Probiotic Properties of Bacillus Strains Isolated from Stingless Bee (Heterotrigona itama) Honey Collected across Malaysia

This study aimed to isolate, identify, and evaluate the probiotic properties of Bacillus species from honey of the stingless bee Heterotrigona itama. Bacillus spp. were isolated from five different H. itama meliponicultures, and the isolates were characterized through Gram-staining and a catalase test. Tolerance to acidic conditions and bile salt (0.3%), hydrophobicity, and autoaggregation tests were performed to assess the probiotic properties of the selected isolates, B. amyloliquefaciens HTI-19 and B. subtilis HTI-23. Both Bacillus isolates exhibited excellent antimicrobial activity against both Gram-positive and Gram-negative bacteria and possessed significantly high survival rates in 0.3% bile solution for 3 h. Their survival rates in acidic conditions were also comparable to a commercial probiotic strain, Lactobacillus rhamnosus GG. Interestingly, the hydrophobicity and autoaggregation percentage showed no significant difference from L. rhamnosus GG, a commercial probiotic strain. The results from this study suggest that B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 isolated from stingless bee honey have considerably good probiotic properties. Therefore, more studies should be done to investigate the effects of these bacteria cultures on gastrointestinal health.


Introduction
Stingless bee species are native to the tropics and subtropics of the world, including Australia, Africa, Southeast Asia, and parts of Mexico and Brazil. They are known for their role as important pollinators of both wild and cultivated flowering plants in different crops and orchards [1]. They produce honey, a natural sweet substance originating from nectar or blossoms that the bees collect, transform, and combine

Isolation of Bacterial Strains from Stingless Bee Honey
Five-milliliter aliquots of stingless bee honey were added to 5 mL of nutrient broth (Oxoid, Basingstoke, UK) and incubated at 37 • C for 24 h. The culture was transferred to a 50-mL centrifuge tube and spun at 1500× g for 15 min, and then the supernatant was discarded. A total of 100 µL of 0.85% saline was added into the pellet and homogenized by vortexing for 10 s. The mixture was then spread onto nutrient agar (Oxoid, Basingstoke, UK) through the spread-plate method. All experiments were done in triplicate. After being dried, the plates were incubated at 37 • C for 16-24 h. The bacterial isolates were streaked onto new plates to obtain a single colony. The colonies and microscopic morphologies were observed. A catalase test and Gram staining were performed according to Patel et al. [25].

Bacterial Strains and Growth Conditions
A total of 23 Bacillus strains isolated from stingless bee honey were included in this study. Pathogenic strains (Escherichia coli, Salmonella thyphimurium, Klebsiella pneumonia, Pseudomonas aeruginosa, and Staphylococcus aureus) were kindly supplied by the Enzyme and Microbial Technology Research Center, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, Malaysia. The strains were maintained at −80 • C in nutrient broth (Oxoid, Basingstoke, UK) with 20% (v/v) glycerol and were propagated three times in nutrient broth for activation prior to experimental use.

