Effect of Farnesol in Trichoderma Physiology and in Fungal–Plant Interaction

Farnesol is an isoprenoid intermediate in the mevalonate (MVA) pathway and is produced by the dephosphorylation of farnesyl diphosphate. Farnesol plays a central role in cell growth and differentiation, controls production of ubiquinone and ergosterol, and participates in the regulation of filamentation and biofilm formation. Despite these important functions, studies of farnesol in filamentous fungi are limited, and information on its effects on antifungal and/or biocontrol activity is scarce. In the present article, we identified the Trichoderma harzianum gene dpp1, encoding a diacylglycerol pyrophosphatase that catalyzes production of farnesol from farnesol diphosphate. We analyzed the function of dpp1 to address the importance of farnesol in Trichoderma physiology and ecology. Overexpression of dpp1 in T. harzianum caused an expected increase in farnesol production as well as a marked change in squalene and ergosterol levels, but overexpression did not affect antifungal activity. In interaction with plants, a dpp1-overexpressing transformant acted as a sensitizing agent in that it up-regulated expression of plant defense salicylate-related genes in the presence of a fungal plant pathogen. In addition, toxicity of farnesol on Trichoderma and plants was examined. Finally, a phylogenetic study of dpp1 was performed to understand its evolutionary history as a primary metabolite gene. This article represents a step forward in the acquisition of knowledge on the role of farnesol in fungal physiology and in fungus-environment interactions.


Introduction
Terpenes are the most abundant secondary metabolites in nature and have a wide variety of physiological functions. Trichoderma species produce a number of these compounds [1,2]. Up to now, several genes involved in their biosynthesis have been characterized and their biological functions described. Among them, the genes belonging to the ergosterol biosynthetic pathway, e.g., erg1, erg7 and erg9, are the best characterized [3][4][5].
Farnesyl diphosphate (FDP) is an essential intermediate in biosynthesis of terpenes and thereby plays a central role in terpene metabolism. FDP acts a primary metabolite that participates in the prenylation of proteins involved in the regulation of cell growth, proliferation, cytoskeletal reorganization, and in the trafficking of intracellular membrane vesicles, which are essential to ensure proper cellular function [6]. In addition, the level of FDP is also related with the production of farnesol, which regulates important processes as fungal filamentation [7], formation of biofilms, or quorum sensing responses, acting in these functions as a tyrosol-antagonistic molecule [8]. Farnesol itself also exhibits a toxic effect MVA, mevalonic acid; hmgR, fppS, erg9, erg1, and erg7, correspond to genes encoding for hydroxymethylglutaryl-CoA reductase, farnesyl diphosphate synthase, squalene synthase, squalene epoxidase, and oxidosqualene cyclase, respectively. A green question mark (?) was included to indicate the gene involved in FDP dephosphorylation whose identification is the main objective of the current study. (B) Chemical structure of farnesol.

Strains Used and Culture Conditions
Trichoderma harzianum CECT 2413 (=T34) was used as a recipient in dpp1 gene overexpression experiments. Trichoderma strains were maintained and sporulated on potato dextrose agar (PDA) medium (Becton Dickinson Co, Sparks, MA, USA). To induce sporulation, strains were incubated at 28 °C for 4-7 days in the dark. Botrytis cinerea strain B05.10 was used in the antifungal and in plant infection assays. Sporulation of this strain was induced by growing it on V8-juice agar medium (2% agar) (Oxoid Ltd., Basingstoke, UK.) for 7 days at 21 °C and a photoperiod of 16 h light/8 h dark. Rhizoctonia solani strain R43 was also used in antifungal assays and was grown on PDA medium at 28 °C for 5-7 days in the dark.
Trichoderma strains generated during the current study and the Trichoderma control strain (T34-43b1.3) were also grown in liquid Trichoderma minimal medium (MMT) [18] to biosynthetic pathway, pointing out the genes analyzed in this study. FDP, farnesyl diphosphate; GGPP, geranyl geranyl pyrophosphate; MVA, mevalonic acid; hmgR, fppS, erg9, erg1, and erg7, correspond to genes encoding for hydroxymethylglutaryl-CoA reductase, farnesyl diphosphate synthase, squalene synthase, squalene epoxidase, and oxidosqualene cyclase, respectively. A green question mark (?) was included to indicate the gene involved in FDP dephosphorylation whose identification is the main objective of the current study. (B) Chemical structure of farnesol.

Strains Used and Culture Conditions
Trichoderma harzianum CECT 2413 (=T34) was used as a recipient in dpp1 gene overexpression experiments. Trichoderma strains were maintained and sporulated on potato dextrose agar (PDA) medium (Becton Dickinson Co, Sparks, MA, USA). To induce sporulation, strains were incubated at 28 • C for 4-7 days in the dark. Botrytis cinerea strain B05.10 was used in the antifungal and in plant infection assays. Sporulation of this strain was induced by growing it on V8-juice agar medium (2% agar) (Oxoid Ltd., Basingstoke, UK) for 7 days at 21 • C and a photoperiod of 16 h light/8 h dark. Rhizoctonia solani strain R43 was also used in antifungal assays and was grown on PDA medium at 28 • C for 5-7 days in the dark.
Trichoderma strains generated during the current study and the Trichoderma control strain (T34-43b1.3) were also grown in liquid Trichoderma minimal medium (MMT) [18] to determine the effect of exogenous farnesol on production of ergosterol, squalene, and in the level of expression of genes related to the terpene biosynthetic pathway.

