The New Zealand Trichoderma strains used in this study were isolated from different locations and chosen for their ability to induce root development in cuttings of commercial ornamental plant species. These strains were sourced from the Microbial Biocontrol Culture Collection (MBCC, Bio-Protection Research Centre, Lincoln University) and represent a range of species as well as strains with previously known biocontrol and growth promotion activity (Supplementary Table 1S). Trichoderma reesei QM6a (Simmons, ATCC® 13631) was provided by The American Type Culture Collection (ATCC). Trichoderma atroviride IMI206040 and T. virens Gv29.8 were kindly provided by Alfredo Herrera-Estrella (Langebio, Mexico) and Charles Kenerley (Texas A&M, USA), respectively. All Trichoderma strains in the MBCC were stored in 25% glycerol at −80°C and were previously identified using morphology and five different genetic markers (tef1α, ACLA1, Calm1, LAS1, and RPB2) under separate research initiatives (Hoyos-Carvajal et al., 2009b; Braithwaite et al., 2016). Strains LU132, LU660 and LU668 were closely related to T. atroviride and were defined as Trichoderma sp. “atroviride B” (Braithwaite et al., 2016).

For all experiments, fungal inoculum consisted of conidia harvested from 7 d old PDA cultures grown at 25°C under a 12/12 light/dark photoperiod. The resulting conidial suspension was filtered through two layers of Miracloth (Merck Millipore) and adjusted to the concentration required for each experiment.

All mutants and transgenic lines are derived from parental Arabidopsis thaliana ecotype Col-0. DR5::GUS (Ulmasov et al., 1997). DR5rev::GFP (Friml et al., 2003), PIN1::PIN1-GFP (Benková et al., 2003), PIN2::PIN2-GFP (Xu and Scheres, 2005), PIN4::PIN4-GFP (Blilou et al., 2005), PIN7::PIN7-GFP (Blilou et al., 2005) and PIN3::PIN3-GFP (Zádníková et al., 2010) transgenic lines were previously described and are listed in Supplementary Table 2S.

Trichoderma strains with plant growth promotion potential were evaluated for their ability to promote growth of the model plant Arabidopsis in sterile soil. Arabidopsis thaliana (Col-0) seeds were surface sterilized in 95% (v/v) ethanol for 5 min and 20% (v/v) bleach for 7 min, washed five times with sterile water, and kept in 2 mL sterile water at 4°C for 2 d to break dormancy and synchronize germination. Seeds were then incubated for 5 d at 25°C under a 16 h light: 8 h dark (16L: 8D) day/night cycle on standard MS medium: 0.2 X MS medium (Murashige and Skoog basal salt mixture, catalog M5524, Sigma-Aldrich) containing 1.0% (w/v) agar (Sigma), then supplemented with 0.6% sucrose and adjusted to pH 7.0 (Lopez-Bucio, personal communication) before being poured into Petri dishes. The resulting seedlings were then transferred to gamma irradiated sterile soil (chemical composition is described in Mendoza-Mendoza et al., 2015) which was previously inoculated with 1 × 10⁶ spores per g soil of Trichoderma spp. (T. trixiae LU297, T. sp. “atroviride B” LU660, T. sp. “atroviride B” LU668, Trichoderma sp. nov LU753, T. sp. “novaeharzianum” LU1328, T. asperellum LU1370, T. atroviride IMI206040). The plants were assessed 4 weeks after planting for survival and the total fresh weight was measured. The experimental design was a complete randomized block with 7 treatments and 1 negative control with 4 replicates, a total of 32 plots. Each plot consisted of 4 pots (50 mm diameter, 50 mm deep) containing a single Arabidopsis plant. The experiment was repeated twice with similar results.

Enhanced growth in the presence of Trichoderma is likely due to changes in the plant root architecture. To investigate how the Trichoderma strains influence root structure, strains were evaluated in an Arabidopsis agar plate bioassay (Contreras-Cornejo et al., 2009). Ten sterilized/stratified seeds were placed on MS agar (section Influence of Trichoderma on Arabidopsis Growth in Sterile Soil) in a row 1.5 cm from the edge of the Petri dish. The Petri dishes were sealed with plastic film to prevent cross-contamination and moisture loss and placed on their edge at an angle of approximately 65° to allow root growth down along the agar surface and upward shoot growth. The seedlings were pre-incubated for 5 d at 25°C under a 16L: 8D day/night cycle (Morange, 2002) and then 1 × 10⁶ Trichoderma spores (strains mentioned in section Influence of Trichoderma on Arabidopsis Growth in Sterile Soil, and T. sp. “atroviride” B LU132; T. virens Gv29.8 and T. reesei QM6a) in a 5 μL aliquot of water were inoculated 5 cm below the growing root tips. Petri dishes were re-sealed with plastic film and incubated for a further 5 d under the same conditions mentioned above. Control plates were inoculated with water only. At the end of the experiment, the length of each primary root was measured and the number of lateral roots counted. The root density was then calculated as a function of lateral root number divided by the primary root length (Contreras-Cornejo et al., 2009). Each Trichoderma strain was assayed in four independent experiments using 10 replicate Petri dishes.

