Introduction

For many decades, the genus Bacillus has been used in different studies in the fields of genetics and biochemistry. During this time, several strains were isolated, with the potential to produce more than two dozens of antibiotics with an impressive array of structures. Therefore, an average of about 4–5% of a Bacillus subtilis genome was employed in antibiotic production. Part of these antibiotics are composed of lipopeptides—LPs (Stein 2005). These compounds have been seen as biological control agents as an alternative to chemical pesticides, which generate strong environmental impacts by selecting resistant pests and contaminating the environment (Torres et al. 2017).

Lipopeptides are biosurfactants that are synthesized non-ribosomally by large multi-enzyme complexes, the non-ribosomal peptide synthetases (NRPS) (Chen et al. 2009). These synthesis mechanisms lead to a great diversity among LPs with regard to the type and sequence of amino acid residues, the nature of the peptide cyclization, and the nature, length, and branching of the fatty acid chain (Ben Abdallah et al. 2015). There are three major families of Bacillus LPs, namely iturin, surfactin, and fengycin (Dimkić et al. 2017). The amphiphilic nature of these molecules provides a capacity of interfering with biological membrane structures, making LPs key factors for the biological control of microorganisms. Lipopeptides are also responsible for systemic resistance induction in plant and biofilm formation (Fig. 1) (Ongena and Jacques 2008).

Fig. 1
figure 1

Effects of LP production by Bacillus species. These molecules can act in two ways, namely by directly acting against different phytopathogens and by inducing the systemic resistance of the plant. For the success of the action of these two abilities, biofilm formation is essential, since it overcomes difficulties in the colonization of the microorganisms

Full size image

Iturins are neutral or monoanionic lipopeptides, and have the chiral sequence LDDLLDL with a restricted number of residues (Asx, Glx, Pro, Ser, Thr, Tyr). Also, they share a common sequence (β-hydroxy fatty acid-Asx-Tyr-Asx) and show a variation at the other four positions (Fig. 2a) (Bonmatin et al. 2003). This family is mainly composed of the following compounds: iturin A and C (Besson et al. 1978, 1986), bacillomycin D (Peypoux et al. 1984), bacillomycin F (Mhammedi et al. 1982), bacillomycin L (Peypoux et al. 1984), bacillopeptin (Kajimura et al. 1995), and mycosubtilin (Peypoux et al. 1986). In terms of activity, iturin has a powerful antibiotic activity against a wide fungal spectrum (Arrebola et al. 2010).

Fig. 2
figure 2

adapted from Ongena and Jacques 2008)

Structures of a iturin variants (n = 9–12), b surfactin variants (n = 9–11; S1 = Val, Leu, or Ile; S2 = Ala, Val, Leu, or Ile; S3 = Val, Leu, or Ile; P1 = Val or Ile), and c fengycin variant (n = 11–14) (

Full size image

The structure of surfactin was elucidated 49 years ago by Kakinuma et al. (1969); it was found to be constituted by a heptapeptide with a chiral central sequence LLDLLDL, interlinked by a β-hydroxy fatty acid to form a cyclic lactone ring structure (Fig. 2b). The main members of the surfactin group are surfactin (Arima et al. 1968), esperin (Thomas and Ito 1969), halobacillin (Trischman et al. 1994), and pumilacidin (Naruse et al. 1990). Surfactin displays antiviral, antimycoplasma, and antibacterial activities (Ongena and Jacques 2008).

The first fengycin was identified by Vanittanakom et al. (1986) and is composed of lipodecapeptides with an internal lactone ring in the peptidic moiety and with a β-hydroxy fatty acid chain that can be saturated or unsaturated (Fig. 2c) (Ongena and Jacques 2008). This family is mainly composed of fengycin A and B (Vanittanakom et al. 1986) and plipastatin A and B (Umezawa et al. 1986). Fengycin mostly displays antimicrobial activity against a range of yeasts and filamentous fungi (Zihalirwa et al. 2017).

This review highlights the recent studies about the ability of LPs to stimulate defense mechanisms of plants and biofilm formation, which is a key factor for the successful biocontrol against organisms that are harmful for plant cultures and a serious economic problem in agriculture worldwide.

