Effects of biocontrol Bacillus sp. strain D5 on the pathogenic Fusarium oxysporum R1 at the microscopic and molecular level in Crocus sativus L. (saffron) corm

Abstract Corm rot of saffron caused by Fusarium oxysporum is a major threat to saffron cultivation the world over. To minimize the ill effects of chemical fungicides, attention has been shifted to the use of biocontrol agents for disease management in a sustainable way. In saffron, various biocontrol agents against corm rot disease have been reported and characterized but no study has been done so far to understand their interaction at the molecular level. The present study was conducted to unravel the mechanism of action of an already characterized native biocontrol agent i.e. Bacillus sp. strain D5 (Bar D5) against F. oxsporum R1 (Fox R1) in the saffron corm. The growth inhibition of Fox R1 was observed in vitro and in planta (saffron corm) by real time imaging. Bacillus sp. strain D5 reduced Fox R1 load in infected corms by 50% as quantified by q-PCR and the colony-forming unit method. Comparative transcriptome analysis revealed upregulation and downregulation of various Fox R1 genes in presence of Bar D5. The genes related to carbon metabolism, cell wall and membrane synthesis, and growth of Fox R1 were significantly downregulated in Bar D5-primed and Fox R1-inoculated corms as compared to only Fox R1-inoculated corms.


Introduction
Saffron ( Crocus sativus L.) is an economically important crop that produces the costliest spice in the world (Mansotra et al. 2023, Spinelli et al. 2023 ).Due to the presence of active compounds such as cr ocin, picr ocr ocin, safr anal, and others in the stigma of saffr on, it is widel y used as a spice and has extensiv e a pplication in the pharmaceuticals , cosmetics , and food industries (Kothari et al. 2021, Ali et al. 2023 ).Recently, it has also been demonstr ated to hav e anti-inflammatory and antivir al potential a gainst the COVID-19 virus (Mentis et al. 2021 ).Other parts of the saffr on plant, suc h as leav es, tunics, spaths, tepals, and petals have also been used in pharmaceuticals (Naim et al. 2022a,b , Lahmass et al. 2017 ).Saffr on pr oduction in Jamm u and Kashmir, India is on the constant decline.In 1997, total production was 15 tonnes, which was reduced to 9 tonnes in 2015, and further reduced to 3.8 tonnes in 2021 (Ganaie andSingh 2019 , Kumar et al. 2022 ).The major reason for this decline is the unavailability of healthy cultivable corms that are free from wound or injury, as saffron is v egetativ el y pr opa gated thr ough corms (Gupta et al. 2021 ).
Corm rot of saffron caused by pathogenic Fusarium oxysporum is the major disease that affects saffron production (Magotra et al. 2021, Bhagat et al. 2022 ).During corm rot disease progression, the pathogen enters the corm through injuries or wounds, multiplies inside the corm, and moves to the fibrous roots, ultimately leading to plant death (Bhagat et al. 2022, Mansotra et al. 2023 ).At pr esent, c hemical measur es ar e used as a pr e v entativ e str ategy for the management of disease due to their easy application and broad spectrum (Gupta et al. 2020 ).The ill effects of these chemicals on the environment as well as humans are well documented (Ma gotr a et al. 2021 ).The excessive use of these chemicals also affects the native microflora, which are important for maintaining the fertility and health of the soil (Pahalvi et al. 2021 ).
The use of native biocontrol agents has become a promising strategy for the control of soil borne diseases and is an ecofriendly a ppr oac h (Ram et al. 2018 ).Among different bacterial genera, Bacillus has gained m uc h importance as a potential biocontrol agent due to the formation of heat and desiccation-resistant spores (Kavitha et al. 2023, Singh et al. 2023 ).Various Bacillus str ains hav e been isolated and c har acterized fr om saffr on plants and fields by our group as plant growth promoting bacteria and biocontr ol a gents a gainst Fox R1, suc h as Bacillus am yloliquef aciens W2 (Gupta and Vakhlu 2015 ), a consortium of three bacilli, i.e.Bacillus thuringiensis DC1, Bacillus megaterium VC3, and B. amyloliquefaciens DC8 (Kour et al. 2018 ) and Bacillus sp.strain D5 (Ma gotr a et al. 2021 ).Among these, Bacillus sp.strain D5 (Bar D5) isolated from the saffron cormosphere was found to be the best, in terms of saffr on gr owth pr omotion and biocontr ol activity a gainst Fox R1 in pots as well as in fields (Ma gotr a et al. 2021 ).Mor eov er, the dr aft genome sequence of Bar D5 r e v ealed the presence of genes related to str ess toler ance and pr oduction of antifungal VOCs (Ma gotr a et al. 2021 ).
The mechanism of antagonism to F. oxysporum by any Bacillus strain or any other biocontrol agent has not been deciphered so far in saffron.In the present study, the antagonist mechanism of action of the Bar D5 antagonist to Fox R1 was studied in planta in real time.Further, the analysis of differentially expressed genes (DEGs) of Fox R1 in the presence of Bar D5 r e v ealed the molecular mechanism of decreased infectivity of Fox R1 to w ar d the saffron plant.Further, Bar D5 was demonstrated to have a long-term inhibition against Fox R1 in comparison to the commonly used c hemical fungicide, carbendazim; ther efor e, Bar D5 can be used to replace these chemical agents in the future.

Corm sample, bacterial and fungal strains, and culture conditions
Saffron corms were procured from the Pampore region (34.0060• N, 74.9238 • E) of the Pulwama district of the Kashmir r egion, Jamm u and Kashmir, during its dormant phase in August 2020.
Already isolated wild type pathogenic F. oxysporum R1 (GenBank accession number of the ITS gene: KF663598 ) causing corm rot in saffr on and EGFP-ta gged F. oxysporum R1 was used in the present study (Gupta andVakhlu 2015 , Bhagat et al. 2022 ).The wild type Fox R1 was cultivated on potato dextrose agar (PDA) plates (Difco, BD Becton, and Dickinson) and EGFP-tagged Fox R1 was cultivated on PDA plates supplemented with 100 μg/ml of hygromycin-B and incubated at 25 • C for 5 days in the Orbitek incubator (Scigenics Biotech, India).Bacillus sp.strain D5 (GenBank accession number of the 16S rRNA gene: KT228251 ) was pr e viousl y isolated from the cormospher e (Ma gotr a et al. 2021 ).Bar D5 was cultur ed on Nutrient Agar (NA; Himedia) plates and incubated at 37 • C for 24 hours in an Orbitek incubator (Scigenics Biotech).

