To isolate B. subtilis from nattō, 250 g of nattō were heated at 80°C for 20 min to inactivate (kill) vegetative bacteria and leave only spores as viable cells (Lombardía et al., 2006). Aliquots from serial dilutions of heat-treated nattō were seeded on Luria-Bertani agar (LBA) plates and incubated for 36 h at 37°C. This procedure was repeated three times independently and in all the repetitions only one type of heat-resistant bacterial isolate was obtained. These nattō isolates and the laboratory B. subtilis 168 strains were cultured on Luria-Bertani (LB) broth except when another broth medium is indicated (Pedrido et al., 2013). The growth media for the other bacteria and fungi used in this work were: trypticase soy broth (TSB) for Staphylococcus aureus; brain heart infusion (BHI) broth for Enterococcus faecalis and Listeria monocytogenes; Chapman broth for MR S. aureus; MacConkey broth for Salmonella enterica, Salmonella typhimurium, and Klebsiella pneumoniae; MacConkey Agar with Sorbitol for EHEC O157:H7; thiosulfate-citrate-bile salts-sucrose (TCBS) broth for Vibrio cholerae; King A broth for Pseudomonas aeruginosa; and potato dextrose agar (PDA) for fungi growth. Schaeffer’s sporulation medium (SM) was used to obtain B. subtilis spores (Pedrido et al., 2013). The efficiency of spore formation was measured by counting viable vegetative and spore cells before and after heat treatment at 80°C, respectively, as previously described (Pedrido et al., 2013). To prepare cell suspensions, B. subtilis (168 and DG101 strains) and Escherichia coli DH5α (used as negative control) were grown in SM or LB broth, respectively, at 37°C with shaking (7,258 × g) during 24 h. The B. subtilis cultures were heated at 80°C for 20 min to kill cells that did not form spores. Then, the bacterial cultures (i.e., E. coli and B. subtilis) were centrifuged (3,578 × g, 15 min at 4°C) and the pellets were brought to a final concentration of 5 × 10⁸ colony forming units per milliliter (CFU/mL) in tris buffer solution (10 mM; pH 7.0) and used immediately.

The presumptive B. subtilis natto isolate was microbiologically characterized using routine nutritional and biochemical tests (see Results section). The molecular typification was performed by the Sequencing Service, Molecular Biology Department of The Pasteur Institute of Montevideo, Uruguay.¹ Briefly, a sample of chromosomal DNA from the presumptive B. subtilis natto isolate was used for PCR amplification using consensus primers for the 16S ribosomal RNA gene from the B. subtilis group. The sequences of the used forward and reverse primers were 5′-AGAGTTTGATCMTGGCTC-3′ and 5′-ATACTAGCGACTCCGACTTC-3′, respectively. The amplification product (~1,500 bp) was purified, sequenced, and compared with the sequences deposited in GenBank using the BLAST tool² and in the Ribosomal Database Project.³ A phylogenetic tree was constructed based on 16S rDNA sequences retrieved from the GenBank database of bacilli and LAB (i.e., Lactobacillus and Bifidobacterium). The best match of the isolate from nattō was with the 16S ribosomal coding gene of the B. subtilis natto BEST195 strain. The 1,407 bp DNA sequence corresponding to the 16S ribosomal gene of the new nattō isolate was deposited in GenBank under the accession number OQ813498.⁴

Five mL aliquots of B. subtilis DG101 or E. coli DH5α (5 × 10⁸ CFU/mL) were independently incubated in the presence of 0, 50, 100, 150, 200, and 250 μ/mL of lysozyme at 37°C. After 20 min (Zheng et al., 1988; Kuwana et al., 2005), an aliquot of each bacterial suspension was taken, and serial dilutions were made in 0.1 M phosphate-buffered saline (PBS) pH 6.2. The corresponding dilutions were seeded with a Drygalsky spatula in LB agar (LBA) plates and after 24 h of incubation at 37°C, the viable cells (CFU) were counted.

To determine the pH resistance of the studied bacteria (B. subtilis DG101 and E. coli DH5α), 10 mL of each bacterial suspension were centrifuged at 3,578 × g for 10 min at 4°C and the cell pellet was reconstituted in the same original volume of sterile distilled water (5 × 10⁸ CFU/mL). The suspensions were brought to different pH (i.e., 1.5, 2.5, and 3.5) with 0.1 M HCl. The bacterial samples were incubated at 37°C for 2 h, considering this is an estimated/average physiological time of gastric emptying with a pH range of 1.5–3.5 (Hellmig et al., 2006; Beasley et al., 2015). The control pH test was carried out at pH 6.2 in 0.1 M PBS. Following the incubation times, serial dilutions of the cells were made in 0.1 M PBS (pH 6.2) to neutralize the acidity of the medium (when necessary), and different aliquots were seeded with a Drygalsky spatula in LBA plates. After 24 h of incubation at 37°C, the viable cells (CFU) were counted. The survival rate was calculated as the percentage of CFU growing in LBA compared to the initial concentration of cells, before exposition to the different pH (Hyronimus et al., 2000).