Antimicrobial Activity Assessment
Antimicrobial activity was assessed using an agar well-diffusion method with slight modifications [28]. The turbidity of bacterial suspensions, adjusted to match the standard McFarland 0.5 (approximately 10 8 colony forming unit, CFU/mL), was spread onto the plate. A 7-mm diameter well was punched aseptically onto the Mueller-Hinton agar (Oxoid, Basingstoke, UK) using the reverse end of a sterile 1-mL pipette tip. Tetracycline (20 µg/mL) was used as a positive control. A total of 100 µL of test agent was seeded into each well. A probiotic strain, Lactobacillus rhamnosus strain GG, was used as the reference strain. After incubation at 37 • C for 16-24 h, the diameter of the clear zone was measured.
2.6. Screening for Probiotic Properties 2.6.1. Acid and Bile Tolerance Acid and bile tolerance were performed according to the method described by Klingberg et al. [29], with slight modifications. Bile tolerance was examined in nutrient broth (Oxoid, Basingstoke, UK) containing 0.3% (w/v) oxgall bile (Sigma-Aldrich, St. Louis, MO, USA). A volume of 100 µL of cell suspensions of Bacillus strains cultured for 18 h (approximately 10 7 CFU/mL) were inoculated into nutrient broth (Oxoid, Basingstoke, UK) without bile and into nutrient broth (Oxoid, Basingstoke, UK) containing 0.3% (w/v) oxgall bile (Sigma-Aldrich, Missouri, US). The mixtures were incubated at 37 • C. Samples were taken at various times (0 h and 3 h), serially 10-fold-diluted using phosphate-buffered saline, PBS (pH 7.4), and plated in duplicate onto nutrient agar (Oxoid, Basingstoke, UK). The plates were incubated at 37 • C for 24 h. After the incubation period, viable bacterial colonies were counted and recorded.
For acid tolerance, the isolates were incubated overnight in nutrient broth at 37 • C. Overnight cultures were harvested by centrifugation (1500× g, 4 • C, 20 min). Harvested cells were washed twice with phosphate-buffered saline (PBS) before being resuspended into nutrient broth (pH 7.0) which acts as control and nutrient broth (pH 2.0), adjusted with 0.1 M HCl. Samples were withdrawn after a time interval of 0 h and 3 h and were serially diluted in phosphate-buffered saline (PBS, pH 7.4) before being plated onto nutrient agar plates and incubated at 37 • C for 24 h. Cell viability was assessed by the plate count method, and the results are expressed as log CFU/mL. Both experiments were performed in triplicate.
The survival rate (SR) was calculated according to the equation below: where N1 (log CFU/mL) is the total viable count of selected species after treatment (3 h), and N0 (log CFU/mL) represents the total viable count of selected species before treatment (0 h). A CFU is a colony-forming unit.

Hydrophobicity
Bacterial adhesion was determined to assess the adherence potential of microorganisms to surface hydrocarbons, which is a measure of adhesion to epithelial cells of the gut. The hydrophobicity of the selected Bacillus isolates was measured according to the method of Kos et al. [30], with some modifications. Following overnight incubation, bacteria were harvested in the stationary phase by centrifugation at 1500× g for 15 min, washed once, and resuspended in phosphate-buffered saline (PBS), pH 7.4, to an absorbance (A = 600 nm) of about 0.25 ± 0.05 (A 0 ) in order to standardize the number of bacteria (10 7 -10 8 CFU/mL). Then, an equal volume of xylene (Fisher Scientific, Waltham, MA, USA) was added. After a 10-min preincubation at 37 • C, the cell suspensions were mixed well through vortexing for 2 min and were incubated at 37 • C for 1 h for aqueous and organic phase separation. The aqueous phase was carefully removed after incubation, and its absorbance was measured at 600 nm (A 1 ). The percentage of bacterial adhesion to solvent was calculated as: Auto-aggregation (%) = 1 − (A 1 /A 0 ) × 100, A 0 = Absorbance at 0 h (600 nm), A 1 = Absorbance at 1 h (600 nm).

Autoaggregation
Autoaggregation assays were performed according to Del Re et al. [31] with certain modifications. Bacteria were grown overnight at 37 • C in nutrient broth (Oxoid, Basingstoke, UK). The cells were harvested by centrifugation at 5000× g for 15 min and washed twice in phosphate-buffered saline (PBS). The initial concentration was adjusted to an optical density (OD) (A = 600 nm) of 0.25 ± 0.05 (A 0 ) to give viable counts of approximately 10 8 CFU/mL. Cell suspensions (4 mL) were mixed by vortexing for 10 s, and autoaggregation was determined over 24 h of incubation at 37 • C. In addition, 1 mL of the upper suspension was transferred to another tube, and the absorbance (A) was measured at 600 nm. The autoaggregation percentage is expressed as Auto-aggregation (%) = 1 − (A t /A 0 ) × 100, A 0 = Absorbance at 0 h (600 nm), A t = Absorbance at 24 h (600 nm).

