Study of Pathogenicity Test, Antifungal Activity, and Secondary Metabolites of Bacillus spp. from Lake Bogoria as Biocontrol of Rhizoctonia solani Kühn in Phaseolus vulgaris L.

The common bean (Phaseolus vulgaris L.) is a yearly herbaceous plant grown for its edible dry seeds. Despite that, pests and diseases have contributed to the decline of common bean production in Kenya. Therefore, the study aimed to identify bacteria from Lake Bogoria, assess the pathogenicity of Rhizoctonia solani Kühn, screen for effective antifungal agents, and determine secondary metabolites for the biocontrol of R. solani. A total of 49 bacteria were isolated, of which 10 isolates had varied mycelial inhibition rates of R. solani in the co-culture technique. The efficacy of volatile compounds of the three selected bacterial strains had varied mycelial growth and percent reduction against R. solani. The pathogenicity assay showed varied plant parameters and biomass of R. solani on common bean plantlets. The molecular characterization based on 16 S ribosomal RNA confirmed the selected bacterial strains' identity with a diversity similar to the Bacillus genus. Gas chromatography-mass spectrometry analysis of secondary metabolites showed different antimicrobial compounds produced by Bacillus subtilis strain TW21. In conclusion, Lake Bogoria harbors useful microbes as biocontrol agents against plant pathogens. The current study discovers the potential biocontrol bacteria isolates from Lake Bogoria as alternative bioagents against R. solani. Therefore, the isolate Bacillus subtilis strain TW21 can be assessed further for toxicological and ecotoxicological studies and registered by the Pest Control Products Board (PCPB), Kenya, as a biocontrol product against common diseases affecting common beans' production.


Introduction
Te common bean (Phaseolus vulgaris L.) is a herbaceous plant grown for its edible dry seeds [1].It is categorized as dry beans, snap beans, and shell beans, with its leaves used as vegetables and fodder for animals [2].Phaseolus vulgaris L is a member of the legumes family Fabaceae with a mutualistic connection with the nitrogen-fxing bacteria Rhizobium [1].Kenya produces about 500,000 metric tons (Mt) of common beans annually Duku et al. [2].Despite that, pests and diseases have contributed to the decline of common beans' production.Pests such as cutworms, bean fies, red spider mites, aphids, pod borers, whitefies, and thrips are among the major factors afecting common beans' production [3].In addition, diseases such as black root rot, damping-of diseases, bean rust, Fusarium wilt, and Rhizoctonia root rot contribute to approximately >70% loss of common bean production [4,5].Biotic factors such as soil infertility and environmental stress also contribute to common beans' low productivity [5].According to Jabnoun-Khiareddine [4], a decline of >13.8% of common beans' production in 2014 was attributed to root and stem rot disease caused by Rhizoctonia fungi.Most Kenyan farmers for over a century have relied on synthetic fungicides as the best alternative in managing plant diseases [6].Tese have been compromised, due to the inefectiveness of fungicides and plant pathogens' adaptation mechanisms [7].In addition, the efect of synthetic products in terms of environmental pollution leaves harmful residues on the plant, hence, expand resilient pathogen strains' repertoire.Tis has initiated the need for other environmentally friendly options in managing plant diseases [8].Terefore, the implementation of integrated pest managment (IPM) in the use of biological control as a substitute management strategy has been encouraged [5,[9][10][11][12].In Kenya, various products are registered by the Pest Control Products Board (PCPB) to be used as biopesticides.For instance, Bacillus thuringiensis var: aizawai sero H7 subtype is used to control thrips and whitefies on French beans, Bacillus subtilis BS-01 is used to control powdery mildew on roses and rice blast in rice, and Bacillus amyloliquefaciens strain QST 713 is used to control cofee leaf rust, black rot, and Alternaria leaf spot in cabbages.In addition, Wekesa et al. [13] reported Bacillus spp.from Lake Magadi, Kenya, to control Rhizoctonia solani and also the use of Bacillus velezensis from Lake Bogoria to control Fusarium solani in common beans [14].
Te antagonism of biological agents against phytopathogenic fungi are mainly through the exudation of active metabolites [15], competition for nutrients [16], induced systemic resistance [17], production of cell wall degrading enzymes [12], and production of microbial volatile compounds (mVOCs) [18,19].Lake Bogoria is characterized as a thermal lake due to its hot springs (>50 °C).Te uniqueness of the lake over the period has been an exploitation center for industrial and agricultural beneft.For instance, endophytic bacteria against Fusarium solani [20] and Bacillus velezensis against Fusarium solani [14] have been characterized from Lake Bogoria.In addition, diferent strains from soda lakes have been reported with the potential to manage plant diseases [21][22][23][24].Te study aimed to isolate bacteria from Lake Bogoria, assess the pathogenicity of R. solani in common bean plants, and screen efective antifungal agents to control R. solani.Te efective strains were assayed for the efcacy of producing volatile compounds and for molecular characterization.Te secondary metabolites produced by the selected antagonistic bacterial strains were identifed using the gas chromatography/mass spectrometry (GC-MS) technique.

