Bacillus amyloliquefaciens was reported to be a plant promoting bacterium that was associated with rhizospheres as they consume nutrients from the plant. Moreover, this group of bacteria could produce secondary metabolites to suppress plant pathogens and also produce compounds to promote plant growth (Idris et al. 2007). The completed genome of 3918 Kb of the FZB42 strain showed a total of 8.5% of genetic material to be associated with the synthesis function of the non-ribosomal secondary metabolites (Chen et al. 2007) and 3693 sequences encoded for proteins. B. amyloliquefaciens FZB42 produces several non-ribosomal peptide products such as lipopeptides; surfactin, fengycin, and bacillomycin D (Koumoutsi et al. 2004) that can inhibit Fusarium oxysporum and polyketides; bacillaene, difficidin and macrolactin that also inhibited F. oxysporum (Chen et al. 2006; Schneider et al. 2007). Moreover, the peptides plantazolicin A, B and cyclic peptides amylocyclicin produced from this bacterium were also reported to inhibit B. subtilis and other closely related bacteria (Kalyon et al. 2011; Scholz et al. 2014). Bacilysin a dipeptide product from B. amyloliquefaciens FZB42 was able to suppress growth of E. amylovora which is the causative agent of fire blight disease (Chen et al. 2009). In this current study, B. amyloliquefaciens N2-4 and N3-8 that were isolated from soil, negative for B. pseudomallei, could inhibit B. pseudomallei and a wide range of human pathogens.

The metabolites with antimicrobial activity secreted from B. amyloliquefaciens N2-4 and N3-8 isolates were produced at the early to mid-stationary phase (12–72 h) when cell density become increased and the activity decreased after 78 h of cultivation. This characteristic fits well to the secondary metabolites that can be induced by multifactor such as stress, starvation or environmental factors and also cell-to-cell communication or quorum sensing which use small peptides as inducer (Kleerebezem and Quadri 2001). Antimicrobial activity of compounds from Bacillus spp. from the Amazon river basin were also capable of producing compounds with antimicrobial activity at the exponential phase and reached their peak at the stationary phase (Motta et al. 2007). The antimicrobial activity of culture supernatant from N2-4 were decreased by heat, abolished when autoclaved and can be completely destroyed by proteolytic enzymes suggesting that the main active compounds against B. pseudomallei are proteins. For N3-8, the heat, autoclaving and proteolytic enzymes digestion could only partially decrease the activity. Therefore, the active compounds should compose of peptides that can be digested by proteolytic enzyme and other compounds that resist to proteolytic enzyme and high temperature conditions. As mentioned, Bacillus spp. were reported to produce both peptides and non-ribosomal peptide antibiotics that can inhibit peptidoglycan synthesis or cause pore formation using some specific molecules such as lipid II or the mannose phosphate transferase system (man–PTS) as a docking molecule (Cotter et al. 2013). Lipopeptides are metabolites that can function as a pore formation or emulsification on the target organism. Several peptides and bacteriocin-like substances from Bacillus spp. were reported to stable at a wide range of temperatures (30–80 °C) (Cao et al. 2011; Hammami et al. 2009; Sutyak et al. 2008). CLI proteins, for example, were stable up to 60 °C and lost activity after being autoclaved at 121 °C (Meng et al. 2012) and a peptide subtilosin was stable after heat at 100 °C (Sutyak et al. 2008). The antimicrobial substances from B. amyloliquefaciens N2-4 and N3-8 were stable up to 25–100 °C and some activity was decreased when they were heated above 80 °C. Moreover, the molecular weights of active fractions of precipitated proteins from both N2-4 and N3-8 were suspected to be less than 11 kDa. This information could indicate their metabolites to contain small peptides similar to bacteriocin antibiotics. Other compounds those were still active after proteolytic enzyme digestion could be the non-peptides group such as lipopeptides or polyketides. These antimicrobial substances were reported to resist a wide range of temperatures up to 100 °C (Pathak and Keharia 2014) and were also resistant to proteolytic enzyme digestion (Zhao et al. 2013).

