Introduction

Cape gooseberry (Physalis peruviana Linn.) has gained importance worldwide due to the presence of bioactive compounds (Muñoz et al., 2021). In Colombia, cape gooseberry is one of the main fruits exported (Fischer & Melgarejo, 2020), but its production has been affected by vascular wilt disease caused by Fusarium oxysporum f. sp. physali (Foph) (Simbaqueba et al., 2018).

The control of Fusarium wilt in cape gooseberry relies on fungicides, but the efficacy is not sufficient. Moreover, there is a risk for the negative impacts on the environment. The use of biological control agents (BCAs) is an important alternative to control soil-borne phytopathogens (Palmieri et al., 2022), and their positive effect in agroecosystems with long-lasting effects has been proven (Brimner & Boland, 2003). Biological control involves antagonistic microorganisms, their metabolites, or cellular products that affect the plant pathogen or reduce disease severity through the induction of defense responses in the host plant (Vinale et al., 2008; Poveda & Eugui, 2022). The success of a BCA depends on its ability to establish at the application site and the capacity to express different modes of action (Saravanakumar et al., 2016).

Some compounds produced by BCAs can directly act on plant pathogens (direct modes of action), having antibiosis effects and parasitism through the production of cell wall degrading enzymes (Fesel & Zuccaro, 2016; Kumari & Srividhya, 2020). The parasitic activity by a BCA against a phytopathogenic fungus for instance, involves a complex process including the growth of the BCA toward the target pathogen, the adhesion to the hyphae and finally the attack and cell wall dissolution mediated by enzymes, such as chitinases and glucanases, which are associated with cell wall penetration (Bailey et al., 2008; Harman, 2000; Vinale et al., 2008).

The Trichoderma genus is widely recognized as efficient biocontrol agents with the ability to express various modes of action to control plant diseases (Harman, 2000; Yao et al., 2023; Mukhopadhyay & Kumar, 2020; Mohamed et al., 2021), and are the most studied antagonistic microorganisms (Kipngeno et al., 2015). Compounds such as xylanase (Calderón et al., 1993) and others produced by strains of Trichoderma spp. induce defense responses in the host plant avoiding the infection or delaying the colonization of the pathogens (Djonović et al., 2006). The competition for nutrients or space by strains of Trichoderma spp. confers long term protection against phytopathogens (Harman, 2000). The root colonization by Trichoderma prevents the entry of the pathogen into the plant by site occupation and the induction of plant defense responses (Daguerre et al., 2014; Mohamed et al., 2021). Antibiosis is also a mechanism of action for several strains of Trichoderma spp. (Degenkolb et al., 2012; Mukherjee et al., 2012; Mukhopadhyay & Kumar, 2020).

In the group of antagonistic bacteria, the rhizosphere associated Bacillus subtilis. species complex such as B. subtilis, B. amyloliquefaciens / B. velezensis, B. licheniformis and B. pumilus, have shown to be efficient in promoting plant growth and controlling phytopathogens (Borriss, 2015). Several strains of beneficial Bacillus spp. produce a wide range of bioactive compounds, where cyclic lipopeptides (CLPs) from the surfactin, iturin and fengycin families are recognized for their potential use in biotechnology and biopharmacy (Cochrane & Vederas, 2014). For instance, iturins and fengycins reduced the growth, and disrupted the cell integrity of Foph Map5, respectively (Moreno-Velandia et al., 2021). In addition to the antimicrobial properties of CLPs, they have also been shown to be involved in the colonization process and motility (Choudhary & Johri, 2009), as well as in the systemic stimulation of plant defense responses (Ongena et al., 2010). Competition for niches or nutrients in Bacillus strains occurs using chemotactic movements to obtain exudates in an efficient way (Lugtenberg & Kamilova, 2009).

On the other hand, the cell wall of fungi is composed of glucans, chitin, and glycoproteins (Pontón, 2008). Mannoproteins represent more than 80% of cell wall and are located on the external surface, β-glucan is located below the surface and is composed of frequent β-1,3 glucan bonds (65% and 90%) and additional β-1,6-glucan bonds (3 -10%), which create a supporting matrix for mannoproteins and provide structural rigidity to the cell wall. Chitin is located interspaced with β-glucan and represents 10 to 20% of the dry weight of the cell wall of filamentous fungi (Fesel & Zuccaro, 2016). The F. oxysporum cell wall is more resistant to CWDEs action than other soil-borne phytopathogens such as Rhizoctonia solani, which has been related to the specificity of BCAs against different pathogens. In addition, F. oxysporum shows a high ratio of laminated chitin compared with other plant pathogens (Sivan & Chet, 1989). Schoffelmeer et al. (1999) found that glycoproteins make up 50-60% of the F. oxysporum biomass, and therefore its high protein and chitin content must be considered in the selection process of antagonists against this pathogen (Sivan & Chet, 1989).

F. oxysporum can also synthesize mycotoxins that can improve its pathogenicity (Prigigallo et al., 2022) such as fusaric acid (FA), which is involved in the development of wilting symptoms through lipid peroxidation, the increase of reactive oxygen species, and plant cell death in the host. FA can also inhibit quorum sensing in Gram-negative bacteria (Prigigallo et al., 2022; Stępień et al., 2019; Tung et al., 2017), can cause toxic effects or act as bacteriostatic or bactericidal on Bacillus spp. (Bacon et al., 2006), and repress the synthesis of compounds related with biocontrol traits in Trichoderma spp. such as chitinases (Lutz et al., 2003). Nevertheless, there are some reports of bacterial species such as Burkholderia ambifaria that can degrade FA (Simonetti et al., 2018) as well as T. harzianum (El-Hasan et al., 2008; Marzano et al., 2013; Sharma et al., 2013).

Two major barriers that limit the functions of biocontrol agents are the poor colonization of the host surface, resulting in an inefficient inhibition of soil-borne phytopathogen growth. It is known that a large number of biotic and abiotic factors acting in the complex rhizosphere environment can influence the activity of BCAs and the outcome of the biological interactions (Mitter et al., 2019). Moreover, it is well known that using single strains to control some plant diseases can result in variable effects of biocontrol, which has been reported in previous studies with the Foph-cape gooseberry pathosystem (Izquierdo-García et al., 2021). To face this variability of biocontrol efficacy the use of mixtures of strains of BCA is a strategy with high potential (Izquierdo-García et al., 2020). Synergies between Trichoderma and antagonistic bacteria cause more benefits than the sum of their individual effects, and this makes them a promising alternative for managing plant diseases (Poveda & Eugui, 2022).

Only a few species of BCA have been explored in consortia against plant pathogens and just recently some studies have focused to identify consortia to mediate induced systemic resistance (Sarma et al., 2015; Patel et al., 2016; Poveda & Eugui, 2022; Hafiz et al., 2022; Zhou et al., 2021). But there are other studies that also report the production of antifungal metabolites (Karuppiah et al., 2019; Li et al., 2020; Ma et al., 2022), lytic enzymes (Radjacommare et al., 2010; Woo et al., 2002), the increase in the abundance of the beneficial microorganism (Wang et al., 2019), and the production of plant growth promoting substances (Kasa et al., 2015; Muhammad-Syafiq et al., 2021; Singh et al., 2020). The modes of action of synergistic consortia remains unknown however in most cases (Firdu et al., 2020, 2021; Negi et al., 2021). The efficacy of combined BCAs depends particularly on the modes of action involved. Therefore, to evaluate the performance of combined BCAs it is crucial to understand the dominant modes of action of each component of the consortium (Xu et al., 2011).