Risk of pesticide resistance around the world and significance of LPs on the pesticide market

The current market for agricultural biologicals is around 2.9 billion dollars. Although it represents a great amount of money, it is still dwarfed by the total agrochemical market, which is around 240 billion dollars. Regarding pesticides, the biological product market is about 2 billion dollars annually, while the chemical product market amounts to 44 billion dollars. Even though still in its starting point, microbiome-based products will have a market size comparable to that of agrochemicals in the next few years. In Europe, it is expected that by 2020, there will be more biopesticides than chemical ones (Singh 2017). The biopesticide markets of Europe and South America are the ones projected to grow most rapidly in the next years, driven by tightening regulatory restrictions, and rapidly emerging disease resistance, respectively (Olson 2015).

A great example of losses due to disease resistance is Brazil, one of the largest producers of soybean in the world, contributing around 30% of the total production worldwide. The country faces significant problems with the Asian soybean rust (Phakopsora phachyrhizi). The chemical control of this fungus began in Brazil in 2002/03, and the total costs to control the disease are estimated to be around 2 billion dollars per year (Kawashima et al. 2016). Fungicide resistance has been observed in the past few years, and the main reason is the frequent exposure of the pathogen to chemical fungicides (Godoy et al. 2016). The problem of resistance to chemical pesticides is also faced by the rest of the world, with a wide range of diseases which developed pesticide resistance over the decades, such as citrus storage rot, cereal leaf spot, apple scab, powdery mildews, potato blight, grape downy mildew, cereal powdery mildew, and others (Lucas et al. 2015).

As a consequence of the adversities faced by chemical pesticides, large companies are working on multiple acquisitions, licensing agreements, and partnerships on the agro-biological market (Table 1). The combination of both government agencies and private sectors, including a large number of start-up companies, are also evidence that the use of microbiome products for agriculture will exponentially increase in the near future (Singh 2017).

Table 1 Recent deals on the biopesticide market
Full size table

Recent studies about Bacillus LPs and their use in agriculture

Production and recovery

In terms of recent scientific studies, B. subtilis is still one of the most used species for LP production among bacteria of the genus Bacillus (Gong et al. 2014; Cawoy et al. 2015; Farace et al. 2015). Another species commonly studied is Bacillus amyloliquefaciens (Chen et al. 2016b; Soares et al. 2016). Different microorganisms and media that are used for LP production at laboratory scale are shown in Table 2.

Table 2 LP production from different Bacillus species at laboratory scale
Full size table

Agro-based low-cost products or wastes are being sought as alternative media to minimize the production costs. For example, cassava wastewater was tested in a pilot scale process and proved to be a good substrate for the production of LPs (Barros et al, 2008). Palm oil mill effluent (POME) (Abas et al. 2013), soy flour, molasses (Yánez-Mendizábal et al. 2012a), soybean meal, wheat flour (Song et al. 2013), and desizing wastewater (Li et al. 2011b) have also been used for the production of LPs by Bacillus species.

One of the biggest challenges in LP production is the intense foam production during fermentation due to its surfactant nature. This characteristic affects LP recovery and purification and impedes the continuous production of these compounds (Coutte et al. 2013). The use of rotating disk bioreactors can promote a non-foaming fermentation and can be a good alternative to pilot and industrial processes. The use of this type of bioreactors showed a higher yield of LPs when compared to others bubbleless reactors (Chtioui et al. 2012). Other alternatives are the use of a bubbleless membrane bioreactor (Coutte et al. 2010), a biofilm reactor (Zune et al. 2014), and solid-state fermentation reactors (Ano et al. 2009). High foam production was also experienced by Barros et al. (2008), who produced LPs at pilot scale using a 40 L batch pilot bioreactor adapted for simultaneous foam collection.

Extraction of the LPs from the supernatant can be performed by two classic methods. The first method consists of adding ethyl acetate to the cell-free supernatant in a 1:1.1 ratio, with the addition of NaCl (30 g/L). The suspension must be homogenized, followed by the collection of the ethyl acetate fraction, which will be dried in a rotary evaporator (Dimkić et al. 2017). Another approach for the extraction of LPs is acid precipitation, which consists of adjusting the cell-free supernatant to pH 2 using HCl, resulting in a precipitate. Resuspension in methanol will provide an LP extract (Asari et al. 2017; Dimkić et al. 2017). Dimkić et al. (2017) demonstrated that acid precipitation, followed by a methanol extraction, led to a reduction of biosurfactant activities to only 23%, which is in accordance with earlier studies. The authors recommend ethyl acetate, in which the hydrophobic residues of LP compounds are probably better dissolved than in methanol.

