Unravelling the Function of the Sesquiterpene Cyclase STC3 in the Life Cycle of Botrytis cinerea

1. Introduction

Sesquiterpenes are C15 terpenes consisting of three isoprene units, found in linear (as farnesene), monocyclic (as zingiberene and humulene), bicyclic (as caryophyllene), tricyclic (as longifolene), and tetracyclic (as punctaporonin C) forms. They are produced by higher plants, marine organisms and fungi [1,2,3]. The biological activity of many of these natural products is related to human health. For instance, arvestolides H and I, and artefreynic acids B, C, and G inhibit the production of nitric oxide [4,5]; chrysanthemulide A has anti-inflammatory activity [6]; 14-O-acetylinsulicolide A, 6β,9α-dihydroxy-14-p-nitrobenzoylcinnamolide and insulicolide A are cytotoxic against renal carcinoma cell lines [7], or santhemoidin A is active against Trypanosoma brucei rhodesiense trypomastigotes [8]. Moreover, sesquiterpenes have been recognized as replacements for petroleum-derived jet-engine fuels [9]. These findings highlight the need to deepen the analysis of sesquiterpene biosynthesis due to their commercial interest.

The sesquiterpene synthases (EC:4.2.3.-) are a group of enzymes that convert the linear C15 precursor farnesyl diphosphate (FDP) into a diverse class of natural products. Many sesquiterpenes are cyclic compounds, and the first reaction in their biosynthesis is the cyclization of FDP at the proximal C6-C7 or the distal C10-C11 double bond. Therefore, the synthases involved in this step are also named sesquiterpene cyclases (STC) [10,11,12,13]. Subsequent modifications such as oxidations or substitution with different functional groups, a lactone ring and rearrangements of the 15-carbon skeleton produce the related sesquiterpenoids [14,15].

In fungi, these compounds constitute the most abundant and diverse group of all categories in the vast system of natural products [16]. Many act as phytotoxins in the interaction between phytopathogenic fungi [17,18] and their host [19]. Botrydial is one example of these non-host-specific toxic sesquiterpenes. It is produced by the phytopathogenic fungus B. cinerea, the causal agent of gray mold, classified as the second most relevant plant pathogen for its economic and scientific importance [20]. Botrydial is secreted during the fungus-host interaction, causing chlorosis and cell collapse, and its role in virulence is closely related to the polyketide botcinin, the second major phytotoxin of Botrytis [21,22]. It triggers the hypersensitive response of plants to pathogens, modulated by salicylic acid and jasmonic acid signaling, that produces a rapid localized cell death [23]. Botrydial also induces phosphatidic acid production, a phospholipid second messenger involved in the induction of plant defense responses [24]. Moreover, this phytotoxin also modulates the interaction of B. cinerea with other fungi and bacteria and their environment [25,26].

Over the last several years, our group has been studying the sesquiterpene synthesis by B. cinerea [27]. We found six Bcstc genes in the B. cinerea B05.10 strain (Bcstc1 or Bcbot2, Bcstc2, Bcstc3, Bcstc4, Bcstc5 or Bcaba5, and Bcstc7) that encode proteins with the two conserved magnesium binding sites required for FDP cyclization. A seventh Bcstc gene, Bcstc6 (Gene ID BofuT4P10000109001), is specific to the T4 strain and absent in B05.10 [28]. We have functionally characterized the sesquiterpene cyclases BcStc1, BcStc5, and BcStc7. Bcstc1 encodes for a pentalenene cyclase, which generates the tricyclic alcohol presilphiperfolan-8β-ol as a precursor of botrydial phytotoxins and other important derived sesquiterpenes [29,30]. On the other hand, BcSct5 is essential for the synthesis of the sesquiterpenoid abscisic acid [31,32,33,34], and BcStc7 is involved in (+)-4-epi-eremophilenol biosynthesis, a member of the family of eremophilene sesquiterpenes [35,36,37]. However, the biosynthesis of these two compounds is not fully understood yet [27]. The role of the three other sesquiterpene cyclases of B. cinerea is still unknown. To date, BcStc2, BcStc3 and BcStc4 remain uncharacterized.

Here, we achieve the functional characterization of the sesquiterpene cyclase 3 (BcStc3). We describe the Bcstc3 expression pattern in vitro and in planta and investigate the putative role of the protein in the virulence and biology of B. cinerea, generating a Bcstc3 knock-out mutant and a new strain that constitutively expresses the Bcstc3 gene.

2. Results

2.1. BcStc3 is a Well Conserved Protein

BcStc3 has 411 amino acid residues with a theoretical molecular weight of 486551.89 g/mol and an isoelectric point of 6.34, estimated using the Expasy Protparam tool (https://web.expasy.org/protparam/). Its aliphatic index is predicted to be 90.73, and the grand average of hydropathicity (GRAVY) is -0.381, indicating that BcStc3 is a thermostable soluble protein, located in the cell cytoplasm, according to DeepLoc 2.0 (https://services.healthtech.dtu.dk/services/DeepLoc-2.0/) prediction. A terpene synthase family 2, C-terminal metal binding domain (Pfam 19086, at 170-336 amino acid residues) was identified using the InterProScan software [38] and as a terpene cyclase-nonplant C1 protein (cd00687, from 147 to 336 residue) according to Conserved Domain Database (CDD) from NCBI [39,40]. Both domains were also found in the BcStc1, BcSct4 and BcStc5 proteins (Table 1), but BcStc3 showed the highest sequence identity percentage with BcStc4 (23.84%). The other two members of the BcStc protein family of B. cinerea B05.10 (BcStc2 and BcStc7) are grouped into the trichodiene synthase or TRI5 family (Pfam 06330 domain), and the percentage identity values with the other four family proteins range between 11.21 to 20.66% (Table 1).

The conservation profile of BcStc3 protein amino acid residues (Figure 1A) was predicted by the Con Surf web server (http://consurf.tau.ac.il), and the conservation score for each residue (Figure 1B) was calculated with Al2CO (http://prodata.swmed.edu/al2co/al2co.php) taking 200 homologous protein sequences obtained doing a BlastP [43] of the BcStc3 protein sequence with the criteria of >30% identity and >70% coverage. The results showed low conservation of residues in the N-terminal region, but different canonical motifs of the class I terpenoid cyclases [44,45] were identified at the C-terminus region (Figure 1B). The two conserved motifs involved in the binding of the magnesium cations were found: the aspartic-rich motif, DDMWE, at amino acid 177, and the NSE/DTE motif, NDLASYDKE, at amino acid 307 (Figure 1). Moreover, the conserved effector triad was identified at R264-D267-I268 (Figure 1), and the RY conserved pair related to the binding of farnesyl pyrophosphate [46,47] was located at the C-terminus of the protein (residues 401 and 402, respectively) (Figure 1). The molecular model of BcStc3 protein was generated using Alphafold [48], finding the α-helical fold (Figure 2) typical for class I terpene synthase [13,49]. As expected, the arrangement of the conserved motifs identified in the sequence defines the catalytic site of the protein (Figure 2). Furthermore, the H-α1 loop, a conserved structural motif that shields the active site and stabilizes the closed enzyme conformation after the substrate binding [50] and the conserved asparagine residue (N329) at the C-terminus of the H-1α loop were also identified in the BcStc3 model (Figure 2).

The putative conservation of BcStc3 in Botrytis species was analyzed. A BlastP search against the NCBI non-redundant protein sequences database restricted to the genera Botrytis (taxid: 33196) and Botryotinia (taxid: 40558) was performed using BcStc3 as a query sequence. The results showed that BcStc3 is a well-conserved protein (Supplementary Figure S1). Homologous proteins were found in Botrytis tulipae, Botrytis sinoallii, Botrytis elliptica, Botrytis deweyae, Botrytis galanthina, Botrytis hyacinthi, Botrytis byssoidea, Botryotinia globosa, and Botryotinia narcissicola, with a percent identity of over 80% and a query coverage exceeding 95%. The multiple sequence alignment showed that the functional motifs identified in BcStc3 are conserved, except in the XP_038737186.1 protein of B. byssoidea, which lacks the DDMWE motif due to an insertion of 30 amino acid residues at the first aspartic residue (Supplementary Figure S1). As a result, the percent identity of this protein (estimated using the SIAS software) with the rest decreased to approximately 40%, while the other proteins exhibited values greater than 90%. Additionally, the XP_038737186.1 protein was positioned on a distinct branch of the phylogenetic tree derived from the sequence alignment of the BcStc3 protein sequence with those of other Botrytis species, using the Neighbor-Joining method (Figure 3).