Overexpression of dpp1 Gene in T. harzianum CECT 2413 (T34)
Construction of pTCdpp1-b1. The homolog to DPP1/LPP1 gene (dpp1) identified in the genome sequence of T. harzianum CBS226.95 v1.0 (Join Genome Institute: jgi.doe.gov/ (accessed on 8 November 2022)) was PCR amplified with the Pfu polymerase (EURx, Gdansk, Poland), using as template the genomic DNA of the T. harzianum T34 strain, and primers Th-dpp1 5F/Th-dpp1 3R (Table S1). The resulting amplified fragment (1137 bp) was phosphorylated with the polynucleotide kinase (Thermo Fisher, Waltham, MA, USA), and, after gel purification, the fragment was ligated to the pBluescript KS+ plasmid (Stratagene, La Jolla, CA, USA) linearized with EcoRV and dephosphorylated. The resulting plasmid, pKS_dpp1 (4098 bp), was digested with endonucleases HindIII-EcoRI, to isolate the dpp1-gene fragment, which was then treated with Klenow, gel purified, and ligated to pTAcbh2 [19] previously digested with NcoI, filled with Klenow fragment (Thermo Fisher) and dephosphorylated with CIAP (Fermentas). The resulting plasmid pTCdpp1 (6495 bp) was digested with BglII-HindIII to isolate the 3723 bp fragment containing the Th-dpp1 overexpression cassette, filled with Klenow fragment, and ligated to the plasmid pJL43b1 [20], previously digested with Acc65I and treated with Klenow to obtain the plasmid pTCdpp1-b1 (8219 bp) ( Figure 2A). Finally, this plasmid was linearized with HindIII prior to its transformation into T. harzianum protoplasts. determine the effect of exogenous farnesol on production of ergosterol, squalene, and in the level of expression of genes related to the terpene biosynthetic pathway.

Overexpression of dpp1 Gene in T. harzianum CECT 2413 (T34)
Construction of pTCdpp1-b1. The homolog to DPP1/LPP1 gene (dpp1) identified in the genome sequence of T. harzianum CBS226.95 v1.0 (Join Genome Institute: jgi.doe.gov/) was PCR amplified with the Pfu polymerase (EURx, Gdansk, Poland), using as template the genomic DNA of the T. harzianum T34 strain, and primers Th-dpp1 5F/Th-dpp1 3R (Table S1). The resulting amplified fragment (1,137 bp) was phosphorylated with the polynucleotide kinase (Thermo Fisher, Waltham, MA, USA), and, after gel purification, the fragment was ligated to the pBluescript KS+ plasmid (Stratagene, La Jolla, CA, USA) linearized with EcoRV and dephosphorylated. The resulting plasmid, pKS_dpp1 (4,098 bp), was digested with endonucleases HindIII-EcoRI, to isolate the dpp1-gene fragment, which was then treated with Klenow, gel purified, and ligated to pTAcbh2 [19] previously digested with NcoI, filled with Klenow fragment (Thermo Fisher) and dephosphorylated with CIAP (Fermentas). The resulting plasmid pTCdpp1 (6,495 bp) was digested with BglII-HindIII to isolate the 3,723 bp fragment containing the Th-dpp1 overexpression cassette, filled with Klenow fragment, and ligated to the plasmid pJL43b1 [20], previously digested with Acc65I and treated with Klenow to obtain the plasmid pTCdpp1-b1 (8,219 bp) ( Figure 2A). Finally, this plasmid was linearized with HindIII prior to its transformation into T. harzianum protoplasts. (control) and six transformants with plasmid pTCdpp1-b1, digested with EcoRI, and hybridized with the 622 bp probe (a) indicated in panel A. Note that in T34 and T34-dpp1.9 (included as a negative control in the Southern), we only observed the wild-type 4.910 bp band (c), corresponding to the endogenous copy of the dpp1 gene. (b) 2,341 bp EcoRI fragment expected for the dpp1-overexpression transformants in the southern experiments.

Transformation of pTCdpp1-b1 Plasmid in T. harzianum T34
Plasmid pTCdpp1-b1 was linearized with HindIII and transformed in the T. harzianum T34 strain following a protoplast-based procedure [20], but using Czapek as regeneration medium (sucrose, 30 g/L; NaNO3, 2 g/L; K2HPO4, 0.5 g/L; MgSO4.7H2O, 0.5 g/L; Figure 2. (A) Schematic representation of pTCdpp1-b1. PgpdA, promoter of the Aspergillus nidulans gpdA gene; ble, phleomycin resistance gene from Streptoalloteichus hindustanus; TCYC1, transcriptional terminator of the Saccharomyces cerevisiae cytochrome C oxidase gene; dpp1, T. harzianum dpp1 gene. Ptadir, promoter of the T. harzianum tss1 gene, and Tcbh2, transcriptional terminator of the T. reesei cellobiohydrolase 2 encoding gene. (B) Southern blot analysis of genomic DNAs from T34 (control) and six transformants with plasmid pTCdpp1-b1, digested with EcoRI, and hybridized with the 622 bp probe (a) indicated in panel A. Note that in T34 and T34-dpp1.9 (included as a negative control in the Southern), we only observed the wild-type 4.910 bp band (c), corresponding to the endogenous copy of the dpp1 gene. (b) 2341 bp EcoRI fragment expected for the dpp1-overexpression transformants in the southern experiments.

DNA Probes Labeling and Southern Hybridization
Genomic DNA purification, labeling of DNA probes, and southern hybridization experiments were performed as previously described [20].

Real-Time qPCR (qPCR) Analysis
Reactions were carried out as described previously [21]. Expression ratios and values of statistical significance were calculated using REST © 2022 software [22], using the α-actin as a housekeeping internal control gene, for both Trichoderma and tomato gene expression analysis. Each sample was analyzed in triplicate.
In Trichoderma strains, expression of genes hmgR, encoding the hydroxymethylglutaryl-CoA reductase; erg1, encoding the squalene epoxidase; erg7, encoding the oxidosqualenelanosterol cyclase; erg9, encoding squalene synthase; and dpp1, characterized in the present study, all involved to the terpene pathway, and additionally laeA, veA and velB, belonging to the regulatory VELVET complex, was analyzed by qPCR.
In the tomato qPCR studies, four different groups of genes were analyzed: (i) genes involved in the salicylate (SA) signaling pathway, including PR1b1, and PR-P2 [23]; (ii) genes belonging to the jasmonate/ethylene (JA/ET) signaling pathways, including PINI, PINII, TomLoxA [23]; ACCS that encodes the enzyme ACC (1-aminocyclopropane 1-carboxylate), the latter catalyzing the synthesis of ACC in the ethylene (a plant hormone) biosynthetic pathway as a derived compound of the methionine cycle; and (iii) genes involved in growth and development, including SUCS (encoding sucrose synthase, which is related to plant growth) and GAI, the product of which is a DELLA protein growth repressor (a plant development-related gene) [21,24].
Oligonucleotide sequences and PCR amplification efficiencies are included in Table S1.

