Complementary Effects of Dark Septate Endophytes and Trichoderma Strains on Growth and Active Ingredient Accumulation of Astragalus mongholicus under Drought Stress

Drought is a major abiotic stress factor affecting plant growth and production, while utilizing beneficial endophytic fungi is one of the most promising strategies for enhancing plant growth and drought tolerance. In the current study, a pot experiment was conducted to investigate the beneficial effects of dark septate endophyte (DSE) (Macrophomina pseudophaseolina, Paraphoma radicina) and Trichoderma (Trichoderma afroharzianum, Trichoderma longibrachiatum) inoculum on Astragalus mongholicus grown in sterile soil under drought stress, alone, or in combination. The addition of Trichoderma enhanced the DSE colonization in roots regardless of the water condition. Under well-watered conditions, M. pseudophaseolina inoculation significantly enhanced the biomass and root length of A. mongholicus. The two DSE and Trichoderma inoculum significantly improved calycosin-7-O-β-D-glucoside content. However, M. pseudophaseolina + T. afroharzianum inoculation better promoted root growth, whereas co-inoculation had higher active ingredient contents compared with single inoculation, except for P. radicina + T. afroharzianum. Under drought stress, DSE and Trichoderma inoculum significantly improved root biomass, root length, calycosin-7-O-β-D-glucoside content, and activities of nitrate reductase and soil urease. P. radicina + T. afroharzianum and P. radicina + T. longibrachiatum better increased root length, and all combinations of DSE and Trichoderma had a greater impact on the increase in formononetin content compared with the single treatments. Additionally, Trichoderma relies on antioxidant enzymes, growth hormones, and the redox system (ascorbic acid–glutathione) to resist drought, while DSE strains have an additional osmotic regulation system in addition to the drought resistance function possessed by Trichoderma, and the effect of co-inoculation (especially M. pseudophaseolina + T. longibrachiatum and P. radicina + T. afroharzianum) on plant physiological parameters was greater than that of single inoculation. This study provides a new research direction for the effects of DSE and Trichoderma on medicinal plant cultivated in dryland.


Introduction
Although drought can destroy the soil structure, reduce soil productivity, limit the absorption of water and nutrients, cause an imbalance in plant infiltration and redox, and eventually lead to plant death [1,2], plant-endophyte symbionts have developed a series of regulatory mechanisms in response to environmental stress during the course of evolution [3]. Endophytic fungi can augment the secretion of host plant endogenous hormones such as ascorbic acid (ASA), proline, and enzymes such as superoxide dismutase, peroxidase, etc., which constitute the total antioxidant system of cells to adapt to drought conditions [4,5]. Therefore, using beneficial endophytic fungi to maintain plant health and quality is an effective way to improve crop yield and productivity. This requires us to select conditions on (I) root structure, (II) plant biomass, (III) active ingredients, (IV) physiological parameters, and (V) soil physicochemical properties. These data can demonstrate whether DSE and Trichoderma can withstand drought stress and whether they have the potential to improve the stress resistance and symbiosis of A. mongholicus in drought-affected lands.

Fungal Isolates and Plant Materials
The two DSE strains used in the experiment were obtained from A. mongholicus roots and two Trichoderma strains were isolated from the rhizosphere soil of Bacopa monnieri and Scutellaria baicalensis grown naturally in farmland habitat in Anguo city, Hebei Province, China, and were preserved in the culture collection of the Laboratory of Mycorrhizal Biology, Hebei University, China. These fungal species were identified based on morphological characteristics ( Figure S1) and phylogenetic analysis of internal transcribed spacer (ITS) and translation elongation factor 1-α (tef1) gene sequences ( Figure S2). All isolates were cultured on the PDA medium for two weeks under dark conditions at 27 • C for subsequent experiments.
Sterile water (5 mL) was placed in a Petri dish containing mature Trichoderma, and the conidia produced were gently scraped and collected with a sterile spatula. The spore suspensions obtained were used as inoculum after measuring the spore concentration (ca. 1 × 10 7 CFU mL −1 ) in a hemocytometer.
Mature seeds of A. mongholicus were collected from the Anguo Medicine Planting Site in Hebei Province, China, and stored at 4 • C.