Briefly, the temperature and salt tolerance assays were performed in 50 mL bottles filled with 10 mL of LB broth supplemented with NaCl at 0.5, 2, 5, 7, and 10%, and B. subtilis DG101 (5 × 10⁵ CFU/mL). Samples were grown at 30, 37, 42, 50, and 54°C with shaking (7.258 × g) for 24 h. After the overnight growth, different dilutions of the salt and temperature-treated cultures were seeded with a Drygalsky spatula in LBA plates and after 24 h of incubation at 37°C, the viable cells (CFU) were counted. The culture that grew in unmodified LB (i.e., LB containing 0.5% NaCl) at 30°C was considered as the control and the final growth yield of the remaining cultures was expressed as a % of the growth yield of the control culture (7.5 × 10⁸ CFU/mL).

Five mL of B. subtilis DG101 or E. coli DH5α suspensions (5 × 10⁸ CFU/mL) were centrifuged at 3,578 × g for 10 min and each pellet was re-suspended in 10 mL of 0.3% bile salt solution (50% sodium cholate and 50% sodium deoxycholate) or in sterile distilled water as a control (Duc et al., 2003). The prepared suspensions were incubated at 37°C for 0, 1, 2, 3, and 4 h. Next, aliquots were taken, and serial dilutions were made in 0.1 M PBS pH 6.2. The corresponding dilutions were seeded with a Drygalsky spatula in LBA plates and after 24 h of incubation at 37°C, the viable cells (CFU) were counted. The survival rate was calculated as the percentage of CFU obtained in LBA compared to the initial cell concentration. The InfoStat 2016 and Sigmaplot 10.0 programs were used to perform the graphs.

B. subtilis DG101 was screened for the production of exoenzymes related to the digestion of nutrients of different natures. Briefly, B. subtilis DG101 was streaked on LBA plates containing one of the following nutrient substrates: gelatin (LBA-gelatin 10 g/L as indicator of protease activity), pectin (LBA-pectine 5 g/L for peptinolytic activity), carboxymethylcellulose (LBA-CMC 10 g/L for cellulolytic activity), Tween-80 (LBA-Tween 80 15 mL/L for lipolytic activity), pythate (LBA-phytate 5 g/L for hydrolytic activity related to release of soluble phosphate), and starch (LBA-soluble starch 10 g/L for amylase activity). The Petri dishes streaked with B. subtilis DG101 were incubated at 37°C for 36 h, and the development of a clear zone around the colonies was indicative of the presence of exoenzyme activity (Lee et al., 2005; Hossain et al., 2020). For the determination of hemolytic activity, suspensions of B. subtilis DG101 were streaked onto blood agar plates (5%–10% sheep blood per plate) and incubated for 24 h at 37°C. The isolates were then examined for the presence of clear zones surrounding the colonies. Clear zones are considered as beta hemolysis, greenish zones as alpha hemolysis, and the absence of hemolysis zones is known as gamma hemolysis (Gerhardt, 1994).

B. subtilis DG101 was incubated in LB broth supplemented with the corresponding metal: cadmium (Cd_(II)), chromium (Cr_(VI)), or arsenite (As_(III)) at 0, 10, 50, 100, 150, 200, 250, and 300 ppm at 37°C with shaking (5 × g). After 20 h, the bacterial cultures were centrifuged and the determination of metal concentration was performed on the cell-free supernatants (i.e., remaining metal in the supernatant). The concentration of Cr_(VI) was determined by a colorimetric reaction based on the formation of a complex between Cr_(VI) and 2,4-difenylcarbazide in an acid solution to a λ = 540 nm, as previously described (Megharaj et al., 2003). The concentration of As_(V) was determined by a colorimetric reaction based on the formation of a red complex, between arsenic hydride with silver diethyl-carbamate, to a λ = 530 nm (Upadhyay et al., 2018). Total chromium (Cr_(VI) plus Cr_(III)), Cd_(II), and As (As_(V) plus As_(III)) were determined by atomic absorption spectrophotometric determination at λ = 357.3 nm, λ = 228.8 nm, and λ = 193.7 nm, respectively (Nakahara et al., 1973; Megharaj et al., 2003; Liu et al., 2006; Cheng and Li, 2009; Guibaud et al., 2009; Singh and Singh, 2014; Upadhyay et al., 2018).