The consortium of Trichoderma virens Gl006 and Bacillus velezensis Bs006 was selected to increase the biocontrol activity against Fusarium wilt of cape gooseberry through synergy (Izquierdo-García et al., 2021). The efficacy was attributed to the interaction between Bs006 supernatant and Gl006 conidia in the rhizosphere (Izquierdo-García et al., 2020). However, the modes of action of the consortium remain unknown.

The aim of this research was to determine whether the modes of action displayed individually by T. virens Gl006 and B. velezensis Bs006 are also found in the mixture of both strains. Direct modes of action were evaluated through the ability to produce CWDEs, CLP´s and gliotoxin, and the ability of the strains to inhibit the production of fusaric acid by Foph in different culture media. Indirect modes of action were evaluated through the induced resistance against Foph in a split root model. The survival of Gl006, Bs006, and Foph in the rhizosphere, and the colonization of plant tissues as endophytes were also studied.

Materials and methods

Microorganisms and culture conditions

The highly virulent strain Map5 of F. oxysporum (Foph-Map5) isolated from cape gooseberry plants (Simbaqueba et al., 2018) and the biological control agents (BCAs) Trichoderma virens (Gl006) and Bacillus velezensis (Bs006) were provided by Agrosavia’s Microorganisms Germplasm Bank. These microorganisms were stored in sterile glycerol (30%) and peptone (0.1%) solution at -80 °C. Foph-Map5 was reactivated on potato-dextrose-agar (PDA, Oxoid®) and the second sub-culture was used for pre-inoculum production in potato-dextrose broth (PDB, Difco®). The inoculum for the bioassays was produced in 300 mL of PDB, in 1000 mL- Erlenmeyer flasks, with an initial concentration of 1 × 106 microconidia mL−1. The inoculated medium was incubated for seven days with continuous agitation (SHKA4000-5 MaxQ 4000, Thermo Scientific®, Canada) at 125 rpm and 25 °C. The fermented broth was filtered through a sterile muslin cloth (0.5 mm mesh) and the harvested microconidia were adjusted by counting in a Neubauer chamber to 1 × 106 microconidia mL−1 with sterile distilled water (SDW) for in vitro tests, and with tap water for greenhouse bioassays.

T. virens Gl006 was reactivated on PDA and the second sub-culture of seven days-old was scraped to harvest the conidia and was homogenized in SDW. B. velezensis (Bs006) was reactivated on Luria Bertani Agar (LBA: Tryptone 10 g, NaCl 10 g, yeast extract 5 g and agar 18 g L−1). The second sub-culture was grown on LBA for 24 h, and a suspension was prepared as a pre-inoculum to start a new culture in LB broth with the initial concentration of 5 × 106 cfu.mL−1, for 48 h under continuous agitation (125 rpm) and 25°C.

Plant material

For greenhouse assays, cape gooseberry seeds (Semicol®) were sown in sterile peat moss in germination trays of 72 alveoli and grown for 60 days. Before sowing, the seeds were washed with tap water to eliminate the attached fungicide, then were disinfected with NaOCl solution (3%) for 20 min and washed three times with sterile distilled water (SDW). The seedlings were transplanted to 600 g of substrate soil:rice husk (3:1 v/v) in black polypropylene bags (one seedling per bag).

Analysis of the enzymatic profile of BCA

Semi-quantitative enzymatic activity on solid medium

Semi-quantitative tests for chitinase, protease and cellulase activity by the BCA T. virens Gl006 and B. velezensis Bs006 were performed on solid medium. To produce the inoculum, Bs006 was grown on LBA and T. virens on PDA. The respective cells and conidia suspensions were prepared in SDW, through scraping the colonies with a sterile loop, then suspensions were adjusted to the desired concentration.

Chitinase activity

The protocol described by Agrawal and Kotasthane (2012) was followed to evaluate the biosynthesis of chitinase by the BCA. Briefly, colloidal chitin was prepared from reactive degree chitin (Sigma®). The hydrolysis of chitin was made by gradually adding 30 g of chitin in 200 mL of HCl 36% at 4 °C, keeping this mixture on continuous agitation (125 rpm) overnight. Subsequently, the extraction of colloidal chitin (neutralization with ethanol) was made by adding 1800 mL of cold (4°C) ethanol (95%) and was kept at 26°C overnight. The mixture was centrifuged (Biofuge Primo R, Thermo Scientific®, Canada) at 3000 rpm for 20 min at 4°C, the supernatant was discarded, and the pellet was resuspended in SDW, and centrifuged again at 3000 rpm for 5 min at 4°C. The pellet was washed until the smell of ethanol disappeared and the final harvested pellet was stored at 4°C until use.

A sterile minimal basal medium was used for chitinase detection (4.5 g colloidal chitin, 0.3 g MgSO4.7H2O, 3.0 g (NH4)SO4, 2.0 g KH2PO4, 1.0 g monohydrated citric acid, 18.0 g Agar, 0.15 g bromocresol purple, 200 µL Tween 80 per L), the pH was adjusted at 5.3 and the medium sterilized in an autoclave (Yamato SM301, USA). The medium was stored at 4°C until use. 10 µL of microorganism suspension was inoculated in the center of the solid medium and incubated at 28 °C. The biosynthesis of chitin by the growing BCA was assessed by the presence of a purple halo around the colony. The diameter of the halo was measured daily for four days with a digital caliper (Fisherbrand®, USA).

Cellulase activity

Using the model described above, the cellulase activity by BCA was evaluated on sterile carboxymethylcellulose solid medium (CMC-agar) (1.0 g KH2PO4, 0.5 g MgSO4·7H2O, 0.5 g NaCl, 0.01 g FeSO4·7H2O, 0.01 g MnSO4·H2O, 0.3 g NH4NO3, 10.0 g carboxymethylcellulose, and 18 g agar per L) (Teather & Wood, 1982). 6-well microplates (Falcon®) were poured with the sterile medium (3 mL per well) and stored at 4°C until use. Just before the inoculation a hole (5 mm) was made in the center of each gelled medium with a stainless-steel borer. Then, 50 µL of each treatment was inoculated and incubated (Memmert®, Germany) at 25°C for three days. After this time, 2 mL of Congo red solution (0.5%) was added, 20 min later 2 mL of NaCl 1M were added and incubated for 20 min. A yellow halo around the colony was the indicator of the positive reaction of cellulase activity, the diameter of this halo was measured as described above.

Protease activity

Protease activity was evaluated on sterile minimal basal medium (0.7 g K2HPO4, 0.3 g KH2PO4, 0.5 g MgSO45H2O, 0.01 g FeSO47H2O, 0.001 g ZnSO4, 0.001 g MnCl2, and 18 g bacteriological agar per L), supplemented with 3% of skimmed milk. 6-well microplates (Falcon®) were poured with this medium and stored at 4 °C until use. The medium was inoculated as described above and incubated for 24 h. The diameter of the hydrolysis halo (semitransparent halo) around the colony of each microorganism was the indicator of a positive reaction, and it was recorded the day after inoculation.