Inoculation Assay
The inoculation experiment was conducted using a completely randomized factorial design (3 DSE inoculations × 3 Trichoderma inoculations × 2 water conditions) with five replicates. The three DSE inoculations were inoculation M, inoculation P, and noninoculated control. The three Trichoderma inoculations were inoculation A, inoculation L, and non-inoculated control. Dual inoculations were inoculation MA, inoculation ML, inoculation PA, and inoculation PL. The two water conditions were well-watered (WW, 65% field water capacity) and drought stress (DS, 30% field water capacity). There were two plants per pot, thus accounting for a total of 90 experimental pots.
Seeds of A. mongholicus were surface-sterilized with 50% (v/v) sulfuric acid for 14 min and then 5% (v/v) sodium hypochlorite for 10 min. The sterilized seeds were rinsed six times with sterile water and then germinated in a constant temperature and humidity incubator at 27 • C for one week. The seedlings were transplanted to sterile plastic pots (9.5 cm diameter, 11.5 cm height) containing 500 g sterile growth medium, which was a 1:1 (w/w) mixture of sand and soil (<2 mm) collected from the growing site of A. mongholicus and autoclaved for 120 min at 121 • C. The growth medium contained 21.38 mg g −1 organic matter, 51.90 µg g −1 available nitrogen, and 18.62 µg g −1 available phosphorus. For DSE inoculation, two 5 mm discs were cut from the DSE colony and inoculated at 1 cm to the roots of A. mongholicus seedlings, and the sterile PDA medium without fungus was inoculated as the control treatment. After 30 days, the seedlings were irrigated with 30 mL of Trichoderma strain spore suspension and an equal amount of sterilized liquid was poured onto the control seedlings [28]. All the above inoculation procedures were carried out in a completely sterile environment, and all the pots were kept in a growth chamber (10 h photoperiod, 27 • C/22 • C (day/night), 60% mean relative humidity). Two months after sowing, half of the seedlings (both control and inoculation treatments) were subjected to WW treatment, and the other half were subjected to DS treatment. The A. mongholicus seedlings were harvested 120 days after sowing. The fresh weights of shoots and roots, root colonization, morphological traits, and active ingredient contents, as well as physiological and soil parameters, were measured 120 days after sowing.

Plant Growth Parameters
Plant shoots and roots were harvested, and the roots were gently washed with deionized water to remove the surface soil. Individual root sections were completely floated in plexiglass trays containing deionized water (approximately 0.3 cm deep) and scanned by a desktop scanner (EPSON Perfection V800 Photo; Epson, Nagano, Japan). The total root length and average root diameter were determined by the WinRHIZO image analysis system.

Physiological Parameter of Plant Leaves
The soluble protein concentration was determined by the Coomassie Brilliant Blue G-250 reagent [29]. Fresh leaves (0.5 g) were homogenized in ice-cold phosphate buffer (50 mM/5 mL, pH 7.0) in an ice bath. The homogenate was centrifuged at 4000× g for 10 min at 4 • C to collect the supernatant. The absorbance of the supernatant was measured at 595 nm using a spectrophotometer (752 N model, Shanghai INESA Instrument Analytical Instruments Co., Ltd., Shanghai, China). Bovine serum albumin was used as a standard.
The proline content was determined based on the method of Bates [30]. Sulfosalicylic acid solution (3%, 5 mL) was used to homogenize fresh leaves (0.5 g) to a homogenate that was filtered through filter paper after being heated in boiling water for 10 min. The filtrate (2 mL) was reacted with acid ninhydrin (2 mL) and glacial acetic acid (2 mL) for 30 min at 100 • C, then terminated in an ice bath. Toluene (4 mL) was used to extract the reaction mixture, and the absorbance was read at a wavelength of at 520 nm.
The glutathione (GSH) content was determined by the dithio-dinitrobenzoic acid (DTNB) method [31]. Fresh leaves (0.5 g) were ground into a homogenate in 5% of 5 mL trichloroacetic acid and centrifuged at 10,000× g for 10 min. Then, 0.25 mL supernatant was added to 2.6 mL NaH 2 PO 4 (150 mm/L) and 0.15 mL DTNB, and the reaction was kept at 30 • C for 5 min. The absorbance was measured at 412 nm with a spectrophotometer. SOD activity was measured by the photochemical reduction method [32]. A potassium phosphate buffer (50 mM/5 mL, pH 7.8) containing chilled EDTA (0.2 mM) and polyvinylpyrrolidone (2%, w/v) was used to homogenize fresh leaves (0.5 g) in an ice bath. The supernatant was collected after the homogenate was centrifuged at 15,000 rpm for 30 min before enzymatic assays. The photochemical reduction of nitroblue tetrazolium was determined at a wavelength of 560 nm using a spectrophotometer.