Dual culture method
The antifungal activity of Bar D5 was tested by dual culture method (Ji et al. 2013 ).In brief, 5 mm agar plug of Fox R1 from a 7-day-old culture was taken and placed on one edge of the Muller Hinton agar plates and on another edge , o vernight grown Bar D5 cultur e was str eaked at a distance of 5 cm.The control plates were without Bar D5 culture .T he plates were incubated at 28 ± 1 • C for 5 da ys .T he photogr a phs of the plates were subjected to ImageJ software for measuring the diameter of Fox R1 in both control and test plates (Abramoff et al. 2004 ).The inhibition percentage was calculated using the formula:

Effect of the cell-free extract of Bar D5 on morphology and growth of Fox R1
The effect of Bar D5 cell-free extract was studied on Fox R1 morphology by visualizing the Fox R1 hyphae under the light microscope (Gupta and Vakhlu 2015 ).Bar D5 was gr own ov ernight in nutrient broth (NB; Himedia) and cell-free extract was collected by centrifuging the broth at 10,000 g for 10 minutes at 4 • C. A volume of 2 ml of cell-free extract was mixed with 8 ml of PDB, and Fox R1 spor es wer e inoculated in it.For control, instead of cell-free extract 2 ml of sterile distilled water was added to 8 ml of PDB.Both test tubes were incubated at 28 • C in the Orbitek incubator shaker (Scigenics Biotech) for 10 days at 180 rpm.The morphology of Fox R1 in control and test was observed under the light microscope using lactophenol cotton blue dye after 10 days of incubation.

Effect of VOCs of Bar D5 on morphology and growth of Fox R1
The production of VOCs by Bar D5 and its antifungal potential w as determined b y the seal plate method (Fernando et al. 2005 ).The Bar D5 was streaked on the bottom dish of the NA plate and incubated at 37 • C overnight in an Orbitek incubator (Scigenics Biotech).The 5 mm plug of Fox R1 was placed at the center of the bottom of the PDA dish and inverted over the NA plate containing Bar D5 culture; both dishes were sealed together with 3 layers of parafilm.The control plates were inoculated with Fox R1 and other plate was uninoculated.The plates were incubated at 28 • C for 7 days in the Orbitek incubator (Scigenics Biotech).After 7 days, the diameter of growth of Fox R1 in the control and test plates was determined using Ima geJ softwar e (Abr amoff et al. 2004 ).The growth inhibition was calculated by using the formula mentioned in the section "Dual culture method."Further, the Fox R1 hyphae from both the plates were stained with lactophenol cotton blue dye and visualized under the light microscope .T hree replicates for all the in vitro experiments were taken and each experiment was performed thrice at different time intervals.

In planta antagonism assays of Bar D5 against Fox R1
The in planta effect of Bar D5 against Fox R1 was e v aluated in pot trials .T he bioformulation of Bar D5 (10 8 CFU/ml) was pr epar ed following the standardized protocol of Magotra et al. ( 2021 ).In brief, Bar D5 was grown in NB for 48 hours and a colony count of 10 8 CFU/ml was maintained.The broth was mixed with double autoclaved calcium carbonate (talc) in a ratio of 1:2 (v/w) and dried for 3-4 days at 35-37 • C. Finally, 1% sterile carboxymethyl cellulose was added to the dried bioformulation.For the priming of saffron corms, slurry of bioformulation was prepared by mixing a 100 gm of dried bioformulation with 100 ml of sterile distilled water.
A total of 72 saffron corms ranging between 8 and 10 gm were tak en and di vided into two sets: contr ol, moc k primed and test, Bar D5 primed.A total of 24 corms were primed with Bar D5 bioform ulation, and r emaining 48 corms were mock primed with calcium carbonate only (Magotra et al. 2021 ).Both sets of corms were dried overnight at room temperature and then sown in an autoclaved sand and soil mixture (1:1 W/W) in pots of 76 mm mouth diameter with 1 corm/pot.The pots were incubated in a plant growth chamber at 25 ± 2 • C under a 16-hour light/8-hour dark cycle for 30 da ys .T he spore suspension of EGFP-labeled Fox R1 (10 12 spores/ml) was prepared as per the standardized protocol of our gr oup (Bha gat et al. 2022 ).After 30 da ys , corms were taken out and an artificial injury was given to each corm with the help of a sterile needle (2 cm × 1 cm) and placed back into their r espectiv e pots .T he 1 ml EGFP-labeled Fox R1 spore suspension (10 12 spores/ml) was added to the pot soil and the pots wer e a gain incubated at the same conditions as mentioned abov e.Further, the corms were taken out at 1, 2, 3, 4, 5, 8, 12, and 20 days postinoculation (dpi) for real time imaging and load quantification of Fox R1 by real time PCR and the colony-forming unit (CFU) method.The experiment layout has been tabulated in Table 1 .
Ta ble 1. Experiment lay out for the quantification of Fox R1 load in differ ent tr eatments using q-PCR and CFU method.

Real time imaging of EGFP-labeled Fox R1 in infected saffron corms with different treatments using confocal microscope
To visualize the effect of Bar D5 on EGFP-labeled Fox R1 inside the saffron corm tissue, the corm samples at 1, 2, 3, 4, 5, 8, 12, and 20 dpi were taken out from their respective pots and washed gently with sterile distilled water to r emov e the soil particles adhered to the corm surface and air dried for 30 minutes .T he corms were dissected into two halves and observed for visual symptoms.Further, the infected part of the corm tissues was excised manually, cut into thin sections with the help of a sterile blade, and placed on the glass slide in a drop of sterile water (Bhagat et al. 2022 ).
The tissue section was cov er ed with a cover slip and directly observed under the confocal microscope at 60X magnification.The tissue was scanned in multiple planes (Z stack) under a confocal microscope (Leica TCS SP8 AOBS, NIPGR, New Dehli).