The selected methodology to test for antimicrobial activity of B. subtilis DG101 against microbial pathogens was adapted from the agar diffusion method (Bauer et al., 1959, 1966; Bonev et al., 2008). Briefly, the antimicrobial activity of B. subtilis DG101 against pathogenic bacteria and fungi (Candida albicans, Fusarium verticillioides, Fusarium graminearum, Aspergillus flavus, Penicillium sp.) was performed on nutrient agar (NA) or PDA plates, respectively. Then, 100 μL of overnight cultures of the different microbial pathogens to be tested were individually poured, using a Drygalsky spatula, on the corresponding type of agar plate (NA or PDA plates). After the seeded microorganisms were totally absorbed in the medium, three sterile paper disks (4 mm in diameter) pre-absorbed with 15 μL of an overnight culture of B. subtilis DG101 grew on LB broth (~2 ×1 0⁶ CFU/paper disk) were loaded equidistantly from each other on each type of agar plate. The NA and PDA plates containing the bacterial and fungi pathogens, respectively, with the B. subtilis DG101-poured paper disks were incubated for 24 h at 37°C or 72–96 h at 30°C, respectively. The clear zone formed around the paper disks indicated the growth inhibition of the tested pathogen. The inhibition halo of each bacterial pathogen produced by B. subtilis DG101 was measured and used to classify the pathogen as sensitive or resistant based on the recommendations of the American Society for Microbiology for antimicrobial control.⁵ For the determination of susceptibility of fungal pathogens to B. subtilis DG101, the fungal growth inhibition index was calculated from measurements of fungal radial growth toward (X1) versus perpendicular to (X2) the bacterial colony according to the formula [1 − (X1/X2)] *100 (Quecine et al., 2015; Bartolini et al., 2019). Paper disks pre-absorbed with sterile water were used as negative controls of antimicrobial activity.

The antibiotic sensitivity of B. subtilis DG101 was tested against the following antibiotics commonly used in hospitals: amikacin, ampicillin/sulbactam, cephalothin, cefepime, cefoxitin, ceftazidime, cilastatin, ciprofloxacin, erythromycin, gentamicin, imipenem, nitrofurantoin, norfloxacin, oxacillin, penicillin, piperacillin/tazobactam, tetracycline, trimethoprim/sulfamethoxazole and vancomycin. All the selected antibiotics are on the lists of recommended antibiotics by the U.S. Food and Drug Administration (FDA), the Center for Drug Evaluation and Research (CDER), the Clinical and Laboratory Standards Institute (CLSI), and The European Committee on Antimicrobial Susceptibility Testing (EUCAST). Briefly, commercial disks containing an appropriate amount of each antibiotic were added to Mueller Hinton agar plates, previously poured with B. subtilis DG101, and incubated under aerobic conditions at 37°C for 24–48 h. The inhibition zones (clear zone, in mm) were measured using a caliper. All experiments were performed in triplicate on the same day. The breakpoint values used for categorizing the microorganisms as resistant were according to the Clinical and Laboratory Standards Institute (2023).

For biofilm formation, B. subtilis DG101 and B. subtilis 168 cultures were grown in LB medium until they reached the stationary phase. Then, 50 μL of each culture was diluted in 2 mL of fresh LBY medium (LB broth supplemented with 4.0% yeast extract; Pedrido et al., 2013) and statically grown at 37°C for 36 h. For quantification of formed biofilm (biofilm mass), we measured the biofilm mass that adhered to the surface of the microtiter plate wells (Pedrido et al., 2013). To this end, B. subtilis (DG101 or 168) cells were grown in 96-well polyvinylchloride (PVC) microtiter plates at 37°C for 36 h in LBY medium. The bacterial inoculum for the microtiter plates was obtained by growing the cells in LB to mid-exponential phase and then diluting the cells to an OD_600 nm of 0.08 in LBY and biofilm formation was monitored by staining with crystal violet (CV) as described by Pedrido et al. (2013). Growth medium and non-adherent cells were carefully removed from the microtiter plate wells. The microtiter plate wells were incubated at 60°C for 30 min. Biofilms were stained with 300 μL of 0.3% CV in 30% methanol at room temperature for 30 min. Excess CV was removed, and the wells were rinsed with water. The CV from stained biofilms was then solubilized in 200 μL of 80% ethanol/20% acetone. Biofilm formation was quantified by measuring the absorbance at 595 nm for each well using a Beckman Coulter Multimode Detector (DTX 880).