Physiological Parameter and Active Ingredients of Plant Roots
Nitrate reductase (NR) activity was assessed by the in vitro method to determine the nitrite nitrogen content [33]. GSH content was determined by the DTNB method (see above). The content of ASA in roots was determined by referring to Kampfenkel [34]. The root indole-3-acetic acid (IAA) concentration was determined by an ELISA kit.
The formononetin and calycosin-7-O-β-D-glucopyranoside content were determined by high-performance liquid chromatography (HPLC) [35]. The dried root samples were ground into powder and passed through a 40-mesh sieve. A 1.0 g sample was extracted in methanol (50 mL) for 180 min in an ultrasonic bath. The weight of the extract solution was made up with methanol and the residue was removed by filtration. The extract solution was concentrated by a RE-52AA pressure rotary evaporator (Automatic Science, Shanghai, China), and dissolved in 5 mL methanol. The extract solution was filtered through a filter (0.22 µm) and transferred to a 2 mL autosampler vial. The filtrate (10 µL) was subjected to separation by HPLC using a reverse-phase C 18 symmetry column (4.6 mm × 250 mm, pore size: 5 µm; Waters Corp., Milford, MA, USA). The mobile phase consisted of acetonitrile (A) and 0.2% formic acid (B). The separation was conducted in the gradient elution mode (0-20 min, 20%→40% A; 20-30 min, 40% A) at 30 • C with a flow rate of 1.0 mL/min. The eluted compounds were detected spectrophotometrically at 260 nm using a 2998 PDA photodiode array detector. The standard of formononetin and calycosin-7-O-β-D-glucopyranoside was purchased from the China National Institutes for Food and Drug Control.

Quantification of DSE Root Colonization
The fungal structures in the roots were observed to assess whether the roots of A. mongholicus were colonized by DSE [36]. Then, 0.5 cm fresh root segments were cleared in potassium hydroxide (10%, w/v) at 100 • C for 1 h and stained in acid fuchsin (0.5%, w/v) at 90 • C for 20 min. Thirty root segments of each sample were randomly selected to be observed by microscopy at 20× and 40× magnification.

Quantification of the N and P Contents of Plants
The 0.2 g dried plant samples were weighed for an elemental nitrogen and phosphorus analysis. Plant total phosphorus (TP) and nitrogen (TN) were measured by digestion with H 2 SO 4 -H 2 O 2 and a whole-element autoanalyzer, respectively [37].

Soil Physicochemical Properties
Soil organic carbon (SOC) was measured by the scorch mass method [38], and soil available phosphorus (SAP) was measured by the sodium bicarbonate leaching-molybdenum antimony anticolorimetric method [39]. Alkaline phosphatase (ALP) activity was measured using p-nitrophenyl phosphate (PNPP) as a substrate [40]. Urease activity was determined using Kandeler and Gerber's method based on the colorimetric determination of ammonium [41]. Soil ammonium (NH 4 + -N) and nitrate (NO 3 − -N) were measured by extraction from the soil with a 0.01 mol liter1 KCl 2 solution (1:100, wt/vol) and determination with a flow-injection autoanalyzer [42].