Quantification of EGFP-labeled Fox R1 in infected saffron corms with different treatments by using CFU method
To quantify the EGFP Fox R1 in infected corm tissue with the different treatments mentioned in Table 1 , Komada medium was used (Komada 1975 ).As EGFP-labeled Fox R1 is resistant to hygromycin, the medium was supplemented with 100 μg/ml of hygromycin-B antibiotic that acted as a selectable marker for the growth of EGFP-labeled Fox R1 as wild-type Fox R1 is susceptible to this concentration of hygromycin.The infected tissues from different tr eatments wer e taken and washed with sterile distilled water under sterile conditions, inside laminar airflow.A total of 100 mg of infected tissue was crushed using a sterile pestle and mortar and then 1 ml of autoclaved distilled water was added.Serial dilution of the suspension was done, and 10 −2 dilution was spread on the Komada media plates and incubated at 28 • C for 5 days in an Orbitek incubator (Scigenics Biotech).After 5 da ys , colonies were counted and the load was determined as the number of Fox R1 colonies per gram of corm tissue in differ ent tr eatments .T he inhibition per centage w as calculated using the belo w mentioned formula.

Inhibition percentage
= ( Fox R1 colonies in the test plate / Fox R1 colonies in the control plate ) × 100 .

Quantification of EGFP-labeled Fox R1 in infected saffron corms with different treatments using q-PCR
The abundance of EGFP-labeled Fox R1 in different treatments (Table 1 ) was quantified by q-PCR using the ITS gene, which has already been characterized for its selectivity and sensitivity in our pr e vious study (Bha gat et al. 2022 ).The crocus 18S rRNA gene was used as the r efer ence gene (RG) for the normalization of data (Table 2 ).The genomic DN A w as extracted from infected tissues to obtain DNA of both Fox R1 and corm from different treatments, as shown in Table 1 at 1, 3,5,8,12, and 20 dpi using the CTAB method (Doyle and Doyle 1990 ). Relative quantification of Fox R1 was done in carbendazim-treated and Fox R1-inoculated corms (Test 1) and Bar D5-primed and Fox R1-inoculated corms (Test 2) compared to Fox R1-inoculated corms (control) at 1, 3, 5, 8, 12, and 20 dpi using a SYBR green dye-based assa y.T he PCR reaction mixture consisted of SYBR green master mix-5 μl (Thermo-Scientific), forw ar d primer-1 μl, r e v erse primer-1 μl, DNA template-1 μl, and MiliQ-2 μl for total reaction volume of 10 μl and the reaction was performed in 8-well strips on a 7500 Real Time PCR System (Applied Biosystems ®).The PCR pr ogr am used was initial denaturation at 95 • C for 10 minutes, follo w ed b y 40 cycles of denaturation at 95 • C for 15 seconds, annealing at 60 • C for 1 minute, and extension at 72 • C for 30 seconds .T he r elativ e quantification was performed using the formula: where TG-test gene, RG-reference gene, Control = corm tissue inoculated with Fox R1, Test1-carbendazime-treated and Fox R1-inoculated saffron corm, Test2-Bar D5-primed and Fox R1inoculated saffron corm.

Tr anscriptome anal ysis of F. oxysporum R1 in saffron corm
The total RNA was isolated fr om contr ol, i.e. moc k-primed corms and inoculated with Fox R1 spore suspension and test, i.e.Bar D5-primed and Fox R1-inoculated corms at 2 dpi using Trizol regent.The quality and integrity of the RN A w as checked by a gar ose gel electr ophor esis on 1.2% a gar ose gel.The integrity of the RN A w as determined b y the 260/280 absor ption r atio by nanodrop.Further, the RNA Integrity Number was determined using an Agilent 2100 Bioanalyzer (Agilent, USA).cDNA libraries wer e pr epar ed using the True-Seq R RNA sample pr epar ation kit v2 (Illumina CA, USA) following the manufacturer's instructions, and the sequencing was done using the Illumina Hiseq 2500 system (2 bp × 150 bp) with default parameters .T he ra w data obtained after sequencing was processed as follows: (1) FastQC of the r aw r eads was performed using the FastQC v0.11.5 tool (Andr e ws 2017 ).The good-quality reads with a phred score Q > 30 were selected and further processed for alignment and assembl y; (2) the r efer ence genome of F. oxysporum f .sp .lycopersici 4287 ( http:// fungi.ensembl.org/Fusarium _ oxysporum/ Info/ Index ) was downloaded from NCBI and raw reads were aligned to it using the software Hisat2 v2.1.0(Kim et al. 2015 ); (3) the aligned reads were further processed for reference-guided transcript assembly and for quantification of the transcripts using the software Stringtie v1.3.5 (Pertea et al. 2015 ).The number of fr a gments per kilobase of transcript sequence per million base pairs sequenced (FPKM) was used to estimate the expression level of each gene in all the samples; (4) the DEGs wer e anal yzed using the Bioconductor package DESeq2 v1.26.1 in the R environment (Varet et al. 2016 ); and (5) for functional annotation of the DEGs, Gene Ontology (GO) and Ky oto Enc yclopaedia of Genes and Genomes (KEGG) w as done using the g: Profiler web server (Raudvere et al. 2019 ).The putative function of all the genes was determined using the Biomart tool of ensemble fungi (Kinsella et al. 2011 ).

Validation of selected DEGs with quantitati v e re verse tr anscription PCR (RT-qPCR)
The gene sequences of 20 genes from the transcriptome data were selected and used for primer design.The primers were designed using the PrimerQuest tool of Integrated DNA Technologies ( http: // www.idtdna.com/SciTools/ SciTools .aspx).T he primer set design has been mentioned in the Table 3 .Total RN A w as extracted and.cDN A w as synthesized from 10 μg of the total RNA using highca pacity cDNA r e v erse tr anscription kit (Applied Biosystems, catalog number 4368814) following the manufacturer's protocol.The q-PCR r eaction mixtur e (10 μl) consisted of SYBR green master mix (5 μl), cDNA template (1 μl), and gene specific primers (0.5 μM each).The q-PCR was performed in 8-well strips using SYBR green-Table 3. List of saffron F. oxysporum R1 genes validated by RT-qPCR along with primer sequence.based assa y (T hermo-Scientific , catalog number 4309155) on 7500 real time PCR System (Applied Biosystems ® model).The r elativ e quantification was done by 2 − CT method (Fleige and Pfaffl 2006 ).