Sampling Site and Sample Collection.
Te samples were collected at fve diferent sites of Lake Bogoria, where 100 ml of water and 50 g of soil were collected in triplicate and separately packed in sterile tubes.Te physiochemical parameters such as temperature, pH, salt, total dissolved solids (TDS), and conductivity were recorded.Te samples were labelled and stored at 4 °C at the Institute for Biotechnology Research Laboratory at Jomo Kenyatta University of Agriculture and Technology for isolation, identifcation, and screening.

Isolation of
Bacteria from Lake Bogoria.Te isolation of bacteria from the soil and water was done according to Wekesa et al. [14].In brief, 9 millilitres of sterile physiological saline (0.85% NaCl) were added to a sterile test tube in which 1 gram of soil was added and then homogenized.After that, the resultant suspensions were vortexed at 150 rpm for one minute.Te suspensions were serial diluted from a concentration ratio of 1 : 9 to a dilution of 10 −4 .On modifed nutrient agar-Himedia, [10.0 g/L peptone, 10.0 g/L beef extract, 5.0 g/L NaCl, and 12.0 g/L agar, 0.01 mg/L cycloheximide, 0.35% (w/v) NaCl] an aliquot of 30 µl was cultured from dilutions of 10 −3 and 10 −4 , according to Hartman [25].Te plates were incubated at 35 °C for 48 hours, and based on colony morphology, distinct colonies were separated and cultured on an isolation medium to obtain pure cultures.Te pure bacteria cultures were characterized using standard microbiological techniques.Cell morphology was done using the Gram stain technique as described by Tripathi and Sapra [26] using a light microscope (MD827S30L).

Culturing and Inoculum Preparation.
Te fungal strain used in this study was Rhizoctonia solani Kühn (ATCC 66150), obtained from Bioline Agrosciences African Limited, Naivasha, Kenya.Te fungal strain was subcultured on potato dextrose agar-Himedia [200.0g/LPotatoes, 20.0 g/L Dextrose and 15.0 g/L Agar] (PDA) medium and incubated at 30 °C for 7 days.Te R. solani inoculum preparation was done as described by Jabnoun-Khiareddine [4].In brief, the mycelium of R. solani was scrapped of from the 7 days old culture from PDA plate and mixed with sterilized distilled water (SDW).It was then stored at 4 °C for further experiment.

In Vitro Assay of Pathogenicity
Test.In vitro pathogenicity assay was done as described by Asaka and Shoda [27].In brief, the forest soil was autoclaved for 60 mins at 121 °C thrice at 12 hours intervals.Approximately 180 g of sterilized soil was packaged in sterilized plastic containers of maximum water-holding capacity.Five days before sowing the common bean seeds, the soil was inoculated with R. solani at 5 : 1 (5 mL of R. solani inoculum: 1 pot), whereby SDW was used as a negative control.Te plant assay was done as reported by Rocha et al. [28], where seeds were surface sterilized by frst socking with 70% (v/v) of ethanol for 5 minutes and then 0.5% of sodium hypochlorite (NaOCl) for one minute.Two seeds were planted in each pot after being washed three times with SDW to remove NaOCl and air dried.Tey were placed in the growth chamber at 8 hours a day and 16 hours a night at 28 °C with 80% relative humidity.Te severity of R. solani symptoms was evaluated two weeks after the initial inoculum using a scale from 1 to 7, as described by Godoy et al. [29].In addition, plant parameters such as shoot length, root length, plant height, and biomass were measured and recorded.