Metabolomic Characterization
For this study, liquid cultures were filtered through Miracloth ® (Millipore Co., Billerica, MA, USA) sterile filters, and both the mycelia and the culture filtrates were used for further analyses.

Quantification of Farnesol
Levels of intra-and extracellular cellular farnesol were quantified following an HPLC procedure. This compound was quantified from both culture broths (=culture filtrates) and mycelia through the following steps. (i) Extraction of farnesol from culture broths: First, 2 mL of broth were extracted with 400 µL of ethyl acetate, mixed and centrifuged for 10 min at 4000 rpm. Supernatant was collected and extracted twice again following the same procedure. Finally, the extract was evaporated in a SpeedVac vacuum concentrator (Thermo Fisher) and resuspended in 50 µL acetonitrile. (ii) Extraction of farnesol from mycelia: Mycelia were frozen at −80 • C, freeze-dried, and pestled to powder in a mortar. Mycelia (50 mg each) were resuspended into 1 mL of Tris-HCl 50 mM, pH 6.8 buffer, extracted with 200 µL of ethyl acetate and processed as described above. (iii) HPLC quantification: A measure of 20 µL of the concentrated extracts was applied to a reversed-phase C18 column of 150 mm × 4.6 mm i.d., 4 µm (octadecyl-silane (ODS)) (YMC Europe GmbH) and eluted with a 1 mL/min flux of mobile phase (90% water/10% acetonitrile at zero time, performing a gradient to reach 100% acetonitrile in 40 min, then returned to 10% acetonitrile at 50 min). Using these conditions, farnesol eluted at 34.7 min and quantification was carried out by comparison of the peak area with a standard curve. For determination of the specific production of farnesol, the total protein concentration from culture supernatants was quantified as described by Bradford (1976) [25].

Quantification of Ergosterol and Squalene
Total intracellular sterols were extracted from the mycelia harvested from the same cultures used for HA quantification. Ergosterol and squalene contents were calculated as reported previously [4,26]. Antifungal assays were made on leaves detached from 4-week-old tomato plants. Leaves were first inoculated with a suspension of B. cinerea B05.10 spores (5 × 10 5 spores/mL) placed in two spots/leaf. Once the spots were dried, 15 µL of 96 h PDB broths collected from the Trichoderma strains analyzed were placed over each spot. Leaves were incubated for four days, and the lesion diameters were measured.

Antifungal Assays on Membranes against Rhizoctonia solani
The ability of dpp1-overexpressed transformants to inhibit growth of the fungal phytopathogen R. solani R43 was assessed by a membrane-based approach that was performed as described previously [27]. Tomato seeds were coated with spores of T34-43b1.3 and T34-dpp1.3 strains as described previously [27]. Seeds were sown in Kekklä 50/50 substrate and incubated at 21 • C, with 50% humidity with a photoperiod of 18 h:6 h light:dark. Plants were watered with fertilizer as needed. After four weeks of growth, plants were collected and parameters such as stem thickness and stem height were recorded. Later, second level leaves of only one of the sides of the plant were infected with B. cinerea spores (spots of 15 µL at a concentration of 5 × 10 5 spores/mL). The infected plants were placed in a humid chamber and incubated for four additional days using the same conditions indicated above.
Leaves from non-infected and infected plants were harvested and used for RNA extraction to analyze the expression of plant defense-related genes: PR1b1 and PR-P2, related to the salicylate (SA) defense signaling pathway, and PINI, PINII and TomLoxA, related to the jasmonate/ethylene (JA/ET) signaling pathways [23].

Tomato-Trichoderma Hydroponic Cultures for Analysis of the Effect of dpp1
Overexpression on the Ability to Colonize Tomato Roots Tomato hydroponic cultures were grown as described previously [19]. Once grown, roots were recovered, rinsed with sterile distilled water, and used for DNA extraction followed by quantitative PCR gene analysis. Relative amounts of Tomato gpd (glyceraldehyde 3 -phosphate dehydrogenase) and T34 α-actin genes were determined for this purpose.

Phylogenetic Analysis
Orthologs of dpp1 homologs from 34 Trichoderma species representing the breadth of known phylogenetic diversity of this genus were used for this study. The nucleotide and predicted amino acid sequences of the orthologs were aligned using MUSCLE as implemented in MEGA version 10.2.6 [29]. The alignments were then subjected to a maximum likelihood analysis (ML) using the program IQ-Tree v. 1.6.7. software [30]. Branch support was determined by bootstrap analysis using 1000 pseudoreplicates. Visualization of the phylogenetic trees was carried out using the program FigTree v1.4.4 (http://tree.bio.ed.ac.uk/, accessed on 8 November 2022). dpp1 nucleotide sequences were retrieved from the genome of the 34 Trichoderma species used in this study. Among them, 15 genomes were already available at the National Center for Biotechnology Information (NCBI) database (Table S2), while the dpp1 orthologs of the remaining 19 Trichoderma species were retrieved from genome sequences generated in previous studies [31] and were deposited at DDBJ/ENA/GenBank under the following accession numbers: OP870016 (for T. albolutescens CBS 119286), OP870017 (T.

Identification of T. harzianum Homologs of DPP1/LPP1
BLASTn analysis using the S. cerevisiae pyrophosphatase genes DPP1 and LPP1 as query sequences revealed the presence of a single homolog designated dpp1 of these genes in the genome sequence of T. harzianum strain CECT2413 (T34). In BLASTp analysis using the deduced amino acid sequence of T34 dpp1 as a query sequence against the GenBank nonredundant database of fungal sequences, the highest BLASTp scores were for putative diacylglycerol diphosphate phosphatases from other Trichoderma species, with percentages of identity ranging from 82% to 99% (E values = 1 × 10 −161 to 0.0) for homologs from T. arundinaceum and T. lentiforme, respectively.