Regulation of plant auxins has been linked to root growth promotion by microorganisms. To investigate if the presence of Trichoderma influences auxin accumulation in the plant, confrontation assays were done using an A. thaliana transgenic line with the auxin responsive DR5 promoter fused to the β-glucuronidase gene (gus) as described by Ulmasov et al. (1997). Assays were conducted as described in section Influence of Trichoderma on Arabidopsis Root Architecture on Agar Plates except A. thaliana DR5::GUS was selected as the plant line. After 5 d interaction with Trichoderma spp. (T. atroviride IMI206040, T. reesei QM6a, T. sp. “atroviride B” LU668, T. virens Gv29.8, T. sp. “atroviride B” LU132, T. sp. “atroviride B” LU660, T. asperellum LU1370) DR5::GUS seedlings were removed and incubated at 37°C overnight in GUS staining solution (50 mM sodium phosphate buffer, pH 7.0, 0.1% Triton X-100, 1.5 mM potassium ferricyanide, 1.5 mM potassium ferrocyanide, and 1.5 mg/mL X-Gluc). After staining, the samples were cleared by incubation with 7% NaOH for 20 min and 60 min with 100% ethanol followed by 60 min hydration with serial ethanol dilutions 70, 60, 40, 20, and 10%. Then the seedlings were incubated for 30 min with 25% glycerol and overnight with 50% glycerol at room temperature. Seedlings were visualized using an Olympus BX51 compound microscope. Images were captured using an Olympus DP70 digital camera system and processed with the software Cell^F (Olympus). The experiment was repeated once with similar results using 10 replicate plants for each repeat.

For the localization of the different auxin transporters, confrontation assays were performed as described above using T. virens Gv29.8 for inoculum and multiple Arabidopsis transgenic lines. These lines correspond to the marker for auxins DR5::GFP, and different auxin transporters fused to the green fluorescent protein: PIN1::PIN1-GFP, PIN2::PIN2-GFP, PIN4::PIN4-GFP, PIN7::PIN7-GFP and PIN3::PIN3-GFP (Supplementary Table 2S). After 5 d interaction, seedlings were removed from the interaction plate and directly mounted on a slide for observation on an Olympus BX51 compound microscope. Images were captured using an Olympus DP70 digital camera system and processed with the software Cell^F (Olympus). The experiment was repeated once using 10 replicate plants each time.

Thin layer chromatography (TLC) was used to further investigate the role of auxins in Trichoderma growth promotion. IAA and its derivatives were isolated and identified from different Trichoderma strains (T. sp. “atroviride B” LU660, T. sp. “atroviride B” LU668, T. sp. “novaeharzianum” LU1328, T. asperellum LU1370, T. sp. “atroviride B” LU132, T. atroviride IMI206040, T. virens Gv29.8, and T. reesei QM6a). Trichoderma spp. spores were inoculated at 1 × 10⁶ spores/mL in 100 mL of 0.2 X MS (pH 7.0) in 500 mL flasks and incubated shaking (120 rpm) for 48 h at 25°C. Mycelia were collected by filtration on sterile filter paper (Whatman 3MM) and equally divided before inoculated in either 0.2 × MS (pH 7.0) with or without 10 mM L-Tryptophan (Trp). Flasks were incubated shaking (120 rpm) for an additional 24 and 48 h at 25°C. The supernatant from each culture was then filtered to remove the fungal material. The supernatants were adjusted to pH 3.0 with HCl and indole derivatives extracted from 30 mL of supernatant mixed with 15 mL ethyl acetate, shaking at 200 rpm for 3 h in darkness. The organic phase was recovered and evaporated in a CentriVap Centrifugal Vacuum Concentrator (Labconco). The dry pellet was dissolved in 50 μL of ethyl acetate and 5 μL of each sample was loaded onto a TLC silica gel 60. The TLC silica gel was sprayed evenly with van Urk and Salkowski reagents (1:3) (Ehmann, 1977) and developed by incubation at 90°C for 7 min. The indole derivatives were separated for 45 min using n-hexane:ethyl acetate:isopropanol:glacial acetic acid (40:20:5:1, v/v) as mobile phase (Chung et al., 2003). Indole derivative compounds were identified against commercial standard compounds (L-Tryptophan, Tryptamine hydrochloride, 3-indole acetonitrile, DL-indole-3-lactic acid, indole-3-acetaldehyde, and indole-3-acetic acid) from Sigma-Aldrich; 25 μg of each indole standard was loaded per lane. To compare IAA production under different nutrient conditions, a second experiment was done using PDB instead of MS medium and strains T. sp. “atroviride B” LU132, T. reesei QM6a, T. atroviride IMI206040, and T. virens GV29.8 were assayed. Both assays were repeated once.