A second BlastP search was conducted, this time excluding the Botrytis and Botrytioinia taxids. The results were filtered based on percent identity (>30%), coverage (>70%) and the bit-score (>50) according to [51], and 168 sequences fulfilled these criteria. Five multidomain proteins (ESZ93961.1, KAH7563860.1, KAF4304151.1, EKG18134.1, and XP_049153155.1) and three other proteins (XP_051299547.1, CAI6339174.1, and KAG7009610.1), lacking the PF19086 domain or being partial annotated, were removed from the analysis. The 160 remaining sequences were distributed in 124 organisms distributed in Dikarya fungi (Supplementary information Table S2). Seven proteins (1.5% of the total sequences) were annotated in three species of the Strophariaceae family (Agarycomycetes class, Basidiomycota) and the resting proteins were distributed in the Lectiomyceta (67.5%) and Sordariomyceta (28.1%) super-classes of Ascomycota (Table 2). Moreover, 60.6% of the total proteins were annotated in Eurotiomycetes and Dothideomycetes species (Table 2). Noteworthy is the absence of BcSTc3-like proteins in subphyla Taphrinomyctina and Saccharomycotina.

The alignment of the first 55 sequences retrieved from the BlastP search was used to construct a phylogenetic tree. Two clades were identified (Figure 4). Clade I includes most of the BcStc3-homologous proteins annotated in the Sordariomycetes fungi together with sequences from Eurotiales (Eurotiomycetes) and Pleorales (Dothideomycetes) organisms (Figure 4). Clade II consists of two groups. One is made up of BcStc3-related proteins from Botryosphaeriales (Dothideomycetes incertae sedis) and Helotiales (Leotiomycetes) fungi and the only member of Xylobotryomycetes class (Figure 4). The BcStc3 falls into the second group, clustered together with the PQE05643.1 protein of Rutstroemia sp. (Helotiales, Leotiomycetes) and the KAI9816264.1 protein of Pycnora praestabilis, the only member of the Candelariomycetes class (Figure 4). All the proteins of Lecanoromycetes fungi, and one of the seven BcStc3-like proteins found in Basidiomycota (the KDR73888.1 protein of Galerina marginata) were also clustered in this second group (Figure 4).

2.2. Expression Analysis of the Sesquiterpene Cyclase Gene Family in B. cinerea

The basal expression of the sesquiterpene cyclase gene family in non-germinated conidia of the UCA992 strain was analysed by qRT-PCR. Results were normalized to actA and B-tub genes and depicted as fold-change relative to the lowest expressed gene (Bcstc2). All family genes were expressed in non-germinated conidia, being the Bcstc5 transcripts the most abundant (approximately 30 times compared with that of Bcstc2), and Bcstc3 and Bcstc4 reaching transcript levels similar to the reference gene (Figure 5A). The Bcstc5 gene was also the highest expressed gene in the non-germinated conidia of the B05.10 strain, but the Bcstc1 mRNAs were the least abundant transcripts, and the Bcstc2 and Bcst3 genes were expressed almost 8-times more than the Bcstc1 gene (Supplementary Figure S2.A).

To evaluate the role of the sesquiterpene cyclases in the vegetative growth of B. cinerea, the expression pattern of the six Bcstc genes was examined at 96 hours after conidial germination in the YGG medium (Figure 5B). All genes were up-regulated related to the basal expression levels of each gene in non-germinated conidia (from 3 times for Bcstc1 to almost 30 times for Bcstc4), except the Bcstc7 gene, which hardly changed its expression (Figure 5B). Surprisingly, an increase of about 2000-fold in the Bcsct7 mRNA levels was observed for the B05.10 strain at the same growth conditions (Supplementary Figure S2.B).

The Bcstc3, Bcstc4, Bcstc1, and Bcstc2 genes were overexpressed after 96 hours of inoculating tobacco leaves with conidia of the UCA922 strain, especially Bcstc2 and Bcstc3 which increased 16 and 21 times, respectively, the mRNA levels related to non-germinated conidia (Figure 5B). However, the Bcstc5 gene was not induced during the host-plant interaction (Figure 5B). Contrary, the Bcstc1 and Bcstc7 genes were significantly up-regulated in planta (almost 700 times respect to their basal level in non-germinated conidia) when tobacco leaves were inoculated with the B05.10 strain (Supplementary Figure S2.B).

2.3. Co-Regulation of Sesquiterpene Cyclase Genes in B. cinerea

The generation of the ΔBcstc3 strain and the overexpressed-Bcstc3 transformant (OvBcstc3) allowed us to analyze the possible co-regulation of the Bcstc gene family. The Bcstc3 gene deletion or its overexpression caused significant but different changes in the expression pattern of the gene family in non-germinated conidia (Figure 6A). The Bcstc3 mutation induced the upregulation of the Bcstc1, Bcstc2, and Bcstc4 genes, but especially of Bcstc7 (over 40-fold changes) related to their expression levels in conidia of the wild-type strain (Figure 6A). However, the relative mRNA transcript levels of Bcstc5 did not change (Figure 6A). On the other hand, the conidia of the OvBcstc3 strain overexpressed the Bcsct5 gene (25-fold) and, to a lesser extent, the Bcstc7 and Bcstc2 genes (7.8 and 6.4-fold, respectively) with respect to the relative abundance of each transcript in conidia of the wild type strain (Figure 6A).

During fungal vegetative growth, the deletion and the overexpression of the Bcstc3 gene also caused changes in the expression pattern of the gene family. After 96 hours in axenic culture, the transcript levels of the Bcstc4, Bcstc5, and, especially, Bcstc7 (24-fold) genes were higher in the mycelia of the ΔBcsct3 and OvBcstc3 strains than in the wild type (Figure 6B). However, the expression level of Bcstc1 hardly changed, and differential expression of Bcstc2 was observed in both modified strains (Figure 6B).

Changes in Bcstc gene family expression during fungus-plant interaction were quite different in the Bcsct3 knock-out and the Bcsct3 overexpressing strains. The Bcstc3 deletion induced the up-regulation of the rest of the genes, especially of Bcstc1 and Bcsct7 (increasing the transcripts levels to 40 and 10 times, respectively, related to the parental strain) (Figure 6C). However, the Bcsct1 gene expression almost did not change in the OvBcstc3 strain (Figure 6C). On the contrary, the overexpression of the Bcstc3 gene caused a significant increase of the Bcstc5 mRNA levels (up to 107 times related to the wild-type strain) while only 2.3 times in the mutant strain (Figure 6C).

2.4. BcStc3 is Related to Fungal Growth and Stress Responses

The potential role of BcStc3 in fungal development was investigated by comparing the growth of the ΔBcstc3 and OvBcstc3 transformants with that of the wild-type strain in solid and liquid YGG media. After four days of growth on YGG agar, the colony radius of both transformants decreased by approximately 14% (Figure 7A), with similar results observed in the estimated growth rate of the colony radius (Figure 7B). In liquid culture, however, the mycelial fresh weight and fungal biomass were only reduced in the OvBcstc3 strain, exhibiting decreases of 15% and 22%, respectively, compared to the UCA922 strain (Figure 7C,D).

The impact of deleting or constitutively expressing the Bcstc3 gene on fungal tolerance to osmotic and oxidative stresses was investigated by cultivating the strains on YGG plates supplemented with 1.4 M sorbitol or 1.5 mM hydrogen peroxide. Compared to the wild type, the ΔBcstc3 strain exhibited a reduction in colony radius of approximately 8% or 17% in media supplemented with peroxide or sorbitol, respectively (Figure 7A). However, both stress agents caused approximately a 25% reduction in the colony radius of the OvBcStc3 strain (Figure 7A).

The effect of sorbitol and hydrogen peroxide on the growth of each strain was evaluated by determining the relative inhibition growth rate, which involved calculating the changes in the colony growth rate for each strain in the presence of each of these agents compared to their growth rate in their absence. The results showed that hydrogen peroxide barely modified the growth rate of the mutant, whereas that of the OvBcstc3 transformant decreased by 34.5% (Figure 7B). The growth rate of the UCA992 strain in the presence of peroxide decreased by 16.6% (Figure 7B), indicating that the deletion or overexpression of the Bcstc3 gene modifies the fungal sensitivity to oxidative stress. On the other hand, the presence of sorbitol in the culture medium led to a reduction in the growth rate of 11.6% and 28.3% in the ΔBcstc3 and the OvBcstc3 strains, respectively, while in the case of the wild type, the reduction was 15.4% (Figure 7B). These results also suggest the involvement of BcStc3 in osmotic stress tolerance.