Effect of Overexpression of T. harzianum dpp1 on Fungal Growth
Eight phleomycin-resistant transformants were recovered following the transformation of T34 protoplasts with plasmid pTCdpp1-b1 ( Figure 2A). In an initial PCR analysis using oligonucleotides TADIR2 and Th-dpp1-E4-1R (Table S1), five of the transformants yielded the 1235-bp amplicon corresponding to the tss1 promoter and dpp1 gene in the overexpression construct.
Genomic DNAs of the five transformants were digested with EcoRI and analyzed by Southern blot, using as a probe a 622 bp internal fragment to T. harzianum dpp1, which was PCR amplified using oligonucleotides Th-dpp1-E4-1F/Th-dpp1-E4-3R (Table S1). The same membrane was used to hybridize with a 201 bp probe corresponding to the ble gene, conferring phleomycin/bleomycin resistance, to confirm the presence of this gene in the genome of the isolated transformant. The ble probe was obtained by PCR amplification using oligonucleotides Phleo-3/Phleo-4 (Table S1) from pJL43b1 plasmid [20]. All five PCR positive transformants gave the expected Southern hybridization 2341 bp band with the 622 bp internal probe to dpp1 gene ( Figure 2B) and with the 201 bp internal to ble gene.

RT-qPCR Analysis of T. harzianum Terpene-Related Gene Expression
These studies were carried out using RNAs purified from 96 h PDB grown mycelia.

Expression of dpp1 in T34 Transformants with Plasmid pTCdpp1-b1
Expression of dpp1 was almost undetectable in the wild-type strain (T34-43b1.3) reaching a relative level of 0.003 (p = 0.000) fold versus the level observed for the actin gene, the latter being a housekeeping gene used as a reference in the qPCR analyses ( Figure 3A). Furthermore, the expression of this gene was strongly increased in all of the strains isolated by transformation with plasmid pTCdpp1-b1, being T34-dpp1.2 and T34-dpp1.3 those with the greatest up-regulation, reaching relative levels of 10.464 (p = 0.000) and 14.312 (p = 0.000) fold versus actin gene, respectively ( Figure 3A). Then, these two transformants were selected for further studies. In addition, when the expression was compared versus the control strain, T34-43b1, levels reached ratios of 3095.7 (p = 0.000) and 6891.5 (p = 0.000) fold for T34-dpp1.2 and T34-dpp1.3, respectively ( Figure 3B).

Effect of dpp1 Overexpression on the Transcription Level of Other Terpene Biosynthetic Genes
The expression of hmgR, encoding for the hydroxymethylglutaryl-CoA reductase, dpp1 (farnesyl pyrophosphatase), erg9 (squalene synthase), erg1 (squalene epoxidase), and erg7 (oxidosqualene cyclase) was also determined by qPCR. Results indicated that the level of expression of other terpene-related genes in these two selected dpp1-overexpressing transformants was only slightly affected, with non-statistically significant or with expression ratios between 0.5 and 2-fold in all cases, in comparison with the control strain, T34-43b.1 ( Figure 3B). Figure 3. qPCR analysis from 96 h PDB grown mycelia to assess the transcription level of: (A) dpp1 gene in the control strain, T34-43b1.3, that was generated from T. harzianum CECT 2413 by transformation with the empty expression vector used for dpp1 overexpression, and in two dpp1-overexpressing transformants T34-dpp1.2 and T34-dpp1.3, relative to the expression of actin gene, used as an internal control (housekeeping). (B) Five genes involved in the terpene biosynthetic pathway in the two selected transformants compared versus expression in T34-43b.1. In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric values were labeled with an asterisk (*) at the bottom of each graphic.

Production of Farnesol by dpp1-Overexpressed Transformants
Production was determined from cultures grown for 24 h. HPLC analyses revealed that the overexpression of dpp1 increased levels of farnesol in transformant T34-dpp1.3 by 868% and 216%, in culture broth and mycelia extracts, respectively, compared to the levels observed in the same samples of the control strain T34-43b1.3 (Tables 1 and 2) (Figures 4 and S1). Despite these results, the observed differences do not correspond with the drastic relative increase in expression of the dpp1 gene in the two overexpressing transformants. gene in the control strain, T34-43b1.3, that was generated from T. harzianum CECT 2413 by transformation with the empty expression vector used for dpp1 overexpression, and in two dpp1-overexpressing transformants T34-dpp1.2 and T34-dpp1.3, relative to the expression of actin gene, used as an internal control (housekeeping). (B) Five genes involved in the terpene biosynthetic pathway in the two selected transformants compared versus expression in T34-43b.1. In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric values were labeled with an asterisk (*) at the bottom of each graphic.

Effect of dpp1 Overexpression on the Transcription Level of Other Terpene Biosynthetic Genes
The expression of hmgR, encoding for the hydroxymethylglutaryl-CoA reductase, dpp1 (farnesyl pyrophosphatase), erg9 (squalene synthase), erg1 (squalene epoxidase), and erg7 (oxidosqualene cyclase) was also determined by qPCR. Results indicated that the level of expression of other terpene-related genes in these two selected dpp1-overexpressing transformants was only slightly affected, with non-statistically significant or with expression ratios between 0.5 and 2-fold in all cases, in comparison with the control strain, T34-43b.1 ( Figure 3B).

Production of Farnesol by dpp1-Overexpressed Transformants
Production was determined from cultures grown for 24 h. HPLC analyses revealed that the overexpression of dpp1 increased levels of farnesol in transformant T34-dpp1.3 by 868% and 216%, in culture broth and mycelia extracts, respectively, compared to the levels observed in the same samples of the control strain T34-43b1.3 (Tables 1 and 2) (Figures 4 and S1). Despite these results, the observed differences do not correspond with the drastic relative increase in expression of the dpp1 gene in the two overexpressing transformants.
Furthermore, the levels of farnesol were much higher in mycelia than in culture broths, indicating that this compound is mainly accumulated intracellularly (Tables 1 and 2).