The activity of Trichoderma indole-derivatives on Arabidopsis auxin accumulation was assessed using an A. thaliana transgenic line with the auxin responsive DR5 promoter fused to β-glucuronidase gene (gus) as described by Ulmasov et al. (1997). DR5::GUS A. thaliana seeds, were surface sterilized and stratified as described in section Influence of Trichoderma on Arabidopsis Growth in Sterile Soil and individually germinated for 6 d in wells of a 96-microtiter plate containing 100 μL of standard MS broth at 25°C under 16 h light/8 h dark settings. MS broth was then replaced with 100 μL fresh standard MS broth containing 10 μL of ethyl acetate containing the dissolved pellet from supernatant extracts from T. sp. “atroviride B” LU132, T. reesei QM6a, T. atroviride IMI206040, or T. virens GV29.8 grown with or without tryptophan for 48 h (section Effect of Trichoderma on Arabidopsis Auxin Accumulation and Transport Using Transgenic Arabidopsis Lines) and incubated for an additional 24 h. Each treatment was allocated 16 wells. The control for the solvent was ethyl acetate, 10 μg/mL of commercial indole-3-acetic acid (sodium salt, Sigma Aldrich) was used as positive control, and distilled water as the negative control. The seedlings were incubated for 24 h at 25°C under 16 h light/8 h dark settings. Seedlings were removed and GUS staining was conducted as described above in section Effect of Trichoderma on Arabidopsis Auxin Accumulation and Transport Using Transgenic Arabidopsis Lines. Ten replicates per treatment were assessed and the entire experiment was repeated once.

The effect of Trichoderma VOCs on plant growth promotion and chlorophyll content was assessed using a split-plate system. Five seeds of A. thaliana (Col-0) were surface sterilized and stratified as described in section Influence of Trichoderma on Arabidopsis Growth in Sterile Soil and then placed onto one side of a 90 mm diameter double compartment Petri plate (Labserv) containing Standard MS medium. Plates were sealed with plastic film and placed on their edge at an angle of approximately 65° and incubated at 22°C under 16 h light/8 h dark settings. After 7 d, 5 μL of a Trichoderma spore suspension (containing 1 × 10⁶ spores) was inoculated into the opposite compartment then the plates were re-sealed and incubated for additional 7 d. The Trichoderma strains used for this experiment were T. atroviride IMI206040, T. virens Gv29.8, T. sp. “atroviride B” LU132, T. reesei QM6a, and T. asperellum LU1370. Fresh weight of shoots, roots and total biomass (roots + shoots) per plant were measured. Shoots or roots of 10 plants were pooled and biomass was measured (n = 7, corresponding to 70 plants). The weight was divided by the number of plants to express the weight per plant. Relative chlorophyll content was quantified from 10 leaves per treatment. Chlorophyll content from the third leaf was determined using a SPAD 502 Plus Chlorophyll Meter. The SPAD values were converted to chlorophyll using the formula described by Ling et al. (2011) (nmol chlorophyll/cm2 = 0.0105 × 2 + 0.4119x + 0.3810). The experiment was carried out twice with 10 plants measured each time.