Quantitative enzymatic activity in liquid media

The biosynthesis of cell wall degrading enzymes (chitinase, protease, cellulase, and β-1,3-glucanase) by the BCAs Gl006 and Bs006 was determined in different liquid media as follows: the synthetic media PDB and LB, a soil solution supplemented with the artificial root exudates (SARE) described by Nihorimbere et al. (2012), the soil solution but now supplemented with cape gooseberry root exudates (SCRE), and a minimal basal medium (MBM) supplemented with Foph-Map5-inert mycelium (autoclaved and lyophilized) (M-Foph). The CWDEs production under co-culture of Gl006 and Bs006 was compared to the response under individual growth of the strains.

While PDB was used for Trichoderma sp. and LB broth for B. velezensis, PDB 50% with LB 50% mixture was used for the co-culture of the two BCA. The soil solution used on SARE and SCRE media was obtained by mixing 500 g of soil:husk (3:1 v/v) with 1000 mL of tap water. It was shaken at 125 rpm for one hour at room temperature (18 °C). Subsequently it was filtered through a sterile muslin cloth (0.5 mm mesh), then through 0.80 µm and 0.20 µm filters (Sartorius®).

SARE medium was prepared according to Nihorimbere et al. (2012), but in this case the soil solution was used as solvent. For SCRE medium, root exudates of cape gooseberry plants were produced as follows. 30 days-old seedlings were uprooted from the nursery-growth substrate and the root was washed with tap water and mild liquid soap. Afterwards the shoot (stem and leaves) was submerged in NaOCl (2%) solution for two seconds and the roots were submerged for five minutes. After these, four consecutive washes with SDW were made to eliminate NaOCl residues. The seedlings were arranged in sterile Falcon tubes (50 mL) (Falcon®) with the roots submerged in 40 mL of filtered soil solution. Plants were incubated at 15 °C for five days (12 h artificial light – 12 h darkness).

The minimal basal medium, corresponds to the MBM described above for protease activity, but supplemented with M-Foph in concentration of 1.2% (w/v) obtained from of the culture of Foph-Map5 in PDB medium kept for seven days at 125 rpm and 25 °C. The culture was filtered through a sterile muslin cloth to harvest the mycelium and lyophilized for 48 h (Thermo Scientific®).

For quantification of enzyme activity, the microorganisms were inoculated in 125 ml-Erlenmeyer with 25 mL of each culture medium at the initial concentration of 5 × 104 conidia.mL−1 for T. virens Gl006, 5 × 106 cells.mL−1 for B. velezensis, and 1.1 × 106 microconidia.mL−1 for Foph-Map5 (keeping the proportion of each microorganism used for bioassays under greenhouse as described by Izquierdo-García et al. (2020)). The liquid media inoculated were incubated (SHKA4000-5 MaxQ 4000, Thermo Scientific®, Canada) for five days at 125 rpm and 28°C. After two- and five-days growth, supernatant samples free of biomass were obtained through centrifugation at 15,000 rpm (Eppendorf 5415C Centrifuge, Germany) for 15 min and filtered by 0,20 µm (Sartorius®), this sample constituted the crude enzymatic extract.

After the incubation time, the viability of the three microorganisms in co-cultures was estimated as cfu.ml−1 through serial dilutions on triplicate Petri dishes with the solid media PDA supplemented with Triton X-100 (0.1% v/v), and Chloramphenicol (250 mg.L−1) for Foph-Map5 and Gl006, and LBA medium for Bs006. For Bs006 samples, a thermal shock was carried out at 90°C for 15 min previous the preparation of dilutions. Fungi were incubated for five days at 28°C and bacteria for 24 h at 30°C. There were three Erlenmeyer by treatment and the assay was performed twice.

Total chitinase activity

To quantify chitinase activity, colloidal chitin was used as substrate and N – Acetyl-D-Glucosamine as standard. The calibration curve of N–Acetyl-D-Glucosamine as a standard to calculate the concentration of reducing sugar was made from a concentration of 0.05 mg.mL−1 to 0.2 mg.mL−1. To determine chitinase activity, 1 mL of supernatant (crude enzymatic extract) with 0.3 mL sodium acetate buffer 1M (pH 4.6) and 0.2 mL of colloidal chitin were mixed and was incubated at 40 °C for 20 h (Memmert BA45, Germany) according to Agrawal and Kotasthane (2012). The mixture was centrifuged at 13,000 rpm (Eppendorf 5415C Centrifuge, Germany) for five minutes, 0.25 mL of supernatant was put in a 1.5 mL-Eppendorf tube and 0.25 mL of dinitro salicylic acid solution (DNS—1%) was added. The mixture was incubated in a water bath at 100 °C for five minutes, and the reaction was stopped by incubating on ice. Samples were put on 96 well microplates (Greiner®) and the absorbance was recorded in a microplate’s reader (Synergy HT, Agilent Biotek, USA—Software Gen5) at a wavelength of 540 nm. To determine the concentration of simple sugars released, the equation of the standard curve with N-acetyl-D-glucosamine (NAG) was used. One unit of chitinase activity (U) was defined as the amount of the enzyme that produces 1 μmol of N-acetyl-D-glucosamine equivalent from colloidal chitin per minute under the test conditions.

Exochitinase activity

The exochitinase activity was evaluated using nitrophenyl N, N-diacetyl β-D-chitobioside as substrate for the detection of chitobiosidase activity, and 4-Nitrophenyl N-acetyl-β-D-glucosaminide for the quantification of β-N-acetylglucosaminidase activity using the chitinase kit ref: CS0980 from Sigma-Aldrich®.

Protease activity

To quantify protease activity casein was used as substrate and tyrosine as standard. The calibration curve of L-tyrosine was made by preparing a solution of L-tyrosine 1.1 mM in ultrapure water, using five points at different concentrations. Each tyrosine standard was added followed by 625 µL of sodium carbonate and 125 µL of F–C reactive. Then it was shaken and incubated for 30 min. Subsequently, the absorbance was measured at 660 nm.

To determine protease activity, 25 µl of filtered extract—throughout 0,22 µm filter—with 130 µL of 0.65% casein solution were mixed in a 1.5 mL-Eppendorf tube. For the control, 25 µL of diluent were added. The mixture was stirred and incubated at 37°C for 10 min (Memmert, Germany). 130 µL of trichloroacetic acid solution (110 mM) was added to stop the reaction and incubated at 37°C for 20 min. Subsequently, it was centrifuged at 14,000 rpm and 250 µL of supernatant was taken to a new tube and 625 µL of sodium carbonate solution (500 mM), and 125 µL of F–C reactive (0.5 M) were added. The resulting mixture was vortexed and incubated at 37°C for 30 min (Memmert, Germany). 200 µL of the reaction mix were added to 96 well microplates, and the absorbance was measured in a spectrophotometer at 660 nm.

β-1,3-glucanase activity

10 µL of crude enzymatic extract was added to 20 µL of laminarin (Sigma®) solution (0.75% w/v) diluted in sodium acetate buffer (50 mM) according to Ramada et al. (2010) and incubated at 45°C for 15 min in water bath (Memmert BA45, Germany), then 100 µL of DNS was added. Reducing sugars were quantified without the addition of laminarin to the enzyme extract, to obtain the difference of reducing sugars obtained from enzymatic reaction. One unit (U) of β-1,3-glucanase activity was defined as the amount of enzyme that produces reducing sugars per minute under the described conditions. There were three biological and three technical replicates to measure the enzymatic activity. The assays were conducted twice.