2.5. BcStc3 is Involved in Virulence

To study the role of BcStc3 in fungal virulence, tomatoes, grapefruits, gerbera petals and detached tobacco leaves were inoculated with conidial suspensions of the wild type, ΔBcstc3 and OvBcstc3 strains. The ΔBcstc3 displayed increased virulence in all the tested hosts compared to the wild type. The disease index in tomatoes and grapefruits increased by 18% and 12%, respectively, compared to the wild-type strain at four days post-inoculation (dpi) (Figure 8A,B, Supplementary Figures S3.A and S3.B). Also, the mean length of the lesions produced by the ΔBcstc3 on gerbera petals was 1.6-fold greater than those produced by the UCA922 strain at four dpi (Figure 8C, Supplementary Figure S3.C), and the lesions grew two times faster compared to the control (Figure 8D,). Finally, the mean diameter of the lesions caused by the mutant strain in tobacco leaves at four dpi and the lesion growth rate were also two times greater than those of the control (Figure 8E and 8F, Supplementary Figure S3.D).

The disease index of OvBstc3-inoculated tomato fruits slightly and gradually decreased over time by 10.3% compared to the wild-type strain four days after inoculation (Figure 8A, Supplementary Figure S3.A). However, there were no discernible differences between grapefruits infected by this transformant or the UCA922 strain (Figure 8B, Supplementary Figure S3.B). The growth rate of the lesions produced by the OvBcstc3 strain in gerbera petals was only 5% lower than the control, but the mean size of the lesions at 2 and 3 days post inoculation were 38% and 22%, respectively, smaller than those produced by the wild-type strain (Figure 8C, Supplementary Figure S3.C). On the other hand, no significant differences were found in the virulence of OvBcstc3 and UCA922 strains when tobacco leaves were inoculated (Figure 8E,F, Supplementary Figure S3.D).

2.6. BcStc3 is Involved in the Germination and Morphology of Conidia

The production of conidia was assessed after growth on a solid tomato extract medium for ten days of incubation at 20°C, and the results showed no significant difference between the control and transformants (Figure 9A). However, the conidial area of the OvBcstc3 strain was almost twice that of the wild type (Figure 9B). The percentage of aggregated conidia (assessed by counting conidial clustered in groups of two or more and by estimating the conidial sedimentation rate) increased by 38% in the conidia produced by the mutant and decreased by 38% in those produced by the OvBcstc3 strain compared to the wild type (Figure 9C). Also, an opposite effect of deletion or constitutive expression of the Bcstc3 gene on conidial germination was observed (Figure 9D). The conidial germination rate of the ΔBcstc3 strain decreased by more than 50% compared to the wild-type strain, while an increase of approximately 17% was observed for the OvBcstc3 strain (Figure 9D).

2.7. Bcstc3 Gene is Involved in ROS and Infection Cushion Production

The BcStc3 protein seems involved in the fungal response to oxidative stress. Therefore, the production of reactive oxygen species (ROS) was studied under different growth conditions. In axenic culture, the OvBcstc3 strain released nearly double the amount of ROS into the medium compared to the wild type, while the Bcstc3 deletion did not affect the ROS generation (Figure 10A). In planta, the infiltration of a DAB solution in tobacco leaves 48 hours after plant conidial inoculation revealed that the ΔBcstc3 strain produced nearly twice the amount of hydrogen peroxide in the lesion areas compared to the wild-type and OvBcstc3 strains (Figure 10B).

On the other hand, as ROS is accumulated in the infection cushions [52], the production of these infection structures was assessed in the three fungal strains in axenic culture conditions. The results showed that the OvBstc3 strain produced more infection cushions than the wild-type, while no change was observed for the Bcsct3 null mutant (Figure 10C).

2.8. Implication of the BcStc3 on the Secondary Metabolism

The mutants were grown on malt agar medium to produce mycelium plugs which were further used to inoculate 40 Roux culture bottles, each containing 150 mL of modified Czapek-Dox medium. Bottles inoculated with six mycelium plugs were incubated for 27 days at 25°C.

The culture media (6 L), from each of the mutant strains, were filtered and the filtrates subjected to liquid-liquid extraction with ethyl acetate and dried over anhydrous sodium sulfate (see experimental for detailed procedures). Evaporation of the solvents yielded crude extracts as yellow oils, 173 mg (ΔBcstc3) and 153 mg (OvBcstc3).

The crude extracts were preliminary fractionated by column chromatography on silica gel eluting with n-hexane-EtOAc mixtures containing increasing percentages of EtOAc (9:1-0:10). The different fractions of interest were purified by semi-preparative HPLC with a mixture of n-hexane-EtOAc-acetone, (7:2:1) as the mobile phase and flow rate 2.5 mL/min. The isolated metabolites (Figure 11) were characterized by extensive spectroscopic and spectrometric mass analyses.

The metabolomic analysis of the null mutant ΔBcstc3 broth yielded the polyketide derivatives 1 (1.8 mg), 2 (2.2 mg) and 3 (1.5 mg) [53], the known eremophil-9-en-11-ols 5 (6.5 mg), 6 (0.5 mg), 7 (3.0 mg), 8 (3.0 mg), 9 (1.3 mg), 10 (2.0 mg), 11 (2.8 mg), 12 (0.8 mg), 13 (0.8 mg) and 14 (0.7 mg) [36,54], the 11-hydroxyeremophil-1(10)-en-2-one 16 (0.5 mg) [35] and the 11,12,13-tri-nor-eremophilene 17 (3.8 mg) [37].

The overexpressed mutant OvBcstc3 broth afforded the reported polyketide derivatives 2 (1.2 mg) and 4 (2.0 mg) and the known compounds 6 (3.4 mg), 7 (5.2 mg), 10 (1.5 mg), 11 (3.6 mg) [36,54], 15 (1.5 mg) [35], 17 (3.4 mg) and 18 (1.1 mg) [37]. The chemical structures were determined by a wide range of spectral analyses using mono- and bidimensional resonance experiments and they were confirmed by direct comparison of their spectroscopic constants with those of authentic samples.

Finally, analysis of the metabolomic profile of both transformants, the null mutant ΔBcstc3 and the overexpressed OvBcstc3, seem to indicate that the production of secondary metabolites was not affected by mutations in the Bcstc3. So, both transformants produced polyketide botcinins, as well as eremophilenol derivatives. Additionally, the ΔBcstc3 biosynthesized the keto derivative 16 and the 11,12,13-tri-nor-eremophilene 17, while the OvBcstc3 mutants produced the keto derivative 15 and the two 11,12,13-tri-nor-eremophilenes 17 and 18, all of them previously reported (Figure 11) [35,36,37]. Unfortunately, no new metabolites were detected/produced by the overexpressed transformant.

3. Discussion

The annotation of the B. cinerea B05.10 genome showed six genes putatively encoding for sesquiterpene cyclases. Three members of this gene family have previously been characterized. Bcstc1/Bcbot2 encodes for a pentalenene cyclase, which catalyzes the first step in the biosynthesis of the botrydial phytotoxin and other important derived sesquiterpenes, generating the precursor tricyclic alcohol presilphiperfolan-8β-ol [29,30]. The deletion of the Bcsct5 gene in the ABA-overproducer strain ATCC58025 let us conclude that BcStc5/BcAba5 is the key enzyme responsible for the sesquiterpenoid abscisic acid biosynthesis [31], and the sesquiterpene cyclase 7 (BcStc7) catalyzes the cyclization of farnesyl diphosphate to eremophil-9-en-11-ols [35,36,37]. Finally, the Bcstc2 gene is related to the cryptic metabolite botrycinereic acid synthesis [55]. Here, the role of another member of this family, the BcStc3 protein, in the biology of B. cinerea is unravelled.