2
International Journal of Microbiology

Antibiosis Assay of Bacterial Isolates against R. solani.
Te antibiosis assay of bacterial isolates was assessed against a test pathogen as described by Aydi et al. [10].Te PDA medium was used to culture fungus and bacteria by placing an active growing plug at the end of the petri dish and streaking bacteria (10 8 CFU/mL) perpendicularly across the plate.In contrast, the pathogen alone was used as a control.Te plates were incubated at 30 °C for 14 days, and the percentage inhibition rate (%I.R) was calculated using the formula as described by Aydi et al. [10].Te mycelium length was measured using a ruler in centimeter (cm).
where I.R is the inhibition rate, C2 is the Colony diameter of the pathogen in control, and C1 is the colony diameter of the pathogen cocultured with bacteria.

Efect of Volatile Compounds by
Selected Bacteria against R. solani.To determine the efcacy of the selected bacteria to produce VOCs, paired disc technique was used as described by Nigris et al. [30] and Nikitin et al. [31].Te selected bacteria were uniformly spread onto Nutrient Agar (N.A)-Himedia medium and a plug (6 mm) of active growing R. solani were punched and placed at the center of the PDA medium plate.Te two plates were sandwiched, whereby PDA medium plate with fungi were placed at the bottom and N.A medium plate with bacteria placed on top.Tey were sealed and incubated at 28 °C for 7 days.Te percentage suppression of the pathogen was determined as described by Garbelotto et al. [32].

Molecular Characterization of the Selected Bacterial
Isolates.Te antagonistic bacterial isolates were identifed by amplifying and sequencing the 16 S ribosomal RNA (rRNA) gene.Genomic DNA was extracted using a bacterial DNA isolation kit (Norgen Biotek Corp. Torold, ON, Canada) according to the manufacturer's instructions.Te genomic DNA was then quantifed using Qubit (Qubit 4 Fluorometric quantifcation, Q33238), and its fnal concentration was adjusted to 100 ng of DNA/µl.Te universal bacterial primers used for the amplifcation were 27 F (5′-AGAGTTTGATCCTGGCTCAG. 3′) and 1492.R (5′-CGGCTACCTTGTTACGACTT-3′).Te Peqlab Primus 96 thermocycler (PEQLAB, Erlangen, Germany) was used for PCR amplifcation.Te PCR solution was prepared by mixing 20.0 μl of master mix, 18.2 μl of PCR water, (10.0 ppm) of 0.4 μl of the 27-F primer, (10 ppm) of 0.4 μl of the 1492-R primer, and 1.0 μl of template DNA (10.0ng/mL) in a fnal volume of 40.0 μl.Te reaction mixtures were subjected to temperature cycling profles as described by Wekesa et al. [14].Te PCR products were confrmed by running a 1.5% agarose gel that was viewed under a UV gel documentation system.Te PCR products were then purifed using QIAquick PCR amplifcation kit (Qiagen, Hilden, Germany), and sent to Macrogen (South Korea) for sequencing.

Sequence Analysis and Phylogenetic Tree Preparation.
ChromasPro V2.0 software was used to edit the Sanger sequencing data of the bacterial isolates, and a nucleotide BLAST search performed on the National Center for Biotechnology Information (NCBI) database.Te sequences were selected based on their similarity to reference genome sequences retrieved from NCBI, and multiple alignments were performed using the MAFT plug-in for Geneious PrimeV2023.1.2software.Maximum likelihood analysis was used to construct the phylogenetic tree for datasets in GeneiousPrime 2023.1.2RAxML plug-in by looking for the highest-scoring ML tree among 1000 iterations under the GTR-gamma model with rapid bootstrapping.Te sequences were submitted to the GenBank with accession PP301463, PP301464, and PP301465.

Extraction of Secondary Metabolites from Efective Bacillus spp.
Following the method outlined by Yehia et al. [33], the crude metabolites were extracted, whereby Bacillus subtilis strain TW21 was cultured in nutrient broth [10.0 g/L Peptone, 10.0 g/L Beef extract, 5.0 g/L NaCl] (N.B)-Himedia, at 28 °C for 3 days.Te supernatant was extracted by centrifuging the sample at 8000 rpm for 30 minutes at 28 °C.Te extracted supernatant was stirred for 8 hours at 28 °C and 100 rpm in an orbital shaker, and then acidifed to a pH of 2.0 using concentrated HCl.Te antifungal chemicals were extracted by adding an equivalent amount of ethyl acetate to the culture broth and shaken for 2 hours at 200 rpm in an orbital shaker.Te culture broth was extracted twice using ethyl acetate and evaporated at 60 °C and 80 rpm to obtain a concentrated antifungal crude extract.Te crude extract was dissolved in 1 ml of methanol: chloroform mixture (1 : 1) and analyzed on GC/MS machine.