Briefly, microtiter plates were coated with 100 μL of each solution containing the following extracellular matrix (ECM) proteins (Sigma-Aldrich): fibronectin, collagen, and gelatine (10 μg/mL in bicarbonate buffer pH 9.6); and subsequently incubated overnight at 4°C. The unfixed proteins were removed, and plates were washed three times with PBS pH 8. After blocking with 0.5% non-fat milk (in PBS), to prevent non-specific bacterial binding, wells were washed again twice with PBS. Finally, bacterial suspensions of B. subtilis DG101 or B. subtilis 168 (100 μL, 1×10⁸ CFU/mL) were added and the plates were incubated on an orbital shaker for 2 h at 37°C. Unbound bacteria were removed by washing the wells five times with PBS. Adhered bacteria to the wells were fixed at 60°C for 20 min and stained with 100 μL/well of CV (0.3% in methanol—H₂O 30% v/v) for 45 min as previously described (Styriak et al., 1999; Rodriguez Ayala et al., 2017a). The wells were subsequently washed extensively with PBS to remove the unbound stain. After adding 100 μL of citrate buffer pH 4.3 to each well and 45 min incubation at room temperature (to release the stain bound to bacteria), the absorbance values (A_595 nm) were determined in a Multiscan enzyme-linked immunosorbent assay reader as described (Sun et al., 2019). The average values of three independent experiments performed in quadruplicate were calculated. The specificity of binding was tested using BSA as the immobilized protein. Data are expressed as the mean absorbance value (A_570 nm), with non-specific binding corrected by subtraction.

The colonization of the GI tract by B. subtilis DG101 was performed according to Hoa et al. (2001). Briefly, B. subtilis DG101 eliminated in feces was monitored in Sprague Dawley rats (male, 6 to 8 weeks old, n = 12) housed individually in cages with gridded floors. Animals were deprived of drinking water for 6 h and then fed on a single dose of 2 × 10⁷ spores of B. subtilis DG101 contained in a volume of 20 mL of drinking water. After this procedure, the animals were fed ad libitum with an unrestricted drinking water supply. Total feces were collected day to day and weighed for 2 weeks. To determine the total number of B. subtilis DG101 viable cells, including spores and vegetative cells in feces, samples were suspended in 10 to 30 mL of 0.1 M PBS with frequent and rigorous vortexing. Serial dilutions were made with PBS, seeded on LBA plates, and incubated at 37°C for 2 days. B. subtilis DG101 colonies were identified by their characteristic colony morphology (see Figure 1). To determine the total number of B. subtilis DG101 spores, the same procedure was used, except that feces were suspended in 0.1 M PBS and incubated at 80°C for 20 min to inactivate (i.e., kill) vegetative cells. The total CFU count was extrapolated for the total weight of feces collected and represented as a percentage of the initial dose.

Three-weeks-old Wistar male rats (n = 14) were housed in metal cages (2 animals/cage) in a temperature-controlled room under a 12 h light/dark cycle. After a 1 week adaptation period on powder chow diet (Gepsa Feed Corporation), the animals were randomly assigned to 2 groups (n = 7): Control group (rats fed ad libitum on powder chow) and Treated group (rats fed ad libitum on powder chow supplemented with B. subtilis DG101 1 × 10¹⁰ CFU/kg of food). On day 120, blood samples were drawn from the tail rats. Serum was separated by centrifuging the blood at 7,500 × g for 15 min. Triglyceride (TG, Cat. 1780101), total cholesterol (TC, Cat. 1220101), high-density lipoprotein-cholesterol (HDL-c, Cat. 1220103), and low-density lipoprotein-cholesterol (LDL-c, Cat. 1220108) were measured using enzymatic reagent kits from Wiener Lab (Wiener Group Co, Argentina).

For protein quantification under normal and dietary restriction conditions, 20 male Wistar rats, 3 weeks old, were housed in metal cages (2 animals/cage) in a temperature-controlled room under a 12 h light/dark cycle. After a 1 week adaptation period on a powder chow diet, the rats were randomly assigned to 4 groups (n = 5): Control group (rats fed ad libitum, consuming an average amount of 26 g of food/day/animal), Treated group (rats fed ad libitum and supplemented with B. subtilis DG101 1 × 10¹⁰ spores/kg of food), Nutritional stress group (rats with a dietary restriction that consisted of ~75% of the average amount of food consumed by day −19.5 g of food/day/animal-under physiological conditions), and Treated nutritional stress group (rats with dietary restriction supplemented with B. subtilis DG101 1 × 10¹⁰ spores/kg of food). After 90 days, blood samples were drawn from the tail rats. Serum was separated by centrifuging the blood at 7,500 × g for 15 min. Total proteins (TP, Cat. 1999736), albumin (ALB, Cat. 1009240), and globulins (TG, Cat. 1690004) were measured using enzymatic reagent kits from Wiener Lab. All aspects of the experiment were conducted according to the guidelines provided by the ethical committee of experimental animal care at the National University of Rosario (Santa Fe, Argentina).