3.1. Bcstc3 Encodes for a Terpene Synthase Family 2, C-Terminal Metal Binding Domain Protein

The Bcstc3 gene encodes for a terpene synthase family 2, C-terminal metal binding domain protein (Pfam 19086), according to the BLAST search against the NCBI non-redundant protein database, the same protein family of the pentalenene cyclase BcStc1/BcBot2 [29,30], the BcStc5 protein involved in the farnesyl diphosphate conversion into 2Z-4E-α-ionylideneethane [31], and the yet-to-be characterized BcStc4. The two remaining genes of the Bcstc family, Bcstc2 and Bcstc7, encode for trichodiene synthases (Pfam 06330). Both enzyme families are members of the class I terpene synthase (cd00868, EC 4.2.3.-) that share a common α-helical fold and use a trinuclear Mg²⁺ cluster to activate the substrate, isoprenoid pyrophosphate precursors, by ionization and cleavage of pyrophosphate, generating several carbocation intermediates conducting to the cyclization of the substrate [56,57,58]. Indeed, the features conserved across this enzyme class are two conserved motifs involved in the magnesium binding: two magnesium ions bound to the enzymes by a conserved aspartate-rich region (D(D/E)XX(D/E)) and the third one by an NSE/DTE ((N,D)D(L,I,V)X(S,T)XXXE motif) [56,57]. Both motifs are localized within the C-terminal or α-domain of the proteins, as in the BcStc3 sequence. The C-terminal region of the BcStc3 showed a higher conservation score containing the two canonical motifs involved in catalysis in class I terpene cyclase. The aspartate-rich motif was identified at position 177, for the first aspartate residue, and at residue 307, for the N of the NSE/DTE motif (Figure 1). Two other motifs involved in catalysis were also identified in the BcStc3 sequence: the conserved effector triad (R264-D267-I268 residues) that also coordinate the ppi, aiding in substrate recognition and is associated with the initiation of the cyclization reaction [47,56,57], and the conserved RY pair (R334-Y335 residues) (Figure 1) that is involved in the PPi recognition through hydrogen bonds [47,59]. The arginine residue in the RY pair of BcStc1/BcBot2 (R373) plays a significant role after the formation of the caryophyllenyl cation [59]. Two of the three residues involved in the selectivity of BcStc1/BcBot2 (W94, F99 and N325 residues) [47] were identified in the sequence of BcStc3 (F174 and N392 residues, Figure 1), probably also related to the selectivity of the enzyme. In addition, the substrate binding to the active site of BcStc1/BcBot2 induced conformational changes of these motifs and the H-α1 loop, too [47]. It has been shown that this loop moves and shields the active site upon the formation of an Mg2+3-PPi complex of BcStc1/BcBot2 and the sesquiterpene cyclases Cop3, Cop4, and Cop6 from the basidiomycete Coprinus cinereus, modulating the selectivity of these enzymes [50]. The loop was also identified in the 3D model of BcStc3, as the conserved asparagine residue (N329 residue) at the C-terminus (Figure 2) that interacts with an aspartic residue of the NSE/DTE motif to place the H-α1 loop at the active [50].

3.2. BcStc3 is a Well Conserved Protein in Botrytis Genus

The BcStc3 protein is a well-conserved protein in the Botrytis genus, especially in species that infect monocots Lliaceae (B. tulipae, B. elliptica), Amaryllidaceae (B. sinoalli, Botryotinia globosa, B. byssoidea, Botryotinia narcissicola, and B. galanthina), Hyancinthacease (B. hyacinthi), and Asphodelaceae (B. deweyae) (Figure 3). The orthologous proteins of Bcstc3 in B. elliptica, B. hyacinthi, B. galanthina, and B. narcissicola had been previously identified [60], but this is the first time that homologous to this sesquiterpene cyclase are identified in B. sinoallii, B. tulipae, B. deweyae, B. byssoidea and Botryotinia globosa (Figure 3, Supplementary Figure S1). The phylogenetic tree of these proteins closely resembled that previously described for Botrytis genes [61,62]. BcSct3 grouped apart in clade I, while the rest formed clade II, with the proteins of B. deweyae, B. elliptica, and B. sinoalli forming a separate subclade according to [61]. Unexpectedly, the BcStc3-homologous protein of B. byssoidea grouped apart (Figure 3), despite this species is phylogenetically close to Botryotinia narcissicola, B. tulipae, and Botryotinia globosa [61]. The sequence alignment confirmed that the aspartic-rich domain DDMWE in the XP_038737186.1 protein is not conserved (Supplementary Figure S1).

3.3. BcStc3 is a Well Conserved Protein in Kingdom Fungi

BcSTc3 protein is also a well-conserved protein among the kingdom Fungi. Homologous proteins to BcStc3 were identified in 124 fungal species, excluding the Botrytis genus. Seven of these proteins were found in three species of the Strophariaceae family and the remaining 153 proteins were distributed along 121 species of different Ascomycota, mainly of the Eurotiomycetes, Dothideomycetes, and Sordariomycetes classes. The ratio of Bcstc3-homologous proteins per organism was slightly higher in Basidiomycota than in Ascomycota. This difference between phyla has been previously described for terpene synthases and may be related to tandem duplication of the gene clusters in Basidiomycota [46,63,64]. Of the 160 BcStc3 homologous proteins, only one has been previously characterized and crystalized, the terpene cyclase AaTPS of Alternaria alternata [65]. AaTPS was characterized as a bifunctional enzyme using the same active site to catalyze terpene cyclization, and aromatic prenylation reactions favored the last ones under alkaline conditions [65]. The prenyltransferase activity might regulate isoprenoid diphosphate concentrations, avoiding toxic levels in the cell [65]. The study of this enzyme allowed to conclude that the bifunctional activity is well conserved among terpene cyclase family proteins and, therefore, it should be kept in mind the possibility that the BcStc3 protein may also function as an aromatic prenyltransferase in a pH-dependent manner.

The phylogenetic tree generated from the alignment of the first 55 sequences retrieved from the BlastP search using the BcStc3 sequence as query grouped BcStc3 with its homologous of Rutstroemia sp, other species of Helotiales order, as B. cinerea. This fungus is the causal agent of bleach blonde syndrome on cheatgrass that secretes diverse phytotoxic secondary metabolites involved in pathogenesis [66]. Surprisingly, these two proteins formed a subclade with the homologous protein of Pycnora preastabilis, a lichen fungus that affects mainly conifer trees. This fungus belongs to the recently established class Candelariomycetes and is characterized by a drastic reduction of the genome size and the number of secondary metabolite biosynthetic gene clusters [67,68]. Moreover, these three proteins clustered with numerous lichenised Ascomycota of the Lecanoromycete and Xylobotryomycete classes and fungi not phylogenetically related as the little brown mushroom Galerina marginata or polymorphic black yeast-like fungi, all of them producers of diverse classes of secondary metabolites, some of them with clinical applications [66,69,70].

3.4. Deletion of the Gene Encoding Sesquiterpene Cyclase 3 (Bcstc3).

This study focused on the deletion of the gene encoding sesquiterpene cyclase 3 (Bcstc3) and the generation of overexpressed Bcstc3 mutants in the B. cinerea UCA992 strain. We replaced the original gene ORF of Bcstc3 (Bcin13g05830) with an ORF conferring resistance to hygromycin in the wild-type strain. Following transformation and selective culturing, DNA extraction and conventional PCR confirmed homologous integration of the replacement fragment at the Bcstc3-5′ and -3′ regions, respectively identified by primer pairs Bcstc3-hi5F/TrpC-T (amplicon size: 1028 bp) and Bcstc3-hi3R/TrpC-P2 (amplicon size: 720 bp). The absence of Bcstc3 alleles in the mutants, demonstrating homokaryotic profiles devoid of parental nuclei containing the Bcstc3 gene, was validated using primers Bcstc3-WT-F/WT-R (absence of amplicon size: 520 bp) (Supplementary Table S1).

Additionally, overexpressed mutants (OvBcstc3) were created by introducing an extra copy of the Bcstc3 gene under the actA gene promoter and the trpC terminator (TtrpC) from Aspergillus nidulans into the BcniaD gene locus, coding for nitrate reductase, in the wild-type strain, alongside a nourseothricin resistance cassette. The homologous integration at BcniaD-5′ and -3′ was confirmed using primer pairs BcniaD-hi5F/PactA-stc3-R (amplicon size: 4243 bp) and BcniaD-hi3R/PactA-amp-F (amplicon size 2174 bp), with the presence of BcniaD alleles in mutants corroborated by BcniaD-WT-F/WT-R primers (absence of amplicon size WT – 780 bp) (Supplementary Table 1).