Blood Hemolysis
The selected Bacillus strains were streaked on Columbia sheep blood agar containing 5% (w/v) sheep's blood (Oxoid, Basingstoke, UK) and incubated at 37 • C for 24 h [32].

Isolation and Preliminary Detection of Bacillus Isolates
Aerobic bacteria were isolated from all samples of stingless bee honey with varied concentrations: the mean values were between 9.7 × 10 0 CFU/g and 3.67 × 10 2 CFU/g. Out of 5 honey samples, a total of 58 isolates of different morphological characteristics were selected and identified as Bacillus species based on early morphological examination. The selected colonies appeared to be circular and creamy and were not pigmented.
The shapes of the colonies were examined on the plates after incubation periods of 24 h at 37 • C. The isolates were initially identified using morphological and biochemical tests. Microscopic characterization proved that 94% of them were Gram-positive and rod-shaped or also known as Bacillus ( Table 1). The Gram-positive and catalase positive isolates were further tested for their tolerance of 7% NaCl, as this is one of the desirable technological properties of probiotic bacteria. Out of 58 isolates, only 23 of them were able to tolerate high concentrations of 7% NaCl (Table 1) and were therefore selected for further identification using 16S rRNA gene sequence analysis.

Molecular Identification through 16S rRNA Gene Sequence Analysis
In general, a 16S rRNA gene sequence analysis of 23 selected isolates revealed that the dominant Bacillus species in this study were B. pumilus (34%) and B. altitudinis (33%), followed by B. megaterium (13%), B. amyloliquefaciens (8%), B. aryabhattai (8%), and B. subtilis (4%) (with 98%-100% similarities). Sequences of the 16S rRNA genes from the 23 new isolates of Bacillus were deposited in the GenBank, National Center of Biotechnology Information (NCBI) database. To further examine the phylogenetic affiliation, the 16S rRNA gene sequences of all isolates were aligned with eight closely related reference sequences ( Figure 1). In general, a 16S rRNA gene sequence analysis of 23 selected isolates revealed that the dominant Bacillus species in this study were B. pumilus (34%) and B. altitudinis (33%), followed by B. megaterium (13%), B. amyloliquefaciens (8%), B. aryabhattai (8%), and B. subtilis (4%) (with 98%-100% similarities). Sequences of the 16S rRNA genes from the 23 new isolates of Bacillus were deposited in the GenBank, National Center of Biotechnology Information (NCBI) database. To further examine the phylogenetic affiliation, the 16S rRNA gene sequences of all isolates were aligned with eight closely related reference sequences ( Figure 1).

Figure 1.
Evolutionary relationships of taxa. The evolutionary history was inferred using the neighbor-joining method. An optimal tree with the sum of branch lengths = 0.34080293 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the maximum composite likelihood method and are in the units of the number of base substitutions per site. The analysis involved 32 nucleotide sequences. All positions containing gaps and missing data were eliminated. There were a total of 1405 positions in the final dataset. Evolutionary analyses were conducted in MEGA7.

Distribution of the Bacillus Species Isolated from Stingless Bee Honey from Different Geographical Locations
In the results, the honey samples from Batang Benar, Negeri Sembilan, and Serdang (Selangor) showed more variation in their Bacillus species compared to the other samples. Even though the species of bacteria and the number of colonies differed between the sites sampled, B. pumilus and B. altitudinis were the most widely distributed, as they were detected in four out of five samples of raw H. Evolutionary relationships of taxa. The evolutionary history was inferred using the neighbor-joining method. An optimal tree with the sum of branch lengths = 0.34080293 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the maximum composite likelihood method and are in the units of the number of base substitutions per site. The analysis involved 32 nucleotide sequences. All positions containing gaps and missing data were eliminated. There were a total of 1405 positions in the final dataset. Evolutionary analyses were conducted in MEGA7.