The VOC profiles of T. atroviride IMI206040, T. sp. “atroviride B” LU132, T. virens Gv29.8 and T. asperellum LU1370 were analyzed using a Shimadzu GCMS-QP2010 (Shimadzu Corporation, Japan) gas chromatograph - mass spectrometer fitted with a Restek Rxi-5 ms fused silica capillary column (30.0 m × 0.25 mm i.d. × 0.25 μm, Bellefonte, PA, USA) and an auto-sampler for solid phase micro-extraction (SPME). Fungal cultures were grown in Petri dishes on standard MS medium at 25°C and 12 h light/12 h darkness for 4 d. Four mycelial plugs (5 mm diameter) were punched out with a sterile cork borer from the outer growing zone of each Trichoderma colony and placed in an amber glass headspace vial (20 mL) sealed with a screw cap containing a blue PTFE/silicone septum (Sigma-Aldrich, Australia). The vials were incubated at 25°C for 24 h before sampling of the headspace VOCs was conducted according to Stoppacher et al. (2010). An SPME fiber coated with 65 μm polydimethylsiloxane/divinylbenzene was used to extract the fungal VOCs from the headspace vial for 30 min without agitation. After injection, the compounds were desorbed for 2 min in a split/splitless injector at 250°C. The oven temperature was held at 40°C for 2 min, then raised to 200°C at 10°C min⁻¹ and 260°C at 25 min⁻¹ and then held at this temperature for 5 min. Helium was used as carrier gas at a constant flux of 1 mL min⁻¹. Compounds were identified by matching their mass spectra and linear retention indices using GCMS solution v. 2.72 (Shimadzu Corporation, Japan) software with NIST 11 and Wiley 10 mass spectrum libraries and by using the software MassFinder4 with a specialized terpenoids library (Hochmuth Scientific Software, Hamburg, Germany). The background of PDA plates without the fungus was extracted analysing the volatiles of four agar medium plugs. Three biological replicates were analyzed for each strain.

Data analyses from fresh weight of Arabidopsis plants grown in sterile soil (Figure 1), effect of Trichoderma on A. thaliana root architecture (Figure 2), and the effect of Trichoderma volatile organic compounds on Arabidopsis plant biomass using a split-plate assay (Figure 7) were analyzed by Analysis of Variance (ANOVA) using the statistical software GenStat Version 18. Differences between treatment means were compared by the unrestricted least significant difference (LSD) analysis at P < 0.05.

Seven Trichoderma strains, chosen for their ability to induce growth promotion in commercial plants, and the reference strain T. atroviride IMI206040, were tested for their ability to promote growth of A. thaliana in sterile soil. After 4 weeks of interaction, three outcomes were evident: an increase in plant biomass, a decrease in biomass, and no effect at all when compared to plants grown in un-inoculated soil. Trichoderma atroviride IMI206040 and T. sp. “atroviride B” LU660 increased Arabidopsis fresh biomass by 72% and 73% respectively, whereas T. asperellum LU1370 significantly inhibited Arabidopsis growth as measured by a reduction of up to 74% (P < 0.05) in biomass with respect to the plants grew in non-inoculated soil. Trichoderma sp. “atroviride B” LU668, T. “novaeharzianum” LU1328, T. trixiae LU297, and Trichoderma spp. LU753 had no significant effect on plant growth (Figure 1).

The influence of Trichoderma inoculation on root architecture was investigated in more detail on MS agar. In the absence of Trichoderma spp., A. thaliana produced long primary roots with few secondary roots and consequently exhibited low root density (Figure 2). When the plates were inoculated with Trichoderma, the primary root length decreased by 13 to 25% and the number of secondary roots increased by 64 to 90%, which in turn resulted in a massive increase in root density (Figures 2B–D). To corroborate if the reduction of primary root length was a common phenomenon during the Trichoderma interaction, additional strains belonging to three different species were added to our analyses: T. virens Gv29.8, which has previously analyzed for their ability to induce A. thaliana secondary roots on MS medium (Contreras-Cornejo et al., 2009), T. sp. “atroviride B” LU132, a commercial biocontrol strain with plant growth promotion capability (Maag et al., 2014), and T. reesei QM6a, a commercial strain used for its capacity to secrete large amounts of cellulases and hemicellulases (Peterson and Nevalainen, 2012; Mukherjee et al., 2013; Strakowska et al., 2014) (Figures 2E–G). Two of these strains, T. sp. “atroviride B” LU132 and T. virens Gv29.8, induced a reduction in primary root length and an increase in the number of secondary roots, while T. reesei QM6a did not induce significant changes with respect to the control (Figures 2E–G). Interestingly, Trichoderma spp. LU753 did not significantly alter root architecture on MS plates even though it was capable of increasing root biomass in willow cuttings (Braithwaite et al., 2009). In trials on ornamental cuttings conducted on commercial nurseries, LU753 did not promote root biomass when applied alone but was effective when applied as part of an effective mixture which included two other Trichoderma strains (M. Braithwaite, pers. comm.) and T. sp. “atroviride B” LU660 promoted lateral root growth on A. thaliana in the plate bioassays (Figures 2B–D) but had no effect on root promotion in the willow screening trial.

A reddening of the leaves was observed for the majority of the Trichoderma treatments indicating strong anthocyanin production. Two strains (T. sp. “novaeharzianum” LU1328 and T. reesei QM6a) did not induce anthocyanin production (Supplementary Figure 1S).