The active compounds produced from both B. amyloliquefaciens N2-4 and N3-8 isolates were able to inhibit several strains of B. pseudomallei from both clinical and environmental sources including antibiotic resistant isolates. Moreover, they could inhibit other Gram-positive and Gram-negative bacteria such as E. coli, S. aureus, E. faecium and C. difficile as shown by the cross-streak method. Surprisingly, these antimicrobial compounds did not inhibit B. thailandensis the non-pathogenic bacteria that is closely related to B. pseudomallei. The difference in lipopolysaccharide (LPS) structure between the B. pseudomallei and B. thailandensis have been reported (Knirel et al. 1992; Perry et al. 1995) together with the difference in their genomes (Brett et al. 1998) may contribute to the difference in their susceptibility to the antimicrobial compounds. Most of B. pseudomallei in Thailand have the LPS genotype A (Tuanyok et al. 2012) and K96243 and 1026b strains with genotype A were susceptible to N2-4 and N3-8 metabolites, however, the SRM117 LPS O-side chain mutant was susceptible to N3-8 but not N2-4. Moreover, M6 and M10, which are biofilm mutants, were more susceptible to N3-8 than N2-4. The non-peptides compounds that are present in N3-8 but much less or none in the N2-4 as observed by heat and proteolytic stability test may be responsible for the differences when the culture supernatants were used to test for the spectrum of inhibition. Therefore, the precipitated proteins from N2-4 and N3-8 were prepared and tested with some pathogenic bacteria by agar well diffusion. Besides S. pyogenes and M. catarrhalis that showed equal inhibition, most of the test organisms were better inhibited by N3-8 than N2-4. Moreover, V. parahaemolyticus showed resistance to N2-4 but not N3-8. In general, most peptide antibiotics can inhibit bacteria in a narrow range or only closely related bacteria. Nevertheless, a bacteriocin-like substance of Bacillus spp. was reported to inhibit a broad-range of bacteria (Guo et al. 2012; Lee et al. 2001; Motta et al. 2007; Xie et al. 2009) that included B. amyoliquefaciens LBM5006 that inhibited L. monocytogenes, B. cereus, Serratia marcescens, E. coli, P. aeruginosa, P. fluorescens, S. cholerasuis, and S. gallinarum (Benitez et al. 2011). Similarly in this present study, the precipitated proteins from B. amyloliquefaciens N2-4 and N3-8 showed a broad range of inhibition. The metabolites of these two isolates should contain different compounds and the non-peptide metabolites in N3-8 may have a synergistic effect against these pathogens.

Several strains of B. amyloliquefaciens were studied to be used as bio-control agents such as the FZB42 strain that was dual-cultured with E. amylovora, both in vitro and in vivo, and showed inhibition activity in the growth of the pathogens (Chen et al. 2009). When B. amyloliquefaciens N2-4 and N3-8 were co-cultured with B. pseudomallei, they could decrease the growth of B. pseudomallei by 5 log₁₀ in 72 h. The time of B. pseudomallei decrease was correlated with the time when the secondary metabolites from N2-4 and N3-8 were produced. Even though metabolites from N2-4 and N3-8 showed a broad spectrum of inhibition against both Gram-positive and Gram-negative pathogens, they did not affect B. thailandensis, a non-pathogenic bacterium from soil. Purification and characterization of both peptides and non-peptides from both isolates and the tests for their spectrum of inhibition may lead to a better knowledge to design a way for controlling B. pseudomallei in soil, and may also be extended to discover some important compounds to attack several problematic pathogens in the near future.

In conclusion, B. amyloliquefaciens N2-4 and N3-8 isolates obtained from soil can produce both peptides and non-peptide metabolites that can inhibit B. pseudomallei and a broad range of other pathogenic bacteria. After purification and characterization, the bacteria themselves or their metabolites could be used as bio-controls to reduce the pathogenic bacteria in soil of endemic B. pseudomallei areas. Moreover, if the compounds are novel and safe, they may be good candidates for the development of new drugs.

