New crosstalk between probiotics Lactobacillus plantarum and Bacillus subtilis

It was reported that oral administration of Bacillus favored the growth of Lactobacillus in the intestinal tract. Here, this phenomenon was confirmed by co-cultivation of Bacillus subtilis 168 and Lactobacillus plantarum SDMCC050204-pL157 in vitro. To explain the possible molecular mechanisms, B. subtilis 168 cells were incubated in simulated intestinal fluid at 37 °C for 24 h, and up to 90% of cells autolysed in the presence of bile salts. Addition of the autolysate to medium inoculated with Lb. plantarum SDMCC050204 decreased the concentration of H2O2 in the culture, alleviated DNA damage and increased the survival of Lb. plantarum, as like the results of exogenous heme addition. These results suggested that the autolysate provided heme, which activated the heme-dependent catalase KatA in Lb. plantarum SDMCC050204. HPLC confirmed the presence of heme in the autolysate. Disruption of the Lb. plantarum SDMCC050204 katA gene abolished the protective effect of the B. subtilis 168 autolysate against H2O2 stress. We thus hypothesized that the beneficial effect of Bacillus toward Lactobacillus was established through activation of the heme-dependent catalase and remission of the damage of reactive oxygen species against Lactobacillus. This study raised new crosstalk between the two frequently-used probiotics, highlighting heme-dependent catalase as the key mediator.

In this study, we performed co-cultivation of B. subtilis 168 and Lb. plantarum SDMCC050204-pL157, analyzed the presence of heme in autolysate of B. subtilis 168, and confirmed the role of the autolysate in activation of catalase (KatA) in Lb.plantarum SDMCC050204. The aim of this study was to propose a new crosstalk model between Bacillus and Lactobacillus strains in the intestinal tract after oral administration of Bacillus.

Materials and Methods
Bacterial strains, culture media, and growth conditions. Bacterial strains and plasmids used in this study are listed in Table 1. B. subtilis 168 and Escherichia coli DH5α were grown aerobically in Luria-Bertani (LB) medium at 37 °C. Lb. plantarum strains were grown in de Man, Rogosa and Sharpe (MRS) broth containing 0.5% glucose at 37 °C in two different conditions, either (i) static cultivation in 100-mL Erlenmeyer flasks containing 50 mL medium, or (ii) aerated cultivation in 300-mL Erlenmeyer flasks containing 50 mL medium with agitation on a rotary shaker at 200 rpm. When appropriate, the following antibiotics were added to the medium: ampicillin (100 μg/mL for E. coli), chloramphenicol (5 μg/mL for Lb. plantarum) and erythromycin (5 μg/mL for Lb. plantarum). When required, 20 μM heme (Sigma, USA) was added to medium. Cell turbidity was monitored by the optical density at 600 nm (OD 600 ).
Co-cultivation of Lb. plantarum SDMCC050204-pL157 with B. subtilis 168. To easily distinguish and enumerate Lb. plantarum easily in co-culture with B. subtilis 168 (which is chloramphenicol sensitive), Lb. plantarum SDMCC050204 was conferred with a chloramphenicol resistance phenotype by transformation of plasmid pL157 by electroporation 21,22 , generating Lb. plantarum SDMCC050204-pL157. Since previous report showed the physical stability of pL157 in host cells without selective pressure 21 , chloramphenicol was not added into the co-culture of Lb. plantarum SDMCC050204-pL157 and B. subtilis 168.
Lb. plantarum SDMCC050204-pL157 and B. subtilis 168 were individually cultivated to reach early stationary phase. The cells were collected by centrifugation at 5,000 × g for 5 min, washed three times with sterile saline solution and resuspended in MRS broth. Then, 5.6 × 10 8 CFU/mL of Lb. plantarum SDMCC050204-pL157 and 6.0 × 10 7 CFU/mL of B. subtilis 168 were inoculated into MRS broth, and cultivated aerobically for 96 h. Viable cell counts of Lb. plantarum SDMCC050204-pL157 were determined on MRS agar supplemented with chloramphenicol.