Importance of Farnesol in the Regulation of Circadian Cycle-Related Genes in T. harzianum
The expression of laeA, veA and velB genes was analyzed by qPCR from 96 h PDB grown mycelia. LaeA/LAE1 (loss of aflR expression A) is a nuclear methyltransferase-domain containing protein that interacts with VeA and VelB to form the VELVET heterotrimeric complex. This complex synchronizes secondary metabolism and fungal development in relation to the light levels [32]. Expression of these three genes did not significantly differ between transformant T34-dpp1.2 and the control strain T34-43b1.3 ( Figure  6). However, the expression of veA and velB was significantly up-regulated in transformant T34-dpp1.3 compared to T34-43b1.3, with relative expression levels that were 2.984 (p = 0.000) and 3.356 (p = 0.000) fold higher in that dpp1-overexpressing transformant, respectively ( Figure 6). Considering that the expression level of dpp1 and production of  Production was determined from mycelia grown for 24 to 96 h. In 96 h mycelia, overexpression of the dpp1 gene resulted in a 92% decrease in squalene levels and an 80% increase in ergosterol levels, compared versus the control strain ( Figure 5). These data indicate that overexpression of the dpp1 gene, and the subsequent increase in the level farnesol production, resulted in a drastic alteration in the level of the other terpenic compounds, which shows the important role of Dpp1 as a regulator of the intracellular FDP levels.
J . Fungi 2022, 8, 1266 11 of 20 farnesol by transformant T34-dpp1.3 were significantly higher than in T34-dpp1.2, these data indicate that the increase in the production of farnesol affected the expression of genes involved in the regulation of circadian cycle, which agrees with the role of farnesol as a key molecule in maintaining fungal homeostasis. However, these data are not conclusive and would require a more detailed analysis in the future.  In the graphic representation, those statistically significant differential expres-

Effect of External Addition of Farnesol on the Squalene-Ergosterol Levels in T34-43b1.3 versus T34-dpp1.3
Strains T34-43b1.3 and T34-dpp1.3 were grown for 96 h in liquid minimal medium amended with 5 µM, 25 µM and 100 µM of farnesol. Mycelia were then collected by filtration, and the concentrations of squalene and ergosterol were quantified. The production of squalene was lower in the T34-dpp1.3 transformant than in the control T34-43b1.3 at all farnesol concentrations. For each strain, there were small but non-statistically significant differences in squalene levels produced at the three farnesol concentrations used ( Figure S2). The levels of ergosterol produced by the dpp1-overexpressed and wild-type strains were similar in all concentrations used except in 25 µM farnesol treatment, where the overexpression strain produced 18.7% more ergosterol than the wild-type strain.
When the expression of terpene biosynthetic genes was analyzed in T34-dpp1.3 transformant versus the control T34-43b1.3 strain, in the presence of 100 µM farnesol, no significant differences were observed compared to the levels detected in the absence of farnesol ( Figure S3). Furthermore, the pattern of expression in MMT (96 h) is clearly different from that observed in PDB (96 h) (see Figure 3). In MMT, at 96 h, all the terpene genes were down-regulated in the T34-dpp1.3 overexpressed transformant compared to T34-43b1.3, being especially marked the down-regulation of the erg7 gene observed both with and without (woF) farnesol ( Figure S3).
These data indicated that exogenous farnesol does not significantly affect the level of transcription of the key terpene genes analyzed. The differences observed in the production of ergosterol in the absence of farnesol almost disappeared when exogenous farnesol was added to the media, indicating that the added farnesol might modify the regulatory network of the farnesol-ergosterol pathways ( Figure S3).

Importance of Farnesol in the Regulation of Circadian Cycle-Related Genes in T. harzianum
The expression of laeA, veA and velB genes was analyzed by qPCR from 96 h PDB grown mycelia. LaeA/LAE1 (loss of aflR expression A) is a nuclear methyltransferasedomain containing protein that interacts with VeA and VelB to form the VELVET heterotrimeric complex. This complex synchronizes secondary metabolism and fungal development in relation to the light levels [32]. Expression of these three genes did not significantly differ between transformant T34-dpp1.2 and the control strain T34-43b1.3 ( Figure 6). However, the expression of veA and velB was significantly up-regulated in transformant T34-dpp1.3 compared to T34-43b1.3, with relative expression levels that were 2.984 (p = 0.000) and 3.356 (p = 0.000) fold higher in that dpp1-overexpressing transformant, respectively ( Figure 6). Considering that the expression level of dpp1 and production of farnesol by transformant T34-dpp1.3 were significantly higher than in T34-dpp1.2, these data indicate that the increase in the production of farnesol affected the expression of genes involved in the regulation of circadian cycle, which agrees with the role of farnesol as a key molecule in maintaining fungal homeostasis. However, these data are not conclusive and would require a more detailed analysis in the future. farnesol by transformant T34-dpp1.3 were significantly higher than in T34-dpp1.2, these data indicate that the increase in the production of farnesol affected the expression of genes involved in the regulation of circadian cycle, which agrees with the role of farnesol as a key molecule in maintaining fungal homeostasis. However, these data are not conclusive and would require a more detailed analysis in the future.

Effect of the Increase in Farnesol Production on the Fungal Phenotype
T. harzianum wild-type strain and the two dpp1-overexpressing transformants were grown in MEA medium containing farnesol (50 μM, 100 μM and 1000 μM). After Figure 6. qPCR analysis of relative expression levels of genes related to regulation of circadian cycle: laeA, veA and velB, in transformants T34-dpp1.2 and T34-dpp1.3 versus levels observed in the control strain T34-43b1.3. In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric values were labeled with an asterisk (*) at the bottom of each graphic.