Bacillus amyloliquefaciens was reported to be a plant promoting bacterium that was associated with rhizospheres as they consume nutrients from the plant. Moreover, this group of bacteria could produce secondary metabolites to suppress plant pathogens and also produce compounds to promote plant growth (Idris et al. 2007). The completed genome of 3918 Kb of the FZB42 strain showed a total of 8.5% of genetic material to be associated with the synthesis function of the non-ribosomal secondary metabolites (Chen et al. 2007) and 3693 sequences encoded for proteins. B. amyloliquefaciens FZB42 produces several non-ribosomal peptide products such as lipopeptides; surfactin, fengycin, and bacillomycin D (Koumoutsi et al. 2004) that can inhibit Fusarium oxysporum and polyketides; bacillaene, difficidin and macrolactin that also inhibited F. oxysporum (Chen et al. 2006; Schneider et al. 2007). Moreover, the peptides plantazolicin A, B and cyclic peptides amylocyclicin produced from this bacterium were also reported to inhibit B. subtilis and other closely related bacteria (Kalyon et al. 2011; Scholz et al. 2014). Bacilysin a dipeptide product from B. amyloliquefaciens FZB42 was able to suppress growth of E. amylovora which is the causative agent of fire blight disease (Chen et al. 2009). In this current study, B. amyloliquefaciens N2-4 and N3-8 that were isolated from soil, negative for B. pseudomallei, could inhibit B. pseudomallei and a wide range of human pathogens.

The metabolites with antimicrobial activity secreted from B. amyloliquefaciens N2-4 and N3-8 isolates were produced at the early to mid-stationary phase (12–72 h) when cell density become increased and the activity decreased after 78 h of cultivation. This characteristic fits well to the secondary metabolites that can be induced by multifactor such as stress, starvation or environmental factors and also cell-to-cell communication or quorum sensing which use small peptides as inducer (Kleerebezem and Quadri 2001). Antimicrobial activity of compounds from Bacillus spp. from the Amazon river basin were also capable of producing compounds with antimicrobial activity at the exponential phase and reached their peak at the stationary phase (Motta et al. 2007). The antimicrobial activity of culture supernatant from N2-4 were decreased by heat, abolished when autoclaved and can be completely destroyed by proteolytic enzymes suggesting that the main active compounds against B. pseudomallei are proteins. For N3-8, the heat, autoclaving and proteolytic enzymes digestion could only partially decrease the activity. Therefore, the active compounds should compose of peptides that can be digested by proteolytic enzyme and other compounds that resist to proteolytic enzyme and high temperature conditions. As mentioned, Bacillus spp. were reported to produce both peptides and non-ribosomal peptide antibiotics that can inhibit peptidoglycan synthesis or cause pore formation using some specific molecules such as lipid II or the mannose phosphate transferase system (man–PTS) as a docking molecule (Cotter et al. 2013). Lipopeptides are metabolites that can function as a pore formation or emulsification on the target organism. Several peptides and bacteriocin-like substances from Bacillus spp. were reported to stable at a wide range of temperatures (30–80 °C) (Cao et al. 2011; Hammami et al. 2009; Sutyak et al. 2008). CLI proteins, for example, were stable up to 60 °C and lost activity after being autoclaved at 121 °C (Meng et al. 2012) and a peptide subtilosin was stable after heat at 100 °C (Sutyak et al. 2008). The antimicrobial substances from B. amyloliquefaciens N2-4 and N3-8 were stable up to 25–100 °C and some activity was decreased when they were heated above 80 °C. Moreover, the molecular weights of active fractions of precipitated proteins from both N2-4 and N3-8 were suspected to be less than 11 kDa. This information could indicate their metabolites to contain small peptides similar to bacteriocin antibiotics. Other compounds those were still active after proteolytic enzyme digestion could be the non-peptides group such as lipopeptides or polyketides. These antimicrobial substances were reported to resist a wide range of temperatures up to 100 °C (Pathak and Keharia 2014) and were also resistant to proteolytic enzyme digestion (Zhao et al. 2013).

