 0.01 1.59 ± 0.04 1.56 ± 0.19 1.73 ± 0.03 2.00 ± 0.18
Vibrio parahaemolyticus ATCC 17802 1.47 ± 0.04 2.13 ± 0.01 1.88 ± 0.17 1.66 ± 0.07 1.63 ± 0.05 1.47 ± 0.00 1.94 ± 0.09
Proteus mirabilis ATCC 25933 1.43 ± 0.02 1.38 ± 0.07 1.38 ± 0.07 1.72 ± 0.16 2.11 ± 0.15 2.70 ± 0.19 1.13 ± 0.06

Data are the mean ± SD (n = 3); -, no activity; ratio of pathogen inhibition = ratio of inhibition zone diameter (mm) to bacterial colony diameter (mm).

Bacterial cell-surface properties

Bacterial cell-surface properties based on hydrophobicity, auto- and co-aggregations were investigated. All strains presented cell-surface hydrophobicity in all three test solvents (Table 4). TpNSP_72 showed the highest hydrophobicity values of 90.00%, 55.20%, and 60.00% in xylene, chloroform, and ethyl acetate, respectively. The highest hydrophobicity, ranging from 70.00% to 90.00%, were found in xylene. The strains tended to show an increased auto-aggregation percentage over time. TpNSP_22 exhibited the highest auto-aggregation values from 1 to 4 h, with 40.36% to 51.77%. The co-aggregation assay of the strains with three different pathogens showed that TpNSP_11 exhibited the highest values of 69.28% and 49.80% with L. monocytogenes ATCC 19111 and B. cereus ATCC 10876, respectively, and TpNSP_122 exhibited the highest value of 33.00% with S. Typhimurium ATCC 13311.

Profiles of cell-surface properties and cell adhesion ability of the selected strains


TpNSP_1 TpNSP_11 TpNSP_22 TpNSP_33 TpNSP_72 TpNSP_84 TpNSP_122
Hydrophobicity (%)
Xylene 88.80 ± 9.62a 80.00 ± 12.44a 82.00 ± 2.83a 87.60 ± 0.57a 90.00 ± 0.57a 88.60 ± 1.98a 70.00 ± 14.14a
Chloroform 35.80 ± 7.07a 45.20 ± 6.22a 38.80 ± 13.01a 46.40 ± 10.75a 55.20 ± 5.09a 38.60 ± 24.04a 37.60 ± 0.57a
Ethyl acetate 37.60 ± 29.98ab 19.00 ± 8.20ab 28.00 ± 25.46ab 51.40 ± 2.55ab 60.00 ± 1.13b 42.20 ± 16.12ab 11.20 ± 1.70a

Auto-aggregation (%)
1 h 19.47 ± 12.21a 29.43 ± 6.63a 40.36 ± 2.25a 27.29 ± 21.67a 33.62 ± 5.62a 20.43 ± 3.12a 24.90 ± 2.64a
2 h 24.40 ± 8.30a 34.75 ± 7.83a 43.85 ± 5.45a 31.11 ± 23.49a 35.73 ± 4.76a 32.53 ± 8.05a 29.28 ± 2.11a
3 h 27.33 ± 8.74a 39.61 ± 4.98a 48.81 ± 8.99a 34.41 ± 21.24a 40.12 ± 4.67a 36.24 ± 7.83a 32.92 ± 4.67a
4 h 29.61 ± 5.53a 42.96 ± 8.27a 51.77 ± 4.00a 39.62 ± 24.32a 42.41 ± 4.35a 42.05 ± 12.82a 37.16 ± 3.43a

Co-aggregation (%)
Listeria monocytogenes ATCC 19111 23.99 ± 6.47b 69.28 ± 2.51e 52.09 ± 4.00d 32.11 ± 2.98c 21.59 ± 2.25b 13.72 ± 13.11a 21.69 ± 2.39b
Salmonella Typhimurium ATCC 13311 14.09 ± 2.50b 25.29 ± 3.69d 28.61 ± 0.60e 21.69 ± 0.28c 21.31 ± 0.72c 10.44 ± 0.18a 33.00 ± 3.74f
Bacillus cereus ATCC 10876 17.18 ± 1.59b 49.80 ± 0.97f 20.23 ± 0.32c 21.16 ± 1.64cd 15.54 ± 4.96b 9.73 ± 0.14a 22.44 ± 2.93d