Statistical Analyses
Two-way analysis of variance (ANOVA) was used to assess the effects of DSE, Trichoderma, and their interactions on the measured parameters. Three-way variance analysis (ANOVA) was performed to examine the effects of DSE, Trichoderma, water, and their interactions on the plant growth parameters, antioxidant parameters, and active ingredient and soil nutrient contents. The correlations among the different variables were analyzed using the Pearson coefficient. Tukey's tests and t-tests were used to compare the mean values to determine significant differences (p < 0.05). SPSS 21.0 software was used in the above analyses. Variation partitioning analysis (VPA) was used to analyze the interaction effects of DSE, Trichoderma, soil nutrients, and water on plant growth with the RStudio package "vegan".

DSE Colonization and Re-Isolation in Roots
After harvesting, DSE hyphae and microsclerotia structures were observed in all stained root segments (DSE inoculation, DSE + Trichoderma inoculation) of A. mongholicus under WW and DS treatment ( Figure S3). No DSE structures were observed in noninoculated plants and single-inoculated plants with Trichoderma strains.
Under WW conditions, co-inoculation with Trichoderma significantly increased DSE hyphal colonization ( Figure S4a). Plants inoculated with L had significantly decreased P microsclerotia colonization ( Figure S4b). Under DS conditions, L inoculation had significantly increased DSE hyphal and total colonization ( Figure S4a,c). Inoculation with A significantly increased P hyphal colonization and total colonization ( Figure S4). The DSE microsclerotia colonization of plants inoculated with L was higher, while the microsclerotia colonization of plants inoculated with A was lower than that of plants inoculated with M alone ( Figure S4b).

Plant Growth Parameters
Under WW conditions, shoot biomass and root biomass improved by 37.5% and 191.02%, respectively, after M inoculation compared to control plants (Figure 1a,b). Under DS conditions, inoculation with M and A increased shoot biomass by 27.54% and 31.88% as compared to the control plants (Figure 1a). Single and dual inoculation significantly increased root biomass compared to nonincubated plants ( Figure 1b). In particular, M inoculation, A inoculation, and MA inoculation led to higher root biomass compared to other inoculation treatments. alone ( Figure S4b).

Plant Growth Parameters
Under WW conditions, shoot biomass and root biomass improved by 37.5% and 191.02%, respectively, after M inoculation compared to control plants (Figure 1a,b). Under DS conditions, inoculation with M and A increased shoot biomass by 27.54% and 31.88% as compared to the control plants ( Figure 1a). Single and dual inoculation significantly increased root biomass compared to nonincubated plants ( Figure 1b). In particular, M inoculation, A inoculation, and MA inoculation led to higher root biomass compared to other inoculation treatments. Under WW conditions, M inoculation significantly increased root length (+29.69%), while PA inoculation significantly increased root length compared to single-inoculated Under WW conditions, M inoculation significantly increased root length (+29.69%), while PA inoculation significantly increased root length compared to single-inoculated plants ( Figure 1c). Inoculation treatments had no effects on root diameter under DS and WW conditions (Figure 1d). Drought significantly enhanced root length compared to plants grown in WW conditions. Under DS conditions, all inoculations led to significantly improved root length compared to non-inoculated plants ( Figure 1c). The root length in plants inoculated with PL grew larger than that in plants inoculated with P or L separately.

Osmotic Materials and Antioxidant Enzyme Activities in Leaves
GSH content and SOD activity were significantly influenced by water, DSE, Trichoderma, and their interaction. Soluble protein content was significantly affected by water, DSE, Trichoderma, and co-inoculation of DSE and Trichoderma. Proline content was significantly impacted by Trichoderma inoculation, interaction of water and DSE, and interaction of DSE and Trichoderma (Table S1). Under WW conditions, M inoculation significantly increased proline content, M and P inoculation increased soluble protein content, DSE and Trichoderma inoculation significantly decreased SOD activity, and M, A, and L inoculation significantly increased GSH content compared with control plants (Figure 2). PL and PA had higher soluble protein content, while ML, PA, and PL had significantly higher SOD activity than in those inoculated with a single strain ( Figure 2). Under DS conditions, P inoculation significantly increased proline content (+15.77%), and plants inoculated with ML had higher proline content than those inoculated with M and L (Figure 2a). Inoculation with DSE and Trichoderma (except for L) enhanced soluble protein content, while the soluble protein content inoculated with MA, ML, and PL was higher than that of plants inoculated with single DSE and Trichoderma (Figure 2b). SOD activity was significantly enhanced by all inoculations; SOD activity after inoculating with MA and PA was higher than that inoculated with a single strain (Figure 2c). GSH content was significantly increased by M, P, and A inoculation compared with control plants. Plants inoculated with MA, PA, and PL had higher GSH content than those inoculated with single DSE and Trichoderma (Figure 2d).