Characterization of Secondary Metabolites by GC-MS.
Te Shimadzu QP-2010 GC-MS (Kyoto, Japan) equipment with a capillary column (inner diameter 0.25 mm and length 30 m) was used to analyze the crude extract of Bacillus subtilis strain TW21.Te GC oven was preheated to 100 °C for 2 minutes, then set to rise to 280 °C at a rate of 10 °C per minute, and fnally held at 280 °C for 13 minutes.A volume of 2 mL of crude extract of Bacillus subtilis strain TW21 was injected into a split ratio of 1 : 0.25 yielding a total volume of 0.5 mL of pure extract.Both the injector and detector ports were heated to a comfortable 200 °C.Te energy level for electron ionization in a GC-MS was 150 eV.Te mass range examined was 20-500 amu at a scan duration of 70 ms.Mass spectrometry was used to analyze the gas chromatogram peaks.Te active ingredients were determined using retention indices and mass spectra compared to the library of mass spectra maintained by the National Institute of Standards and Technology.

Data Analysis.
Te pathogenicity and coculture data were analyzed using statistical analysis software (SAS) version 8.0 software, while all graphs were analyzed using GraphPad-Prism version 6.0.
International Journal of Microbiology

Isolation of Bacteria.
A total of 49 bacterial isolates were obtained from 5 sampling points of Lake Bogoria.Te isolates were characterized by colony and cell morphology (Figure 1).Te shape of colonies with round had 29 isolates, irregular (11), and penctiform (9) isolates.In addition, the elevation of colonies recorded 32 isolates with raised elevation, while 17 were fat.Te colony margin varied from entire (19), wavy (6), lobate (4), irregular (5), and flamentous (3).Te colony size in millimetre showed that most of the isolates had medium size (28), followed by large (11), and small (10).Te colony's surface varied from smooth to rough, with the largest number having a smooth surface.Tere were more Gram-positive (34) isolates than Gram-negative (15) isolates (Figure 1).

Test for Pathogenicity of Rhizoctonia solani.
Tere was no necrotic lesion detected on the roots and stem of control treatment (Figure 2(b)), whereas the R. solani treatment exhibited lesions ranging from up to 30 mm in length to a large discoloured dry lesion (Figure 2(c)).In addition, the roots were long and healthy in the control treatment compared to R. solani treatment (Figure 2).Furthermore, the germination rate of R. solani was not aggressive, with the lowest seed germination rate (63.00 ± 0.58%b), compared to the control (94.00 ± 0.58%a).Te fungal treatments significantly afected the plant height of common bean plantlets compared to the control.Tere was a signifcant diference (P < 0.05) amongst R. solani and control in plant parameter and biomass (Table 1).
Te common bean plantlet biomass difered signifcantly (P < 0.05) upon treatments, with R. solani treatment having lower biomass weights than the control.From the results (Table 1), there was a signifcant diference between test fungal and control in plant biomass.Te severity of the common bean plantlets indicated that R. solani had the highest number of dead plants (37.50 ± 1.20b%) compared to the control (0.00 ± 0.00a%) (Table 1).

Antibiosis of Bacterial Isolates against Selected Test
Pathogen.A clear zone of inhibition was observed in ten out of forty-nine isolated bacteria against R. solani, indicating the antifungal ability of the bacterial isolates when in direct contact (Table 2).However, the inhibition rate difered depending on the isolate interacting with R. solani.Isolates mainly drove the variation, as post hoc analysis revealed variation among the isolates (Figure 3).In contrast, much higher variation was observed amongst the isolates (F 10 �124.72,P � 0.05, Table 2), and ANOVA analysis showed that BW21 was the most inhibitory strain, followed by BW07 and BW20, which showed modest levels of inhibition (P < 0.05).Te isolate BW35 had the smallest inhibition zones against R. solani (P < 0.05).Te results also suggest that Lake Bogoria isolates inhibit the mycelium growth of R. solani in vitro, which varied amongst the isolates (Table 2).Te bacterial isolates with >30.00% inhibition rate were selected for further analysis.