Cell adhesion (%)
Caco-2 20.63 ± 2.58d - 23.42 ± 0.81e 17.21 ± 0.79c 41.67 ± 1.13f 26.23 ± 2.07g 13.29 ± 0.21b
HT-29 46.83 ± 0.78c 40.16 ± 2.74b 33.68 ± 1.27a 39.07 ± 0.99b 49.08 ± 3.80c 45.56 ± 1.46c 48.21 ± 0.58c

Data are the mean ± SD (n = 3); different superscripts indicate significantly difference (P < 0.05) in different strains; cell-surface properties were evaluated by hydrophobicity in different solvents, auto-aggregation at different incubation periods (h), and co-aggregation with different pathogens at 2 h of incubation; adhesion ability of bacteria was assessed using Caco-2 and HT-29 cells.

Cell adhesion

The ability to adhere to intestinal cells is a prerequisite for probiotics. Adhesion assays were provided for detecting bacterial adhesion to Caco-2 and HT-29 cells. In Table 4, TpNSP_11 could only adhere to HT-29 cells (40.16%) but not to Caco-2 cells. By all strains, adhesion ability to HT-29 cells (33.68–49.08%) was stronger than that of Caco-2 cells (13.29–41.67%).

Bacterial viability and NSP-hydrolytic activity in simulated gastrointestinal tract conditions

As shown in Fig. 3, all selected strains, except TpNSP_11, showed considerable survival in all three treatment phases under simulated digestive system conditions and retained their ability to produce NSP-hydrolytic enzymes in the final phase. The survival rates of all strains in GP, EP-I, and EP-II were approximately 81–97%, 71–96%, and 65–87%, respectively. Strains TpNSP_22, TpNSP_84, and TpNSP_122 showed high survival rates of ≥ 85% in all phases. TpNSP_1 showed the highest survival rate in GP and EP-I, with ≤ 4% decrease. In the final phase EP-II with 2-h incubation, all promising strains, except TpNSP_11, could produce xylanase and pectinase, whereas only TpNSP_22 and TpNSP_122 showed cellulase activity. Using the agar well diffusion method, all strains exhibited all tested enzyme activities in the EP-II with another 24-h incubation.

Fig. 3. Survival rate (%) of the bacterial isolates in the simulated gastrointestinal tract system including three treatment phases: gastric phase (GP); enteric phase-I (EP-I); and enteric phase II (EP-II). Data are the mean ± SD (n = 3). Different letters on the graph bars indicate significant difference (P < 0.05).

Our study presents knowledge of potential probiotic bacteria from termite guts, with the isolation and characterization of bacteria in the genus Bacillus with probiotic properties and the capacity to degrade NSPs. This finding broadly supports the works of other studies in microbial technological area. The Bacillus strains isolated in the present study could reduce NSPs, which are the main components of plant cell walls and highly abundant in plant-based raw materials. Although NSPs are considered dietary fiber, these feed ingredients showed negative effects as anti-nutritional components. The isolated strains showed enzyme-producing capacity useful for future applications. Their produced fibrolytic enzymes could eliminate cellulose, xylan, and pectin, which are commonly present in plant-based feed. Moreover, the ability of probiotics to produce NSP-hydrolytic enzymes is advantageous. According to Maas et al. (2021), the activities of phytase and xylanase in NSP digestion can result in the generation of short-chain fatty acids, which are a source of energy for the probiotic Bacillus amyloliquefaciens; the digested NSP promotes growth performance and nutrient use in Nile tilapia.