Osmotic Materials and Antioxidant Enzyme Activities in Leaves
GSH content and SOD activity were significantly influenced by water, DSE, Trichoderma, and their interaction. Soluble protein content was significantly affected by water, DSE, Trichoderma, and co-inoculation of DSE and Trichoderma. Proline content was significantly impacted by Trichoderma inoculation, interaction of water and DSE, and interaction of DSE and Trichoderma (Table S1). Under WW conditions, M inoculation significantly increased proline content, M and P inoculation increased soluble protein content, DSE and Trichoderma inoculation significantly decreased SOD activity, and M, A, and L inoculation significantly increased GSH content compared with control plants (Figure 2). PL and PA had higher soluble protein content, while ML, PA, and PL had significantly higher SOD activity than in those inoculated with a single strain ( Figure 2). Under DS conditions, P inoculation significantly increased proline content (+15.77%), and plants inoculated with ML had higher proline content than those inoculated with M and L ( Figure 2a). Inoculation with DSE and Trichoderma (except for L) enhanced soluble protein content, while the soluble protein content inoculated with MA, ML, and PL was higher than that of plants inoculated with single DSE and Trichoderma (Figure 2b). SOD activity was significantly enhanced by all inoculations; SOD activity after inoculating with MA and PA was higher than that inoculated with a single strain (Figure 2c). GSH content was significantly increased by M, P, and A inoculation compared with control plants. Plants inoculated with MA, PA, and PL had higher GSH content than those inoculated with single DSE and Trichoderma (Figure 2d).

Antioxidant Parameters and IAA Contents in Roots
GSH content was significantly influenced by all treatments and their interaction. NR activity and ASA content were significantly affected by all inoculations and their interaction (Table S1). Under WW conditions, M, A, and L inoculation significantly increased IAA content, while MA inoculation significantly improved IAA content compared with plants inoculated with M or A alone (Figure 3a). ASA contents of single-inoculated plants (except for M inoculation) were significantly increased (Figure 3b). GSH content was significantly improved in A. mongholicus by M and A inoculation, while PL inoculation significantly increased GSH content compared with single-inoculated plants (Figure 3c). All inoculated plants (except for A inoculation) significantly enhanced NR activity compared with the control plants, while PA inoculation had higher NR activity than P and A inoculation (Figure 3d). Under DS conditions, inoculation with P and A significantly increased IAA content, while inoculation with M and L had significantly decreased IAA content compared to control plants. Plants inoculated with ML and PA had higher IAA content than those inoculated with M, P, and L (Figure 3a). ASA content was significantly enhanced after inoculation with M and L, and significantly reduced after inoculation with P and A compared with the control plants. The ASA content of plants inoculated with MA and PA was larger than in those inoculated with M, P, and A (Figure 3b). The GSH contents of inoculated plants (except for L inoculum) were significantly enhanced compared with the control plants, and the GSH contents in plants inoculated with ML and PL were higher than in plants inoculated with a single strain (Figure 3c). The NR activity of plants inoculated by M, P, and A was significantly enhanced compared with the control plants ( Figure 4d). The NR activity of plants inoculated with ML and PA was higher than that of plants inoculated with M, P, and A (Figure 3d).

Active Ingredient Contents in the Roots
Calycosin-7-O-glucoside and formononetin contents were significantly impacted by all tested treatments and their interaction (Table S1)

Active Ingredient Contents in the Roots
Calycosin-7-O-glucoside and formononetin contents were significantly impacted by all tested treatments and their interaction (Table S1). Under WW conditions, all inoculated plants had significantly increased calycosin-7-O-glucoside contents compared with the control plants (Figure 3e). The formononetin content of plants with M and P were significantly enhanced compared with the control plants (Figure 3f). Inoculation with MA and PL led to a higher calycosin-7-O-glucoside and formononetin content, respectively, than single inoculation. Under DS conditions, all single-inoculated plants had significantly increased calycosin-7-O-glucoside contents compared with the control plants (Figure 3e). The formononetin content of plants with M inoculation was significantly increased compared to control plants. The content of formononetin in co-inoculated plants was higher than that in single-inoculated plants (Figure 3f).