The influence of Trichoderma spp. in the induction of auxins was assessed using DR5::GUS plant seedlings and Trichoderma strains on MS plates. DR5: GUS activity was profoundly visible in the primary and secondary root tips in the control seedlings (Figure 3). Unlike the control, GUS was weakly expressed in the primary root tip when A. thaliana DR5::GUS line interacted with T. atroviride IMI206040 and T. reesei QM6a. No GUS activity was observed in the primary roots with the remaining strains (Figure 3, PR). When secondary roots from A. thaliana DR5::GUS were analyzed, greater activity was observed in seedlings interacting with T. virens Gv29.8, T. sp. “atroviride B” LU668 and LU660, in comparison to the control and other Trichoderma strains (Figure 3, SR). Unexpectedly, T. asperellum LU1370 completely induced an inhibition of DR5 expression in secondary roots (Figure 3) but induced lateral root promotion on plates (Figure 2).

Although auxin is synthesized in the plant apices of shoots and roots, it is transported throughout the plant via the phloem, forming concentration gradients and accumulating in different tissues (Friml et al., 2003). PIN proteins operate as efflux carriers for auxin localization (Petrásek et al., 2006). These proteins are polarly localized in plants, including A. thaliana (Chen et al., 1998; Luschnig et al., 1998; Müller et al., 1998; Shin et al., 2005). To assess if polarity localization of PIN proteins is altered in A. thaliana by the interaction with T. virens Gv29.8, reporter lines of A. thaliana containing different PIN proteins fused to GFP were used. Expression of GFP modulated by DR5 was inhibited by the presence of T. virens, in contrast to the control where the typical localized GFP signal was observed (Figure 4). When the localization of PIN transporters in the primary roots (PIN1::PIN1:GFP, PIN2::PIN2:GFP, PIN3::PIN3:GFP, and PIN7::PIN7:GFP) were analyzed on MS medium, specific localization occurred in the different lines; however in the presence of the fungus, expression of PIN2 and PIN3 was completely inhibited and a diffuse localization of PIN1 and PIN7 was observed (Figure 4).

Changes in growth promotion activity by Trichoderma species on Arabidopsis plantlets may depend on their capacity to produce auxins. Therefore as a first step to investigate the role of auxins indole derivatives from Trichoderma were extracted and separated on TLC plates. On standard MS medium two T. sp. “atroviride B” strains (LU668 and LU132), T. reesei QM6a and T. sp. “novaeharzianum” LU1328 synthesized a compound that migrated to the same Rf as the commercial 3-indole acetic acid (IAA) (Figure 5). Interestingly, four strains did not produce IAA on MS standard medium (T. atroviride IMI206040, T. virens Gv29.8, T. sp. “atroviride B” LU660 and T. asperellum LU1370). In the absence of Trp, no production of indole derivatives was observed, with the exception of T. virens which produced a compound that migrated to the same Rf as IAA but with a different color; this was present but less evident when Trp was added (Figure 5A). Additionally T. virens synthesized additional compounds in standard MS independently of the presence of Trp.

Four strains from the first experiment were used to assess the IAA production in PDB: T. atroviride IMI206040, T. virens Gv29.8, T. sp. “atroviride B” LU132, and T. reesei QM6a. All strains were able to synthesize a molecule that migrated to the same Rf as commercial IAA (Figures 5A,B).

Strains T. sp. “atroviride B” LU132, T. reesei QM6a, T. atroviride IMI206040 and T. virens GV29.8 were selected for further analysis on the basis of their variable pattern on the TLC plates. Culture supernatants were assessed for their bioactivity on auxin accumulation in the reporter line A. thaliana DR5::GUS. Arabidopsis seedlings exhibited significantly more GUS activity after a 12-h treatment with 5 μM of commercial 3-indol-acetic acid than the control seedlings (Figure 6A). For the supernatants produced in MS medium, only those generated from T. reesei QM6a in the presence of tryptophan resulted in a clear increase in auxin accumulation in both the leaves and roots (Figures 6B,C). All culture supernatants produced in PDB in the presence of tryptophan resulted in a clear increase in auxin accumulation in both leaves and roots (Figures 6B,C).

To assess the effect of mVOCs emitted by Trichoderma in growth promotion, co-cultivation experiments were done using a split-plate assay to keep both organisms physically separated. After 7 days of co-cultivation VOCs emitted by four Trichoderma strains (T. sp. “atroviride B” LU132 and IMI206040; T. virens Gv29.8 and T. asperellum LU1370) showed a significant increase in shoot, root and total biomass compared with the control seedlings (Figure 7). mVOCs emitted by T. reesei QM6a had no effect in increasing biomass in both shoots and roots with respect to the control.