Metabolites production in co-culture

The production of CLPs by Bs006, gliotoxin by Gl006 and FA by Foph was analyzed in the filtrates of co-cultures harvested from LB and PDB, MBM-MFoph, SARE and SCRE media. The detection and quantification of the compounds were carried out by HPLC analysis using the patterns of fengycin (SMB00292 Sigma-Aldrich®), surfactin (S3523 Sigma-Aldrich®), and iturin A (I1774 Sigma-Aldrich®) all from Bacillus subtilis; the pattern of gliotoxin from Gliocladium fimbriatum (G9893 Sigma-Aldrich®), and the patterns of fusaric acid from Gibberella fujikuroi (F6513 Sigma-Aldrich®).

Determination of enzymatic activity in soil

Cape gooseberry seeds were sown on peat moss in 72-well germination trays and the resulting seedlings were rooted for 60 days under greenhouse. The substrate colonized by the roots was carefully discarded and the seedlings were transplanted in non-sterile substrate (soil:rice husk 3:1). Suspensions of the BCA [T. virens (1 × 106 conidia.mL−1) and B. velezensis (1 × 108 cfu.mL−1)] (30 mL per plant) were applied immediately by soil drenching, as single and as consortium treatments, together with the pathogen Foph-Map5 (2.22 × 106 microconidia.mL−1). Plants inoculated only with Foph-Map5 and plants free of both pathogen and BCA were included as negative control and absolute control, respectively. Rhizosphere samples were taken at one, two and seven days after application of treatments and stored at -20 °C until its processing. The protocol described by Jackson et al. (2013) was followed to obtain the enzyme extract from the soil. Briefly, 6 g of soil sample was poured into a sodium acetate buffer solution (6 mL—pH 5.1) (ratio 1:1 v/v), the sludge obtained was the enzyme extract used for the enzymatic activity tests (preserving granules of soil within each reaction) following the protocols described previously. Each sample was vigorously vortexed for 30 s. There were three biological replicates and three technical replicates for each enzyme assay chitinase, protease, and β-1,3-glucanase, and each test was driven twice. The survival of the applied microorganisms was determined by counting of colonies from washed soil dilutions plated on solid media as described previously.

Induction of resistance

To determine the ability of Gl006, Bs006, as single treatments and as a consortium, to induce resistance against vascular wilt of cape gooseberry, a bioassay was conducted according to the split root model described by Moreno-Velandia et al. (2009). A half portion of the plant root into the substrate was inoculated with 30 ml-suspension of the BCAs (Gl006 at 1 × 106 conidia.mL−1 and Bs006 at 1 × 108 cfu.mL−1). Three days after inoculation of BCA the other half of the root was inoculated with 30 mL of Foph-Map5 (2.22 × 106 microconidia.mL−1). Microorganisms applied in an individual way were used as controls for the effect of the consortium. Untreated plants growing in the presence of the pathogen, and plants untreated and free of pathogen were included as negative control and absolute control, respectively. The incidence of the Fusarium wilt disease, expressed as the percentage plants showing typical symptoms of the disease, and the disease severity according to the scale described by Moreno-Velandia et al. (2019) were recorded every four days since the detection of the first symptoms. The disease intensity index was calculated with the equation \(DII=\left[\left(\sum Si*Ni\right)\div \left(5*N\right)\right]*100\), where Si is the severity level of the symptoms, Ni is the number of plants in each severity class Si, 5 represents the number of classes of the severity minus 1, and N is the total number of plants of the experimental unit (Hervás et al., 1998). The area under the disease incidence and intensity index progress curves were calculated through the trapezoidal integration method described by Shaner and Finney (1977). The percentage reduction of the area under disease progress curve due to the treatments was calculated through the equation \(E=\left(\left(a-b\right)\div a\right) x 100\), where a represents the value of the disease in the negative control and b represents the disease in a given treatment (Abbott, 1925). Detection of disease reduction due to the biocontrol treatments was considered as a result of the systemic induced resistance effect, since the treatments and the pathogenic inoculum were spatially separated.

Colonization of plant tissue and survival of BCA in the rhizosphere

The colonization and survival of BCAs in plant tissue and substrate were evaluated through two independent assays in the seedbed and transplant stages, respectively. 5 mL-suspension 1 × 106 conidia.mL−1 for T. virens Gl006 and 1 × 108 cfu.mL−1 for B. velezensis Bs006 were applied per plant in seedbeds, and 30 mL in transplants. Samples of peat moss (seedbed) and soil (transplant) at time zero (T0), samples of substrate and plant tissue at 15 days after emergence of seedlings (on seedbed) and at seven and 21 days after transplant) were taken to estimate colonization and survival of the microorganisms. Samples of plant tissue (stem and root) were washed with detergent solution and disinfected with 3% sodium hypochlorite for 10 min, followed by four consecutive washes with SDW. Then 7 cm of stem and main root were weighed in analytical balance, and then macerated in 7 mL of sterile Tween 80 solution (0.1% v/v). 1 mL of the suspension was plated on solid culture media PDA supplemented with Triton X-100 and Chloramphenicol, for counts of Foph-Map5 and Gl006 colonies, and on LBA for Bs006 previously exposed to thermal shock (90°C for 15 min) (Memmert BA45, Germany). For soil or peat samples processing, 1 g of each sample was taken to 9 mL of sterile Tween 80 solution (0.1% v/v), then serial dilutions were made and plated on each solid culture media.

Statistical analysis

Data were analyzed for normality according to Shapiro Wilk’s test (P > 0.05) and homoscedasticity with the Levene’s test (P > 0.05). The significant effect of the treatments on the response variables was determined by one-way ANOVA using a generalized linear model (GLM) of SAS. Duncan’s multiple range, Fisher’s least significant difference (LSD) and Tukey’s tests (α = 0.05) were used for comparing media among treatments. All the statistical analysis were carried out with the S.A.S software (version 9.4: SAS Institute, Cary, NC).

Results

Enzymatic profile of single BCAs and as consortium

Semiquantitative enzyme activity

While T. virens Gl006 synthesized only chitinase, B. velezensis Bs006 synthesized protease and cellulase in solid medium, (Fig. 1). A higher concentration of microorganisms generated a larger halo. When both microorganisms were inoculated in consortium on the culture media, cellulase activity of Bs006 was inhibited; the chitinase activity of Gl006 decreased significantly under the combination of high concentrations of Bs006 (107 and 108 cfu.mL−1) with the lowest concentration of Gl006 (105 conidia.mL−1); and the protease activity of Bs006 was reduced by the presence of Gl006 at all concentrations tested (Table 1 and Fig. 1).

Fig. 1
figure 1

Hydrolysis halos produced by T. virens Gl006 or B. velezensis Bs006 in different culture media, evaluating semi quantitative enzymatic activity. A. chitinase purple halo in chitin medium, after four days of inoculation. B. Protease transparent halo in skim milk. C. Cellulase yellow halo in CMC medium

Full size image
Table 1 Effect of co-culture of T. virens Gl006 and B. velezensis Bs006 in solid medium on the enzymatic profile
Full size table

Quantitative enzymatic activity

Co-culture enzymatic activity

When the production of cell wall degrading enzymes (CWDEs) was evaluated in the liquid cultures LB and PDB, β-1,3-glucanase activity was detected in PDB cultured with T. virens Gl006. However, glucanase activity was not detected in the co-culture of Gl006 with B. velezensis Bs006, suggesting an inhibitory effect due to the presence of Bs006. On the other hand, total chitinase activity was detected in PDB cultured with Gl006 alone and in co-culture with Bs006. Chitinase activity was not detected in LB broth cultured with Bs006, suggesting that the enzymatic activity detected in co-culture was produced by Gl006 (Table 2). Protease activity was detected only in LB broth cultured with Bs006, but not in the media cultured with Gl006. Cellulase activity was detected in LB broth cultured with Bs006, but it was significantly reduced in co-culture with Gl006 (Table 2).