The active compounds produced from both B. amyloliquefaciens N2-4 and N3-8 isolates were able to inhibit several strains of B. pseudomallei from both clinical and environmental sources including antibiotic resistant isolates. Moreover, they could inhibit other Gram-positive and Gram-negative bacteria such as E. coli, S. aureus, E. faecium and C. difficile as shown by the cross-streak method. Surprisingly, these antimicrobial compounds did not inhibit B. thailandensis the non-pathogenic bacteria that is closely related to B. pseudomallei. The difference in lipopolysaccharide (LPS) structure between the B. pseudomallei and B. thailandensis have been reported (Knirel et al. 1992; Perry et al. 1995) together with the difference in their genomes (Brett et al. 1998) may contribute to the difference in their susceptibility to the antimicrobial compounds. Most of B. pseudomallei in Thailand have the LPS genotype A (Tuanyok et al. 2012) and K96243 and 1026b strains with genotype A were susceptible to N2-4 and N3-8 metabolites, however, the SRM117 LPS O-side chain mutant was susceptible to N3-8 but not N2-4. Moreover, M6 and M10, which are biofilm mutants, were more susceptible to N3-8 than N2-4. The non-peptides compounds that are present in N3-8 but much less or none in the N2-4 as observed by heat and proteolytic stability test may be responsible for the differences when the culture supernatants were used to test for the spectrum of inhibition. Therefore, the precipitated proteins from N2-4 and N3-8 were prepared and tested with some pathogenic bacteria by agar well diffusion. Besides S. pyogenes and M. catarrhalis that showed equal inhibition, most of the test organisms were better inhibited by N3-8 than N2-4. Moreover, V. parahaemolyticus showed resistance to N2-4 but not N3-8. In general, most peptide antibiotics can inhibit bacteria in a narrow range or only closely related bacteria. Nevertheless, a bacteriocin-like substance of Bacillus spp. was reported to inhibit a broad-range of bacteria (Guo et al. 2012; Lee et al. 2001; Motta et al. 2007; Xie et al. 2009) that included B. amyoliquefaciens LBM5006 that inhibited L. monocytogenes, B. cereus, Serratia marcescens, E. coli, P. aeruginosa, P. fluorescens, S. cholerasuis, and S. gallinarum (Benitez et al. 2011). Similarly in this present study, the precipitated proteins from B. amyloliquefaciens N2-4 and N3-8 showed a broad range of inhibition. The metabolites of these two isolates should contain different compounds and the non-peptide metabolites in N3-8 may have a synergistic effect against these pathogens.

Several strains of B. amyloliquefaciens were studied to be used as bio-control agents such as the FZB42 strain that was dual-cultured with E. amylovora, both in vitro and in vivo, and showed inhibition activity in the growth of the pathogens (Chen et al. 2009). When B. amyloliquefaciens N2-4 and N3-8 were co-cultured with B. pseudomallei, they could decrease the growth of B. pseudomallei by 5 log₁₀ in 72 h. The time of B. pseudomallei decrease was correlated with the time when the secondary metabolites from N2-4 and N3-8 were produced. Even though metabolites from N2-4 and N3-8 showed a broad spectrum of inhibition against both Gram-positive and Gram-negative pathogens, they did not affect B. thailandensis, a non-pathogenic bacterium from soil. Purification and characterization of both peptides and non-peptides from both isolates and the tests for their spectrum of inhibition may lead to a better knowledge to design a way for controlling B. pseudomallei in soil, and may also be extended to discover some important compounds to attack several problematic pathogens in the near future.

In conclusion, B. amyloliquefaciens N2-4 and N3-8 isolates obtained from soil can produce both peptides and non-peptide metabolites that can inhibit B. pseudomallei and a broad range of other pathogenic bacteria. After purification and characterization, the bacteria themselves or their metabolites could be used as bio-controls to reduce the pathogenic bacteria in soil of endemic B. pseudomallei areas. Moreover, if the compounds are novel and safe, they may be good candidates for the development of new drugs.