Category
For the quantification of Fox R1 genes, its actin gene was used as r efer ence gene for the normalization of gene expression.

Sta tistical anal ysis
Results wer e expr essed as the mean ± standard deviation.The data was statistically analyzed by ANOVA using IBM SPSS Statistics version 26.The Multiple Duncan range test was performed to anal yze differ ences between mean v alues at significant le v el ( P < .05).All the experiments were replicated independently three times at three different time intervals.

Results
In vitro antagonism of Bar D5 against Fox R1 In the dual culture assay, the diameter of growth of Fox R1 in control and test was found to be 6.36 ± 0.62 cm and 3.9 ± 0.47cm, r espectiv el y and the inhibition per centage w as found to be 38.6% ( P = .00)in the presence of Bar D5 (Fig. 1 ).The incubation of Fox R1 spores in the cell-free extract of Bar D5 (test) resulted in structural deformities in hyphae such as thickening of cell wall (Fig. 2 B), changes in the morphology, and shortening of hyphae (Fig. 2 C).The hyphae were shrunken and round in structure (Fig. 2 C).Howe v er, in the contr ol, F. oxysporum R1 hyphae wer e gr owing normall y, elongated, visible with numerous spores, and the formation of a germ tube from the hyphae was also observed (Fig. 2 A).Further, Bar D5 was accessed for the production of antifungal v olatiles b y the seal plate method.The VOCs produced by Bar D5 not only inhibited the growth of Fox R1 but also caused degenerativ e c hanges in the morphology.The diameter of growth of Fox R1 in control was 8.4 ± 0.2 cm, and in test, the diameter was 6.49 ± 0.12 cm (Fig. 3 E).A 22.7% ( P = .00)reduction in zone due to inhibition of growth was observed after 7 days of incubation.In addition, a str ong r eduction in the pur ple pigment was observ ed.The Fox R1 in the presence of volatiles of Bar D5 a ppear ed off-white compar ed to the control (Fig. 3 A and B).The cottony texture of Fox R1 was also not observed in the presence of volatiles of Bar D5 compared to the control (Fig. 3 C and D).Further, on micr oscopy, structur al deformities were observed such as coiling of hyphae (indicated by the r ed arr ow in Fig. 4 B), fr a gmentation of hyphae (indicated by the y ello w arro ws in Fig. 4 B and C), swelling of hyphae, and shrinkage of cytoplasm (indicated by the red arrow in Fig 4 C) as compared to control (Fig. 4 A).This indicated that Bar D5 has the potential for the production of VOCs that have an inhibitory effect on Fox R1.

In planta antagonism of Bar D5 against Fox R1 in saffron corm in real time
In planta antagonism of Bar D5 against Bar D5 was studied using thr ee a ppr oac hes sim ultaneousl y: r eal time ima ging, qPCR and culture-based method.

Real time imaging of saffron corms infected with EGFPlabeled F. oxysporum R1
The infected corms were dissected and observed for the visual symptoms caused by Fox R1 in control and test.The difference in the intensity of the symptoms was observed in the presence and absence of Bar D5 (Fig. 5 ).In control, the symptoms were more intense and ranging from y ello wing to dark browning of the corms around the injury point.Ho w ever, in presence of Bar D5, the symptoms of the corm rot were less intense (Fig. 5 ).
Further, the thin sections of the infected corms tissue were visualized under the confocal microscope at 1, 2, 3, 4, 5, 8, 12, and 20 dpi.At 1 dpi, in mock-primed corms, Fox R1 hyphae were higher in number and extended compared to Bar D5 primed wherein v ery fe w and distorted hyphae were observed (Fig. 6 A and B).At 2 and 3 dpi, in mock-primed corms the Fox R1 were growing optimally, ho w ever, in Bar D5-primed corms at 2 and 3 dpi comparativ el y fe w er hyphae w er e observ ed.The hyphae wer e v ery short and swollen compared to the control at 3 dpi (Fig. 6 C-F).A similar tr end was observ ed at 4, 5, 8, and 12 dpi in contr ol, wher e hyphae were elongated and in test samples less in n umber, shrunk en and stressed (Fig. 6 G-N).Interestingly, at 20 dpi in the test sample, the formation of c hlamydospor es occurr ed, as indicated by the arr ows (Fig. 6 P).

Quantification of Fox R1 load in saffron corm
Quantification of Fox R1 load in the absence and presence of Bar D5 was done in saffron corms by real time PCR.Since confocal micr oscopy ima ges r epr esent qualitativ e data; quantification of Fox R1 was done to get a complete picture.Fox R1 load was quantified in control, test 1 and test 2 using Fox R1 specific ITS primers.Test 1 was taken into consideration to draw a comparison between biocontrol Bar D5 and the chemical fungicide carbendazim to compare the effect of biocontrol with the commonly used fungicide.Up to 5 dpi the load of Fox R1 was less in test 1, i.e. carbendazimtreated corms as it was 4.31-, 4.06-, and 3.1-fold less compared to the control (Fig. 7 ).Ho w ever, at 8, 12, and 20 dpi Fox R1 load was least in test 2 i.e.Bar D5-primed corms, the load was 2.38-, 2.3-, and 2.29-fold less compared to the control but in test 1 the load was 1.85-, 1.55-, and 1.5-fold less compared to control (Fig. 7 ).This indicated that Bar D5 has a longer-term effect compared to commercial fungicides.