Distribution of the Bacillus Species Isolated from Stingless Bee Honey from Different Geographical Locations
In the results, the honey samples from Batang Benar, Negeri Sembilan, and Serdang (Selangor) showed more variation in their Bacillus species compared to the other samples. Even though the species of bacteria and the number of colonies differed between the sites sampled, B. pumilus and B. altitudinis were the most widely distributed, as they were detected in four out of five samples of raw H. itama honey collected from different geographical locations ( Figure 2). Our results coincide with a study on Argentine honeys, where the presence of B. pumilus, together with B. cereus and B. laterosporus, was found among the 70 samples examined [33]. Recently, Bacillus spp. were also reported as the most frequently isolated bacteria in honey, making up 67% of total isolates [34].
itama honey collected from different geographical locations ( Figure 2). Our results coincide with a study on Argentine honeys, where the presence of B. pumilus, together with B. cereus and B. laterosporus, was found among the 70 samples examined [33]. Recently, Bacillus spp. were also reported as the most frequently isolated bacteria in honey, making up 67% of total isolates [34].

Antimicrobial Test against Pathogenic Bacteria
The antagonistic activity of the isolates in this study was evaluated against Gram-positive and Gram-negative pathogenic bacteria: S. aureus, B. cereus, S. thyphimurium, E. coli, K. pneumonia, and P. aeruginosa. The results were compared to a commercial probiotic strain, Lactobacillus rhamnosus GG ( Table 2). Nineteen isolates that were Gram-positive bacteria, rod-shaped, catalase-positive, and were able to grow in the presence of 7% NaCl were selected. These strains exhibited inhibitory effects against at least one of the tested pathogens, except for B. pumilus HTI-3, B. megaterium HTI-16, B. megaterium HTI-17, B. megaterium HTI-18, B. aryabhattai HTI-21, and B. aryabhattai HTI-22. Two isolates with the most excellent antagonistic activity against the tested bacteria were B. amyloliquefaciens HTI-19 and B. subtilis HTI-23, where the degree of inhibition spectrum of B. amyloliquefaciens HTI-19 were almost comparable to L. rhamnosus GG (Figure 3)). B. amyloliquefaciens HTI-19 was able to inhibit the growth of all pathogenic bacteria in this study except for K. pneumoniae, while B. subtilis HTI-23 could inhibit four out of six pathogenic bacteria. The remaining strains exhibited remarkable but lower antagonistic effects in comparison to B. amyloliquefaciens HTI-19 and B. subtilis HTI-23.
(a) (b) Figure 2. Distribution of the Bacillus species isolated from stingless bee (Heterotrigona itama) honey from different geographical locations.

Antimicrobial Test against Pathogenic Bacteria
The antagonistic activity of the isolates in this study was evaluated against Gram-positive and Gram-negative pathogenic bacteria: S. aureus, B. cereus, S. thyphimurium, E. coli, K. pneumonia, and P. aeruginosa. The results were compared to a commercial probiotic strain, Lactobacillus rhamnosus GG ( Table 2). Nineteen isolates that were Gram-positive bacteria, rod-shaped, catalase-positive, and were able to grow in the presence of 7% NaCl were selected. These strains exhibited inhibitory effects against at least one of the tested pathogens, except for B. pumilus HTI-3, B. megaterium HTI-16, B. megaterium HTI-17, B. megaterium HTI-18, B. aryabhattai HTI-21, and B. aryabhattai HTI-22. Two isolates with the most excellent antagonistic activity against the tested bacteria were B. amyloliquefaciens HTI-19 and B. subtilis HTI-23, where the degree of inhibition spectrum of B. amyloliquefaciens HTI-19 were almost comparable to L. rhamnosus GG (Figure 3)). B. amyloliquefaciens HTI-19 was able to inhibit the growth of all pathogenic bacteria in this study except for K. pneumoniae, while B. subtilis HTI-23 could inhibit four out of six pathogenic bacteria. The remaining strains exhibited remarkable but lower antagonistic effects in comparison to B. amyloliquefaciens HTI-19 and B. subtilis HTI-23.