N and P Content in the Shoots and Roots
Plant TN content was significantly influenced by all treatments and their interactions, and plant TP content was significantly influenced by all treatments, except for water condition (Table S1). Under WW treatment, plant TN and TP contents were significantly increased by P and L inoculation, while A inoculation only significantly enhanced plant TP content compared with control plants ( Figure S4). Under DS treatment, inoculation with MA and ML had an activity effect on plant TN content compared with M, A, and L inoculation ( Figure S4a). Inoculation with M and L significantly increased the plant TP content compared to control plants ( Figure S4b).

Soil Factors
Soil urease activity and NO 3 − -N content were significantly influenced by all treatments and their interactions, while water condition had no effect on NH 4 + -N content. SOC and SAP contents were significantly impacted by DSE and Trichoderma inoculation. ALP activity was significantly affected by water condition, Trichoderma inoculation, the interaction between water and DSE, and the interaction between DSE and Trichoderma (Table S1).
Under WW conditions, L inoculation significantly improved soil urease activity, while MA and ML inoculation increased soil urease activity more than single inoculation.
All single inoculations significantly enhanced SAP content but decreased NH 4 + -N content and NO 3 − -N content ( Figure 4). Under DS conditions, inoculation with M significantly enhanced ALP activity (Figure 4a). Single inoculation (except for P inoculation) significantly increased soil urease activity (Figure 4b). Co-inoculation of PL significantly decreased SOC content (Figure 4c). SAP content was significantly improved by all single inoculations except for A, and MA, ML, and PL inoculation led to higher SAP content than single DSE or Trichoderma inoculation (Figure 4d). Inoculation with M, P, and A significantly increased NH 4 + -N content, but it was decreased by PA inoculation (Figure 4e). NO 3 − -N content was significantly improved by A inoculation, but was decreased by M and P inoculation (Figure 4f).

Correlation Analyses of Growth Parameters and Chemical Constituents
Pearson correlation analyses showed a correlation between soil nutrient elements and growth parameters of A. mongholicus inoculated with DSE, Trichoderma, as well as DSE-Trichoderma under DS conditions (Table 1 and Tables S2-S4). Under DSE inoculation, ALP and urease activity were positively correlated with root biomass, root length, formononetin content, and NR activity. SAP and NH 4 + -N contents were positively correlated with Calycosin-7-O-glucoside and IAA contents. NO 3 − -N content was negatively correlated with root biomass, root length, calycosin-7-O-glucoside content, and NR activity. Under Trichoderma inoculation, shoot biomass, root biomass, and IAA content were positively correlated with urease activity, NH 4 + -N, and NO 3 − -N contents. Root length and calycosin-7-O-glucoside content were positively correlated with urease activity and NH 4 + -N content. However, SAP content was negatively correlated with shoot biomass and IAA content. Under co-inoculation, SAP content was positively correlated with root biomass, root length, and calycosin-7-O-glucoside content. NH 4 + -N content was positively correlated with root biomass, calycosin-7-O-glucoside content, and NR activity. Urease activity was positively correlated with calycosin-7-O-glucoside content, formononetin content, IAA content, and NR activity. Nevertheless, NO 3 − -N content was negatively correlated with root length and IAA content.

Variation Partitioning Analysis of Growth Parameters and Chemical Constituents
VPA was used to quantify the association between DSE species, Trichoderma species, water condition, soil nutrients, and plant growth ( Figure 5). Co-inoculation explained 40.1% of soil factors, which was higher than that of DSE inoculation (36.1%) and Trichoderma inoculation (20.3%), indicating that co-inoculation may make a greater contribution to soil nutrition (Figure 5a). The effect of water condition on soil factors was smaller than that of DSE inoculation and co-inoculation (Figure 5a1,a2). Co-inoculation (9.7%) explained more plant growth and physiological parameters than DSE inoculation (3.8%) and Trichoderma inoculation (1%) (Figure 5b). The amount of medicinal ingredients explained by co-inoculation (23.3%) was greater than that by DSE inoculation (16%) and Trichoderma inoculation (18.5%) (Figure 5c). For the contents of active ingredients, the effect of interaction between soil and inoculation was greater than that of soil or inoculation alone (Figure 5c).