Fox R1 load quantification in saffron corms under different treatments by CFU method
This method was done to complement and validate the results obtained from confocal microscopy and q-PCR.The results obtained were in sync with real time PCR data.In control, the load of Fox R1 k e pt on increasing up to 20 dpi.The maximum inhibition of Fox R1 was 74.6% and 70.0% at 5 dpi in test 1 and test 2, respectiv el y, compar ed to the control (Fig. 8 ).In test 1, the load of Fox R1 was less compared to test 2 up to 5dpi.After 5 dpi, the load of Fox R1 increased in test 1 and was higher as compared to test 2. At 8, 12, and 20 dpi, the maximum inhibition was recorded in test 2 compared to pathogen control (Fig. 8 ).In the test 1, the inhibition per centage w as 73% at 1 dpi but 43.6% at 20 dpi.Ho w e v er, in test 2, at 1 dpi the inhibition was 50.1% and 49.7% at 20 dpi.This indicated that the Bar D5 on av er a ge r educed the Fox R1 load by 50% in infected corms.

Transcriptomic profiling of Fox R1 in Bar D5-primed corms
To understand the molecular mechanism of action of Bar D5 against Fox R1, in planta transcriptomic analysis was performed to establish their interaction in a natural wa y.T he expression of DEGs of Fox R1 was studied in control vs. test at 2 dpi.In planta interaction was conducted at 2 dpi as it represents the midpoint for interaction between initial time period i.e. 0 dpi to maximum defense response in saffron at 5 dpi.The raw reads obtained were processed for quality check and alignment with the reference genome of Fox, which resulted in 2.81% alignment in the control and 1.42% alignment of Fox R1 in the test sample .T he number   of r aw r eads ma pped to the r efer ence genome has been tabulated in Table 4 .During the antagonism, 3689 significant genes based on ( P < .05) of Fox R1 were expressed.A total of 1241 genes of Fox R1 were upregulated and 1282 genes were downregulated in presence of Bar D5.All the DEGs were subjected to GO func-tional annotation analysis.A total of 2553 GO terms were assigned of which 813 (31.8%) were related to molecular function, 1370 (53.6%) wer e r elated to biological process and 370 (14.4%) wer e r elated to the cellular component.The 96 significantly enriched GO terms (sorted by adjp-value < .05)were identified.The  top 10 enriched GO terms in each category were plotted against the number of genes associated with each GO term (Fig. 9 ).The most enriched GO terms related to molecular function were structural constituent of ribosome (GO:0003735), purine ribonucleoside triphosphate binding (GO:0005198), in cellular component was protein-containing complex (GO:0032991) intracellular anatomical structure (GO:0005622), in biological process was cellular process (GO:0009987), and metabolic process (GO:0008152).

DEGs of Fox R1 essential for its growth, de velopment, and pa thogenicity affected b y Bar D5
The genes of Fox R1 that are related to growth, morphology, and pathogenicity were selected for further analysis.A total of 60 such genes were identified by subjecting the total data to Pathogen-Host-Interaction database.Among 60 genes, 48 genes were found to be significant with P < .05.A total of 40 genes were found to be differ entiall y expr essed with log2 fold change ≥ 0.5 and 23 genes were found to be differentially expressed with log2 fold change ≥ 1 in control vs. test.All the 60 genes identified have been enlisted in the Table 5 .A total of 3 DEGs encoding different chitin synthase enzymes of Fox R1 were found to be downregulated by Bar D5, which includes chitin synthase D (FOXG_04179), chitin synthase 4 (FOXG_00113), and chitin synthase V (FOXG_04162).Another important gene encoding 1,3-beta-glucan synthase (FOXG_03721) responsible for cell wall construction was significantly downregulated in presence of Bar D5 including erg3 (FOXG_10530), and phospholipase A2 (FOXG_04155) (Table 5 ) (Fig. 11 ).The DEGs related to reacti ve o xygen species (ROS) deto xification wer e identified suc h as catalase (FOXG_03915), glutathione S-tr ansfer ase (FOXG_05962), thioredo xin (FOXG_00069), e po xide hydr ol yse (FOXG_08657), and per oxir edoxin.In the pr esent study, gene encoding per oxir edoxin of Fox R1 was upregulated in presence of Bar D5; ho w e v er, genes encoding other enzymes were downregulated (Table 5 ).
The gene of Fox R1 encoding alcohol dehydrogenase (FO XG_10292) w as do wnr egulated that plays important r ole in the carbon and energy metabolism of fungi.In contrast, the enzymes such as glutamine synthetase (FOXG_05182), transcription factors (TFs) Fnr1 ( FOXG_03165) wer e upr egulated in presence of Bar D5 (Table 5 ).Further, two DEGs related to ATPbinding cassette (ABC) transporter and five genes related to major facilitator superfamily (MFS) transporter were downregulated.The gene encoding MFS toxin efflux pump (FOXG_03950) and MFS peptide transporter (FOXG_15681) was also downregulated in presence of Bar D5.Bar D5 significantly downregulated the expression of genes encoding for plant cell wall degrading enzymes (PCWDEs) such as pectin ly ase (FO XG_19077), xylanase (FOXG_15742), and endo pol ygalactur onase (FO XG_13051).Ho we v er, the expr ession of two genes encoding exo pol ygalactur onase (FOXG_08862) and pectin methyesterase (FOXG_12330) were significantl y upr egulated.The v arious gene encoding differ ent TFs in Fox R1 was downregulated in presence of Bar D5.The expression of genes for TF Ebr1 (FO XG_05408), Xnlr (FO XG_03748), Sge1 (FO XG_10510), Snt2 (FO XG_01993), and FTF2 (FOXG_09390) were significantl y downr egulated in pr esence of Bar D5.In addition, the DEG encoding MAP kinases Pbs2 (FO XG_03107) w as significantly downregulated but the expression of two other genes encoding MAP kinases [Hog1 (FOXG_06318) and Fmk1 (FOXG_08140)] was found nonsignificant.Another important gene cnb1 (FOXG_01489) that encodes regulatory subunit B of calcineurin, a heterodimeric calcium/calmodulin dependent protein phosphatase involved in the regulation of chlamydospores formation was upregulated (Table 5 ).