Lifespans for the wild-type C. elegans N2 strain (Bristol strain) were monitored at 20°C as previously described (Cogliati et al., 2020). Briefly, worm embryos were isolated by exposing hermaphrodite adult worms to alkaline hypochlorite treatment for 3 min, and synchronized eggs were allowed to develop. In all cases, L4/young adult worms (n = 100) were used at time zero for lifespan analysis; they were transferred to fresh nematode growth medium agar (NGMA) plates previously seeded with E. coli OP50 or B. subtilis DG101 or B. subtilis 168 cells each day until the death of the last worm of each assay (Cogliati et al., 2020). Worms were considered dead when they ceased pharyngeal pumping and did not respond to prodding with a platinum wire. Worms with internal hatching were removed from the plates and excluded from lifespan calculations.

B. subtilis 168- or B. subtilis DG101-fed N2 (10 days-old adults) worms were collected, washed three times with M9 buffer (Cogliati et al., 2020), and seeded in 10 cm Petri dishes prepared with NGMA without food for 1 h. Then, approximately 75 worms were placed in the center of 6 cm plates prepared with NGMA. After all the animals were transferred to the center of the assay plates, 2 μL of attractant or repellent were seeded 2 cm from the center of the plate, and 2 μL of solvent (control) in which the attractant or repellent was diluted were seeded equidistantly. Both the attractant, repellent, and solvent (control) were added with a 1-μL drop of 1 M azide. The plates were incubated for 1 h at 25°C. Then, worms found at each end of the plates were counted, and the chemotaxis index (CI) was calculated (Bargmann et al., 1993). The attractant compounds used for the assays were 0.5% diacetyl (DA, Sigma-Aldrich) diluted in ethanol and 1 mM isoamyl alcohol (IAA; Sigma-Aldrich) diluted in water, 1 mM 2,3-pentanedione and 1 mM 2-butanone (Sigma-Aldrich). The repellent compounds used for the assays were 0.3% octanol and 15% 2-nonanone (Sigma-Aldrich). The CI was defined as the number of worms at the attractant or repellent location—the number of worms at the control location divided by the total number of worms on the plate (Bargmann et al., 1993).

All experiments were performed in triplicate and the mean values along with the standard deviation are given. All the data from laboratory experiments were analyzed separately for each experiment and were subjected to an analysis of variance (ANOVA test). Significant effects were determined by the F value (p ≤ 0.05). For C. elegans survival assays, experiments were performed at least three times in duplicate. Mean survival days, standard error of the mean (S.E.M.), intervals of mean survival days with 95% confidence, and equality p-values to compare averages were calculated by log-rank and Kaplan–Meier tests using the OASIS program. The S.E.M. values are used in the figures; p < 0.1 was considered statistically significant.

The genetic pathways (e.g., the phosphorelay signaling and the Spo0A/SinR/AbrB/LuxS regulatory circuit) controlling the multicellular behavior of the B. subtilis natto strain described here have been previously reported (Lombardía et al., 2006; Pedrido et al., 2013; Grau et al., 2015: Rodriguez Ayala et al., 2017a). Now, we present a short discussion of the biofilm, sliding, and adherence proficiencies of the natto isolate plus a profound and complete description of its in vitro and in vivo probiotic properties. The bacterial nattō isolate (see Material and methods) showed a complex colony architecture, resembling the sophisticated colony (biofilm) architecture of wild (undomesticated) B. subtilis isolates (Branda et al., 2001), (Figure 1A). Accordingly, the biofilm physiognomy of the nattō isolate remarkably differed from the simple and undifferentiated colony (biofilm) architecture of domesticated B. subtilis isolates (Branda et al., 2001), i.e., 168 strain (Figure 1B). The confirmation that the nattō isolate belongs to B. subtilis was achieved through Sanger sequencing and alignment of the 16S ribosomal RNA (rRNA) encoding gene (Figure 2, see Material and methods for details). The nattō isolate exhibited strong homologies (~99.7% similarity and ~99.5% identity over its full length) to the 16S rRNA sequences belonging to different B. subtilis strains, and significant homology with other Gram-positive probiotic bacteria of the genus Lactobacilli and Bifidobacterium (Figure 2). Interestingly, the 16S rRNA gene of the nattō isolate showed the strongest homology with the BEST 195 B. subtilis natto strain previously characterized (Kamada et al., 2014). Therefore, the nattō isolate (Figures 1A, 2) was given initially the name Bacillus subtilis RG4365 (Roberto Grau; Lombardía et al., 2006; Pedrido et al., 2013; Grau et al., 2015; Rodriguez Ayala et al., 2017a) and more recently, because of its approval for consumption as a human probiotic, B. subtilis DG101 (Doctor Grau; Cardinali et al., 2020; Rodriguez Ayala et al., 2021, 2022a, 2022b). Both strain designations (i.e., RG4365 and DG101) refers to the same nattō strain. The 1,407-base sequence fragment of the 16S rRNA gene of B. subtilis DG101 was deposited in GenBank (accession number OQ813498).