Efect of Volatile Compounds by
Selected Bacteria against R. solani.Te production of VOCs of the selected bacterial isolates showed varied signifcant diferences in the length of the mycelium growth across the treatments.However, the control had maximum mycelium growth of 84.00 mm with the lowest percentage reduction of 0.00% (Figure 4).Isolate BW021 had the lowest mycelium growth (41.00 mm) and the highest percentage reduction of 51.19%.Isolate BW07, on the other hand, recorded the same mycelium growth (Figure 4).Isolate BW07 recorded slightly lower mycelium growth compared to BW20; however, it had a higher percentage reduction compared to isolates BW07 and BW21.

Molecular Characterization of Selected Antagonistic
Isolates.Te molecular sequencing performed on three selected bacterial isolates showed the highest similarity identity >99.77% with the Bacillus genus (Figure 5).Among the identifed strains are Bacillus sp. and Bacillus velezensis (Figure 5).Te phylogenetic analysis of the isolates classifed into two groups.Te frst group comprises B7 and B21, afliated with Bacillus sp. and Bacillus subtilis.Te second group comprises B20, afliated with Bacillus velezensis strain QH03-23, with a similarity index of 99.89% (Table 3).

Discussion
Control of R. solani has faced many challenges since no lasting control strategies from synthetic, semisynthetic, or biological products are currently used for their control and management.Tese pathogens attack more than 500 legume species contributing to low-yield production among small-scale farmers [7,45].Te results from pathogenicity showed that R. solani afected common bean plantlet germination rates, severity, plant biomass, and plant parameter.In addition, a reddish-brown lesion was observed on the plantlets' roots indicating the efect of R. solani on the bean plantlets.Tese fndings agree with previous studies on the pathogenicity of R. solani, which is associated with stunting 4 International Journal of Microbiology growth due to the ability of the pathogen to afect the plant's roots limiting uptake of nutrients [4,5,10,46,47].
Te BLAST result based on 16 S rRNA showed taxonomic classes of the isolated Bacillus strains.Tese fndings agree with Lake Bogoria's earlier investigations, which found that Firmicutes are the preponderant bacterial kingdom [23].In addition, Bacillus spp. is one of the most common aerobic, eubacterial alkaliphiles in soda lakes and other natural environments [48].
Bacteria isolates obtained from Lake Bogoria showed high inhibition activity on the mycelial growth of R. solani in coculturing and by production of volatile compounds.Te mechanism can be due to the production of various lytic enzymes involved in cell degradation during antagonism [12].Some isolates had high inhibitory activity, while others showed limited activity, indicating the types of antifungal metabolites produced may vary [5].
Diferent bacteria, such as Bacillus spp., are well known to produce various antibiotics biocontrol agents of plant diseases.In addition, endophytes bacteria have been reported to have antibiosis efects against R. solani [28] using antibiosis as the most important mechanism to limit plant pathogen invasion.It also inhibited the development of plant pathogenic organisms by producing secondary metabolites [12,49].Diferent studies have been carried out in biological control, aiming to fnd the best solution for the control and   Te data are average of three replicates.Results of the comparative efect of R. solani are shown as mean values (±SE).Following signifcant one-away ANOVA, subsequent Tukey's honest signifcant diference (HSD) test at P < 0.05.Means in a column followed by the same letter do not signifcantly difer.
International Journal of Microbiology Te percent inhibition rate (%IR) was calculated after 14 days.Mean values (±SE) in a column followed by the same letter do not signifcantly difer according to the Tukey HSD test (P < 0.05).Te data are an average of three replicates.International Journal of Microbiology management of R. solani.For instance, Belete [50] reported native Bacillus isolates to eradicate black root rot diseases initiated by F. solani in faba beans; Jabnoun-Khiareddine [4] reported the application of fungal and bacterial agents to control root rot disease in pepper.In addition, Aydi et al. [10] reported using endophytic bacteria from Datura stramonium to manage Fusarium wilt disease in tomatoes, and lastly, Mahmoudi and Naderi [11] reported the antifungal and biocontrol properties of chitinolytic bacteria in control of Fusarium root rot