Table 2 Cell wall degrading enzymes (CWDEs) produced by T. virens Gl006 and B. velezensis Bs006 in Luria Bertani and Potato Dextrose liquid media
Full size table

The evaluation of CWDEs in minimal basal medium supplemented with mycelium of F. oxysporum f. sp. physali strain Map5 (Foph), showed a lower production of glucanase and chitinase by Gl006 both when grown alone as well as in co-culture with Bs006. β-1,3-glucanase was detected in the minimal basal medium supplemented with Foph-mycelium from the second day after inoculation of Gl006, but the enzyme was not detected in the medium without Foph-mycelium (Table 3), suggesting that β-1,3-glucanase and total chitinase activities were induced by Foph mycelium. In co-culture, β-1,3-glucanase activity was detected later, and at lower concentrations than detected in the individual culture of Gl006 (Table 3). Contrasting results were observed with the total chitinase activity, which was identified early in co-culture conditions, and at higher values than found in individual culture of Gl006 (Table 3). On the other hand, no activity of glucanase and chitinase was detected in the minimal basal medium inoculated with Bs006 both with and without adding Foph mycelium (Table 3). The exochitinases, chitobiosidase and β-N-acetylglucosaminidase were detected in some treatments in the absence of Foph mycelium. Chitobiosidase activity was detected at both the second and fifth day after inoculation of the medium in all treatments, except for Bs006 inoculated in the medium without addition of Foph mycelium. Detected chitobiosidase activity in the M-Foph control suggests no chitobiosidase activity by Bs006. The chitobiosidase activity detected in co-culture was due only to Gl006, but this was lower than chitobiosidase in the single Gl006 cultures, suggesting a negative effect by Bs006 on Gl006 enzymatic activity. Detected β -N-acetylglucosaminidase activity in minimal medium supplemented with Foph mycelium also was attributed to Gl006 since single cultures of Bs006 did not show this enzymatic activity. In addition, the presence of Foph mycelium in the medium stimulated the β -N-acetylglucosaminidase by Gl006 since its level was very low in absence of Foph mycelium.

Table 3 Cell wall degrading enzymes (CWDEs) produced by T. virens Gl006 and B. velezensis Bs006 in minimal basal broth supplemented with Foph-Map5 inert mycelium (M-Foph)
Full size table

The co-culture of Gl006, Bs006, and Foph-Map5 in the artificial root exudates medium showed no activity of total chitinase but it showed exochitinase activity (chitobiosidase and β-N-acetylglucosaminidase). Chitobiosidase activity was observed on the second day of growing in the single cultures of BCA and in the co-culture in the absence of Foph (Table 4). While Bs006 showed chitobiosidase activity in the presence of Foph, the presence of the pathogen in the culture media inhibited this enzymatic activity by Gl006 and by the co-culture of Gl006 with Bs006 (Table 4). The samples from the fifth day after growing showed an increase in chitobiosidase activity by Gl006 as single culture, in consortium with Bs006, and in the presence of Foph (Table 4).

Table 4 Cell wall degrading enzymes (CWDEs) profile produced by T. virens Gl006 and B. velezensis Bs006 alone and in co-culture in soil solution broth with artificial root exudates in the presence or the absence of Foph-Map5 (Foph)
Full size table

The evaluation of the enzymatic activity in the medium with the natural root exudates harvested from cape gooseberry plants detected no β-1,3-glucanase and chitinase activity. Chitobiosidase activity was identified at the second day after the inoculation in the consortium, both with and without Foph presence. The highest values were found in co-culture with Foph (Table 5). On the fifth day after incubation, chitobiosidase activity was detected only in the cultures of single Gl006, and the co-culture of Bs006 with Foph. The β-N-acetylglucosaminidase activity was identified in the culture of single Gl006, and in the co-culture of Gl006 with Bs006 in the absence of Foph. At day five it was detected in all treatments, except in the culture of single Bs006 and in consortium with Gl006, indicating a negative effect of Bs006 on the production of β-N-acetylglucosaminidase by Gl006. The protease activity was detected on both days of sampling, being higher on the fifth day than on the second day after inoculation. On the second day of incubation the protease activity was not detected in the culture of Gl006 + Foph nor in the culture of single Foph. It was found that the highest values were found in the control (Table 5).

Table 5 Cell wall degrading enzymes (CWDEs) profile produced by T. virens Gl006 and B. velezensis Bs006 alone and in co-culture in soil solution broth with cape gooseberry root exudates in the presence or the absence of Foph-Map5 (Foph)
Full size table

Population of Gl006, Bs006 and Foph-Map5 in co-culture

When inoculations of the single strains T. virens Gl006 and B. velezensis Bs006 were made in the minimal basal medium supplemented with inert mycelium of Foph-Map5, both BCAs showed similar population density at both sampling times, the second and the fifth day after incubation. However, under co-culture of the BCAs in the presence of inert mycelium of Foph it was observed that population density of Bs006 was significantly lower than the population in the single culture. The population of Gl006 was significantly lower in the medium without mycelium of Foph than in the medium with it. In contrast, the population of Bs006 was similar among the two conditions of the culture medium at two days after inoculation, but at the fifth day after incubation the population of Bs006 was higher under co-culture with Gl006 than in the single culture, in absence of mycelium of Foph (Fig. 2).

Fig. 2
figure 2

Population density of T. virens (Gl006) and B. velezensis (Bs006) under co-culture in the minimal basal medium (MBM) supplemented with mycelium of F. oxysporum f. sp. physali strain Map5 (MFoph). Gl006 was inoculated at 5 × 104 conidia.mL−1 and Bs006 at 5 × 106 cfu.mL.−1. Columns with the same letter are not significantly different according to the Tukey test (α = 0.05)

Full size image

The evaluation of the microorganism’s population in the medium supplemented with the natural root exudates from cape gooseberry showed that Bs006 had a similar population density among single and co-culture inoculation with Gl006, as well as in presence and absence of Foph in the culture medium. The population density of Bs006 was similar on both sampling dates (Fig. 3). The behavior of Gl006 was like that of Bs006. Moreover, no reduction of the BCAs populations was observed in the co-cultures with and without Foph. Concerning the population of Foph, it was found that Foph was reduced to less than 10 cfu.mL−1 during the first 48 h of co-culture with each biocontrol agent (Bs006 and Gl006). However, this effect was transient since at the fifth day Foph kept its population in all co-cultures (Fig. 3).

Fig. 3
figure 3

T. virens (Gl006), B. velezensis (Bs006) and F. oxysporum f. sp. physali strain Map5 (Foph-Map5) populations in co-culture in liquid soil solution medium and cape gooseberry root exudates. Gl006 was inoculated at 5 × 104 conidia.mL−1, Bs006 at 5 × 106 cfu.mL−1 and Foph-Map5 at 1.1 × 105 microconidia.mL.−1. Columns with the same letter are not significantly different according to Tukey test (α = 0.05)

Full size image

In the media supplemented with artificial root exudates, the population density of Bs006 was similar in the single and the co-culture with Gl006 both in the presence and absence of Foph at 48 h of incubation. However, at five days after incubation the population of Bs006 was reduced significantly in the co-cultures with Gl006 and with Foph (Fig. 4). Viability of Gl006 was affected in this culture medium during the first 48 h of incubation and its density was reduced significantly when was in co-culture with Bs006 and Foph. However, five days after incubation its viability was restored under co-cultures (Fig. 4) suggesting a fungistatic effect. In contrast, the population of Foph decreased to 50 and less than 10 cfu.mL−1 when this was co-cultured with Gl006 and Bs006 respectively. The population of Foph-Map5 growing alone, was high and similar in both sampling times (Fig. 4).