Antimicrobial Test against Pathogenic Bacteria
The antagonistic activity of the isolates in this study was evaluated against Gram-positive and Gram-negative pathogenic bacteria: S. aureus, B. cereus, S. thyphimurium, E. coli, K. pneumonia, and P. aeruginosa. The results were compared to a commercial probiotic strain, Lactobacillus rhamnosus GG ( Table 2). Nineteen isolates that were Gram-positive bacteria, rod-shaped, catalase-positive, and were able to grow in the presence of 7% NaCl were selected. These strains exhibited inhibitory effects against at least one of the tested pathogens, except for B. pumilus HTI-3, B. megaterium HTI-16

Tolerance to Acidic Conditions and Bile Salts
The effect of simulated gastrointestinal conditions on the viability of B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 in comparison to L. rhamnosus GG is presented in Table 3. After exposure to acidic conditions (pH 2.0) and 0.3% bile salt solution for 3 h, the survival rates of B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 were found to be >85%. In addition, both isolates exhibited significantly high survival rates in 0.3% bile salt solution compared to L. rhamnosus GG.

Cell Adhesion Activity of Bacillus Species
Autoaggregation is a probiotic characteristic that pertains to the entrapment of bacteria in an aggregated form, which allows for the stability of microbial strains in the gastrointestinal tract (GIT), resulting from lesser exposure to inhospitable intestinal conditions [35]. After 24 h of incubation, B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 showed autoaggregation abilities of 84.13% and 57.51%, respectively (Table 4). Interestingly, the autoaggregation and hydrophobicity percentage of both Bacillus species showed no significant difference from L. rhamnosus GG.