3.5. Bcstc3 is Differentially Expressed during Fungal Development and Pathogenesis

The Bcstc3 gene is not only expressed in non-germinated conidia but also up-regulated during fungal vegetative growth and during plant-host interaction up to 7 and 16 times compared to the expression level in no-germinated conidia (Figure 5). These results suggest a significant role of BcStc3 in the biology of B. cinerea. Actually, Bcstc3, along with the Bcsct1 and Bcstc2 genes, is one of the most induced genes in planta, and with Bcstc4, in axenic culture (Figure 5). This expression pattern, however, differs in some aspects from that in the B05.10 strain (Supplementary Figure S2). Overall, the family gene expression is higher in the B05.10 strain than in the UCA922, as it has been reported previously in different culture conditions [35,54]. Although, in ungerminated conidia of both strains, the Bcsct5 mRNA was the most abundant, the least expressed genes were Bcstc2 and Bcstc1 in the UCA922 and B05.10 strains, respectively (Figure 5 and Supplementary Figure S2). On the other hand, during vegetative growth and in planta, the Bcstc7 gene did not change its basal expression level in the UCA922 strain compared to ungerminated conidia, but this gene was the most overexpressed of the gene family in the B05.10 strain in axenic culture (more than 2000 times respect the expression level in ungerminated conidia) and the second one (more than 600 times) after Bcstc1 (770-fold increase), in planta (Supplementary Figure S2).

Furthermore, the expression pattern is differentially regulated in a fungal strain-dependent manner and points to a functional specialization through differential expression patterns. Variations between strains were also observed in the gene family expression when the B05.10 and UCA922 strains were grown in sublethal levels of copper sulphate [54]. The generation of two new strains, the null Bcstc3 mutant and the overexpressed-Bcstc3 strain, allowed us to study the putative co-regulation of the Bcstc gene family in B. cinerea (Figure 6). In ungerminated conidia, the deletion of the Bcstc3 gene caused the down-regulation of Bcsct5, but the contrary effect was observed when the Bcstc3 gene was overexpressed, suggesting a close co-regulation of Bcstc3 and Bcstc5 genes (Figure 6). Bcstc1, Bcstc2, Bcstc4 and Bcstc7, to a greater or lesser extent, were upregulated in ungerminated conidia of both transformants compared to the wild type (Figure 6). These findings suggest that the relative Bcstc3 mRNA abundance affects the expression of the other Bcstc genes and point to the relevance of this gene family in conidial morphogenesis and/or viability. During vegetative growth, Bcsct4, Bcstc5 and Bcstc7 were up-regulated in both strains, and the Bcstc1 level almost did not change compared to the basal expression in non-germinated conidia of each strain (Figure 6).

3.6. BcStc3 is Involved in Fungal Development and Tolerance to Osmotic and Oxidative Stress

The putative role of the BcStc3 protein in the biology of Botrytis was ruled out by studying the fungal phenotypic changes due to the deletion or the constitutive expression of the Bcstc3 gene in the ΔBcstc3 and OvBcstc3 strains, respectively. The growth rate of both strains was reduced by 14% compared to the wild type on solid medium (Figure 7). However, the biomass of the OvBcstc3 strain decreased by 21% when cultured on a liquid medium, while the mutant strain grew as the wild type (Figure 7). Probably, the secondary metabolite/s related to the BcStc3 enzymatic activity could affect the fungal ability to expand on a solid substrate, resulting in a reduction in growth rate in a solid medium. On the other hand, the putative terpenoid deficiency may be compensated by activating alternative metabolic pathways that enable growth and development in a liquid environment. In fact, we found that the deletion or the constitutive expression of the Bcstc3 gene changed differentially the expression pattern of the rest of the Bcstc family genes during axenic culture, especially of Bcstc2 (Figure 6). The role of Bcstc7 and Bcstc1/Bcbot2 in mycelial development has already been ruled out [30,35], but it yet to be established whether Bcstc2, Bcstc4 and Bcstc5 are involved in the fungal growth. Recently, it has been shown that the terpene cyclase-like 2 CcPtc1 of the necrotrophic phytopathogenic fungus Cytospora chrysosperma plays a relevant role in the development avoiding the accumulation of trehalose 6-phosphate that has a cytotoxic effect [71]. The specific role of Bcstc3 in vegetative growth of B. cinerea remains to be elucidated.

A strong correlation has been identified between the regulation of secondary metabolism and stress triggers and the involvement of the secondary metabolites in environmental adaptation and stress tolerance in filamentous fungi [72,73]. The profile of secondary metabolites produced and the regulatory cell components involved in each stress response vary depending on the type of stress [74,75]. Thus, the putative role of Bcstc3 to overwhelm stress conditions was investigated by studying the effect of sorbitol and hydrogen peroxide on mycelium development when the protein was lacking or overproduced. The osmotic agent sorbitol caused a 15% reduction of the growth rate of the wild-type strain, but lack or overproduction of the BcStc3 protein slightly reduced or increased twice the sensitivity to this agent, respectively (Figure 7). We hypothesize that the secondary metabolites generated as end-products of the pathway in which Bcstc3 is involved may affect the ability of B. cinerea to respond effectively to osmotic stress conditions. Moreover, the reduced biomass in the liquid culture of the OvBcstc3 strain compared to the ΔBcstc3 mutant and the wild type may be related to the differential tolerance to osmotic stress of the fungal strains (Figure 7). The overproduction of BcStc3 caused an increase in osmotic sensitivity that triggered a reduction of fungal biomass when the OvBcstc3 strain was cultured in a liquid medium. On the contrary, the lack of Bcstc3 increased tolerance to osmotic stress enough that the mycelial of the mutant strain development was not affected (Figure 7).

It has been extensively demonstrated that the induction of secondary metabolism is associated with oxidative stress protecting fungi from ROS damage [72,76]. Lycopene, lutein, perillic acid and eremanthin are examples of terpenes that act as direct antioxidants through free radical scavenging mechanisms, although these secondary metabolites may also act as indirect antioxidants by enhancing the antioxidant status (enzymatic and non-enzymatic) [77,78].

Therefore, BcStc3 is not related to ROS production and, probably, its absence activates alternative metabolic pathways or the overexpression of other antioxidant proteins in response to external hydrogen peroxide that enables scavenging the excess of the oxidant. The enhanced BcStc3 production may cause the accumulation of the secondary metabolites related to the enzymatic activity of the protein that triggers the production of ROS or alternative metabolic pathways that alter the redox balance within the cell. The additional stress caused by the overproduction of these compounds could also overload the antioxidant defence systems of the cell, resulting in an accumulation of reactive oxygen species. In Aspergillus flavus, oxidative stress upregulates the production of aflatoxins and the terpenoid kojic acid, along with the expression of monooxygenase genes [79,80]. Kojic acid binds free Fe cations preventing non-enzymatic ROS formation or directly reacts with ROS, reducing oxidants in the cell [80]. However, the overproduction of kojic acid causes iron deficiencies that may affect the regulation of siderophore biosynthetic genes and other signalling pathways [80]. In the basidiomycete Ganoderma lucidum, the decreased production of the triterpenoid ganoderic acid was accompanied by an increase in fungal resistance to oxidative stress (increasing intracellular ROS levels and inhibiting the enzymatic activity of the antioxidant systems) [81]. Overall, although the specific role of BcStc3 in the fungal response to oxidative stress remains the subject of further studies, it seems that BcStc3 is not involved in ROS production but in regulating detoxification systems.

3.7. BcStc3 is Involved in Conidial Morphogenesis and Infection Cushions Production

Secondary metabolism gene clusters exhibit increasingly dynamic and differential expression during asexual growth, conidiation, and sexual development in filamentous fungi [82,83]. The BcStc3 protein does not seem to be related to the production of conidia since the deletion or overexpression of the Bcstc3 gene did not cause any change in the ability to produce these asexual reproductive structures (Figure 9). However, BcStc3 plays a role in conidial morphogenesis, as the size and possibly the composition/structure of the surface of conidia changed when the BcStc3 protein was overproduced (Figure 9). Additionally, the protein is also involved in conidial germination. Its deficiency resulted in a 50% reduction in the germination rate, while overproduction led to a 17% increase compared to the wild type.