Autolysis of B. subtilis in simulated intestinal fluid. B. subtilis 168 spores (2.0 × 10 7 CFU/mL) were incubated in LB medium for 9 h to reach late logarithmic phase to early stationary phase. The vegetative cells were collected, washed and resuspended in simulated intestinal fluid (SIF) at an OD 600 of 17.00 ± 0.46. The bacterial suspensions were incubated at 37 °C with agitation (200 rpm) to simulate peristalsis, and aliquots were taken to determine the OD 600 at 0, 12 and 24 h. The SIF contained 0.5% NaCl, 1 g/L pancreatin (Sigma, USA), and 0%, 0.05%, 0.1% or 0.3% pig bile salts (Sigma, USA), pH 8.0 23 . cultivation of Lb. plantarum with the autolysate of B. subtilis 168. To prepare autolysate of B. subtilis 168, vegetative cells were resuspended in 0.1% NaCl to one-tenth of the initial culture volume and stored at 4 °C for 7 days. Then, the intact cells and cell debris were removed by centrifugation at 13,000 × g for 20 min. The resultant supernatant was filtered with a 0.22 μm membrane and then heated at 100 °C for 15 min to inactivate any proteinic enzymes to obtain the autolysate. Culture medium was prepared by mixing an equal volume of two-fold concentrated MRS broth with the autolysate, or with 0.1% NaCl as a control. Then, 5.6 × 10 8 CFU/mL Lb. plantarum SDMCC050204 were inoculated into the culture medium and incubated aerobically for 60 h.
Analysis of catalase activity and H 2 o 2 concentration. Cell pellets from 2 mL of Lb. plantarum SDMCC050204 culture were collected, washed three times with sterile saline solution, and resuspended in 50 μL saline solution. The cell suspensions were mixed with 20 μL of 30% H 2 O 2 solution, and the air bubble formation was determined 24 25 . Specifically, 100 mL of autolysate was concentrated to 10 mL by lyophilization (Thermo Savant, USA). Then, 30 mL of 90% aqueous acetone containing 5% HCl (v/v) were added to the autolysate. The mixture was vortexed for 10 min at room temperature, and then centrifuged at 13,000 × g for 10 min. The heme-containing supernatant was recovered, while the pellets were extracted with another 10 mL of acidic acetone. After centrifugation, the two fractions of supernatant were combined followed by evaporation in a vacuum rotary evaporator (Thermo Savant, USA) to remove the organic phase and a vacuum freeze-dryer to remove the water phase. The dry residue was dissolved in 3 mL of distilled water and the pH was adjusted to 12.0 for the transformation of heme to soluble hematin. The solution was filtered with a 0.22 μm membrane for the further detection.
Hematin was detected by high-performance liquid chromatography (HPLC; Shimadzu, Japan) using an XBridge BEH300 C18 reverse phase column (150 × 4.6 mm; Waters, USA) with a flow rate of 0.6 mL/min. The column was equilibrated with solvent A. Separation of the hematin was effected with a gradient of 20% to 70% solvent B over 40 min. The column effluent was monitored by photo-diode array detection at 398 nm. Solvent A was 0.1% trifluoroacetic acid (TFA) (v/v) in water; solvent B was 0.1% TFA (v/v) in acetonitrile. Hematin sample as prepared above (50 μL) was injected into the column for analysis. 10 μL of 30 μg/mL hematin (Sigma, USA) was used as the standard sample. Molecular cloning techniques were performed essentially as described previously 26 . Taq polymerase, restriction enzymes and T4 DNA ligase were used according to the manufacturer's instructions (TaKaRa, Japan).
Disruption of the katA gene was carried out using a single-crossover integration strategy 27 . A DNA fragment containing the katAk was PCR amplified from Lb. plantarum SDMCC050204 genomic DNA with primers laX-1 (5′-ACCTCTAGAGACGTGCGCGGTTTT-3′; the underlined bases indicate an XbaI site) and laX-2 (5′-CAGAAGCTTAGGCTCGTAGTTGACC-3′; the underlined bases indicate a HindIII site). The PCR products were digested with XbaI/HindIII and then ligated into the compatible end of pUC-erm, yielding pUC-erm-kat. pUC-erm-kat (200 ng) was introduced into competent Lb. plantarum SDMCC050204 cells by electroporation. Transformants were selected on MRS agar containing erythromycin at a final concentration of 5 μg/mL, generating the mutant Lb. plantarum SDMCC050204ΔkatA. Statistical analysis. Statistical analysis was performed using data from three technical replicates by unpaired two-tailed Student's t-test. P values of <0.05 were considered statistically significant; P values of <0.01 were considered statistically highly significant.