Discussion
The interaction between microorganisms and plants in nutrient absorption, gr and defense of the system is an integral part of the ecosystem [43]. Microorganism considered to be natural partners of plant defense mechanisms under adverse cond [44]. Although many studies showed that DSE and Trichoderma can promote the g of plants, few studies have been performed on the growth-promoting mechanis plants co-inoculated with DSE and Trichoderma under drought stress. In this study, t DSE hyphae and microsclerotia were found in all treated root samples, indicatin DSE could effectively colonize the roots of A. mongholicus, even under drought cond Inoculation of T. longibrachiatum and T. afroharzianum promoted the colonization pseudophaseolina and P. radicina, indicating that Trichoderma and DSE could coexist roots of A. mongholicus.
In this study, the root biomass and root length of A. mongholicus under drought were better than those under well-watered conditions. It has been proven that allo

Discussion
The interaction between microorganisms and plants in nutrient absorption, growth, and defense of the system is an integral part of the ecosystem [43]. Microorganisms are considered to be natural partners of plant defense mechanisms under adverse conditions [44]. Although many studies showed that DSE and Trichoderma can promote the growth of plants, few studies have been performed on the growth-promoting mechanisms of plants co-inoculated with DSE and Trichoderma under drought stress. In this study, typical DSE hyphae and microsclerotia were found in all treated root samples, indicating that DSE could effectively colonize the roots of A. mongholicus, even under drought conditions. Inoculation of T. longibrachiatum and T. afroharzianum promoted the colonization of M. pseudophaseolina and P. radicina, indicating that Trichoderma and DSE could coexist in the roots of A. mongholicus.
In this study, the root biomass and root length of A. mongholicus under drought stress were better than those under well-watered conditions. It has been proven that allocating greater biomass to roots is a key mechanism to enhance drought tolerance in plants [45].
Under drought conditions, DSE and Trichoderma inoculation had no significant effects on shoot biomass, but DSE and Trichoderma inoculation made significant contributions to the root biomass and root length of A. mongholicus. Li et al. [46] found that DSE inoculation can regulate the root biomass and root architecture of Hedysarum scoparium to improve the performance of plants under drought conditions. T. harzianum significantly increased root growth and development in maize and several other crop plants [47]. The application of DSE or Trichoderma led to the development of deep and extensive roots to penetrate farther into the soil, which helped to absorb deep-seated water and nutrients to enhance drought tolerance in plants [48,49]. Under drought stress, the effect of P. radicina + T. longibrachiatum on root length was better than that of P. radicina and T. longibrachiatum alone. He et al. [28] found that three DSE strains inoculated with Trichoderma viride at different concentrations in Glycyrrhiza uralensis and dual inoculation can enhance the performance of G. uralensis by promoting root development. Wang et al. [50] found that the co-inoculation of Funnelliformis mosseae and Sinorhizobium medicae improved the nutrient absorption and growth of alfalfa, which could be an effective way to improve productivity. In addition, DSE and Trichoderma inoculation significantly enhanced calycosin-7-O-glucoside content, and the effect of co-inoculation on the content of active ingredients was greater than that of single inoculation. Li et al. [51] found that DSE inoculation enhanced the epigoitrin content of Isatis tinctoria under drought. The withanolide A content of T. viride-treated plants increased by co-inoculation of T. viride, and Sarocladium kiliense modulated the expression of the withanolide biosynthetic pathway genes of Withania somnifera [52]. Thus, DSE and Trichoderma inoculation can be used to stimulate the production of more active ingredients in medicinal plants.