The role of other members of the Bcstc gene family of B. cinerea in conidiation has also been studied with different results. The deletion of the Bcstc7 gene encoding for a sesquiterpene cyclase abolished the (+)-4-epi-eremophil-9-en-11-ols biosynthesis [35]. The role of these secondary metabolites in conidiation had been confirmed as the external addition of eremophilenols increased the formation of conidiophores [35]. Nevertheless, the deficiency of the Bcbot2 protein involved in the botrydial biosynthesis did not affect conidial morphogenesis [30]. Finally, the deletion of the Bcstc5/Bcaba5 gene did not induce any change in conidiation of the ABA-overproducing strain ATCC58025, which is known to be deficient in conidia and sclerotia formation [84].

The relationship between terpene biosynthesis and conidiation has been shown in other filamentous fungi. The Ccptc gene deletion, which encodes for a terpene cyclase in the phytopathogen fungus Cytospora chrysosperma, reduced the conidial production [71]. The Trt1 terpene cyclase of the opportunistic fungus A. terreus is involved in the biosynthesis of the mycotoxin terretonin, expressed in the conidia and activated during the germinating stage [85]. In this work, we studied the relative expression levels of the Bcstc gene family in conidia of the ΔBcstc3 and the OvBcstc3 strains compared to those in the wild-type strain (Figure 6). The deletion or overexpression of Bcstc3 led to differential changes in the expression of the remaining genes within the family. The most upregulated gene in the mutant strain was Bcstc7 (Figure 7). Bcstc5 showed the highest upregulation in the OvBcstc3 strain but was repressed in the ΔBcstc3 strain (Figure 9). These different expression patterns of the Bcstc genes likely contribute to the variations observed in the germination rate, conidial size, and aggregation between both transformants (Figure 9). Additionally, terpenes can influence conidial morphogenesis and germination, interacting with specific receptors and cellular signaling pathways in filamentous fungi. Numerous transcriptional factors in B. cinerea involved in the regulation of the secondary metabolism also control conidiation and terpene production. The transcriptional factor BcReg1 and the regulator of the abscisic acid biosynthesis BcLae1 are two examples of factors upregulating conidial morphogenesis and biosynthesis of the sesquiterpene botrydial and the polyketide botcinic acid [86,87]. On the contrary, the BcVel1 and BcLTF1 factors differentially regulate the Bcstc gen family and repressed conidiation [88,89]. Finally, recently it has been described the role of the polyketide synthase PKS15 of the entomopathogenic fungus Beauveria bassiana in the cell wall formation. The pks15 mutation caused an alteration of mannan and glucan organization in the wall, generating more elongated conidia than the wild type with smoother surfaces [90]. The enzyme and its metabolite may act in cellular signaling, and it may also be possible that the PKS15 metabolite could directly be a spore wall component.

The infection cushions are specialized structures produced by certain pathogenic fungi formed by the aggregation of fungal hyphae and specialized cells at the site of infection [91]. The infection cushions enable the fungus to efficiently infect the plant by promoting the penetration and colonization of the host tissues, playing a crucial role in the pathogenicity and spread of fungal diseases in plants [92]. They serve as physical barriers that shield the fungal hyphae from the damaging effects of ROS produced by the host plant, displaying mechanisms such as the upregulation of antioxidant enzymes or the synthesis of protective metabolites [52]. We found that the overexpression of the Bcstc3 gene resulted in a twofold increase in hydrogen peroxide production compared to both the wild-type and the mutant strains, which exhibited similar behavior to each other (Figure 10), and in a reduced tolerance to oxidative stress (Figure 7). Although the Bcstc1/Bcbot2 and Bcboa6 genes encoding for the core enzymes of the BOT and BOA biosynthetic gene clusters, respectively, have been ruled out as players in the development of infection cushions [93]. Choquer et al. (2021) have shown that both clusters are upregulated in these infective structures [28]. However, in that study, the expression of the Bcstc3 gene was not detected [52]. Altogether, these results suggest that the increased infection cushion production by the OvBcstc3 strain is more likely related to enhance ROS generation than a direct role of the protein.

3.8. BcStc3 in Involved in Virulence

Surprisingly, the ΔBcstc3 strain exhibited a higher infection efficiency than the wild type in grape, tomato fruits, in gerbera petals and tobacco leaves (Figure 8). The Bcstc3 overexpression restored infectivity to wild strain levels in grapefruits and tobacco leaves infection but caused a slight decrease in the disease index in tomato fruits compared to that of the UCA922 strain (Figure 8). In planta, the Bcstc3 deletion induced significant changes in the expression pattern of the remaining Bcstc genes, particularly the upregulation of Bcstc1/Bcbot2, which displayed a 40-fold increase in relative expression compared to the wild-type strain (Figure 6). BcSct1/BcBot2 is involved in the synthesis of botrydial, a sesquiterpene that triggers the hypersensitive response on plant tissue, inducing accumulation of ROS and phenolic compounds and causing chlorosis and cell collapse [94]. The overexpression of the Bcstc1/Bcbot2 gene during infection of tobacco leaves by the ΔBcstc3 transformant may be related to the increased production of ROS in the host interaction and the enhanced virulence of the ΔBcstc3 strain. Similarly, the null mutant of the bcpks13 encoding for a polyketide synthase involved in the melanin synthesis caused larger lesions and accumulated ROS more rapidly and abundantly than the wild type at the infection site [95]. The mutant strain did not show any advantage in the expression of ROS synthesis genes compared with the wild type, indicating that infection by the mutants could induce much stronger plant responses compared with the wild-type strain [95]. The lack of the BcStc3 protein and the BcStc1/BcBot2 overproduction may modify the oxidative balance at the infection site, increasing the fungal capability to penetrate and invade host plant tissues.

During the ΔBcstc3-tobacco leaves interaction, the Bcstc7 gene was also upregulated, 10-fold compared to its expression level in the wild-type strain (Figure 6). This structural similarity between capsidiol and eremophilenols has suggested that the fungal secondary metabolites may module the plant defense response to facilitate the attack of the plan tissue [35,36]. The upregulation of Bcsct7 in the ΔBcstc3 strain in planta may also be involved in the increased virulence of this strain. Probably, the overexpression of Bcstc1/Bcbot2 in the Bcstc3 mutant strain might also be accompanied by the upregulation of the BOA biosynthetic cluster affecting the virulence of this transformant, although it should be confirmed experimentally. BcLae1, a methyltransferase involved in the ABA biosynthesis control, is a positive regulator of the BOT and BOA clusters but represses the Bcstc3, Bcstc4 and Bcstc5 expression [86,88]. Moreover, Bcbot2, Bcstc3 and Bcstc4 are not related to a light response but are down-regulated by the action of BcLtf1, a modulator of the transcriptional responses to light, while Bcstc5 is upregulated [89]. It cannot be ruled out that the alterations of the pathway in which BcStc3 participates might cause the biosynthesis of other secondary metabolites that could act as effectors, increasing the infective capacity of the fungus and/or modulating the plant defense response.

4. Materials and Methods

4.1. Bioinformatic Analysis

The BcStc3 sequence was used as a query to perform a BlastP search against the non-redundant protein sequence database at the National Centre for Biotechnology Information (NCBI) [43]. Homologous proteins were selected from the resulting list following the criteria: percent identity (>30%), coverage (>70%), and the bit-score (>50) [51]. Sequence identities were calculated by the Sequence Identity And Similarity (SIAS) module from [http://imed.med.ucm.es/Tools/sias.html] [96]. Phylogenetic analyses were performed by the neighbor-joining method with 1000 bootstrap replicates using the Mega X software (Version 11) [97]. The predicted BcStc3 3-D model was obtained from the alphafold tool (https://alphafold.ebi.ac.uk/) [48,98], using the default parameters. The resulting protein structure was then visualized with PyMOL Molecular Graphics System (Version 1.8, Schrödinger, LLC).