B. subtilis DG101 was initially characterized by Gram staining, sporulation proficiency, and basic biochemical and physiological tests whose results correlated well with the expected biochemistry and physiology attributes of bacilli (Table 1A). B. subtilis DG101 was proficient to grow in the presence of high salt concentrations at different growth temperatures (Table 1B) and in the production of several exoenzymes (Table 1C) involved in the digestion of diverse types of macromolecules (i.e., polysaccharides, complex lipids, proteins). The in-situ production of these exoenzymes in the host intestine could play a positive role in improving food diet digestion. Importantly, B. subtilis DG101 lacked hemolysis activity (Table 1C) and was sensitive to antibiotics routinely used to treat infections (Table 1D).

B. subtilis DG101 was able to grow in the presence of high concentrations of toxic metals (i.e., Cd _(II), Cr_(VI), or As_(III)) (Figure 3). The minimal inhibitory concentrations (MICs) of B. subtilis DG101 growth in the presence of each metal were ~ 300 ppm for Cd _(II) and Cr_(VI); and ~250 ppm for As_(III) (Figure 3A). These MIC values were significantly higher than the metal limits allowed to be present in drinking water: 5 ppb, 5 ppm, and 10 ppb for Cd _(II), Cr_(VI) or As_(III), respectively (Nakahara et al., 1973; Singh and Singh, 2002; Megharaj et al., 2003; Liu et al., 2006; Cheng and Li, 2009; Guibaud et al., 2009; Upadhyay et al., 2018). The proficiency of B. subtilis DG101 to grow satisfactorily in the presence of 50 ppm of each metal (Figure 3A) correlated with the significant removal of Cd _(II), Cr_(VI), and As_(III) from the liquid phase of the culture (i.e., cell-free supernatant) (Figure 3B). The presence of Cd _(II), Cr_(VI), and As_(III) in the cell-free supernatant of B. subtilis DG101 that grew overnight in the presence of 50 ppm of each metal was reduced more than ~80%, ~ 60%, and ~85%, respectively (Figure 3B). The growth of B. subtilis DG101 in the presence of each metal correlated with the absorption of Cd _(II) to the cell envelope, and the reduction or oxidation of Cr_(VI) to Cr_(III) and As_(III) to As_(V), respectively (data not shown). The proficiency of B. subtilis DG101 to transform toxic metals into less hazardous species in humans (i.e., conversion of Cr_(VI) to Cr_(III) and As_(III) to As_(V)) or a less available form (absorption of Cd_(II) to the cell envelope) opens the potential use of this probiotic bacilli for environmental bioremediation uses and/or prevention/treatment of food/water metal poisoning.

The growth rate (μ) of B. subtilis DG101 cultured in LB broth with shaking at 37°C was ~0.0231 min⁻¹ with a generation time (t_g) of 30 min (Figure 4A). This growth showed two distinctive features that distinguish B. subtilis DG101 from other B. subtilis strains previously described (e.g., strains 168 and NCIB3610; Branda et al., 2001). First, as a key difference with the swimming proficiency of the B. subtilis 168 and NCIB3610 strains (Branda et al., 2001), B. subtilis DG101 was not motile in the liquid medium and this phenotype correlated with the absence of visible flagella after sample staining with a specific dye for flagella visualization (Table 1A and data not shown). Associated with the observed non-motility phenotype in liquid broth with shaking, B. subtilis DG101 showed a tendency to form long strings (longer than 50 μm each) constituted of individual cells lined up one after the other (Figure 4B). The tendency to form long cords of cells was observed during the entire log phase of growth and reached the maximal expression of cords formation at the transition from T₋₁ to T₀ (the end of the log phase of growth, Figure 4B). At this developmental time (T₀) in LB broth, a small portion (~ 15%) of the culture initiated the sporulation program, and later (approximately 2–3 h after T₀) the cellular cords were completely dissembled, and free vegetative cells and sporangia (similar to the ones of B. subtilis 168 or B. subtilis NCIB3610, data not shown) were released into the culture medium (data not shown). Surprisingly, B. subtilis DG101 was able to perform surface-associated motility similar to the flagella-dependent swarming motility exhibited by B. subtilis NCIB3610 (Branda et al., 2001) but by a flagella-independent mechanism of sliding motility (Grau et al., 2015; Figure 4C).