Effect of the Increase in Farnesol Production on the Fungal Phenotype
T. harzianum wild-type strain and the two dpp1-overexpressing transformants were grown in MEA medium containing farnesol (50 µM, 100 µM and 1000 µM). After incubation for 7 days at 28 • C, no differences in growth were observed between the wild type and the two dpp1 overexpressing transformants ( Figure S4). Furthermore, the growth of all three strains analyzed was inhibited at farnesol concentrations higher than 50 µM ( Figure S4), which is a similar effect to the inhibition of yeast growth observed in the current study, i.e., C. albicans and S. cerevisiae, when increasing concentrations of farnesol were used ( Figure S5).
As indicated above, the highest level of dpp1 expression coincided with higher levels of farnesol production in transformant T34-dpp1.3. However, those farnesol levels could be affected by phosphorylation of the produced farnesol, e.g., by the activity of the farnesol kinase enzyme, which has not yet been identified but likely occurs in T. harizanum.
In addition, an analysis of the phenotypic characteristics was performed by growing the control strain T34-43b1.3 and the two dpp1-overexpressing transformants T34-dpp1.2 and T34-dpp1.3, in five different solid media to compare colony shape, sporulation rate, pigmentation, and growth rate. The media used in this study were OA, CMA, PDA, SNA and MBA. As result, after 14 days of growth at 28 • C, no differences were found between the three species analyzed ( Figure S6).   Lower panels). In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric (Upper panels) and the relative expression of five tomato defense-related genes: PR1b1 and PR-P2, related to the salicylic acid signaling pathway, and PINI, PINII and TomLoxA related to the jasmonate/ethylene signaling pathways, and three development-related genes: SUCS, ACCS, and GAI, as determined by quantitative PCR (Lower panels). In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric values were labeled with an asterisk (*) at the bottom of each graphic.

Effect of dpp1 Overexpression in T. harzianum on Plant Development
Tomato seeds were coated with T34-43b1.3 and T34-dpp1.3 conidia and then incubated as described in the Section 2. After four weeks of incubation, growth of the resulting plants was assessed by measuring stem diameter and length. We did not observe significant differences in tomato plant growth following the treatment of seeds with the T. harzianum strains. Treatment with T34-43b1.3 (control strain) resulted in an average stem diameter of 8.0 ± 0.4 cm and stem length of 22.5 ± 1.9 cm (n = 14), whereas treatment with the dpp1-overexpressed strain resulted in an average stem diameter of 7.7 ± 0.6 cm, and stem length of 19.0 ± 2.2 cm (n = 14). These results indicate that the overexpression of dpp1 in T. harzianum did not affect plant growth during the plant-fungus interaction.

Effect of dpp1 Overexpression on Resistance of Tomato Plants to Botrytis cinerea
Comparison of healthy tomato plants that had not been inoculated with B. cinerea but had been treated with the control strain (T34-43b1.3) or a dpp1 overexpression strain (T34-dpp1.3) of T. harzianum revealed that dpp1 overexpression caused the down-regulation of all defense-related genes examined. For example, the expression rations range from 0.050 (p = 0.000) and 0.179 (p = 0.000) fold for the PR-P2 and TomLoxA genes, respectively ( Figure 8). A possible explanation for these results is that increased intracellular farnesol concentrations and/or changes in squalene and ergosterol levels in the dpp1 overexpression strain contributed to the observed down-regulation. There are previous reports that indicate the importance of ergosterol/squalene levels of Trichoderma on the ability of the fungus to colonize plant roots and on the regulation of the plant defense response [28]. with the T. harzianum T34 wild-type strain. These results indicate that increased dpp1 expression reduced the ability of T. harzianum to colonize tomato roots.  (right panel). Tomato actin gene has been used as a housekeeping-reference gene in this study. In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed Figure 8. qPCR analysis of expression of five tomato defense-related genes. The relative expression of these genes has been analyzed in leaves grown from tomato seeds coated with T34-dpp1.3 versus leaves from seeds inoculated with T34-43b1.3, in both cases without Botrytis cinerea B05.10 infection (left panel), or leaves from the same plants after infection with B. cinerea (right panel). Tomato α-actin gene has been used as a housekeeping-reference gene in this study. In the graphic representation, those statistically significant differential expression data (p < 0.05) are squared with dashed lines and numeric values were labeled with an asterisk (*) at the bottom of each graphic.
The expression of plant defense related genes in health and B. cinerea-infected plants differed markedly. In infected plants, PR1b1 and PR-P2 genes were relatively up-regulated, with ratios of 6.3 (p = 0.000) and 13.5 (p = 0.000), respectively. In contrast, among the genes related to the JA/ET signaling pathway, only PINI was up-regulated [3.4 (p = 0.000) fold], whereas PINII and TomLoxA were down-regulated (Figure 8). The up-regulation of SA-related genes (PR1b1, PR-P2) has been also previously observed in response to necrotrophic pathogens or to volatile organic compounds produced by these pathogens [28]. A possible explanation for the higher level of expression of SA-related genes in T34-dpp1.3treated plants in the presence of B. cinerea could be that the increased farnesol production and/or concomitant deregulation of ergosterol biosynthesis could induce a higher level of sensitization of plants against fungal plant pathogens.

Effect of Increased Farnesol Production on Root Colonization
Roots of hydroponic grown tomato plants were inoculated with T34-43b1.3 (control strain) and T34-dpp1.3 to determine the effect of dpp1-overexpression in T. harzianum on the tomato root colonization ability. For this analysis, qPCR was used to monitor the levels of the tomato glyceraldehyde 3 -phosphate dehydrogenase gene (gpd) versus the T. harzianum α-actin gene (act1). The analysis revealed that the relative level of T. harzianum act1 in roots from T34-dpp1.3 treated plants was only 52% of the level in roots of plants treated with the T. harzianum T34 wild-type strain. These results indicate that increased dpp1 expression reduced the ability of T. harzianum to colonize tomato roots.
3.9. Effect of dpp1 Overexpression on the Trichoderma Antifungal Activity 3.9.1. Antifungal Activity against B. cinerea B05.10 on Tomato Leaves This analysis was performed using 96 h grown PDB culture broths from T34-43b1.3, T34-dpp1.2, and T34-dpp1.3. Two 15 µL aliquots of a B. cinerea spore suspension, 5 × 10 5 spores/mL, were deposited onto tomato leaves, each aliquot at a different position in the leaf. Once the aliquots had dried, a 15 µL aliquot of a culture broth from one T. harzianum strain was deposited on each dried spot of B. cinerea spores. Leaves were then incubated on a humid atmosphere for 72 h, after which the size of any resulting lesions was measured. No significant differences were observed in the sizes of lesions resulting from the treatment of broth of the wild-type strain and the two dpp1 overexpressing transformants ( Figure S7). Thus, the overexpression of dpp1 did not have a significant effect on the antifungal activity of T. harzianum against the pathogen B. cinerea.