Discussion
The amount of aerobic bacteria detected in the fresh honey of stingless bees could be considered relatively low, with a mean value of 1.7 x 10 2 CFU/g, compared to its other byproducts, such as beebread and propolis (1.83 × 10 6 CFU/g) [12,37]. This is supported by the results from a previous study, as the presence of aerobic bacteria in honey was also detected in the range of 5.7 × 10 0 to 52.8 × 10 4 CFU/g [38]. The reason for this is that most bacteria are not able to multiply in honey due to the physicochemical properties of honey itself, such as high osmolarity, high sugar concentration, low pH, and the presence of many agents, including hydrogen peroxide and phytochemicals [13,39]. These conditions provide a stressful environment for bacteria, thus preventing the growth or even survival of different types of bacteria in honey. Therefore, a high number of aerobic bacteria could indicate contamination during processing, handling, or storing.
In a previous study by Esawy et al. [8], the strains isolated from honeybees were rod-shaped, Gram-positive, motile, and spore-forming. All of the isolates were moderately thermophilic and were preliminary identified as Bacillus spp. The results are in agreement with the results obtained in our study, where most of the bacterial isolates were also Gram-positive and rod-shaped. Potential probiotic species, B. amyloliquefaciens and B. subtilis, were also isolated in this study. Previously, both Bacillus species had been isolated from the gut and honey of Apis mellifera [40]. Probiotic bacteria were commonly selected from the Gram-positive bacteria, as the cell surface structures of Gram-positive microbes can ensure effective bacterial adhesion to the intestinal cell wall [41]. This characteristic is really important to ensure the successful colonization of the host.
A detailed analysis of the 16S rRNA gene sequences of the isolates exhibited significant diversity even in the case where bacteria were isolated from the same species of stingless bee. Environmental factors such as nectar, water, and pollen might be responsible for the diversity of the strains. A recent study found that the highest microbial diversity was found in multifloral honey [13]. Recently, the presence of B. altitudinis in stingless bee honey, H. itama, was reported for the first time [12]. B. altitudinis had been previously found in Apis mellifera honey together with other Bacillus isolates, namely B. licheniformis, B. safensis, B. zhangzhouensis, and B. xiamenensis [13]. This showed that B. altitudinis and B. pumilus have a niche in both honey samples of H. itama and A. mellifera. Interestingly, this species has been identified as one of the starter culture strains in rice wine [42]. Thus, the presence of these species in honey might suggest the roles of B. altitudinis in the fermentation of both A. mellifera and H. itama honeys.
B. amyloliquefaciens HTI-19 and B. subtilis HTI-23, which are associated with fermentation products, were successfully isolated from stingless bee honey. These species, together with B. methylotrophicus, B. safensis, and B. vallismortis, have been previously detected in A. mellifera honey, Korean traditional soy sauce, and the fermented seed condiment Kantong [8,43,44]. In fact, an assessment of cultivable microorganisms in honey has reported B. amyloliquefaciens to be the most prevalent strain among 13 species isolated from 38 honeys [45]. The 16S rRNA gene sequence of the B. subtilis HTI-23 isolate exhibited 99% sequence similarity to the B. subtilis strain BSFLG01 isolated from the black soldier fly larval gut, which is known as an invading species that has caused the mass infestation of domesticated stingless bees in Malaysia [46]. Hence, it was assumed that B. subtilis is a natural inhabitant in the honey.
Although it appears that there was no correlation between the microbial diversity and the geographical origin, the distinction of Bacillus strains found in the H. itama honey may be explained by the uses of the tubular proboscis of the bees while collecting nectar from various floral sources [47]. During the feeding process, the external surfaces of the bee's frontal organs are in close proximity to the nectar, and bacteria are then inoculated into the honey, which confirms the role of bees as bacterial vectors. Strains of B. amyloliquefaciens ssp. plantarum and B. methylotrophicus that have plant growth-promoting abilities are frequently isolated from plant material and/or soil [48]. However, bees might contribute to the presence of these bacteria strains in stingless bee honey during the pollination of different plants. Hence, it was hypothesized that Bacillus strains isolated from H. itama honey might come from floral sources, transferred by the H. itama bee during its foraging flight [49].
Another purpose of this study was to select the bacteria that exhibited excellent antimicrobial activity against pathogenic bacteria. The successful selection of antimicrobial producers from honey has been reported by several different authors [13]. For example, Manhar et al. [50] reported that B. amyloliquefaciens AMS1 inhibited the growth of L. monocytogenes and K. pneumoniae, but did not affect the growth of B. cereus, Yersinia enterocolitica, and Salmonella entericatyphimurium. In contrast with our study, the B. amyloliquefaciens strain HTI-19 exhibited a wider antimicrobial spectrum against pathogens, as it can inhibit the growth of S. thyphimurium and B. cereus. Many attempts were have been made to prevent the growth of B. cereus in food products, because B. cereus is known to be one of the major threats to food safety. Further characterization of B. amyloliquefaciens HTI-19 will be particularly helpful in food industries. It has been reported that B. amyloliquefaciens strains were able to inhibit the growth of a variety of fungal pathogens because of their ability to produce a vast array of antibiotics, such as bacillomycin, zwittermicin, bacilysin, difficidin, and fengycin [51,52].