Various Bacillus species with ability to produce the NSP-degrading enzymes have been previously investigated from many termites. The termite T. propinquus used in the present study is a widespread species in Thailand and it has been used as an isolation source for protease-producing B. amyloliquefaciens strains Tp-5 and Tp-7 with high level of soymilk degradation, which were firstly applied for improving amino acid profile and nutritional value of fermented soybean meal (Nualkul et al., 2022). In the present study, the seven selected isolates were molecularly identified as closely related strains of B. subtilis, B. velezensis, B. siamensis, and B. amyloliquefaciens in the B. subtilis group. The species in the B. subtilis group presented limited variation in 16S rRNA gene and the multilocus sequence typing of the housekeeping genes gyrA, gyrB, and rpoB were used for differentiating closely related Bacillus species (Jeyaram et al., 2011). The cultural morphological characteristics of the isolates conformed to the characteristics of the genus Bacillus and the Bacillus sp. TpNSP_11 showed different characteristics from the other six strains. Additionally, the selected Bacillus species in our study were physiologically classified as individual strains, based on the profiles of functional properties. Recently, many strains in the B. subtilis group have been previously isolated from various sample sources and widely used as potential probiotic in animal feed and functional foods (Elshaghabee et al., 2017). For example, B. subtilis CM3.1 from intensive shrimp ponds showed good effects on water quality and the growth performance of whiteleg shrimp (Liptopenaeus vannamei) (Truong et al., 2021); B. velezensis LB-Y-1 with multi-enzyme production property displayed as a potential probiotic for poultry (Li et al., 2023); xylanase-expressing B. amyloliquefaciens R8 could improve growth performance and enhance immunity against Aeromonas hydrophila in Nile tilapia (Oreochromis niloticus) (Saputra et al., 2016); and B. siamensis B44v from Thai pickled vegetables displayed a high potential as a probiotic in catfish (Meidong et al., 2017).

In vitro determinations have long been used for probiotic evaluation, and a variety of bacterial strains can affect their characteristics and potentials. Prior to demonstrations of other probiotic properties, the promising candidates must be safe. In the present study, the Bacillus strains were susceptible to almost all antibiotics; ampicillin had highest impact on all isolated strains. In terms of hemolysis, two strains showed gamma-hemolysis (no digestion), whereas the other five strains showed alpha-hemolysis (partial digestion). The results are in agreement with the findings of Zulkhairi Amin et al. (2019), who reported that B. amyloliquefaciens HTI-19 showed alpha-hemolysis, whereas B. subtilis HTI-23 presented gamma-hemolysis. Similarly, Lee et al. (2017) reported that B. amyloliquefaciens and B. subtilis were most susceptible to ampicillin. The bacteria with antibiotic susceptibility can restrict and inhibit the spread of infections at large dosages. In the present study, the evaluated hemolytic activity and antibiotic susceptibility of the strains indicated preliminary safety for use in animal feed biotechnology. However, additional evaluations are recommended such as pathogenicity, toxigenicity, genetic stability and transferability for future study (Sanders et al., 2010).

Probiotics generally have a certain level of stress tolerance. All promising candidates in the present study, except TpNSP_11, could survive in both of harsh gut conditions with the presence of 0.3% bile salts (~66–91% survival) and a high acidity at pH 2.5 (~64–95% survival). The presence of intestinal bile salts and gastric acid largely impedes the survival of probiotic bacteria in host gastrointestinal system; generally, the pH of gastric acid fluctuates from 1.5 to 4.5, and the bile salt concentration in the upper intestine ranges from 0.03% to 0.30% (w/v) (Tokatlı et al., 2015). Bile salts are antibacterial compounds causing cell membrane disruption, protein denaturation, iron and calcium chelation, and DNA damage (Urdaneta and Casadesus, 2017). Moreover, in the present study, all Bacillus strains tolerated phenol (up to 0.4% v/v) and salt stress (at least 5.0% NaCl) affecting growth of bacteria. Phenolic compounds are hazardous to food, endogenous proteins, and gut flora, and high salt stress negatively affects microbial physiology, metabolism, and enzyme synthesis and activity (Fonseca et al., 2021).