After extraction, a finer purification can be executed for further LP characterization. Thin-layer chromatography (TLC) and high-performance thin-layer chromatography (HPTLC) are alternatives of LP purification. Liquid chromatography (LC) and high-pressure liquid chromatography (HPLC) are also extensively employed, especially using a reverse phase. In this case, C18 columns are highly used. For complete identification, mass spectrometry is recommended (Table 3). Figure 3 presents a summary of the main ways to produce, recover, and identify LPs.

Table 3 Methods of purification and identification of LPs
Full size table
Fig. 3
figure 3

Different alternatives for the production, recovery, and identification of LPs. The production of these molecules can be performed at smaller scales, such as flasks, as well in different bioreactors (STR, rotating discs, and membrane bioreactor). The recovery is made by two classic methods: (1) acid precipitation, followed by resuspension in precipitate in methanol; (2) addition of ethyl acetate and drying in a rotary evaporator after extraction of the ethyl acetate fraction. Identification can be performed by different analytical methods of chromatography and mass spectrometry

Full size image

LPs and their producers’ application in agriculture

Lipopeptides can be used as antimicrobials against a wide range of organisms including bacteria, fungi, oomycetes, and viruses (Raaijmakers et al. 2010). When it comes to important agricultural crops, such as soybean, wheat, maize, and potatoes, there are several studies about the use of LP producers in the biocontrol of pathogens as well as about the direct application of LPs. As mentioned above, P. pachyrhizi is one of the greatest challenges in soybean production. The use of B. subtilis QST-713 Serenade from Bayer (producer of lipopeptides), coupled with Bacillus pumilus, showed a positive effect on soybean exposed to the Asian rust. In particular, B. subtilis was able to reduce disease severity by 98.6% in tests with detached leaves and by 23% under field conditions (Dorighello et al. 2015). In wheat, Bacillus LP efficacy was observed against Zymoseptoria tritici (Mejri et al. 2017), Gaeumannomyces graminis var. tritici (Zhang et al. 2017; Yang et al. 2018), and Fusarium graminearum (Gong et al. 2015). The latter was also inhibited by Bacillus LPs in maize (Chan et al. 2009). In potatoes, Bacillus biosurfactants were effective against Fusarium solani (Mnif et al. 2015). Different antimicrobial actions of LPs towards different organisms are presented in Table 4.

Table 4 Antimicrobial action of LPs against different pathogens
Full size table

Induced systemic resistance in plants

Bacillus species promote an enhanced defensive capacity to the plant against a wide spectrum of fungi, bacteria, and viruses; this phenomenon is known as induced systemic resistance (ISR). The defense mechanism may be activated through a similar way as the response against pathogenic microorganisms with incompatible interactions (García-Gutiérrez et al. 2013). Induced systemic resistance is often represented by jasmonic acid/ethylene (JA/ET)-dependent signaling pathways. According to Rahman et al. (2015), although ISR is typically independent of salicylic acid (SA), some rhizobacteria may trigger the SA-dependent signaling pathway. Cyclic LPs are key contributors to ISR-eliciting activity (Rahman et al. 2015). The mechanisms triggered by ISR are related to biochemical changes, including reinforcements of plant cell walls, production of antimicrobial phytoalexins, and synthesis of pathogenesis-related (PR) proteins, such as chitinases, β-1,3-glucanases, or peroxidases (García-Gutiérrez et al. 2013). An ISR scheme can be seen in Fig. 4.