Fig. 4
figure 4

T. virens (Gl006), B. velezensis (Bs006) and F. oxysporum f. sp. physali strain Map5 (Foph-Map5) populations in co-culture in soil solution medium and artificial root exudates. Gl006 was inoculated at 5 × 104 conidia.mL−1, Bs006 at 5 × 106 cfu.mL−1 and Foph-Map5 at 1.1 × 105 microconidia.mL.−1. Columns with the same letter are not significantly different according to Tukey test (α = 0.05)

Full size image

Enzymatic activity and population of microorganisms in soil

All four enzymatic activities were detected in soil samples (Table 6; Fig. 7). The β-1,3-glucanase activity was detected at both sampling times (first and second day after inoculation). However, the β-1,3-glucanase activity was significantly lower in soil samples treated with Gl006 and Foph than in the control. On the other hand, total chitinase activity was detected from the first day of sampling in the control and in soil samples inoculated with Gl006, but not in soil samples treated with Bs006 and with Foph. However, total chitinase activity was detected in these last two treatments at two days after inoculation. At this time, the soil samples treated with Gl006 showed significantly higher chitinase activity than the samples with the consortium of BCAs, and the Foph treatment (Table 6). At seven days after inoculation the chitinase activity decreased in most treatments, except on soil inoculated with Bs006, where it tended to be higher. Exochitinase-chitobiosidase activity was detected in all treatments. However, this tended to be higher in non-inoculated soil. β-N-acetylglucosaminidase activity was similar in all soil samples (Table 6). Protease activity was detected in Bs006 treatments on day one after inoculation. However, protease activity was also detected in the other treatments on day two, except on soil inoculated with the three microorganisms (Gl006, Bs006, and Foph), where it was detected at seventh day after inoculation. The non-treated soil showed high protease activity at seven days after inoculation and remained high throughout the experiment.

Table 6 Cell wall degrading enzymes (CWDEs) profile produced by T. virens Gl006 and B. velezensis Bs006 alone and in co-culture in soil inoculated with biological control agents and F. oxysporum Foph-Map5
Full size table

The population of Foph had decreased significantly by 63% in the soil samples treated with the consortium of Gl006 and Bs006 at seven days after inoculation, indicating a synergistic interaction (supplementary material – Table S1).

Production of CLPs and other metabolites in co-culture

With respect to the CLPs synthesized by Bs006 in the minimal medium supplemented with the natural cape gooseberry root exudates, surfactin was detected under the detection limit (< 0.1 µg.mL−1) only in the culture of Bs006 + Foph at two and five days after inoculation, also iturin A was detected at five days (< 0.05 µg.mL−1), production of surfactin detected in this medium by Bs006 was doubtful since in cape gooseberry exudates similar concentrations were detected (Table 7). On the other hand, in the medium supplemented with the artificial root exudates, surfactin and fengycin were detected at the second and fifth day after incubation with similar values among the sampling dates. In addition, there was no significant difference among the concentration of these CLPs under single culture of Bs006 and the co-culture with Gl006. Iturin A was detected in the single culture of Bs006, but only at five days after incubation, and at low concentrations (< 0.05 µg.mL−1) (Table 7).

Table 7 Metabolites produced by T. virens Gl006 and B. velezensis Bs006 alone and in coculture and in the presence or the absence of Foph-Map5 in different culture media
Full size table

In minimal basal medium supplemented with Foph mycelium, traces of fusaric acid (FA) were detected only in the co-culture of Bs006 with Gl006. In this medium fengycin and surfactin were also detected from the second day after incubation in the single culture of Bs006, but the coculture with Gl006 reduced the biosynthesis of these CLPs. Iturin A was not detected in this culture medium under any of the culture combinations (Table 7).

The highest concentrations of surfactin and fengycin produced by Bs006 were detected in the LB-PDB medium at similar concentrations in the single culture of Bs006 and in the co-culture with Gl006. At five days after incubation the CLPs were detected only in the single culture of Bs006 but not detected in the co-culture of Bs006 with Gl006 (Table 7).

Gliotoxin was not detected in samples of Gl006 alone and in consortium in co-culture with Foph or its inert mycelium (data not shown). Furthermore, FA was detected at 25,4 µg.mL−1 in the PDB medium cultured with Foph (5 × 107 microconidia.mL−1) used as inoculum for the greenhouse bioassay (Table 7).

Induced resistance

Application of cells and supernatant from Gl006 and Bs006 liquid cultures to plant root showed a reduction of the disease, which can be attributed to induced systemic resistance. The treatments with cells of the BCAs, both singly and as consortium, significantly decreased the incidence of Fusarium wilt compared with the negative control (Fig. 6a). The application of Gl006 supernatant also significantly reduced the disease (Fig. 6a). The area under incidence progress curve (AUDPC) were similar for the three cell-based treatments. However, only the Gl006 cells-based treatment reduced the area under severity progress curve (AUDPC) (Fig. 6b). In this experiment, the consortium-based treatment did not show a synergistic or additive effect in the effectiveness against Fusarium wilt (Supplementary material – Table s2).

Colonization of plant tissues and survival of microorganisms in the soil

Population of microorganisms on seedbeds

At 15 days after application of the consortium of B. velezensis Bs006 and T. virens Gl006 to cape gooseberry plants on seedbeds or in the phase of transplantation, the density of both microorganisms were equal to that at the beginning of the experiment (8 × 104 and 2 × 104 cfu.g−1 of substrate, respectively), suggesting that the presence of one did not affect the population of the other BCA in the substrate (supplementary material – Figure s2a). However, it was found that the population of Bs006 decreased on day 15 after single application. On the other hand, Bs006 was isolated as endophyte from plant stems (Supplementary material – Figure S1b) and from the main root (Supplementary material – Figure S1c), but this population was not affected significantly by the addition of Gl006 in the consortium treatment.

Population of microorganisms after transplantation

The population of Foph decreased at seven days after the inoculation (dai) into the soil under treatment with the consortium of T. virens Gl006 and B. velezensis Bs006, compared with the population in the negative control. However, at 21 dai the population of Foph had increased and showed similar values in all treatments (Fig. 5). The population density of the BCA was not affected at seven dai as single and consortium treatments. Interestingly, the population of Gl006 was higher in the consortium treatment compared with the single application. Similar results were observed for Bs006 at 21 dai (Fig. 5).

Fig. 5
figure 5

T. virens (Gl006), B. velezensis (Bs006) and F. oxysporum f. sp. physali strain Map5 (Foph-Map5) populations in soil samples inoculated individually and in consortium. 0 dai: initial time and 7 dai: final time. Soil samples correspond to samples used for quantifying enzymatic activity. Columns with the same letter are not significantly different according to Tukey test (α = 0.05)

Full size image

While Bs006 was endophytic in all treatments both in stem and root tissues sampled at seven dai (Supplementary material – Figure S2a and Figure S2b), Gl006 was isolated only from the main root at 21 dai, when the strain was applied as single treatment (Supplementary material – Figure S2b). On the other hand, Foph was isolated from the inner tissues of the stem and the main root of cape gooseberry at 21 dai, in single treatment of Bs006 (supplementary material – Figure S2a and Figure S2b).