Some VOCs emitted by microbes increase the efficiency of photosynthesis and chlorophyll content in A. thaliana by modulating the sugar content (Zhang et al., 2008). In this study, a clear difference in color was observed in leaves exposed to Trichoderma in comparison to the controls, with the exception of T. reesei QM6a which showed no difference (Figure 7E). Analysis of the chlorophyll content supported these observations; all strains except T. reesei QM6a induced a significant increase in chlorophyll content (Figure 7F).

The mVOC profiles of the four strains identified as promoting growth in the absence of physical contact (section Growth Response of A. thaliana to Trichoderma spp. Volatile Organic Compounds (VOCs).) were analyzed by GCMS. Trichoderma sp. “atroviride B” LU132 and T. atroviride IMI206040 differed in emission of mVOCs when grown on standard MS medium; T. sp. “atroviride B” LU132 produced 6-PP in large quantities whereas T. atroviride IMI206040 emitted only minor quantities of this compound but higher quantities of 1-Octen-3-ol and 3-Octanone (Tables 1, 2). For T. virens Gv29.8, 29 sesquiterpenes were identified and of these β-Elemene and ε-Amorphene were significantly overrepresented in the mixture (Table 3). Of the mVOCs emitted by T. asperellum LU1370, 1,3-Octadiene, Limonene and β-Eudesmol + Valerianol were the most abundant in the mixture (Table 4).

In nature, multiple symbiotic relationships form between plants and microorganisms and these relationships can be manipulated to improve plant productivity. One such trait is plant growth promotion. It has been reported that several species of Trichoderma have the ability to induce growth promotion on diverse crop plants regardless of the place from which they were isolated (Harman et al., 2004; Mastouri et al., 2010; Lee et al., 2016). The Trichoderma strains tested in this study were collected from both free-living and plant symbiotic sources and, as has been reported, there appeared to be no correlation between isolation source and the ability to promote plant growth. However, our work clearly demonstrates that growth promotion is not a universal trait of all Trichoderma. Similar to that observed by Lee et al. (2016), we observed promotion, no effect at all, and detrimental effects on plants. Further, the variability in plant growth promoting potential appears to be more strongly influenced by environmental parameters than the choice of plant host. This work highlights the need for greater research into the environmental control of successful plant health enhancement.

Detailed biochemical studies on growth promotion, by necessity, require the use of in vitro systems, however the choice of system can directly affect the outcome and will not necessarily reflect how the microorganism behaves in nature. Originally isolated from healthy Hydrangea roots, when tested in sterile soil in this study, T. asperellum LU1370 significantly reduced Arabidopsis fresh weight. Interestingly, this strain also significantly reduced the survival of Lophomyrtus cuttings by 40% (data not shown) and had no effect on the growth parameters of tomato seedlings (M Braithwaite, unpublished data.). However, when T. asperellum LU1370 was tested on agar plates, a positive effect on Arabidopsis growth was observed.

The inhibitory effects of Trichoderma on plants have previously been reported, particularly with respect to seed germination on diverse plants including lettuce, onion and chicory (Ousley et al., 1993; Celar and Valic, 2005). Trichoderma spp. synthesize a plethora of secondary metabolites with diverse biological function in plants (Reino et al., 2008). For example, trichosetin, a secondary metabolite isolated from dual cultures of T. harzianum-Catharanthus roseus callus (Marfori et al., 2002), inhibited root and shoot growth in five plant species (Oryza sativa, Vigna radiate, Medicago sativa, Capsicum frutescens, and Lycopersicon esculentum) by damaging the cell membrane (Marfori et al., 2003). Roots treated with trichosetin were mostly dead, indicating a clear phytotoxic effect of this compound (Marfori et al., 2003). Additional compounds with negative effects on plant growth include Trichocaranes (A, B, and C) (Macías et al., 2000), konionginins (B, C, E, and G) (Cutler et al., 1989; Parker et al., 1995), cyclonerodiol (Cutler et al., 1991), and a laevorotatory form of harzianopyridone (Cutler and Jacyno, 1991). The last compound caused necrosis in bean, tobacco and corn in a concentration dependent manner (Cutler and Jacyno, 1991). Trichoderma virens also synthesizes negative plant growth promoters such as viridiol, a potent herbicidal compound, which is effective for weed control (Héraux et al., 2005).