4.2. Organisms, Media and Culture Conditions

The Bcstc3 knock-out mutant and the OvBcstc3 strain were generated in this work from the parental B. cinerea strain UCA992, obtained from Domecq vineyard (Jerez de la Frontera, Cádiz, Spain) and deposited in the Mycological Herbarium Collection of the University of Cádiz. All fungal strains were routinely grown in YGG medium (2% glucose, 0.5% yeast extract, 0.3% Gamborg’s B5 medium (Duchefa Biochemie), and 1.5% agar, when needed) at 20°C for three days. Tomato agar plates (25% homogenized tomato fruits (w/v), 1.5% agar, pH 5.5) were used for conidia production. Potato dextrose agar (PDA) (Sigma-Aldrich) medium (1/4) was used to estimate the infection cushion production. The conidial stock suspensions were maintained in 10% glycerol at -80°C. For DNA extraction from mycelium, fungal strains were incubated in Petri dishes containing completed medium (CM) (10 g glucose, 2 g casein peptone, 1 g casamino acids, 1 g yeast extract, 50 mL salt solution, 1 mL vitamin solution (0.5 g biotin, 50 g nicotinamide, 16 g p-aminobenzoic acid and 20 g pyridoxine hydrochloride per liter) and 2 mL microelement solution (1 g FeSO₄·7H₂O, 0.15 g CuSO₄·5H₂O, 1.61 g ZnSO₄·7H₂O, 0.1 g MnSO₄·H₂O, 0.1 g (NH₄)₆Mo₇O₂₄·4H₂O, per liter), per liter, pH 5.0) overlaid with cellophane for 3 to 4 days at 20°C.

For the S. cerevisiae strain FY834 [35,99] transformation, plates of SD-uracil medium (D-glucose 20 g, Difco^® yeast nitrogen base without amino acids 6.7 g, Clontech –Ura DO supplement 0.77 g, agar 16 g, per liter, pH 5.8) were used, and they were incubate for 3 or 4 days at 30°C.

N. tabacum var. Havana plants were maintained in a growth chamber at 22°C, 70% humidity under a light/dark cycle of 14 h light/10 h dark. Gerbera jamesonii, grapes and tomato fruits were purchased from local groceries.

4.3. Standard Molecular Methods for Gene Inactivation and Overexpression

Fungal genomic DNA was isolated following the protocol of Mansfield (1985) [100]. Bacterial plasmids were isolated using the GeneJET PCR Plasmid Purification Kit (Thermo Scientific). NanoDrop 2000c spectrophotometer (Thermo- Scientific) was used for checking purity and concentration of DNA samples. Phusion High-Fidelity DNA Polymerase (Thermo-Scientific) and Go-Taq DNA Polymerase (Promega) were used for conventional PCR. Agarose gel electrophoresis and restriction enzyme digestions were performed using standard procedures [101]. For sequence analysis, DNASTAR Lasergene package (DNASTAR, Inc.) programs were used. Oligonucleotides (Supplementary Table S1) were from Metabion International AG.

4.4. Generation of the ΔBcstc3 and OvBcstc3 Strains

For the generation of Bcstc3 (Bcin13g05830) deletion mutants, two regions flanking the Bcstc3 ORF of 910 bp and 630 bp were amplified using primer pairs Bcstc3-5F/-5R and Bcstc3-3F/-3R, respectively (Supplementary Table S1). The two DNA fragments were assembled with a hygromycin resistance cassette obtained by PCR using plasmid pCSN44 as template and primer pair Hph-F/-R and with the EcoRI/XhoI-digested plasmid pRS426 [102,103] by yeast recombinational cloning using the strain S. cerevisiae FY834 [35,99]. The resulting Bcstc3-replacement fragment (3029 bp) was amplified with primer pair Bcstc3-5F and Bcstc3-3R and used to transform protoplasts of the B. cinerea UCA992 strain generated as described previously by Schumacher (2012). Transformants were selected by hygromycin resistance [102]. Homokaryotic mutant strains were generated using a single-spore isolation protocol as described by Gonzalez-Rodriguez et al. [99] and Schumacher [102] and confirmed by PCR amplification with the primer pairs Bcstc3-WT-F and Bcstc3-WT-R (Supplementary Table 1). To confirm the homologous integration of the replacement cassette at the Bcstc3 locus, fungal DNA was isolated and amplified with the primer pairs Bcstc3-hi5F /TrpC-T and TrpC-P2/Bcstc3-hi3R to detect targeted integration at 5′ or 3′ ends, respectively (Supplementary Table S1).

For the generation of the Bcstc3-overexpressed strain (OvBcstc3), the ORF of Bcstc3 (1758 bp) was amplified using primers TtrpC-Bcstc3-F and PactA-Bcstc3-R with DNA from the parental strain UCA992 serving as the template. The amplicon was then assembled with a fragment containing the actA promoter (PactA), obtained by PCR using primers PactA-amp-F and PactA-PtrpC-R (Supplementary Table S1), and the plasmid pNDN-AGT [68] as the template, followed by digestion with SpeI and NotI enzymes of the same plasmid in yeast FY834 [35,99], resulting in the vector pNDN-Bcstc3-OVER. This resulting vector contains the ORF Bcstc3 gene under the control of the actA gene promoter and the trpC terminator (TtrpC) from A. nidulans [67], along with a nourseothricin resistance cassette (PtrpC::nat1), flanked by fragments of the B. cinerea BcniaD gene, encoding nitrate reductase. Protoplasts of the UCA992 strain were transformed with the Bcstc3Ov-replacement fragment (6400 bp), obtained by PCR amplification using primer pair BcniaD-5F and BcniaD-3R (Supplementary Table S1) and vector pNDN-Bcstc3-OVER as template. Transformants were selected for nourseothricin resistance, and heterokaryons were generated. The targeted integration of the Bcstc3 expression cassette at the BcniaD locus was checked by PCR with primer pairs BcniaD-hi5F/PactA-stc3-R and BcniaD-hi3R/PactA-amp-F, and heterokaryotic mutants were confirmed by PCR amplification with the primer pairs BcniaD-WT-F and BcniaD-WT-R (Supplementary Table S1).

4.5. Quantitative Assessment of Gene Expression via qRT-PCR

The fungal mycelium from B. cinerea UCA992 wt, ΔBcstc3, OvBcstc3 and B. cinerea B05.10 wt used for total RNA extraction was collected by filtration from 30 mL of YGG medium inoculated with 5 µL of 105 conidia/mL and incubated at 20°C for four days in darkness. To study in planta gene expression, detached leaves of N. tabacum var. Havana plants were inoculated with 5 µL droplets of a conidial suspension of 10⁵ conidia/mL in TGGK solution (60 mM KH₂PO₄, 10 mM glycine, 0.01% Tween 20, 100 mM glucose) of B. cinerea UCA992 wt, ΔBcstc3, OvBcstc3 and B. cinerea B05.10 wt and incubated for 96 hours at 22°C in darkness with 70% humidity. The infected areas were cut and frozen at -80°C until use. Mycelia and infected plant tissues were homogenized with a drill, and total RNA was extracted using Trizol Reagent (Sigma-Aldrich, T 9424), following the manufacturer’s instructions.

cDNA synthesis from 1 µg of total RNA was performed using iScript cDNA Synthesis Kit (Bio-Rad, USA) according to the manufacturer’s instructions. qRT-PCR was carried out on an iCycler iQ system (Bio-Rad, USA) using the iQ SyBR Green Supermix (Bio-Rad, USA) and primers listed in Supplementary Table S1. The actA (Bcin16g02020) and B-tub (Bcin01g08040) genes of B. cinerea, encoding for actin and tubulin, respectively, were used as internal controls to normalize the cDNA samples. The relative amount of mRNA was calculated by the ΔΔCt method [104] for the mean of three independent determinations of the threshold cycle (Ct). The results were represented as mean values of 2^{(ΔΔCt ±SD)} normalized to the expression of each gene in non-germinated conidia.

4.6. Phenotypic Characterization of Fungal Transformants

4.6.1. Vegetative Growth and Tolerance to Stress Agents

Petri dishes containing YGG-agar medium were inoculated with 10 µL droplets of a conidial suspension of 5 x 10⁶ conidia/mL and incubated at 20°C to test for altered fungal growth. The colony radii were measured daily for four days, and the radial growth rate was calculated by plotting the colony radius over time, which fitted to a linear model (Pearson’s correlation coefficient value (r² ≥ 0.98)). Results are presented as the mean values of the growth rates in cm/day ± standard deviation of at least six independent experiments. To assess the sensitivity of fungal strains to stress conditions, the YGG-agar medium was supplemented with 1.4 M Sorbitol or 1.5 mM H₂O₂ for osmotic or oxidative stress, respectively. The relative growth inhibition rate was calculated as follows: Inhibition rate(%)=[1 -(T₄ – T₃)/(C₄ – C₃)]× 100, where T and C represent the radius of the colonies (cm) in treated and control plates, respectively, and 4 and 3 the days of incubation at 20˚C [105].