Fig. 4
figure 4

Systemic resistance induced in plants by LPs. These molecules work as an elicitor, triggering the plant’s systemic resistance, by activating a JA/ET response or an SA-dependent pathway. As a result, the plant produces a defensive compound

Full size image

The LPs from B. amyloliquefaciens mediated plant defense gene expression against R. solani in lettuce. In this case, with the presence of the bacteria in the plant, there was a higher expression of the gene PDF 1.2, which encodes for defensin (host defense peptide). The same response was not observed using surfactin-deficient mutants, showing the important role of surfactin in the ISR (Chowdhury et al. 2015). For grapevine, gene expression analysis suggests that mycosubtilin (iturin family) activated the SA and JA signaling pathways, whereas surfactin mainly induced an SA-regulated pathway. Both LPs were responsible for a local long-lasting enhanced tolerance to the pathogen B. cinerea in grapevine leaves (Farace et al. 2015). Surfactin and iturin also played a significant role in the plant defense response of strawberry against C. gloeosporioides. The LPs played a major role in the expression of chitinase and β-1,3-glucanase in strawberry leaves (Yamamoto et al. 2015). Other recent studies reported Bacillus LPs as inducers of defense responses of several plants including rice (Chandler et al. 2015), Arabidopsis (Kawagoe et al. 2015), tomato (Abdallah et al. 2017), and maize (Gond et al. 2015).

Biofilm formation induced by LPs

Colonization of biocontrol microorganisms and their maintenance in the plant area are important factors and major challenges. Changes in environmental conditions, such as temperature and relative humidity, are decisive for the colonization of these organisms. Bacillus species evolved a mechanism to overcome these challenges by developing a multicellular behavior known as biofilm formation (Zeriouh et al. 2014). The biofilm structure is initially formed by a process called swarming, which is a rapid and massive migration of cooperating groups of bacteria. Basically, a group of cells forms ‘buds’ at the edge of the original colony, which are then abruptly released forming initial monolayer dendrites. Surfactin is considerably involved in the swarming process, since in a biofilm structure, it was mainly located in the mother colony and along the edges of the dendrites (Debois et al. 2008). In addition, bacillomycin D, a member of the iturin family, also played a role in the expression of the genes involved in biofilm formation of B. amyloliquefaciens (Xu et al. 2013). The biofilm formation of B. subtilis is less robust in strains with null mutation in the gene srfAA, which is responsible for encoding part of the NRPS, consequently forming the surfactin molecule. Biofilm formation in the root helped to increase the local concentration of LP in root-surrounding areas with further stimulation of biofilm formation and antimicrobial action (Chen et al. 2013). The importance of biofilm formation to the success of the biocontrol activity and the role of the surfactin in this process have also been confirmed by other authors such as Aleti et al. (2016), Bais et al. (2004), and Luo et al. (2015).

NRPS engineering

Non-ribosomal peptide synthetases are composed of multi-modules that are responsible to recognize, activate, modify, and link the amino acid intermediates to the product peptide. These multi-enzymes are capable of synthesizing a variety of peptides by adding unusual amino acids, including d-amino acids, β-amino acids, and hydroxy- or N-methylated amino acids. The multi-modular property is another feature that leads these enzymes to produce a great variety of products. Each module is composed of specific domains that catalyze different enzymatic activities (Roongsawang et al. 2010). These modules are responsible for incorporating specific amino acids in the molecules and consist of three major catalytic domains: condensation (C), adenylation (A), and thiolation (T) domains. Finally, the last module frequently also contains the thioesterase (Te) domain (Gao et al. 2018).

Based on the NRPS’s features presented above, studies are being developed on engineering these enzymes to create new LPs. One of the first studies in the field was done by Symmank et al. (2002), who created a lipohexapeptide by genetically engineering the surfactin biosynthesis using a combination of in vitro and in vivo recombination. In their work, a complete amino acid incorporating module was eliminated, creating a modified peptide synthetase. The remaining modules, which are adjacent to the deletion, were recombined at different highly conserved sequence motifs that are characteristic of amino acid-incorporating modules of peptide synthetases. Gao et al. (2018) stated that the entire deletion of an NRPS module, which was a producer of plipastatin, caused the inactivation of the enzyme. However, the authors observed that individual domain deletion (A and T domains) leads to the creation of three novel plipastatin derivatives. Jiang et al. (2016) studied the subunits of the A domain, SrfA-A (responsible for the Glu1-Leu2-Leu3 portion of surfactin) and SrfA-B (responsible for the Val4-Asp5-Leu6 portion of the LP). The authors knocked-out the modules SrfA-A-Leu3, SrfA-B-Asp, and SrfA-B-Leu from surfactin NRPS in B. subtilis, and three novel surfactin products were produced individually, lacking amino acid Leu-3, Asp-5, or Leu-6. Both [∆Leu3] surfactin and [∆Leu6] surfactin presented reduced toxicity, and [∆Asp5] surfactin showed greater inhibition when compared to native surfactin against B. pumilus and Micrococcus luteus. Also, [∆Leu6] surfactin showed a significant antifungal activity against Fusarium moniliforme. Liu et al. (2016) were able to create novel LPs by shifting the selectivity of the donor COM domain (communication-mediating domain, essential for coordinating intermolecular communication within NRPSs complexes). Using this technique, and reprogramming the plipastatin biosynthetic machinery, five new LPs were identified. All of the molecules showed antimicrobial activity against five fungal species (Rhizopus stolonifer, F. oxysporum, Aspergillus ochraceus, Penicillium notatum, and A. flavus). Gao et al. (2016) claim to be the first authors to report truncated cyclic LP production and module skipping by simply moving the TE domain forward in an NRPS system. The authors stated that the plipastatin TE domain could be used to rationally manipulate the ring size of macrocyclic products and could be a potential domain in the engineering of peptide synthetases for generating many new analogues of active peptides.