Discussion

Direct interaction between BCA and plant pathogens is a process that takes place in several stages. In the case of biocontrol strains of Trichoderma spp. and the targeted fungi, the process takes place in two stages. In the first, BCA produces hydrolytic enzymes and at the second stage detects the presence of molecules resulting from enzymatic degradation and activates the cascade of gene expression (Vinale et al., 2008; Yao et al., 2023). Lu et al. (2004) observed in R. solani-bean/cucumber pathosystems, that T. atroviride started to colonize the soil within the first 24 h and at this first stage showed the expression of the genes nag1 (exochitinase – N-acetyl-β-D-glucosaminidase) and ech42 (endochitinase). In addition, the gene expression in T. atroviride was constant until the seventh day after inoculation of BCA into the soil. After this time the intensity of gene expression decreased (Lu et al., 2004). In the current study we found that enzymatic activity by T. virens Gl006 remained high in the soil in the two sampling dates, the first and the seventh day after inoculation, agreeing with Lu et al. (2004).

During co-culture of T. atroviride and R. solani in minimal medium, Lu et al. (2004) found that the higher gene expression of nagI and ech42 occurred at 48 h after incubation. This also happened in minimal medium supplemented with colloidal chitin, whereas in minimal basal medium controls without supplement, the higher gene expression was observed at 72 and 96 h, being prolonged but in lower concentration. In this study T. virens Gl006 presented β-N-acetylglucosaminidase activity in minimal basal medium supplemented with mycelium of Foph and, interestingly, this activity was not affected when Gl006 was co-cultured with B. velezensis Bs006 which allows us to suggest that the gene expression model described by Lu et al. (2004) for T. atroviride – R. solani could apply also in the system T. virens Gl006 – F. oxysporum f. sp. physali and could suggest that B. velezensis does not affect the biosynthesis of chitinases in T. virens. However, these hypotheses need to be tested in further (Figs. 6 and 7).

Fig. 6
figure 6

a Effect of application of T. virens Gl006 (Gl) and B. velezensis Bs006 (Bs) alone and in consortium on incidence progress of vascular wilt caused by F. oxysporum f. sp. physali strain Map5 (Foph) in the split root model. Area under disease progress curve (AUDPC-Incidence) was calculated based on the incidence of the disease during 50 days after application of treatments. Soil was inoculated with Foph at 1 × 104 cfu.g−1 of soil in one portion of the split root. Bars on columns represent standard deviation (n = 4). Columns with the same letter are not significantly different according to Duncan multiplate range test (α = 0.05). Con: Conidia, Cel: Cells, Sup: Supernatant. b Effect of application of T. virens Gl006 (Gl) and B. velezensis Bs006 (Bs) alone and in consortium on severity progress of vascular wilt caused by F. oxysporum f. sp. physali strain Map5 (Foph) in the split root model. Area under disease progress curve (AUDPC-Severity) was calculated based on the incidence of the disease during 50 days after application of treatments. Soil was inoculated with Foph at 1 × 104 cfu.g−1 of soil in one portion of the split root. Bars on columns represent standard deviation (n = 4). Columns with the same letter are not significantly different according to Duncan multiplate range test (α = 0.05). Con: Conidia, Cel: Cells, Sup: Supernatant

Full size image
Fig. 7
figure 7

Consortium arsenal model proposed against Fusarium wilt of cape gooseberry. According to our findings, preventive application of Bs006 alone allows to enter in plant tissues and induced plant resistance, whilst Gl006 at this point can induce resistance but can enter to the plant, but when they are together as a consortium Bs006 enter to plant tissues and both induce plant resistance. When they are applied at transplant, Bs006 applied alone in Foph presence could facilitate Foph entry, at this point Bs006 can synthesize CWDE like chitobiosidase, protease, and β 1,3 – glucanase and other compounds like CLP´s that are not enough to control the disease. In consortium, compounds produced by Bs006 and Gl006 can stop Foph entry and reduced its population in soil for a while using direct modes of action. Further studies are needed to know what is activating in the plant by the consortium, how is Gl006 blocking Foph entry and what Bs006 compounds could help Foph to enter in plant tissues faster than when Foph is alone

Full size image

Hirsch (2018) found that the most effective strain of B. subtilis to control Fusarium spp. showed chitinase activity and volatile metabolites. Although the synthesized enzymes by biological control agents are capable of degrading cell walls of fungal hyphae in vitro, this does not prove that these enzymes are responsible for the biocontrol activity in planta (O’Brien, 2017). We found that the population of Foph significantly decreased under in vitro co-culture with the BCAs and its effect was more prolonged in treatments and culture media where enzyme activity was detected, so this would be related to the quantity of enzyme activity and to the repertoire, which suggests that prolonged production of these enzymatic activities can be responsible for a more prolonged effect against Foph. Similar results were found in soil samples from the in planta bioassay, suggesting that such biocontrol traits were active also in the soil.

Considering that the co-culture of the BCA and Foph in the liquid culture media supplemented with cape gooseberry root exudates showed a transient fungistatic effect on Foph, but the population of Foph in the soil samples started to decrease at seventh day after incubation, the transient effect could be related to the protease activity detected. These results agree with Han et al. (2015) who found that strains more effective as BCA are those with protease and chitinase activity. Proteases play an important role in lysis because they bind to the layer of mannoproteins of cell walls and open the structure, exposing internal glucan and chitin microfibers, thus facilitating chitinase and glucanase activity.

In the soil samples from the greenhouse we observed that the control treatment showed a high β-1,3-glucanase and protease activity compared to other treatments. This may be explained from an initial stimulus when transplanting, leading to exudates of cape gooseberry roots Plants deposit around 40% of photosynthetically fixed carbon in the roots, being a stimulus for soil microbiota that feed on that carbon source, leading to an increase in population (Pieterse et al., 2014). This stimulus on the microbial population could stimulate also the enzymatic activity from the soil microbiota.

According to Bhatia et al. (2018), the use of interkingdom co-cultures like fungus and bacteria, can increase their functionality and effectiveness. In this study the consortium of T. virens Gl006 and B. velezensis Bs006 was effective in decreasing the population of Foph in soil supplemented with cape gooseberry root exudates. However, chitinase and β-1,3-glucanase were not detected in this medium which is surprising as this medium is quite similar to the in vivo conditions, where the enzymes were detected. This could have two reasons, firstly, in fact total chitinase was not produced in detectable concentration trough this protocol and thus, this activity was responsible for decreasing the population of Foph in the soil; secondly, it could be due to the protocol used for production of the enzyme extract of the two sample types. The protocol used the supernatant from liquid cultures as enzyme extract and the sediment that contained some soil granules was discarded, whereas the protocol proposed by Jackson et al. (2013) suggests that for processing soil samples it is important to keep the granules inside the enzymatic reaction because they may have trapped enzymes that are involved in the evaluated activities. Probably due to the low values of the chitinase and β-1,3-glucanase activity, these were not detected, whilst the technique applied for chitobiosidase and y β-N-acetylglucosaminidase is more sensitive because the chitinase kit (Sigma®) contains pure substrates for the detection of chitobiosidase, and β-N-acetylglucosaminidase activity, whereas in the quantification of total chitinase activity colloidal chitin was used.