Most of the antioxidant activity is based on probiotic administration. The Bacillus strains in the present study displayed antioxidant property, with a relative scavenging activity against hydroxyl (~52–65%) and DPPH (~45–58%) radicals, which is a positive characteristic of probiotic. Kadaikunnan et al. (2015) previously reported that B. amyloliquefaciens VJ-1 exhibit the highest hydroxyl and DPPH radical-scavenging activities, with inhibition rates of ~57% and ~67%, respectively. One of the essential features of probiotics is antimicrobial activity against pathogens. All selected strains in the present study showed broad-spectrum antimicrobial activities and have considerable potential for use as biocontrol agents against foodborne and gastrointestinal pathogenic bacteria. Members of Bacillus have been previously described as a goldmine of antimicrobial compounds active against a wide range of microorganisms. In the previous report by Ramachandran et al. (2014), B. subtilis RLID 12.1 showed antibacterial activity against pathogens such as S. aureus, E. coli, P. aeruginosa, and Pseudomonas vulgaris.

Characterization of bacterial cell surface aids in the demonstration of bacterial colonization and adhesion to gastrointestinal epithelial cells, and their interactions further prevent pathogen colonization in the gut (Somashekaraiah et al., 2019). In the present study, all Bacillus strains showed positive results regarding hydrophobicity as well as auto- and co-aggregations and these probiotic properties are advantageous. According to Zeng et al. (2021), bacterial cells can adhere to gastrointestinal epithelial cells via a weak non-covalent contact, and the force between them is positively related to hydrophobicity. Aggregation provides a barrier to prevent pathogens from colonizing the gut surface (Liu et al., 2021). In addition, ability of the Bacillus strains to adhere to the intestinal cells, HT-29 and Caco-2, was evaluated in the present study, and the adhesion ability to HT-29 cells (up to ~49%) was stronger than that of Caco-2 cells (up to ~42%). Comparable results have been reported for probiotic candidates such as B. subtilis CM1 and CM2 showing adherence of 49.66% and 47.35% to HT-29 cells (Daneshazari et al., 2023). Aside from the contact mechanism between bacteria and surface components of intestinal cells, cell type can also influence bacterial adhesion to epithelial cells (Fonseca et al., 2021).

In our study, simulations of gastrointestinal tract conditions were performed as a conclusive experiment to evaluate bacterial survival as well as the ability of the bacteria to produce NSP hydrolytic enzymes in the gastrointestinal system. All Bacillus strains, except TpNSP_11, survived under the simulated conditions, with high survival rates up to ~97%. In addition to the survival ability, the potential strains could generate NSP digestion enzymes in the simulated enteric phase (EP-II; pH 7.4) containing bile salts and pancreatin. The results are in agreement with the findings of the previous study for the probiotics B. coagulans, and B. subtilis, with a survival rate > 80% in simulated gastrointestinal tract conditions (Soares et al., 2019). The potential strains can produce enzymes under the stressful condition, and this enteric phase simulated an environment in which bacteria may colonize in the gut (Han et al., 2021).

In summary, the present study reports a knowledge of potential probiotic bacteria from termite guts. The intestinal extreme environment of the termite Termes propinquus is a valuable source of NSP hydrolytic Bacillus spp. possessing safety and beneficial probiotic properties. This in vitro study explores a scientific finding in the terms of microbiology and biotechnology, and benefits for future research to reduce antinutritional effects of dietary NSPs in animal digestion.


This research and innovation activity is funded by National Research Council of Thailand (NRCT). This research is supported in part by the Graduate Program Scholarship from The Graduate School, Kasetsart University. The corresponding author acknowledges the financial support provided by the International SciKU Branding (ISB), Faculty of Science, Kasetsart University. We thank Assoc. Prof. Dr. Kiattawee Choowongkomon for supporting HT-29 and Caco-2 cells, and Assoc. Prof. Dr. Bundit Yuangsoi, Ms. Maneeploy Nualkul, and Mr. Kittipong Chanworawit for helping.