Validation of transcriptomics data by q-PCR
In total, 20 genes were selected from the transcriptome data for validation using q-PCR.A total of six genes encoding for TF, five genes encoding for PCWDEs, two genes encoding for G-proteins and cell wall synthesizing enzymes each, one gene encoding for eac h ar ginosuccinate l yase, mitoc hondrial carrier pr oteins, calcineurin B subunit, and alcohol dehydrogenase enzyme (Fig. 12 ).The actin gene was used as the RG for the normalization of data.The effect of Bar D5 on the genes of Fox R1 was e v aluated using Ct method.The results of q-PCR were consistent with the RNA seq data except or one gene , i.e .pgx1 gene that has shown upregualtion in RNA seq data but downregulation in q-PCR results.The similar expression pattern of the remaining 19 genes supports the results of RNA seq (Fig. 12 ).

Discussion
Corm rot of saffron caused by the pathogen F. oxysporum is one of the most de v astating diseases that affect its production worldwide (Mansotra et al. 2023 ).The disease spreads from the sowing of injured, wounded, or infected corms in soil infested with F. oxysporum as its c hlamydospor es can persist in soil for many years (Gupta et al. 2021, Bhagat et al. 2022, Mansotra et al. 2023 ).Although c hemical mana gement is emplo y ed for disease mana gement, in r ecent years attention has shifted to the use of an ecofriendl y a ppr oac h i.e. the use of biocontr ol a gents (Mawar et al. 2021 ).Biocontr ol a gents employ m ultiple str ategies for disease contr ol, suc h as competition for food and niches, the secretion of inhibitory compounds, and the induction of systemic resistance in the host plant (Boro et al. 2022 ).
Although Bar D5 exhibited mild antifungal activity in dual culture, it has been reported to decrease the disease incidence in pot trials by 71% and by 57% in field trials (Ma gotr a et al. 2021 ).In the present study, Bar D5 was demonstrated to secrete volatiles in the seal plate assay, as indicated by the growth inhibition and change in morphology of Fox R1.Pr e viousl y, tw o genes for v olatile compounds (VOCs) such as 2, 3 butanediol and acetoin were identified fr om the dr aft genome of Bar D5 (Ma gotr a et al. 2021 ).VOCs produced by Bacillus species such as Bacillus sp.strain B44 and Bacillus cereus MH778713 have been reported to inhibit the growth of disease causing F. oxysporum, hence acting as antifungal com-pounds (Gotor-Vila et al. 2017, Jangir et al. 2018, Ramirez et al. 2022 ).
Additionall y, the cell-fr ee extr act of Bar D5 caused structur al deformities in Fox R1 (Fig. 2 ).Structural deformities in Fox R1 have also been reported by cell-free extracts of B. amyloliquefaciens W2 and Burkholderia gladioli E39CS3 (Gupta andVakhlu 2015 , Ahmad et al. 2022 ).The cell-free extracts of various species belonging to the gener a Bacillus , Streptom yces , and Tricoderma have been reported to cause structural abnormalities in various phytopathogens (Kubicek et al. 2001, Trejo-Raya et al. 2021 ).The effect of Bar D5 was e v aluated a gainst Fox R1 in Bar D5-primed saffr on corms, and c hlamydospor e formation of Fox R1 was observed at 20 dpi, indicating the fungistatic mode of action of Bar D5 (Fig. 6 P).Chlamydospor es ar e asexual r esting cells that ar e formed in r esponse to stress.Warma and Aitken ( 2018 ) have reported the formation of c hlamydospor es of F. oxysporum in the decaying leaf sheath of two differ ent v arieties of banana, Lady Finger and Cav endish at 70 dpi and 80 dpi, r espectiv el y.
Further, to c hec k the effect of the D5 on the growth of Fox R1 in planta , the load of Fox R1 was determined in the absence and presence of Bar D5 in saffron corm, and it was observed that up to 5 dpi carbendazim was better, but after 5 dpi, Bar D5 efficientl y r educed the load of Fox R1 in infected samples (Fig. 7 ).The results of Fox R1 load in the presence and absence of biocontrol were further complemented by the CFU method and a similar trend of inhibition of Fox R1 was observed (Fig. 8 ).Bacillus subtilis SQR9, Trichoderma asperellum , and B. am yloliquef aciens sub sp.plantarum XH-9 hav e also been demonstrated to reduce the load of F. oxysporum in planta , (Cao et al. 2011, Sar av anakumar et al. 2016, Wang et al. 2018 ).In the present study, carbendazim was effectiv e in contr olling Fox R1 initially, but Bar D5 was found to be a better management strategy in the long run (Fig. 7 and 8 ).Similar to the present study, Gupta et al. ( 2020 ) hav e demonstr ated the effects of different biocontrol agents and carbendazim against F. oxysporum causing corm rot in saffron fields.After 250 days of corm sowing, only 4% inhibition was reported in carbendazim treated samples, but 80% with the biocontr ol a gent T. asperellum.This clearl y indicated that biocontr ol a gents ar e better in the long run compar ed to commercial fungicides.