in safower and the impact of biocontrol on Rhizoctonia diseases on potatoes.Lastly, Chen et al. [51] reported the efect of bichar and Bacillus subtilis to efectively reduce disease incidence and disease index in radish plants afected by R. solani.Te production of VOCs has attracted growing attention as a biocontrol mechanism since there is interaction between microorganisms and the environment.From our fndings, the mycelium growth of R. solani was inhibited by bacterial strain TW7, TW20, and TW21.Tis is consistent with the previous study of bacterial VOCs inhibiting fungal and bacterial plant pathogens [52][53][54].Te study on Bacillus VOCs antagonistically against phytopathogen has been also reported by Bruisson et al. [55].Other research has reported Bacillus spp.ACB-65 and Bacillus spp.ACB-73 produced volatile chemicals that inhibited Phyllosticta citricarpa by 86% [56].
In this work, secondary compounds such as pyrrolo (1,2-a) pyrazine-1,4-dione, hexahydro-, 9-octadecenol, 1-propanol, 2,2-dimethyl-acetate, butanoic acid, 2methyl-, N, NDimethyl, 3-heptanone, 5-ethyl-4-methyl-, phenol, and benzoic acid were among the most signifcant antifungal chemicals found.Te fndings agree with Surya et al. [57], who reported the antimicrobial activity of fatty acid salts of N-N-dimethyl.Te research done by Bharose and Gajera [58] also reported B. subtilis strain JNDKHGn-29-A to produce antifungal metabolites such as bis (2-ethylhexyl) phthalate and pyrrolo [1,2-a]pyrazine-1,4-dione. Our fndings are also consistent with the fndings by Wu et al. [59], who found that Bacillus spp.produce secondary metabolites with antifungal activity such as bacillomycin, fengycin, iturin and sufactin in control of R. solani in peppers.According Wu et al. [59], these compounds afect the spore germination and membrane permeability of Fusarium, Rhizoctonia, and Alternaria.Te extraction of the lipoprotein has been reported to have a strong inhibitory efect on the growth Figure 5: Te evolutionary relationship of the selected bacterial isolates using maximum likelihood-based in the GTR-gamma model.A total of ten nucleotide sequences were selected based on their similarity to reference genome sequences retrieved from NCBI, and multiple alignments were performed using the MAFT plug-in.Maximum likelihood analysis was used to construct the tree using RAxML plug-in with rapid bootstrapping of 1000 iterations under the GTR-gamma model.A scale bar of 0.02 was used.8 International Journal of Microbiology of R. solani [30].Finally, Li et al. [60] reported Bacillus subtilis SL-44 to produce L,D-transpeptidase and D-alanine carboxypeptidase which play an important role in the peptidoglycan cross-linking and also glutamate which are used for cell wall synthesis.In addition, the metaboic analysis of Bacillus subtilis SL-44 showed the presence of athyl ester, L-pyroglutamic acid, and L-alanosine which are used to synthesise drugs in control of fungal diseases.Our results also agrees with Jangir et al. [61] who found antifungal activity against Fusarium oxysporum from compounds in Bacillus sp., including N, N-dimethyl-1,2-benzene dicarboxylic acid and 9-octadecenoic acid.

Conclusion
Lake Bogoria harbors diverse microbes with diferent morphological characteristics.Te pathogenicity test showed the ability of R. solani to illicit necrotic lesions and impact on plant biomass.Te antifungal activity indicated the ability of the bacteria isolates to inhibit the growth of R. solani in vitro and capability to suppress the mycelium through production of secondary metabolites.Tis has been quantifed through GC-MS and identifed metabolite compounds with both antifungal and antibacterial activities.Terefore, further research is required to address the toxicological and ecotoxicological studies of Bacillus subtilis TW21 to be registered as a biocontrol product.

Figure 1 :
Figure 1: Morphological characteristics of the isolated bacterial colonies.

Table 1 :
Comparative efects of R. solani on common bean plantlets observed 14 days of inoculation.

Table 2 :
Antifungal activity of Lake Bogoria bacterial isolates on mycelium growth of R. solani.

Table 3 :
Te molecular identifcation based on BLAST analysis of 16S rRNA genes sequencing with corresponding GenBank accession number of 3 selected bacterial strains.

Table 4 :
Te retention times (RT), chromatographic relative area percentages, molecular weight, and functional activity of the secondary metabolites generated from Bacillus subtilis strain TW21 identifed by gas chromatography/mass spectrometry solvent-free solid injection.