Conflict of Interest

The authors declare that they have no conflicts of interest.

Author Contributions

PD conceived the study, supervised and designed the project, performed some experiments and analyzed the data, took the lead in writing the manuscript and contributed to termite collection. PT participated in bacterial isolation and characterization, worked out almost all the experiments, collected and analyzed the data, and wrote the manuscript with input from all authors. All authors reviewed and approved the final manuscript.

Data Availability

All sequences of 16S rRNA, gyr A, gyr B and rpo B genes were submitted and recorded in the NCBI database and corresponding accession numbers were provided in the results. In case of requirement, please contact the corresponding author for any detailed question.

  1. Bedford MR and Partridge GG. Enzymes in Farm Animal Nutrition. 2nd edn. Wallingford, Oxforshire, UK.
  2. Bernardeau M and Vernoux JP. 2013. Overview of differences between microbial feed additives and probiotics for food regarding regulation, growth promotion effects and health properties and consequences for extrapolation of farm animal results to humans. Clin. Microbiol. Infect. 19, 321-330.
    Pubmed CrossRef
  3. Chandran C and Keerthi T. 2018. Probiotic potency of Lactobacillus plantarum KX519413 and KX519414 isolated from honey bee gut. FEMS Microbiol. Lett. 365, fnx285.
    Pubmed CrossRef
  4. CLSI, Clinical, Laboratory Standards Institute. CLSI Supplement M100. 30th edn. Clinical and Laboratory Standards Institute, Wayne, Pennsylvania, USA.
  5. Cutting SM. 2011. Bacillus probiotics. Food Microbiol. 28, 214-220.
  6. Daneshazari R, Rabbani Khorasgani M, Hosseini-Abari A, and Kim JH. 2023. Bacillus subtilis isolates from camel milk as probiotic candidates. Sci. Rep. 13, 3387.
    Pubmed KoreaMed CrossRef
  7. De Figueirêdo RE, Vasconcellos A, Policarpo IS, and Alves RR. 2015. Edible and medicinal termites: a global overview. J. Ethnobiol. Ethnomed. 11, 29.
    Pubmed KoreaMed CrossRef
  8. Deng J, Zhang X, Sun Y, Mi H, and Zhang L. 2021. Effects of different types of non-starch polysaccharides on growth, digestive enzyme activity, intestinal barrier function and antioxidant activity of rainbow trout (Oncorhynchus mykiss). Aquac. Rep. 21, 100864.
  9. Elshaghabee FMF, Rokana N, Gulhane RD, Sharma C, and Panwar H. 2017. Bacillus as potential probiotics: status, concerns, and future perspectives. Front. Microbiol. 8, 1490.
    Pubmed KoreaMed CrossRef
  10. Elzeini HM, Ali ARAA, Nasr NF, Hassan M, Hassan AAM, and Elenany YE. 2021. Probiotic capability of novel lactic acid bacteria isolated from worker honey bees gut microbiota. FEMS Microbiol. Lett. 368, fnab030.
    Pubmed CrossRef
  11. Fonseca HC, de Sousa Melo D, Ramos CL, Dias DR, and Schwan RF. 2021. Probiotic properties of lactobacilli and their ability to inhibit the adhesion of enteropathogenic bacteria to Caco-2 and HT-29 cells. Probiotics Antimicrob. Proteins 13, 102-112.
    Pubmed CrossRef
  12. Han S, Lu Y, Xie J, Fei Y, Zheng G, Wang Z, Liu J, Lv L, Ling Z, and Berglund BBerglund B, et al. 2021. Probiotic gastrointestinal transit and colonization after oral administration: a long journey. Front. Cell Infect. Microbiol. 11, 609722.
    Pubmed KoreaMed CrossRef