Results
Detection of survival of Lb. plantarum SDMCC050204-pL157 co-cultured with B. subtilis 168 in vitro. It was reported that oral administration of Bacillus favored the growth of Lactobacillus in the intestinal tract 29,30 . To confirm this result in vitro, Lb. plantarum SDMCC050204-pL157 was incubated with or without B. subtilis 168. The H 2 O 2 concentrations and cell viability of Lb. plantarum SDMCC050204-pL157 were detected. As shown in Fig. 1a,b, when Lb. plantarum SDMCC050204-pL157 was cultivated alone under aerobic conditions for 96 h, about 2 mM H 2 O 2 was detected in the cultures, and the viable cell numbers showed a significant decrease, from 5.62 ± 0.2 × 10 7 CFU/mL at 0 h to 0 CFU/mL at 96 h. In contrast, in the static culture, very little H 2 O 2 was detected, and the viable cell counts were 1.0 ± 0.4 × 10 9 CFU/mL at 96 h, suggesting that Lb. plantarum SDMCC050204-pL157 suffered from H 2 O 2 stress under aerobic conditions, causing cellular damage.
To test whether B. subtilis could decrease the H 2 O 2 level in aerobic culture of Lb. plantarum, Lb. plantarum SDMCC050204-pL157 as well as two other Lb. plantarum strains SDMCC050276-pL157 and SDMCC050277-pL157 were respectively co-cultured with B. subtilis 168. The results showed that the H 2 O 2 was not detectable in the co-cultures (Fig. 1a), as that of the static culture, and the viable cell counts of SDMCC050204-pL157 was 8.4 ± 0.1 × 10 8 CFU/mL after 96 h incubation, close to that in the static culture (Fig. 1b). Similar phenomena were observed for co-cultivation of B. subtilis 168 with strains SDMCC050276-pL157 and SDMCC050277-pL157 (data not shown). These results confirmed that B. subtilis 168 could provide bioactive molecules to protect Lb. plantarum strains from H 2 O 2 stress.

Autolysis of B. subtilis 168 in SIF.
B. subtilis is prone to autolysis due to environmental stressors or the regulated processes that occur at different stages of the cell life 31,32 . To find out the bioactive molecules, B. subtilis 168 cells were suspended in SIF, and the OD 600 was determined at 12 and 24 h, respectively. As shown in Fig. 2 (Fig. 3), suggesting that the heated autolysate contributed biomolecules that activated catalase activity in Lb. plantarum SDMCC050204. The same phenomenon was obtained when addition of exogenous heme to the medium instead of the heated autolysate. However, nor did these results when Lb. plantarum SDMCC050204 was cultivated in the medium without the heated autolysate or heme (Fig. 3). Thus, we concluded that the autolysate of B. subtilis 168 supplied heme, which in turn resulted in activity of the heme-dependent catalase KatA in Lb. plantarum SDMCC050204. The H 2 O 2 concentrations in cultures of mutant SDMCC050204ΔkatA and the wild type SDMCC050204 were compared after supplementation with the autolysate or heme. As stated above, about 2 mM H 2 O 2 was detected in the culture of strain SDMCC050204 in MRS medium, the same was observed in culture of the mutant SDMCC050204ΔkatA. When the autolysate was added, no H 2 O 2 was detected in culture of strain SDMCC050204, while as much as 3.6 mM H 2 O 2 was present in culture of the mutant SDMCC050204ΔkatA (Fig. 5b). Heme addition resulted in similar comparison of H 2 O 2 levels in the cultures of the wild type and the mutant. Moreover, the mutant strain exhibited an earlier and sharper decline in viable cell numbers (Fig. 5c). These results indicated that SDMCC050204ΔkatA lost the ability to produce active catalase even in the presence of the B. subtilis autolysate or heme, highlighting the heme-dependent catalase as the key component mediating interaction between the Bacillus sp. and the Lactobacillus. No significant differences were observed between the addition of the autolysate and exogenous heme in respect to H 2 O 2 concentration or cell viability (Fig. 5b,c). highly-reactive molecules that cause cellular damage 17,28 . To examine DNA integrity, chromosomal DNA was extracted from cells of Lb. plantarum SDMCC050204 and SDMCC050204ΔkatA cultivated aerobically for 12, 36 and 60 h (Fig. 6). DNA damage was obviously alleviated in strain SDMCC050204 by the addition of heme or the B. subtilis 168 autolysate, but not in strain SDMCC050204ΔkatA. These results demonstrated that the activation of the heme-dependent catalase KatA efficiently protected Lb. plantarum strains against oxidative stress, and consequently helped maintain the DNA integrity.