Antifungal Assays on Membranes against R. solani
Metabolites secreted by the three Trichoderma strains assayed in this study, T34-43b1.3, T34-dpp1.2, and T34-dpp1.3, did not have any significant inhibitory effect on R. solani growth ( Figure S8). Thus, similar to the results with B. cinerea shown above, the overexpression of dpp1 did not affect the antifungal activity of T. harzianum against the pathogen R. solani.

Phylogenetic Analysis of Trichoderma dpp1 Orthologs
The phylogenetic relationships of dpp1 orthologs in Trichoderma was assessed using full-length coding or deduced amino acid sequences of orthologs from 34 Trichoderma species representing the known breadth of phylogenetic diversity within the genus. The topologies of the resulting nucleotide and amino acid trees were similar to one another ( Figure 9) and to a previously reported species tree inferred from concatenated alignments of 20 housekeeping genes from the same set of species [31]. The presence of a dpp1 homolog in all 34 Trichoderma species examined and the mirroring of the dpp1 and species trees is consistent with dpp1 having an essential role in primary metabolism.

Phylogenetic Analysis of Trichoderma dpp1 Orthologs.
The phylogenetic relationships of dpp1 orthologs in Trichoderma was assessed using full-length coding or deduced amino acid sequences of orthologs from 34 Trichoderma species representing the known breadth of phylogenetic diversity within the genus. The topologies of the resulting nucleotide and amino acid trees were similar to one another (Figure 9) and to a previously reported species tree inferred from concatenated alignments of 20 housekeeping genes from the same set of species [31]. The presence of a dpp1 homolog in all 34 Trichoderma species examined and the mirroring of the dpp1 and species trees is consistent with dpp1 having an essential role in primary metabolism. Figure 9. Maximum likelihood phylogenetic trees inferred from full-length dpp1 exon (A) or Dpp1/DPP1 amino acid (B) sequences retrieved from the genome of 34 Trichoderma species that represent the known phylogenetic breadth of the genus. The trees were inferred as described in the Materials and Methods section. The numbers on each branch are bootstrap values derived from 1,000 pseudoreplicates. The names of previously described subgeneric lineages [31,33] are indicated in red type at the right of the figure. The location of the T. harzianum dpp1 gene and Dpp1/DPP1 protein in the trees has been highlighted with a green rectangle. The best substitution models used to generate these phylogenetic trees were determined by the IQ-Tree software as TIM3+F+I+G4 and JTTDMut+I+G4 for the nucleotide and amino acids alignments, respectively, on the basis of the optimal AICc (Akaike Information Corrected) and BIC (Bayesian Information) criteria.

Discussion
Farnesol is an isoprenoid intermediate of the mevalonate pathway produced by the dephosphorylation of FDP (Figure 1). FDP plays a central role in cell growth and differentiation; it controls the production of ubiquinone and ergosterol but also participates in the regulation of filamentation and biofilm formation [7,8]. Figure 9. Maximum likelihood phylogenetic trees inferred from full-length dpp1 exon (A) or Dpp1/DPP1 amino acid (B) sequences retrieved from the genome of 34 Trichoderma species that represent the known phylogenetic breadth of the genus. The trees were inferred as described in the Section 2. The numbers on each branch are bootstrap values derived from 1000 pseudoreplicates. The names of previously described subgeneric lineages [31,33] are indicated in red type at the right of the figure. The location of the T. harzianum dpp1 gene and Dpp1/DPP1 protein in the trees has been highlighted with a green rectangle. The best substitution models used to generate these phylogenetic trees were determined by the IQ-Tree software as TIM3+F+I+G4 and JTTDMut+I+G4 for the nucleotide and amino acids alignments, respectively, on the basis of the optimal AICc (Akaike Information Corrected) and BIC (Bayesian Information) criteria.