The increase of the Bs006 population density in the minimal basal medium co-cultured with Gl006 could be related to adherence of Bs006 to Gl006 structures which was observed in a previous study (Izquierdo-García et al., 2020). This could facilitate the movement of bacteria in the soil (Warmink & Van Elsas, 2009; Warmink et al., 2011). Other authors have also found increased population growth of a BCA in the presence of another microorganism in the consortium. For example, Dandurand and Knudsen (1993) reported that Pseudomonas fluorescens 2–79-B46 increased the population density of T. harzianum on pea seeds.

In this study we reported the biosynthesis of fusaric acid (FA) by Foph strain Map5, which is an important mycotoxin not only involved with the pathogenicity of F. oxysporum, but it can also affect the biocontrol traits of BCA (Bacon et al., 2006; Lutz et al., 2003; Prigigallo et al., 2022; Ruiz et al., 2015; Venkatesh & Keller, 2019). FA was detected in the culture of Foph in PDB medium in which the inoculum for the bioassays was produced. Therefore, the in planta bioassay in the present study, with artificial inoculation of Foph, cape gooseberry plants were in contact with 0.5 µg.mL−1 or 0.5 ppm FA.

In culture media under in vitro conditions, FA was detected only in the minimal basal medium containing mycelium of Foph inoculated with the co-culture of Gl006 and Bs006, but not in the single culture of Bs006. Moreover the CLPs biosynthesis was not affected by Fusaric acid in this treatment. In contrast, some studies have found that Fusaric acid can affect BCA growth, or the production of some metabolites. For example, Bacon et al. (2006) found that FA had a bactericide or a bacteriostatic effect on some strains of Bacillus spp. at 100 and 200 µg.mL−1 and only few of them were tolerant to FA. FA can also suppress the biosynthesis of 2,4-diacetylphloroglucinol by P. fluorescens (Notz et al., 2002). In the present study it was noticed that Bs006 was tolerant to FA (1.1 ppm), which did not affect the population of Bs006 to a significant extent. To the best of our knowledge there are not reports describing the mechanisms through which Bacillus could degrade FA. In contrast, Burkholderia ambifaria can use FA as a sole carbon and energy source (Simonetti et al., 2018).

If Bs006 have some mechanism to degrade FA, this could be suppressed in the consortium with Gl006, since FA was detected in these treatments. Some studies have found that Trichoderma harzianum can tolerate FA at 500 ppm but cannot reduce the toxin to any significant extent. Moreover, FA can reduce chitinases produced by T. harzianum (Sharma et al., 2013). In this study the concentration of chitinases produced by Gl006 was high in coculture with Bs006, where FA was detected at 1.1 ppm Thus FA did not affect chitinase activity by Gl006 and it could tolerate FA at this concentration in the presence of Bs006.

Induced systemic resistance is associated with BCA in soil pathogen control (Pieterse et al., 2014; Poveda & Eugui, 2022). In this study, the three BCAs cells-based treatments decreased the incidence of cape gooseberry vascular wilt in the split root model. However, there was no increased control of the disease using the consortium treatment compared to the single strain treatments. Gl006 was more effective against the disease, which could be due to the Gl006-elicitors that can induce defense responses in the plants (Djonović et al., 2006). In the case of T. virens specific elicitors of defense like Sm1 (a protein elicitor of non-enzymatic origin) has been reported in corn against Colletotrichum graminicola and is secreted in early stages of the interaction, activating the production of jasmonic acid (Djonović et al., 2006).

The induced resistance due to Bacillus spp. has been associated with the volatile compound 2,3-butanediol (Choudhary et al., 2007). The induced response is elicited by different microorganisms and therefore there are several routes promoted in the plant. In case of Bacillus spp. strains the activated route is dependent on the strain and the host plant, and in some cases, also on the target pathogen (Choudhary & Johri, 2009). Cyclic lipopeptides such as surfactin also have been associated with this mode of action (Ongena & Jacques, 2008), which were detected in the culture broth of Bs006.

In the present study B. velezensis was isolated from the inner tissue of cape gooseberry, which is consistent with our previous report (Moreno-Velandia et al., 2019). Beneficial endophytic microorganisms commonly enter the root through natural openings of emerged lateral roots or using root hairs and the apical zone as entry points, using cell wall degrading enzymes (Pieterse et al., 2014). However, it is not unreasonable to think that the application of cells and supernatant of Bs006 could facilitate the entry of Foph, because of the cellulases in the supernatant from Bs006 liquid culture (Di Pietro et al., 2003). However, we also found that when Bs006 was applied in consortium with Gl006, no colonies of Foph were obtained from the inner tissues of the plant, which could be attributed to the production of phenolic compounds that inhibit cellulases (Ximenes et al., 2011). Thus, it is well known that Trichoderma spp. produces phenolic compounds or can induce its synthesis in the host plant (Nawrocka & Malolepsza, 2013). Considering that in solid medium Gl006 inhibited cellulase activity by Bs006, and in liquid co-culture the cellulase activity significantly decreased, these phenolic compounds could inhibit cellulases in the liquid culture of Bs006 with Gl006 or those synthesized by Foph, blocking the entry of Foph to plant tissue.

These results suggest that synergistic or additive effects of the consortium could be attributed to a combination of direct and indirect modes of action. Direct modes of action are mainly involved with the synergistic effect of the consortium, considering the significant reduction of population density of Foph in the soil. Previous studies reported that the supernatant from 48 h-old culture of Bs006 contains the CLPs surfactins, iturins and fengycins. Fengycins showed the strongest direct action against Foph-Map 5 disrupting the integrity of the cell wall (Moreno-Velandia et al., 2021) which was reported also by Ongena and Jacques (2008). It can be assumed that these mechanisms are also present in the consortium of Bs006 and Gl006 since fengycins and surfactins were detected in the co-culture. The two BCA evaluated in this study, alone and in consortium, induced resistance in cape gooseberry against Foph.

We found that in soil the enzyme activity was minimal compared to the co-culture in minimal basal medium supplemented with the mycelium of Foph, but when we evaluated the medium with root exudates we found values closer to those found in planta which is why these media would be used to evaluate enzyme activity under laboratory conditions.

The predominant mode of action of B. velezensis Bs006 controlling Foph could be indirect as an inducer of resistance, since it was observed that its application individually was not effective in decreasing the viable population density of Foph (Fusarium oxysporum f. sp. physalis) In the case of T. virens Gl006, a promising mechanism was observed concerning to indirect modes of action of cells and supernatant as an inducer of resistance. Its effectiveness was maintained when it was applied simultaneously with Foph into the soil (Izquierdo-García et al., 2020) but Gl006 cannot control the population density of Foph. Direct and indirect modes of action could be complementary in the control of Foph.

These results show that the synergistic effects by the consortium was presented only in decreasing the population density of Foph in the soil of the in planta bioassay. Although, we did not evaluate CLPs in soil, we found that the consortium is capable of synthesizing fengycin and surfactin in all media evaluated, so we could suggest that these CLP’s could be produced by Bs006 under in vivo conditions as well. The consortium showed control of Foph related to the induction of resistance in the split root model. The consortium has direct and indirect modes of action which may be complementary and are translated into a synergistic effect controlling the population density of Foph in the soil.