Interaction of Bar D5 and Fox R1 at transcriptome level
To unr av el the inter action of Bar D5 and Fox R1 at the molecular le v el, the tr anscriptomes of F. oxysporum R1 in the absence and presence of Bar D5 in the saffron corm were compared.The results of transcriptomics confirmed that Bar D5 affects the growth, morphology, and pathogenicity profiling of Fox R1 as genes of Fox R1, related to these were downregulated in presence of Bar D5 (Fig. 11 ).
It was observed that a number of genes related to cell wall synthesis , cell membrane , carbon metabolism, and plant cell wall degr adation wer e downr egulated in Fox R1 in the presence of Bar D5.In the compar ativ e tr anscriptomics anal ysis, thr ee DEGs encoding differ ent c hitin synthase (CHSs) enzymes of Fox R1 were downregulated (Table 5 ).CHSs are the k e y enzymes involved in chitin synthesis and can also trigger an innate immune response in host plant (Li et al. 2016 ).Another important gene encoding 1, 3-beta-glucan synthase, was significantly downregulated in the presence of Bar D5. 1, 3-beta-glucan synthase is a glucosyltransferase enzyme that is involved in the synthesis of beta-glucan, a major component of fungal cell walls (Ruiz-Herr er a and Ortiz-Castellanos 2019 ).The fungal plasma membr ane, pr esent next to the cell wall, is the place where multiple cellular processes of importance occur (Athanasopoulos et al. 2019 ).Er goster ol is one of the major components of the cell membrane of fungus , pla ys an important role in the permeability and fluidity of the membrane, and was downregulated in the present study (Rodrigues 2018, Mohid et al. 2022 ).The erg3 gene has been reported to be involved in nutritional differentiation and virulence regulation in Fusarium graminerum (Yun et al. 2014 ).Similar to the present studies, researc hers hav e also r eported the downr egulation of these genes in many fungal pathogens in the presence of biocontrol agent (Han et al. 2021 ).In contrast to the present study, in response to the biocontr ol a gent B. am yloliquef aciens the genes r elated to cell wall synthesis and cell membrane was significantly upregulated in Sclerotinia sclerotiorum, suggesting a robust defense response of this pathogen (Yang et al. 2020 ).
Another downregulated gene in the present study is phospholipase A2 (Table 5 ), an important enzyme that is involved in the hydr ol ysis of phospholipids into various signaling molecules such as phosphatidic acid, free fatty acids, diacylgl ycer ol, and so on.These signaling molecules regulate diverse processes in cells such as growth and development, homeostasis, signal transduction, and so on in fungi (Zhu et al. 2016, Barman et al. 2018 ).The subfamily A2 of phospholipase has been reported to play an important role in signal transduction, homeostasis, and virulence (Roy et al. 2021 ).
Mor eov er, catalase-and glutathione-encoding genes were significantl y downr egulated in the pr esence of Bar D5.The catalase enzyme is a type of antioxidant enzyme that regulates the production of free radicals.It breaks down the hyderogen peroxide into oxygen and water molecules (Valenzuela-Cota et al. 2019 ).The glutathione-encoding gene encodes antioxidant enzymes that catalyze the conjugation of glutathione to various nonpolar molecules (Dhokane et al. 2016 ).Two genes for thioredoxin and one for thioredoxin reductase were also downregulated.Thior edoxins ar e ubiquitous proteins that reduce oxidized cysteine residues by cleaving disulfide bonds, protecting proteins from o xidati ve aggregation and inacti vation.Thioredo xin reductase regulates intracellular o xidati ve potential by reducing thioredoxins (Zhang et al. 2020 ).Ho w e v er, in the present study, one gene for per oxir edoxin was upr egulated.Per oxir edoxin also plays an important role by regulating the peroxide level in the cell.In the presence of Bar D5, overall the antio xidati ve potential of Fox R1 is hampered, as documented by the downregulation of various genes involved in the detoxification of ROS (Table 5 ) (Fig. 11 ).
Nitrogen is one of the essential elements r equir ed for the gr owth and de v elopment of fungi and enables them to survive nutrition depletion and different environmental niches (Tudzynski 2014 ).The expression of TFs Fnr1 (Fusarium nitrogen regulator-1) that regulate nitrogen catabolism and pathogenicity in F. oxysporum was upregulated in the presence of Bar D5.In addition, two genes encoding two main enzymes involved in the nitrogen metabolism pathwa y, i.e .glutamate dehydrogenase and glutamine synthetase, w ere identified.Ho w ever, gene encoding glutamine synthetase was upregulated and gene encoding glutamate dehydrogenase was downregulated.In addition to nitrogen, another k e y factor for gr owth and de v elopment is carbon.The gene encoding alcohol dehydrogenase was highly down regulated (Table 5 ).Alcohol dehydrogenase plays a central role in carbon and energy metabolism and is necessary for fungal infection in plants .T he downregulation of this gene implies that the carbon metabolism of Fox R1 has been negativ el y affected by Bar D5.Expression of the F. oxysporum adh1 gene is reported to occur mainly during the initial periods of colonization of the plant, as during the invasion of roots, this gene helps F. oxysporum survive under hypoxia conditions (Escobosa et al. 2011 ).
The toxic compounds produced by the host or other microorganisms in the vicinity of the fungi hamper the growth of the fungi.T hese toxins , when accumulated in the cell, inhibit the growth and morphology of fungi.The toxins are removed from the cell with the help of transporters that efflux the antifungal compounds (Perlin et al. 2014 ).The two transporter families involved in efflux are the ABC superfamily and the MFS.ABC transporters function in the efflux of toxic compounds and fungicides and are also the center for nutrient transport (Fravel et al. 2003, Kumari et al. 2021 ).MFS, the main promoter super family of membr ane tr ansporters, is a tr ansporter that is mor e selectiv e than ABC transporters (Colemam and Mylonakis 2009 ).Two genes of Fox R1 related to the ABC transporter and five genes related to the MFS tr ansporter wer e downr egulated (Table 5 ).The genes encoding the MFS toxin efflux pump and MFS peptide transporter were totall y r epr essed in the pr esence of Bar D5.The downregulation of DEGs related to ABC and MFS transporters has also been reported in F. oxysporum by B. subtilis HSY21 (Han et al. 2021 ).Ho w e v er, upregulation of DEGs related to the ABC transporter in S. sclerotiorum and Fusarium pseudograminearum in response to Bacillus revealed the r obust r esponse of the pathogen (Yang et al. 2020, Zhang et al. 2022 ).