Discussion
Farnesol is an isoprenoid intermediate of the mevalonate pathway produced by the dephosphorylation of FDP (Figure 1). FDP plays a central role in cell growth and differentiation; it controls the production of ubiquinone and ergosterol but also participates in the regulation of filamentation and biofilm formation [7,8].
In yeast, reductions in FDP dephosphorylation found in dpp1-, and also in lpp1deleted strains were explained by the involvement of the corresponding proteins, Dpp1 and Lpp1, in avoiding the toxic accumulation of isoprenoid pyrophosphates, including FDP. Such accumulation could result from diverse processes, e.g., down-regulation of squalene synthase activity, which catalyzes the first committed step in ergosterol biosynthesis (i.e., conversion of FDP to squalene) [34]. Overexpression of the FDP synthase gene (ERG20) in a yeast mutant lacking squalene synthase activity led to a dramatic increase in the level of dolichol [35], indicating that the FDP pool should be shunted toward dolichol, but the dephosphorylation of FDP may also be important [13,17].
The purpose of the current study was to assess the effect of farnesol production on the physiology and ecology of Trichoderma. To do this, we identified a homolog (dpp1) of the yeast DPP1 and LPP1 genes in T. harzianum T34, which is a well-characterized and considered a model in the biological control studies. An initial gene expression analysis revealed that dpp1 exhibits very low levels of expression in wild-type T. harzianum. This low-level expression of dpp1 in the wild type could explain why previous attempts to silence dpp1 expression in Trichoderma species were unsuccessful (Gutiérrez S. unpublished results). In subsequent analyses, introduction of an overexpression construct into T34 increased dpp1 expression by over 1000-fold. Farnesol production was increased in the two selected transformants (T34-dpp1.2 and T34-dpp1.3) compared to the level observed in the untransformed strain (Tables 1 and 2), with T34-dpp1.3 reaching the highest levels in agreement with the higher level of dpp1 gene expression observed in this transformant. The increase in farnesol production in T34-dpp1.3 was accompanied by a marked decrease in squalene production and a moderate increase in ergosterol production. However, the huge increase in the level of dpp1 gene expression does not match the comparatively moderate increases in farnesol production. This discrepancy could result from the toxicity of the increased levels of farnesol, but other factors as RNA instability could also explain these results. Furthermore, due to the reversibility of the FDP dephosphorylation reaction, farnesol can be detoxified by channeling it to other terpenoids. Indeed, we observed increased levels of ergosterol in the dpp1 overexpression strains ( Figure 5). The increased farnesol production in dpp1 overexpressed transformants resembles the increase in FDP production observed in yeast as result of a mutation in the ERG9 gene, which encodes the enzyme, squalene synthase, that catalyzes the first committed step in ergosterol biosynthesis from FDP [36].
The previously reported effects of farnesol on yeast morphogenesis, as well as in other processes critical for fungal survival, such as the regulation of quorum-sensing and protein prenylation, has also been observed at a transcriptomic level [37], where changes were observed in the expression of genes belonging to several functional categories, including iron transport, cell wall synthesis, drug resistance, and cell cycle progression. In the current study, overexpression of the dpp1 gene had only minor effects on the expression of ergosterol biosynthetic genes. In contrast, dpp1 overexpression in T. harzianum significantly up-regulated expression of the velvet complex genes veA and velB but not laeA. This heterotrimeric complex, consisting of the LaeA, VeA and VelB proteins, synchronizes secondary metabolism and fungal development in relation to the light levels [32]. Thus, the results of the current study suggest that farnesol affects fungal physiology beyond the morphogenesis and quorum-sensing regulation.
Exogenously added farnesol reduced the growth of tomato plantlets and up-regulated SA-related defense genes. This phytotoxic effect was also observed in studies of bean plants [38], where the addition of exogenous farnesol at concentrations between 10 and 100 mM caused reduced development of the root system, suggesting that these farnesol concentrations were toxic to bean plants [38]. Furthermore, a similar up-regulation of SArelated genes was also found in tomato plants inoculated with Trichoderma strains producing the mycotoxin harzianum A (HA), which would indicate that farnesol, similarly to HA, acts as a plant sensitizing metabolite (microbial-associated molecular patterns = MAMP or pathogen-associated molecular patterns = PAMP) [27,39]. These results suggest that exogenously added farnesol is perceived as a toxic compound by plants, and as result, they activate their response in a manner similar to how they respond to Trichoderma phytotoxic secondary metabolites, e.g., trichothecenes and aspinolides [27,28].
In addition to farnesol, other terpene analogs, e.g., squalene and ergosterol, have been deeply characterized regarding their importance on the interaction of T. harzianum with plants. Thus, studies performed using pure squalene/ergosterol at concentrations similar to those found in the T. hazianum T34 mycelium resulted in an up-regulation of genes related to the jasmonate/ethylene (JA/ET) signaling pathways and no effect on SA-related genes [21]. The effects observed for ergosterol/squalene contrast with those found in the present study in plants grown in media amended with farnesol or in plants inoculated with the strain T34-dpp1.3, overexpressing dpp1 gene, in which it was observed that an increase in farnesol production resulted in an up-regulation of SA related genes, with a smaller or no effect on expression of JA/ET related genes.
The results discussed above would indicate that fungal farnesol, ergosterol and squalene play important roles in the fungal-plant interaction, acting as MAMPs or PAMPs. Thus, being their relative amounts found in a given Trichoderma strain essential for the detected plant response, with concentrations up-regulating SA related genes, e.g., increase in farnesol production, and others up-regulating JA/ET related genes, e.g., high ergosterol/low squalene production [21].

Conclusions
The functional characterization of dpp1 overexpression strains of T. harzianum indicate that the corresponding enzyme is involved in the dephosphorylation of FDP to produce farnesol. Therefore, the dpp1-encoded enzyme is most likely a diacylglycerol pyrophosphatase. However, the observed levels of farnesol were much lower than those expected based on the high levels of dpp1 gene expression reached by the transformants. These data suggest that farnesol levels are tightly regulated in order to avoid the toxic effects associated with its cellular accumulation. The increase in farnesol production was sufficient to have a marked effect on fungal physiology that was exemplified by changes in levels of squalene/ergosterol, but they do not have a significant effect on growth or in the antifungal activity of the dpp1 overexpressing strains. Finally, the increase in farnesol production: (i) had a plant sensitization-like effect, in such a way that tomato plants inoculated with a dpp1 overexpression transformant strain exhibited an up-regulation of plant defense genes related to the salicylate signaling pathway when a phytopathogenic fungi was present, and (ii) resulted in a reduction in the ability to colonize tomato roots compared to the wild-type strain. The results of the current study provide additional insights into the role of farnesol in physiology and environmental interactions in the biological control agent T. harzianum.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8121266/s1, Table S1: Oligonucleotides used in the current study; Table S2: Genomic sequences of Trichoderma spp. retrieved from NCBI database for their use in the phylogenetic analysis carried out in the current study; Figure S1: HPLC chromatograms showing farnesol production in culture broths; Figure S2: Level of squalene and ergosterol produced in mycelia; Figure S3: Analysis by qPCR of expression of genes involved in ergosterol biosynthetic pathway in a dpp1 overexpressing transformant; Figure S4: Effect of farnesol on growth of wild-type strain and dpp1-overexpressing transformants; Figure S5: Toxicity of farnesol against Candida albicans and Saccharomyces cerevisiae; Figure S6: Effect of dpp1 overexpression on growth in different media; Figure S7: Antifungal assay on tomato leaves of culture broths from wild-type strain and two dpp1overexpressing transformants; Figure S8: Effect of the Trichoderma strains analyzed in the current study on growth of the phytopathogen Rhizoctonia solani strain R43.