Under sessile conditions of growth (i.e., without shaking) in liquid and solid medium, B. subtilis DG101, but not B. subtilis 168, formed (liquid and solid) robust, sophisticated, and long-lasting hydrophobic biofilms (Branda et al., 2004; Pedrido et al., 2013; Grau et al., 2015; and Figures 4D,E). The proficiency in biofilm formation depended on the adherence ability of the bacterium to different substrates (Pedrido et al., 2013). For example, B. subtilis DG101 showed a strong and specific adherence to diverse proteins present in the extracellular matrix of the human intestinal mucosa (e.g., collagen and fibronectin), an adhesive property that would be important for the competitive exclusion of pathogen adherence (VidyaLaxme et al., 2014; Rodriguez Ayala et al., 2017a; and Figure 4F).

The antimicrobial activity of B. subtilis DG101 was assessed by measuring the growth inhibition zone diameter of different pathogenic bacteria and fungi that grew in the presence of the probiotic (see Material and methods for details). The results shown in Table 2 indicate the effectiveness of B. subtilis DG101 in controlling the growth of pathogenic Gram-positive and Gram-negative bacteria (Table 2A) as well as the proficiency of the probiotic to inhibit the growth of yeasts and mycotoxin-producer filamentous fungi (Table 2B).

A probiotic bacterium must be able to remain alive in the hostile environment of the host GI tract (Cutting, 2011). To this end, the probiotic must resist the lytic activity of the lysozyme during its transit through the buccal cavity and the low pH and bile salts present in the stomach and intestine, respectively (Zheng et al., 1988; Duc et al., 2003; Kuwana et al., 2005; De Vecchi and Drago, 2006; Hellmig et al., 2006; Beasley et al., 2015). As shown in Figure 5A, B. subtilis DG101 spores were resistant to the lysozyme’s hydrolysis at similar concentrations present or secreted in the human buccal cavity (Figure 5A), the low pH that resembled the exposure and transit through the stomach acidity (Figure 5B), and bile salts concentrations that correlated with physiological bile concentrations present in the human GI tract (Figure 5C).

The in vitro stress resistance of B. subtilis DG101 to GI-simulated conditions (Figures 5A–C) correlated well with its proficiency in in vivo gut colonization in animal models (Figure 5D). In this assay, 60 days-old Sprague Dawley male rats were fed with a single dose of 2 × 10⁷ spores of B. subtilis DG101 present in drinking water (20 mL), and the fecal elimination of the probiotic strain was quantified day to day for 2 weeks (see Material and methods for details). B. subtilis DG101 started to be present in the feces of inoculated animals approximately 15 h after the feeding, and after ~40 h, the CFU of B. subtilis DG101 in animal feces passed the initial inoculum of the probiotic (Figure 5D), indicating that the ingested spores were able to germinate and grow (data not shown). After 2 weeks, the total CFU present in the feces of B. subtilis DG101-fed animals (a total of 3.0 × 10⁸ CFU, 10% vegetative cells, and 90% spores, data not shown) exceeded by ~1.5 log the total CFU of B. subtilis DG101 ingested with the drinking water (2.0 × 10⁷ CFU). Interestingly, the B. subtilis DG101-fed animals continued with the fecal elimination of the probiotic (spores and vegetative cells, data not shown) at a rate of ~1 × 10³ to 5 × 10³ CFU/gram of fecal matter during 3 months after the end of the experiment (the CFU of spore-forming indigenous bacilli different to B. subtilis DG101 in rat feces was around 1 × 10¹ to 1 × 10² CFU/gram of fecal matter, data not shown). Overall, these results show that B. subtilis DG101 is able to persist, grow, and colonize the GI tract.

To analyze the potential beneficial effects of B. subtilis DG101 on blood biochemical parameters, the food of Sprague Dawley male rats, from day 1 after weaning, was supplemented with a daily dose of 1 × 10¹⁰ CFU of B. subtilis DG101/kg food (treated group). After 120 days, blood samples were taken from the animal tails to analyze and compare the lipid levels present in the serum with the lipid levels present in animals of similar age fed without the probiotic supplementation (control group). The analysis included the determination of triglycerides, total cholesterol, HDL cholesterol, and LDL cholesterol. As shown in Figure 6A, the animals of the treated group (compared with the control group animals) did not modify their triglyceride content although they showed a significant reduction of the levels of total cholesterol and LDL cholesterol (“bad cholesterol”). Accordingly, the animals of the treated group showed higher serum levels of HDL cholesterol (“good cholesterol”) compared with the control group animals (Figure 6A). These results provide experimental evidence of the beneficial effect of B. subtilis DG101 on serum lipid homeostasis.