Studies on engineering of new LP molecules are an important effort to bypass current and future problems of resistance against these molecules. Lipopeptides resistance against a B. subtilis surfactin by Streptomyces sp. has already been detected. Based on previous study, this species was able to secret an enzyme that could hydrolyse and confer resistance to aerial growth inhibition (Hoefler et al. 2012). The dissemination of the use of these biocontrol agents, along with its produced molecules, could be a triggering agent to new cases of resistance. Therefore, continuous advances and discoveries of new molecules are indispensable for the endurance of this biological technology.

Recent technologies

In terms of recent technologies, major companies such as Bayer and Dupont showed interest in using cyclic LPs as biocontrol agents [e.g. WO2016044529-A1 (Curtis and Thompson 2016), WO2012162412-A2 (Guilhabert-Goya and Margolis 2012), WO2015184170-A1 (Kijlstra et al. 2015), WO2013126387-A2 (Weber et al. 2013)]. In the past 10 years, some technologies related to genetic modified organisms (GMOs), processes, purification and characterization, and formulation have been developed and protected (Table 5).

Table 5 Different patents on the use of LPs as biocontrol agents
Full size table

Regarding new GMOs, the patent WO2017125583-A1 (Lereclus et al. 2017) claims a genetically engineered Bacillus thuringiensis capable of overexpressing the krsE gene, responsible for the production of the LP kurskatin. The molecule showed antimicrobial activity against Galactomyces geotrichum and B. cinerea. Another molecular technology was developed in the patent CN102492639-A (Li et al. 2011a), which claims a GMO, Bacillus spp., with a high yield of antimicrobial LPs (iturin and surfactin) and an antifungal activity increased by 20–60%. Genetic engineering is also applied in the patent WO2008131014-A1 (Keenan et al. 2008); this work alleges an engineered LP synthetase polypeptide useful in synthesizing novel LPs, which are effective against insect or microbial pathogens.

New processes are also being developed, such as in the patent CN101041846-A (Gong et al. 2007), which claims a new pathway for the production of surfactin, using a medium containing soluble starch, sodium nitrate, potassium dihydrogen phosphate, magnesium sulfate, potassium chloride, ferrous sulfate heptahydrate, manganese sulfate, copper sulfate pentahydrate, and yeast extract. For LP purification, the invention employs an acid precipitation at 4 °C by adding hydrochloric acid until pH 2, followed by freeze drying to obtain a purified surfactin preliminary product. The invention CN103865855-A (Huang et al. 2014) also consists of a technology using a different medium composition from standard. For that, the inventors used a mutant B. subtilis, able to use glycerol as the main carbon source, for producing LPs. The use of some residues or alternative substrates as carbon source can also be seen in the patent WO2010039539-A2 (Jarrell et al. 2010), which states a medium for growing Bacillus cells constituted by cellulosic material as its main carbon source. The cellulosic material is comprised of soybean hulls, which contain cellobiose, xylose, xylan, or a combination thereof. The employment of soybean as substrate in semi-solid state fermentation was developed in the invention WO2016179735-A1 (Lu 2015); the inventors claim a high surfactin yield. The patent CN104232499-A (Wang et al. 2013) claims a production of LPs using Bacillus licheniformis in a 10 L bioreactor at 27–40 °C, a pressure of 0.03–0.08 MPa, and an aeration rate of 3–9 L/min.