The biosynthesis of secondary metabolites is modulated by environmental conditions. The general effect of T. asperellum LU1370 on plant growth in soil was an inhibition in plant size without causing plant death, by contrast, on plate assays the general effect was stimulatory, suggesting that either different molecules are produced in both environments (soil and plates) or the proportion of specific metabolites are differently accumulated in these conditions. The dual effect of specific metabolites have been documented before, for example, 6-pentyl- 2H-pyran-2-one (6-PP) induces plant growth promotion, depending on the concentration (Garnica-Vergara et al., 2015). At high concentrations 6-PP has an inhibitory effect and at low concentrations it promoted growth in wheat seedlings (Vinale et al., 2008). However, Lee et al. (2016) recently reported no correlation between 6-PP and growth promotion. Trichoderma asperellum IsmT5 synthesizes 6-PP (Kottb et al., 2015) but we did not identify this molecule in T. asperellum LU1370, although it may be induced by the plant as was recently reported for T. atroviride (Garnica-Vergara et al., 2015). The role of 6-PP in the strains presented in this study is unknown.

It is known that some microorganisms have the ability to produce phytohormones like IAA (Robinson et al., 1998; Contreras-Cornejo et al., 2011; Hilbert et al., 2012; Fu et al., 2015).

In bacteria, indole acetic acid (IAA) biosynthesis varies significantly depending on the media used (Spaepen et al., 2007). In fungi, the most well-known mechanism of IAA biosynthesis is tryptophan-dependent, and consequently the addition of this amino acid to the media results in the production of higher levels of IAA (Robinson et al., 1998; Chung et al., 2003; Spaepen and Vanderleyden, 2011; Hilbert et al., 2012). An example of how media determine the production of IAA is shown in Figure 5 where T. virens and T. atroviride synthesize IAA in PDB supplemented with L-Tryptophan but not in MS supplemented with the same amino acid. The ambient pH was different between the PDB and MS media, pH 5.8 ± 0.2 and pH 7.0 respectively. MS media used in this study contained 0.6% sucrose while PDB formulation contained 2% glucose (dextrose), these differences may have an impact on the mechanisms of synthesis of IAA. In fungi, environmental factors, including pH and temperature have an impact on IAA biosynthesis (Yu et al., 1970; Strzelczyk et al., 1992; Bose et al., 2013; Sun et al., 2014). Carbon and nitrogen sources have been shown to be essential factors influencing fungal and bacterial IAA biosynthesis (Yurekli et al., 2003; Shokri and Emtiazi, 2010); whether they play an intrinsic role in fungal IAA production remains to be explored.

By comparing MS medium supplemented with or without L-tryptophan, Salas-Marina et al. (2011) reported a ~2-fold increase of IAA derivatives secreted by T. atroviride IMI206040 and T. virens Gv298. This does not contradict to findings because the authors reported a total amount of indole derivatives using the Salkowski's reagent, which is a colorimetric method to detect general indole substances (Perley and Stowe, 1966). In this study, T. atroviride IMI206040 secreted multiple indole derivatives, however, the Rf differed from the commercial IAA standard. Furthermore, the indole derivatives induced in MS from T. atroviride IMI206040 and T. virens Gv29.8 did not display auxin activity. Further experimental work is required to understand the mechanisms of IAA synthesis by Trichoderma and how this molecule impacts plant physiology.

Currently IAA is thought to be a diffusible signal involved in interspecies communication (Fu et al., 2015); however, the final outcome of this communication is dependent on the organisms interacting and is also influenced by the environment where the interaction occurs. For example, while plant growth promotion induced by T. virens Gv29.8 and T. atroviride IMI206040 has been related to their ability to secrete IAA (Contreras-Cornejo et al., 2009), some other studies have observed no relation between IAA production and growth induction, at least when plants are tested in soil. Hoyos-Carvajal et al. (2009a) identified that of 101 Trichoderma strains collected from Colombia, 60% produced IAA-derivatives in vitro but only 18% had the ability to induce plant growth promotion in beans. In the present study, some Trichoderma strains produced significant quantities of IAA in vitro and also increased Arabidopsis seedling shoot fresh weight in soil experiments, providing a possible link between the two systems. However, the link between IAA production and plant growth promotion was not consistent for all strains. For instance, in the in vitro agar plate tests, five of the Trichoderma strains were similar to T. atroviride IMI206040 and T. virens Gv29.8 with respect to increases in lateral root growth and root density, but not all isolates produce IAA in liquid MS supplemented with L-Trp. A similar lack of correlation between growth promotion and 6-PP production was recently observed in Trichodema (Lee et al., 2016).

Several authors have reported that bacteria, microalgae, fungi, and plants exchange IAA as a signaling molecule that affects their physiology (Fu et al., 2015). The rice blast fungus Magnaporthe oryzae synthesizes IAA specifically in the area of the infection hyphae (Tanaka et al., 2011), but it is not clear whether this occurs to manipulate the host plant or to benefit its own self. Furthermore the endophytic fungus Piriformospora indica synthesizes IAA not for growth promotion but for plant colonization (Hilbert et al., 2012), suggesting that additional mechanisms must exist for this process. Further molecular studies on Trichoderma are required to understand the mechanisms of synthesis in the fungus and how this molecule affect the interaction with plant roots.