Discussion
It is generally considered that Bacillus strains are beneficial for the survival and growth of Lactobacillus in animal intestinal tracts on the basis of the "Biological Oxygen-Capturing Theory [9][10][11][12]33 ". According to this theory, the anaerobic environment generated by the growth of Bacillus strains plays major role in promotion of the growth of Lactobacillus 14 . Others have stated that catalase, subtilisin and surface proteins produced by Bacillus helped Lactobacillus inhabit the same niche 15,16 . In this study, we found that B. subtilis was prone to autolysis, particularly in the presence of bile salts, which offered the possibility to release heme that could be beneficial to other microbiota, including the core Lactobacillus in the intestinal tract. Our work focused on exploring the key factors linking the B. subtilis autolysate and Lb. plantarum, and the critical roles of heme and KatA, a heme-dependent catalase, were consequently demonstrated. This work thus proposed a novel crosstalk model between Bacillus and Lactobacillus in the intestinal tract, which would shed new light on the complicated interactions of different bacterial species in the gut microbiota.
ROS, including − O 2 , H 2 O 2 and ⋅ HO , are generated as by-products of normal human cellular metabolic activities 34 . Alcohol, chronic infections and inflammatory disorders stimulate the production of ROS, and thus the intestinal tract is a key source of ROS 34,35 . Excessive accumulation of ROS results in oxidative stress, leading to intracellular biological macromolecular damage 34 . Meanwhile, the metabolism of Lb. plantarum is impacted by oxidative stress. H 2 O 2 could induce the activity of pyruvate oxidase (POX), which converts pyruvate into acetate,  www.nature.com/scientificreports www.nature.com/scientificreports/ accompanied by the production of extra H 2 O 2 36 . Here, to imitate the ROS pool in the animal intestinal environment, aerobic cultivation was carried out to subject Lb. plantarum cells to oxidative stress, as 2 mM H 2 O 2 was detected in the aerobic culture of Lb. plantarum SDMCC050204-pL157. We also found that Lb. plantarum was sensitive to oxidative damage from H 2 O 2 , agreeing with previous reports 37 . H 2 O 2 damage to the Lb. plantarum cells mainly resulted from the lack of effective antioxidant systems. Although ROS erasers, including catalase, superoxide dismutase and NADH peroxidase, can be produced in some Lactobacillus strains by genetic and physiological analysis, the enzymatic activities are low 38,39 . In particular, catalase is commonly inactive in Lactobacillus, because the main cofactor heme is absent 18,[40][41][42] . Therefore, mechanisms to cope with oxidative stress from H 2 O 2 in the intestinal tract are of great importance for Lactobacillus. Our results here demonstrated that co-cultivation with B. subtilis significantly decreased the level of H 2 O 2 and enhanced the survival of Lb. plantarum cells (Fig. 1).
Bacillus are complex organisms that exist as vegetative cells or metabolically inert spores or as part of a multicellular biofilm when encountering extreme environments 43 . When the environment is deficient in nutrients,   Chromosomal DNA was prepared from cells taken at 12, 36 and 60 h, and separated by electrophoresis on a 1% agarose gel to observe the extent of DNA degradation. Distance of migration reflects the degree of degradation, as indicated.