Next, it was interesting to analyze the performance of B. subtilis DG101-fed animals under nutritional stress conditions (dietary restriction). It is known that dietary restriction (or caloric restriction) represents nutritional stress that decreases the level of serum proteins (Alexopoulos et al., 2004). Therefore, we were intrigued to evaluate if B. subtilis DG101 could manage this nutritional stress (e.g., maintenance of a physiologic level of plasmatic proteins) to benefit host health. To answer this question, 30 days-old male Wistar rats were divided into four groups as indicated in the Material and methods section. As shown in Figures 6B,C, the subgroup of animals fed with the supplement of B. subtilis DG101 and under nutritional stress showed higher levels of total serum proteins and albumin than the respective group of animals under nutritional stress and without B. subtilis DG101 supplementation (p < 0.01). The other main protein component of the serum, apart from albumin, are the globulins which are divided into alpha-1, alpha-2, beta, and gamma globulins (Abbas et al., 2011). As shown in Figure 6D, the serum concentration of globulins of rats complemented in their diet with B. subtilis DG101 was significantly higher in animals with or without nutritional stress. Interestingly, there were no significant differences in the levels of total serum proteins, albumin, and globulins in B. subtilis DG101-fed animals without nutritional stress and B. subtilis DG101-fed animals under nutritional stress (Figures 6B–D). These findings suggest that hosts colonized by B. subtilis DG101 were well-equipped to respond to the stressful condition of nutrient limitation. Overall, these results provide experimental evidence of the beneficial effect of B. subtilis DG101 on the maintenance of serum protein levels in healthy individuals under physiological conditions of growth or subjected to nutritional stress.

One interesting property of some probiotic strains is the attribute to extend the host lifespan in different animal models (e.g., flies, mice, and worms), and we wondered if B. subtilis DG101 would possess this anti-aging property. To this end, we assayed the lifespan extension of the animal model Caenorhabditis elegans (Cogliati et al., 2020) when its gut was colonized by different bacterial strains including B. subtilis DG101. One-day-old wild-type worms of the Bristol N2 strain were grown on NGMA plates seeded with E. coli OP50 or B. subtilis DG101 (the ingested bacteria that survived the transit from the worm grinder to the gut constituted the gut flora of the animal). The NGMA plates seeded with the different bacteria were incubated at 20°C and the worm survival was monitored every day as described (see Material and methods and Donato et al., 2017). As shown in Figure 7A, N2 worms fed on B. subtilis DG101 gained ~45% in lifespan extension (mean survival 33.22 ± 0.82 days) when compared with the lifespan extension of the N2 worms colonized by the lab routine and the non-probiotic strain E. coli OP50 used to feed C. elegans (mean survival 18.37 ± 0.63 days, p < 0.001). Figure 7A also shows the lifespan extension of N2 worms fed the domesticated and non-probiotic B. subtilis 168 (mean survival 20.34 ± 0.71 days) compared with the lifespan of N2 worms fed E. coli OP50 (p < 0.001). Importantly, B. subtilis 168 and E. coli OP50 were much less effective than B. subtilis DG101 in prolonging the life expectancy of C. elegans when colonizing its gut.

In addition to an extension of the host life expectancy, a healthy life extension (i.e., a host aging accompanied by a delay in the behavioral decay associated with the age progression) is also desirable (Cogliati et al., 2020). One sensory behavior we examined was chemotaxis. The age-synchronized wild-type control strain (N2) of C. elegans was cultured at 20°C from egg hatching, using B. subtilis 168 or B. subtilis DG101 as a food source. Then, worms of 10 days of age, were starved of food (168 or DG101) for 1 h before to compare their chemotactic response toward the attractants. As shown in Figure 7B, B. subtilis 168-colonized N2 worms exhibited a normal chemotactic response toward DA (CI = 0.69 ± 0.02; n = 75), IAA (CI = 0.75 ± 0.02; n = 75), 2,3-pentadione (CI = 0.65 ± 0.01; n = 75), and 2-butanone (CI = 0.41 ± 0.02; n = 75) (Figure 7B, attractants, white bars). Importantly, the colonization of C. elegans gut by B. subtilis DG101 significantly improved the behavioral responses (chemotaxis index) of this host model organism to the chemical attractants [DA (CI = 092 ± 0.01; n = 75), IAA (CI = 0.88 ± 0.02; n = 75), 2,3-pentadione (CI = 0.90 ± 0.02; n = 75), and 2-butanone (CI = 071 ± 0.03; n = 75)] (Figure 7B, attractants, black bars).

Accordingly, N2 worms fed on B. subtilis DG101 significantly improved the response repellent stimuli [2-nonanone (CI = −0.71 ± 0.01; n = 75), and octanol (CI = −0.91 ± 0.02; n = 75) (Figure 7B, repellents, white bars)] when compared with worms colonized by B. subtilis 168 2-nonanone (CI = −0.41 ± 0.02; n = 75) and octanol (CI = −0.66 ± 0.01; n = 75) (Figure 7B, repellents, black bars). Therefore, B. subtilis DG101 possesses the potential attribute to extend the host’s healthy lifespan.