  13. Hongoh Y, Deevong P, Inoue T, Moriya S, Trakulnaleamsai S, Ohkuma M, Vongkaluang C, Noparatnaraporn N, and Kudo T. 2005. Intra- and interspecific comparisons of bacterial diversity and community structure support coevolution of gut microbiota and termite host. Appl. Environ. Microbiol. 71, 6590-6599.
    Pubmed KoreaMed CrossRef
  14. Jeyaram K, Romi W, Singh TA, Adewumi GA, Basanti K, and Oguntoyinbo FA. 2011. Distinct differentiation of closely related species of Bacillus subtilis group with industrial importance. J. Microbiol. Methods 87, 161-164.
    Pubmed CrossRef
  15. Kadaikunnan S, Rejiniemon TS, Khaled JM, Alharbi NS, and Mothana R. 2015. In-vitro antibacterial, antifungal, antioxidant and functional properties of Bacillus amyloliquefaciens. Ann. Clin. Microbiol. Antimicrob. 14, 9.
    Pubmed KoreaMed CrossRef
  16. Lee S, Lee J, Jin YI, Jeong JC, Chang YH, Lee Y, Jeong Y, and Kim M. 2017. Probiotic characteristics of Bacillus strains isolated from Korean traditional soy sauce. LWT 79, 518-524.
  17. Li C, Li S, Dang G, Jia R, Chen S, Deng X, Liu G, Beckers Y, and Cai H. 2023. Screening and characterization of Bacillus velezensis LB-Y-1 toward selection as a potential probiotic for poultry with multi-enzyme production property. Front. Microbiol. 14, 1143265.
    Pubmed KoreaMed CrossRef
  18. Liu C, Xue WJ, Ding H, An C, Ma SJ, and Liu Y. 2021. Probiotic potential of Lactobacillus strains isolated from fermented vegetables in Shaanxi, China. Front. Microbiol. 12, 774903.
    Pubmed KoreaMed CrossRef
  19. Lo Curto A, Pitino I, Mandalari G, Dainty JR, Faulks RM, and John Wickham MS. 2011. Survival of probiotic lactobacilli in the upper gastrointestinal tract using an in vitro gastric model of digestion. Food Microbiol. 28, 1359-1366.
    Pubmed CrossRef
  20. Maas RM, Verdegem MCJ, Debnath S, Marchal L, and Schrama JW. 2021. Effect of enzymes (phytase and xylanase), probiotics (B. amyloliquefaciens) and their combination on growth performance and nutrient utilisation in Nile tilapia. Aquaculture 533, 736226.
  21. Meidong R, Doolgindachbaporn S, Jamjan W, Sakai K, Tashiro Y, Okugawa Y, and Tongpim S. 2017. A novel probiotic Bacillus siamensis B44v isolated from Thai pickled vegetables (Phak-dong) for potential use as a feed supplement in aquaculture. J. Gen. Appl. Microbiol. 63, 246-253.
    Pubmed CrossRef
  22. Nualkul M, Yuangsoi B, Hongoh Y, Yamada A, and Deevong P. 2022. Improving the nutritional value and bioactivity of soybean meal in solid-state fermentation using Bacillus strains newly isolated from the gut of the termite Termes propinquus. FEMS Microbiol. Lett. 369, fnac044.
    Pubmed CrossRef
  23. Nwagu TN, Ugwuodo CJ, Onwosi CO, Inyima O, Uchendu OC, and Akpuru C. 2020. Evaluation of the probiotic attributes of Bacillus strains isolated from traditional fermented African locust bean seeds (Parkia biglobosa), "daddawa". Ann. Microbiol. 70, 20.
  24. Ramachandran R, Chalasani AG, Lal R, and Roy U. 2014. A broad-spectrum antimicrobial activity of Bacillus subtilis RLID 12.1. Sci. World J. 2014, 968487.
  25. Sanders ME, Akkermans LM, Haller D, Hammerman C, Heimbach J, Hörmannsperger G, Huys G, Levy DD, Lutgendorff F, and Mack DMack D, et al. 2010. Safety assessment of probiotics for human use. Gut Microbes 1, 164-185.