In addition, the growth of S. aureus was successfully inhibited by most of the Bacillus strains in this study. Since this study used cell-free supernatant for the antimicrobial activity assay, the potential antimicrobial metabolites produced were bacteriocin, hydrogen peroxide, and lactic and propionic acid [53]. Despite a few Bacillus spp. being known as toxin producers, some of the Bacillus strains are already considered to be safe probiotic bacteria. These include B. endophyticus, B. subtilis, B. amyloliquefaciens, B. pumilus, and B. licheniformis [8,43,50]. Bacillus subtilis has also been shown to have a broad spectrum of antimicrobial activities over diverse pathogenic fungal and bacteria [54].
Tolerance to low-acidic gastric and bile-rich intestinal environments is one of the essential properties required for probiotic cultures in order to function effectively in the intestines, because such conditions provide a stressful environment for bacteria [55]. The results obtained after 3 h established the possibility that the strain can survive under acidic conditions that exist in the human gut (pH 2-5), as the transit time of the food along the human gut is a maximum of 3 h [10]. This result suggests that B. amyloliquefaciens species have high levels of survival in simulated gastric juices (pH 2.0), as previously reported by Wang et al. [56]. Bile salts have been reported to inhibit bacterial growth by disrupting cell membranes. Some studies have observed that some Bacillus spp. are weakly tolerant or sensitive to bile salt concentrations [56]; however, the present results showed that the survival rates of B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 in 0.3% bile salt solution were significantly higher (p < 0.05) than for L. rhamnosus GG. Tolerance to bile salt enables a probiotic strain to survive, grow, and exert itself during gastrointestinal transit [36].
The adherence ability of probiotic bacteria to intestinal epithelial cells involves various types of interactions, including hydrophobicity and autoaggregation [57]. The ability to adhere to epithelial cells and mucosal surfaces is considered to be a prerequisite for ideal probiotics. In this study, xylene was chosen as an apolar solvent because it reflects cell surface hydrophobicity and hydrophilicity [58]. As the results showed, both strains exhibited high hydrophobicity with xylene, indicating good bacterial adhesion to hydrocarbons. Patel et al. [25] have reported that the autoaggregation activity of B. subtilis DET6 is about 60%, which is in agreement with our study. These properties are crucial for probiotic cultures in colonizing epithelium cells in the digestive tract to prevent elimination by peristalses and to become functionally effective in intestinal balance [10]. Autoaggregation is also strongly correlated with cell adhesion to the digestive tract, which is responsible for the probiotic characteristics of bacteria [30]. The results showed that the two probiotic strains had high cell hydrophobicity and autoaggregation, indicating good cell adhesion ability.
The antibiotic susceptibility of probiotics should be measured for safety purposes. Antibiotic resistance gene transmission can occur due to transposons, plasmids, and bacterial gene mutations, leading to new antibiotic-resistant strains [59]. An antibiotic susceptibility test indicated that B. amyloliquefaciens HTI-19 and B. subtilis HTI-23 were sensitive to all antibiotics included in this study. Resistance to a given antibiotic can be inherent to a bacterial species or genus. In addition, γ-hemolysis and α-hemolysis are considered to be safe, and β-hemolysis is considered to be harmful [60] as β-hemolysis is an indication that bacteria contain cytotoxic phospholipases [61].

Conclusions
The results of the present research demonstrated that honey of different geographical origins in Malaysia can be considered as a reservoir of bacteria with antimicrobial activities, with potential for use as probiotic cultures. Interestingly, B. amyloliquefaciens HTI-19 not only showed a broad range of antimicrobial activities that could inhibit both Gram-positive and Gram-negative bacteria, but also was able to inhibit the growth of B. cereus and S. thyphimurium, which had not been inhibited previously by a different strain of B. amyloliquefaciens species in other studies. Two Bacillus strains (B. amyloliquefaciens HTI-19 and B. subtilis HTI-23) that were isolated from stingless bee honey possess great potential as probiotics for human and animal use and as fermentation starter cultures. This was supported by positive probiotic characteristics such as high survivability in the artificial modified digestive tract system, wide antimicrobial spectra, and safety confidence with regard to antibiotic susceptibility and nonhemolytic activity. The current findings suggest that these strains may exhibit the ability to remain viable after exposure to stressful environments in the gastrointestinal tract of humans and animals, thus being able to be functionally effective in the intestine. As probiotic effects on certain noncommunicable diseases have proven to be strain-specific, further investigation into these isolates may lead to the discovery of new beneficial probiotic strains that can be used in the therapeutic field.