Altogether, our findings suggest that IAA production seems not to be a key determinant for plant growth promotion. However, in this stage we cannot discard the possibility that plants are triggering IAA production in Trichoderma as occurs in other microorganisms including Xanthomonas axonopodis (Costacurta et al., 1998) and Magnaporthe oryzae (Fu et al., 2015).

Cell polarity has been crucial for evolution of multicellular organisms (Offringa and Kleine-Vehn, 2013) and is one of the fundamental aspects of development (Dhonukshe, 2009). In plants, PIN proteins are integral membrane proteins implicated in the directional efflux of IAA (Luschnig and Vert, 2014). These proteins can alter their localization from one cell to another by developmental and environmental conditions (Friml et al., 2002, 2003, 2004; Kleine-Vehn et al., 2008). Our results showed that Trichoderma spp. inhibit primary root elongation and increase lateral root numbers (This study; Salazar-Badillo et al., 2015). Our results showed that in the primary tip, Trichoderma promotes a reduction in the accumulation of DR5::GUS and DR5::GFP expression, indicating a reduction of DR5 activity (Figures 3, 4). When the T. virens Gv29.8 and Arabidopsis interaction was analyzed, we observed decreased expression or delocalization of auxin PIN receptors. This reduction in the accumulation of PIN receptors might be the cause of the reduction of A. thaliana root length in the interaction with Trichoderma. Recently it was reported that Trichokonin VI (TK VI), a peptaibol produced by T. longibrachiatum SMF2, inhibits primary root growth on Arabidopsis by suppressing cell division and cell elongation, and disrupting root stem cell niche maintenance (Shi et al., 2016). Contreras-Cornejo et al. (2015) recently reported that the mitogen-activated protein kinase 6 (MPK6) has an important role in the induction of root length during the interaction with T. atroviride. Mutants of the mpk6 have an enhanced root inhibition when interacting with Trichoderma. The fact that Trichoderma spp. strongly activates DR5 in the secondary roots also suggests that additional molecules that work at distance might be involved, such as volatile organic compounds emitted by the fungi as was recently reported in T. atroviride (Garnica-Vergara et al., 2015).

Microbial volatile organic compounds have been described as important inducers of growth promotion in bacteria and fungi, including mycorrhizae and Trichoderma (Garnica-Vergara et al., 2015). The mechanisms behind microbial-induced plant growth promotion are complex, and, in the case of the strains discussed in the present work, may involve IAA but not as a key determinant. Although their exact mechanism is not known, the results from this study clearly demonstrate that VOCs are able to actively promote plant growth in the absence of physical fungus-plant contact. Four out of five strains tested in this study were able to increase plant biomass and chlorophyll content when physically separated from the plant. However, the wide and varied array of VOCs produced by the various strains suggests that non-contact induction may be due to a number of diverse compounds and mechanisms. Garnica-Vergara et al. (2015) reported that 6-pentyl-2H-pyran-2-one (6-PP) from T. atroviride is a key element in growth promotion. Interestingly, Garnica-Vergara et al. (2015) observed that higher concentrations of the pyrone induce a reduction of the primary root length. However, the concentrations used can be far from those observed in natural conditions. Furthermore, Lee et al. (2016) reported no correlation between 6-PP production and growth promotion.

In our study, we observed an atonishing diversity of VOCs produced by different strains, suggesting the existence of diverse mechanisms of induction. However, we cannot exclude the possibility of and influence of CO₂ in plant growth promotion as has been suggested before (Ramadan et al., 2015), although Garnica-Vergara et al. (2015), and Kottb et al. (2015) discarded this possibility by using pure compounds instead of the fungi. In addition, Lee et al. (2016) recently reported that the amount of CO₂ generated by Trichoderma spp. is insufficient to induce Arabidopsis growth promotion and the effects observed are attributable to VOCs synthesized by Trichoderma spp.

In summary, this study provides evidence that the communication between plants and Trichoderma involves the recognition of fungal-derived molecules such us auxins and VOCs, however this communication is strongly dependent on the environment. Many studies have been conducted on the effect of Trichoderma metabolites. However, there is a lack of strong evidence of a correlation between growth promotion and the production of a particular compound, suggesting a complex interplay of molecules is required. Future molecular studies examining the roles of specific genes and pathways may shed further light on the role of fungal-derived molecules in promotion of plant growth.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.