In order to evaluate fungal biomass accumulation in liquid culture, 30 mL of YGG medium was inoculated with 10 µL of a conidial suspension of 5 x 10⁶ conidia/mL and incubated at 20°C without shaking for four days. Mycelia were harvested by filtration, washed with sterile water, and dried for 1 hour at room temperature to obtain the fresh weight. The dry weight was estimated by incubating the mycelium at 50°C until a constant weight was achieved (dry weight). The experiment was conducted in triplicate, and the results are presented as the mean values of the biomass (g) ± standard deviation.

4.6.2. Conidial Production and Germination

Tomato agar plates (25% homogenized tomato fruits (w/v), 1.5% agar, pH 5.5) were inoculated with agar plugs containing young mycelium (0.5-cm YGG-agar cubes), incubated for three days at 20°C in dark conditions, and exposed for 16 h to near-UV light at a distance of 30–40 cm. After seven more days in darkness, conidia were collected as described by van der Vlugt-Bergmans et al. [101] and quantified using a hemocytometer under a bright-field microscope (Olympus BX-50). The results are presented as the mean values ± standard deviation of six individual experiments.

To assess the conidial germination rate, aliquots of 20 µL of a 10⁵ conidia/mL suspension in YGG medium were placed on a sterile glass slide and incubated at 20°C in the dark and humidity conditions. After 6 hours, bright-field pictures of conidia were taken with an Olympus BX-50 microscope, and in each field examined, at least 100 conidia were counted. Conidia were considered germinated when the elongating germ tube was longer than the conidial longitudinal diameter. Conidial germination rate values are expressed as the percentage of germinated conidia to the total conidia counted.

To study conidial aggregation, the percentage of conidia clustered in groups of two or more was calculated from random bright-field pictures of 10000 conidia in YGG incubated on a sterile glass slide for 6 hours at 20°C in high humidity conditions. On the other hand, conidial aggregation was also estimated from the conidial sedimentation out of the suspension. The optical density of a 10⁸ conidia/mL suspension in water was measured at 600 nm, and of the remained suspended conidia after 2 hours at room temperature. The optical density values were used to estimate the conidial concentration using a standard curve prepared from dilutions of a conidial suspension of known concentration. The conidial aggregation was calculated using the following equation: % aggregated conidia = 100 x ((initial conidia concentration- final conidia concentration)/initial conidia concentration).

4.6.3. Virulence Assays

Detached tobacco leaves, gerbera petals, and tomatoes and grapefruits were inoculated with 5 µL droplets containing 2500 conidia in TGGK solution (60 mM KH2PO4, 10 mM glycine, 0.01% Tween 20 and 100 mM glucose), incubated in the dark at 20°C under conditions of high humidity, and photographed every 24 hours. The diameter and length of the lesions on tobacco leaves and gerbera petals, respectively, were measured using ImageJ software [106], and the lesion growth rate was calculated and expressed in cm/day or mm/day, as indicated in Figure 8.

A semi quantitative scale with four grades according to the severity of disease symptoms was used to evaluate the virulence of the fungus in fruits. The results are presented as the percentage of fruits in each disease grade to the total number of infected fruits (>30). Disease scores from 0 to 4 were subsequently converted to a percentage fruit disease index (%DI) as follows: [∑(disease rank × number of fruits)/(highest disease rank × total number of fruits)] ×100.

4.6.4. Reactive Oxygen Species and Infection Cushions Production

Quantitative determination of H₂O₂ was performed incubating 24 mg of fresh mycelium with 1 mL of DAB solution for 5 hours at room temperature. The mycelium was harvested, and the optical density of the supernatant was measured at 471 nm. The absorbance values were compared with a standard curve prepared with known peroxide concentrations. The results were presented as ng of H₂O₂/mg of mycelium.

In vivo ROS was analyzed in tobacco leaves inoculated with 5 µL droplets of a 5 × 10⁵ conidia/mL conidial suspension in TGGK solution and incubated for 48 hours at 20°C. Leaf discs, including the infected area, were cut and vacuum infiltrated for 1 hour with 1 mg/mL of DAB solution, pH 3.7. Discs were boiled in ethanol for 10 minutes to eliminate chlorophyll and photographed. ROS production was measured by ImageJ software, and values were expressed as the percentage of brown pixels detected around the infection point.

The production of infection cushions was assessed following the semiquantitative method described by Sekulska-Nalewajko et al. in 2016 [107].

4.7. Statistical Analysis

Statistical Analysis was performed with the SPSS v 24 (IBM) software package. The normal distribution of data was analyzed with the Kolmogorov-Smirnov (N>50 samples) or Shapiro Wilks (N<50 samples) tests. Depending upon the results from normality tests, statistical significance was analyzed with the T-test or the Mann-Whitney test for comparison of normally distributed or nonparametric data, respectively. Statistical tests were significant if p-value < 0.05.

4.8. Metabolomic Characterization of the ΔBcstc3 and Overexpressed OvBcstc3 Mutant Strains

For the metabolomic characterization, the fungal strains were grown on malt agar medium (20 g/L D-glucose, 20 g/L malt extract, 1 g/L peptone, 20 g/L agar, pH 6.5–7) at 25°C to produce mycelium plugs (0.8 mm) that were further used to inoculate medium for metabolite production. All studied strains were fermented in 40 Roux culture bottles (1000 mL), each containing 150 mL of modified Czapek-Dox medium (50 g glucose, 1 g yeast extract, 1 g K₂HPO₄, 2.5 g NaNO₃, 0.5 g MgSO₄⋅7H2O, 0.01 g FeSO₄⋅7H₂O, 0.005 g CuSO₄⋅5H₂O, pH 6.5-7, 1L of water). Bottles were inoculated with six mycelium plugs per bottle and incubated for 27 days at 25°C on surface culture under day-light.

Then 6 L of culture medium from each of the mutant strains were filtered under vacuum to remove mycelium, and the filtrates were saturated with NaCl and subjected to liquid−liquid extraction with ethyl acetate (EtOAc × 3) and dried over with anhydrous sodium sulphate. After filtration, solvents were evaporated at reduced pressure to yield the crude extracts as yellow oils, 173 mg from (ΔBcstc3) and 153 mg from (OvBcstc3). For metabolites isolation and characterization, the crude extracts were initially fractionated by column chromatography on silica gel, with a mixture of ethyl acetate/hexane containing increasing percentages (10−100%) of ethyl acetate. Different fractions were further purified by HPLC with a mixture of acetone, ethyl acetate and hexane. The metabolites purified were subjected to extensive spectroscopic analysis by ¹H-NMR and ¹³C-NMR using 1D and 2D NMR, HRMS and IR techniques and their optical activity measured (α) for identification purposes.

5. Conclusions

This study has advanced our understanding of the sesquiterpene cyclase BcStc3’s role in B. cinerea, shedding light on the molecular biology and pathogenicity of this phytopathogenic fungus. Through the generation and characterization of BcStc3-deficient and overexpressed mutant strains, we have unveiled the crucial involvement of BcStc3 in regulating the expression of genes within the sesquiterpene cyclase family and its impact on virulence, stress response, and secondary metabolite production.

Although the quest for new metabolites stemming from BcStc3 activity was not fulfilled, our findings significantly contribute to the phenotypic characterization of B. cinerea. We have identified notable differences in oxidative and osmotic stress tolerance, conidial morphology, and infection cushion formation. These results emphasize BcStc3’s importance not just in specific metabolite biosynthesis but also in the fungus’s adaptation and survival under challenging environmental conditions and during host-pathogen interactions.

Furthermore, this work has illustrated the complexity of co-transcriptional regulation among Bcstc family genes, providing insights into the genetic networks that modulate B. cinerea’s pathogenicity. The detailed examination of interactions between BcStc3 and other secondary metabolism components lays a valuable foundation for future research aimed at unravelling the molecular mechanisms behind fungal pathogenicity and stress resistance.

In summary, our discoveries significantly expand the current understanding of B. cinerea biology and offer groundwork for developing disease control strategies by targeting key metabolic pathways. This study highlights the importance of comprehensive phenotypic characterization and gene expression analysis in revealing critical gene functions in complex biological contexts.