    Pubmed KoreaMed CrossRef
  26. Saputra F, Shiu YL, Chen YC, Puspitasari AW, Danata RH, Liu CH, and Hu SY. 2016. Dietary supplementation with xylanase-expressing B. amyloliquefaciens R8 improves growth performance and enhances immunity against Aeromonas hydrophila in Nile tilapia (Oreochromis niloticus). Fish Shellfish Immunol. 58, 397-405.
    Pubmed CrossRef
  27. Sargautiene V, Nakurte I, and Nikolajeva V. 2018. Broad prebiotic potential of non-starch polysaccharides from oats (Avena sativa L.): an in vitro study. Pol. J. Microbiol. 67, 307-313.
    Pubmed KoreaMed CrossRef
  28. Soares MB, Martinez RCR, Pereira EPR, Balthazar CF, Cruz AG, Ranadheera CS, and Sant'Ana AS. 2019. The resistance of Bacillus, Bifidobacterium, and Lactobacillus strains with claimed probiotic properties in different food matrices exposed to simulated gastrointestinal tract conditions. Food Res. Int. 125, 108542.
  29. Somashekaraiah R, Shruthi B, Deepthi BV, and Sreenivasa MY. 2019. Probiotic properties of lactic acid bacteria isolated from Neera: a naturally fermenting coconut palm nectar. Front. Microbiol. 10, 1382.
    Pubmed KoreaMed CrossRef
  30. Sreena CP and Sebastian D. 2018. Augmented cellulase production by Bacillus subtilis strain MU S1 using different statistical experimental designs. J. Genet. Eng. Biotechnol. 16, 9-16.
    Pubmed KoreaMed CrossRef
  31. Thongaram T, Kosono S, Ohkuma M, Hongoh Y, Kitada M, Yoshinaka T, Trakulnaleamsai S, Noparatnaraporn N, and Kudo T. 2003. Gut of higher termites as a niche for alkaliphiles as shown by culture-based and culture-independent studies. Microbes Environ. 18, 152-159.
  32. Tokatlı M, Gülgör G, Bağder Elmacı S, Arslankoz İşleyen N, and Özçelik F. 2015. In vitro properties of potential probiotic indigenous lactic acid bacteria originating from traditional pickles. BioMed Res. Int. 2015, 315819.
    Pubmed KoreaMed CrossRef
  33. Truong QP, Phan TC, Vu HH, Pham TT, Huynh TG, and Vu NU. 2021. Isolation of potential probiotic Bacillus subtilis CM3. 1 and its effect on water quality and growth performance of whiteleg shrimp Litopenaeus vannamei in the mekong delta, Vietnam. AACL Bioflux 14, 3347-3357.
  34. Tsegaye B, Balomajumder C, and Roy P. 2019. Isolation and characterization of novel lignolytic, cellulolytic, and hemicellulolytic bacteria from wood-feeding termite Cryptotermes brevis. Int. Microbiol. 22, 29-39.
    Pubmed CrossRef
  35. Urdaneta V and Casadesus J. 2017. Interactions between bacteria and bile salts in the gastrointestinal and hepatobiliary tracts. Front. Med. 4, 163.
    Pubmed KoreaMed CrossRef
  36. Zeng Z, He X, Li F, Zhang Y, Huang Z, Wang Y, Li K, Bao Y, Iqbal M, and Fakhar MFakhar M, et al. 2021. Probiotic properties of Bacillus proteolyticus isolated from Tibetan yaks, China. Front. Microbiol. 12, 649207.
    Pubmed KoreaMed CrossRef
  37. Zulkhairi Amin FA, Sabri S, Ismail M, Chan KW, Ismail N, Mohd Esa N, Mohd Lila MA, and Zawawi N. 2019. Probiotic properties of Bacillus strains isolated from Stingless Bee (Heterotrigona itama) honey collected across Malaysia. Int. J. Environ. Res. Public Health 17, 278.
    Pubmed KoreaMed CrossRef

June 2024, 60 (2)
Full Text(PDF) Free
Supplementary File

Social Network Service

Author ORCID Information

Funding Information