Effect of Bar D5 on the pathogenicity profiling of Fox R1
Bar D5 has significantl y downr egulated the expression of genes encoding PCWDEs, such as lipase, alpha-amylase, pectin lyase, xylanase, and endopol ygalactur onase (Table 5 ).Ho w e v er, the expression of two genes encoding exopol ygalactur onase and pectinmethy esterase w as significantly upregulated.All these enzymes play an important role in virulence, as there is a correlation between PCWDEs and the pathogenicity of fungi.Similar to the present study, Zhang et al. ( 2022 ) have also reported the downregulation of genes of F. pseudograminearum encoding PCWDEs except for pectinmethyester ase, whic h was upr egulated in presence of the biocontrol agent Bacillus velezensis YB-185.The downregulation of various CWDEs of F. oxysporum (so y abean root rot agent) has been reported by biocontrol agent B. subtilis HSY21 (Han et al. 2021 ).
Further, the various gene encoding TFs in Fox R1 were found to be downregulated in the presence of Bar D5.The genes encoding TFs Ebr1, Xnlr, Sge1, Snt2, and FTF2 wer e significantl y downr egulated in the presence of Bar D5.All these TFs have an important role in the virulence and pathogenicity of different f .sp . of F. oxysporum .The TF Xnlr regulates the expression of the gene encoding PCWDE xylanase (Zuriegat et al. 2021 ).The TF Sge1 regulates the inv asiv e gr owth, virulence, and activities of extr acellular amylase and cellulase in F. oxysporum f .sp .lycopersici (Rispail and Di Pietro 2010 ) and Snt2 regulates the growth parameters such as hyphae growth, septation, and conidiation in F. oxysporum f .sp .melonis (Denisov et al. 2011 ).The downregulation of these genes implies that the infection efficiency of Fox R1 is hampered.The gene encoding MAP kinases (Pbs2) was significantly downregulated; ho w e v er, the expr ession of genes encoding MAP kinases (Hog1 and Fmk1) was found to be nonsignificant.The MAP kinase Fmk1 and Hog 1 play an important role in the pathogenicity of F. oxysporum , as Fmk1 regulates the growth, colonization, and expression of genes involved in cell wall synthesis, while Hog 1 is involved in o xidati ve stress and osmolarity response of F. oxysporum (P ar eek and Rajam 2017 ).
Another important gene cnb1 was found to be upregulated and encodes regulatory subunit B of calcineurin, a heterodimeric calcium/calmodulin-dependent protein phosphatase involved in the regulation of chlamydospores formation and the virulence of F. oxysporum f .sp .lycopersici (Hou et al. 2020 ).The upregulation of cnb1 gene can be correlated with the formation of c hlamydospor es both in in vitro and in planta experiments.
The comparison of the DEGs in F. oxysporum R1 in the presence and absence of biocontrol clearly indicates downregulation of the genes of pathogens important for gr owth, mor phology, and pathogenicity and upregulation of genes involved in nitrogen metabolism.Ov er all, the negativ e effect of Bar D5 on Fox R1 is confirmed by the results of comparative transcriptomics, q-PCR and CFU/g of infected corm quantification.

Conclusion
Though biocontr ol a gents ar e used to contr ol plant diseases for sustainable cultivation, a detailed study of the mechanism of their action on pathogenic agents is lacking altogether in some cases, such as the tripartite interaction between saffron-Fox R1 and Bar D5.In the present study, the effect of Bar D5 on the growth of the pathogenic Fox R1 has been clearl y demonstr ated in vitro as well in planta .Further, the genes up and downregulated have also been identified.This work can be further complimented by the studies, wherein selected fungal genes from the DEGs will be mutated, and the effect of this mutation will be c hec ked on its ability to cause disease.
growth of the Fox R1 in the control plate , and Test = growth of the Fox R1 in the test plate .

Figure 2 .
Figure 2. Effect of cell-free extract of Bar D5 on Fox R1 morphology after 10 days of incubation.(A) Control; Fox R1 grown in PDB, (B) and (C), Test; Fox R1 grown in the cell-free extract of Bar D5.Arrows in (B) represents the thickening of cell wall whereas in (C) represents shortening of hyphae.Each bar r epr esents 50 μm.

Figure 3 .
Figure 3.Effect of Bacillus sp.strain D5 volatile compound on F. oxysporum R1 (Fox R1) by sealed plate method.(A) and (C) Control (B) and (D) Test (Fox R1 grown in presence of volatile compound), after 7 days of incubation, complete loss of pigmentation was observed.Bar r epr esents 1 cm.(E) Bar plot r epr esents the diameter of Fox R1 growth in control and test.Error bar represents the standard deviation.

Figure 4 .
Figure 4. Micr oscopic ima ges of Fox R1 gr own in pr esence of Bar D5 volatiles.(A); contr ol, (B) and (C), test.Structur al deformities wer e observ ed as indicted by arrows.Yellow arrow indicates the fr a gmentation of hyphae whereas red arrow in (B) and (C) represents the coiling of hyphae and swelling of hyphae, r espectiv el y.Eac h bar r epr esents 50 μm.

Figure 6 .
Figure 6.Confocal images of the corm tissue at different dpi under the confocal microscope at 60X.C-mock-primed corms inoculated with EGFP-tagged Fox R1, T-bar-primed corms inoculated with Fox R1 at different dpi.(A, C, E, G, I, K, M, and O) are the control images and (B, D, F, H, J, L, N, and P) are the test images.In image (p) arrow indicates the chlamydospores formed at 20 dpi in test.Each bar represents 10 μm scale.

Figure 9 .
Figure 9. GO-enriched terms identified in Fox R1 under the influence of Bar D5 (C vs. T) by GO analysis.The GO terms are divided into three categories (molecular function, cellular components, and biological process) based on the function of the DEGs .T he y -axis r epr esents the number of DEGs and their positioning in x -axis is based on enrichment (adjp-value < .05).

Figure 10 .
Figure 10.Enriched KEGG pathways identified in Fox R1 under the influence of Bar D5 (C vs. T).The y -axis r epr esents the number of DEGs involved in particular pathway and their positioning on x -axis is on the basis of (adj P < .05).

Figure 11 .
Figure 11.Pictorial r epr esentation of gene/pathw ays of Fox R1 affected b y Bar D5. ↓ Indicated do wnr egulated of genes and indicated ↑ upr egulation of genes.

Figure 12 .
Figure 12.Validation of the expression patterns of genes of Fox R1 selected from transcriptome by RT-qPCR.Gene expression values in different comparisons were obtained by normalizing the values.Fox R1 actin gene was used as RG.Values are expressed as mean ± sd ( n = 3).

Table 2 .
Prime pair used for the in planta quantification of EGFP-labeled F. oxysporum R1.

Table 4 .
The number of reads obtained in each sample and the % alignment obtained with F. oxysporum f.sp .lycopersici reference genome.

Table 5 .
List of F. oxysporum R1 genes effected by biocontrol agent Bacillus sp.strain D5.