Different inventions were developed for LP formulation and composition. Seeking to produce an environmentally compatible, storable, and long-acting agent against phytopathogenic microorganisms, the inventors of the patent WO2012130221-A2 (Borriss 2012) developed a product containing B. amyloliquefaciens spores. The formula contains at least 1 × 1010 spores/mL and an LP concentration of at least 1 g/L. In the patent WO2016044529-A1 (Curtis and Thompson 2016), the Bayer Company claims a product with a large action range, with the following composition: (1) recombinant exosporium-producing Bacillus cells that express a fusion protein comprising at least one plant growth-stimulating protein or peptide as well as a targeting sequence, exosporium protein, or exosporium protein fragment; (2) at least one biological control agent selected from the group consisting of the following LP-producing strains: B. subtilis, B. amyloliquefaciens, Bacillus firmus, and B. pumilus in a synergistically effective amount. Bayer also protected other technologies such as the patents WO2012162412-A2 (Guilhabert-Goya and Margolis 2012) (comprising a synergistic fungicidal combination of a polyene fungicide and at least one Bacillus LP) and WO2015184170-A1 (Kijlstra et al. 2015) (comprising a strain of B. subtilis or B. amyloliquefaciens and one of several compounds in a synergistically effective amount, which includes LPs from the surfactin, iturin, and fengycin families). Dupont is another major company that is interested in Bacillus LPs for agricultural application, covered by the patent WO2013126387-A2 (Weber et al. 2013). The invention claims an anti-contaminant composition comprising a cell-free fermentation product of one or more B. subtilis strains, containing LPs selected from surfactin, bacillomycin, and fengycin groups and combinations thereof.

Concerning the separation and purification of LP molecules, the extraction method of the patent CN105861602-A (Chen 2016) is based on the classic acid precipitation mentioned earlier in this work. It starts by the preparation of a fermentation broth of B. subtilis at 28 °C in a shaker for 72 h. The broth must be then centrifuged, with the pH adjusted to 2.5–3.0, and set at 4 °C overnight. After centrifugation, the precipitate is extracted by methanol (organic solvent), and finally, the solution is centrifuged again for crude LP methanol extraction and concentrated in a rotary evaporator to 1/8 of its original volume. The last step consists of a size-exclusion chromatography using a Sephadex G-100 column for separation to obtain the single antifungal substance. The inventors of the patent CN101851654-A (Liu et al. 2009) also extracted the LPs using methanol. However, RP-HPLC (Column YMC ODS-A 250 mm X 10 mm) was used to prepare and obtain a pure antifungal LP. Another approach can be seen in the patent CN101724014-A (Lin et al. 2009), which uses a 45–55% ammonium sulphate saturation to precipitate these compounds and later purifies the LPs from the iturin, fengycin, and surfactin families through Sephadex G-25 molecular sieve chromatography, cellulose DEAE-52 anion exchange chromatography, and FPLC 300SB-C18 column chromatography successively. The authors allege a distribution range of the molecular weight of the extracellular antibacterial LP between 1000 and 2200 Da.

Conclusions

Different species of Bacillus produce LPs with a large spectrum of antimicrobial activity. One of the most used species is B. subtilis, a widely adopted bacterial model organism for laboratory studies and one of the most understood prokaryotes in molecular and cellular biology. Therefore, studies relying on LP production by these organisms are provided with a huge literature to base and start a research project. Surfactin, iturin and, fengycin, families of LPs that are produced by Bacillus, are strong antimicrobial agents capable of acting synergistically against pathogens in a direct and indirect way by affecting and killing them or by activating an induced systemic resistance and biofilm formation, respectively. These properties of Bacillus LPs make them an excellent option for biocontrol systems, which are increasingly being demanded by both tightening regulatory restrictions and rapidly emerging disease resistance. Bacillus LPs present an important group of biomolecules that will inexorably contribute to the development of a cleaner and more sustainable agriculture in the next decades.

Author contribution statement

ROP, LPSV, CF, VTS, and CRS all contributed with ideas and the overall structure of the review. ROP wrote the initial draft and draw the figures. All authors read and approved the final manuscript.