Drought stress causes excessive production of reactive oxygen species (ROS) in plants, leading to oxidative stress that damages plant cells and ultimately reduces plant growth and development [53]. When stressed, plants produce effective enzymes such as SOD to remove ROS and protect cells from oxidative damage [54]. Our results showed that both DSE and Trichoderma inoculation had positive effects on SOD activity, and M. pseudophaseolina + T. afroharzianum and P. radicina + T. longibrachiatum inoculation had higher SOD activity than single inoculation. Li et al. [51] revealed that inoculation with DSE can enhance SOD activity under drought stress. Inoculation with Trichoderma resulted in overexpression of SOD metabolic pathway genes and significantly increased SOD activity in drought-stressed rice plants as compared to control plants [55]. Singh et al. [56] found that increasing SOD activity can improve ROS scavenging ability in rice plants with co-inoculation of Trichoderma asperellum and Pseudomonas fluorescens. These results indicated that inoculation with DSE and Trichoderma can improve antioxidant enzyme activity and alleviate the oxidative damage induced by drought. In response to osmotic stress caused by drought, osmotic regulation is also an important mechanism for plants to regulate water potential to adapt to drought stress [57]. Soluble proteins, as compatible osmotic regulators, can promote osmotic regulation and plant tolerance to dehydration. Although drought resulted in decreased protein synthase activity and protein content in plants, inoculation with DSE and Trichoderma significantly improved the soluble protein content in this study, and the effect of co-inoculation was higher than that of individual inoculation. These results suggest that DSE and Trichoderma inoculation can promote the production of soluble proteins to alleviate the negative effects of drought [58,59]. Proline plays an important role in osmotic regulation, protein property stability, and cell structure stability under drought stress [4]. Valli and Muthukumar [60] also found that DSE inoculation could significantly enhance proline accumulation in tomato shoots. Rawat et al. [61] indicated that T. harzianum inoculation increased proline content to reduce the oxidative damage caused by salt stress. Rezaei-Chiyaneh et al. [62] found that co-inoculated plants under severe water stress produced the highest proline contents in ajowan leaves. Thus, DSE and Trichoderma inoculation could promote the biosynthetic activity of proline to regulate osmotic pressure under water deficiency. GSH is an important antioxidant composed of glutamic acid, cysteine, and glycine, and the sulfhydryl structure contained in it can combine with reactive oxygen species, thereby reducing the oxidative damage of cells [63]. ASA is a water-soluble antioxidant organic small molecule widely present in plants. In order to maintain redox balance in stress, the ASA-GSH cycle, as the main nonenzymatic antioxidant to maintain redox dynamic balance in cells, is the main defense system against ROS damage in cells [64]. Sadeghi et al. [65] found that inoculation with endophytic fungi significantly increased GSH and ASA content in citrus under drought to induce drought tolerance in the host. In the present study, A. mongholicus inoculated with DSE and Trichoderma under drought stress obtained a large amount of GSH and ASA, which adjusted the redox state and prevented cell damage. Whereas inoculation with T. afroharzianum and M. pseudophaseolina + T. longibrachiatum decreased the GSH content in leaves, inoculation with P. radicina and T. afroharzianum decreased the ASA content in roots. Mastouri et al. [66] also discovered similar results in maize inoculated with Trichoderma isolate T-22. This may be due to the conversion of reduced GSH and ASA to a more oxidized form with the increase of ROS, which directly or indirectly removes ROS [64,65]. These results show that DSE and Trichoderma inoculation increased enzymatic and nonenzymatic antioxidant reactions as well as metabolites induced by inoculation, which acted as agents of quenching and contributed strongly to alleviating drought in plants.
Plant hormones play an important role in plant growth and development and in the interaction between plants and endophytes [67]. IAA, the most abundant auxin in nature, plays a central role in beneficial root-microbial interactions and is generally thought to be involved in balancing growth and defense responses, coordinating many key processes in plant growth [68]. Although drought reduced IAA levels, IAA content was enhanced by inoculating P. radicina and T. afroharzianum. Additionally, the IAA content of plants co-inoculated with M. pseudophaseolina + T. longibrachiatum and P. radicina + T. afroharzianum was higher than that of individual inoculation under drought conditions. It was found that the key enzymes encoding IAA synthesis were significantly upregulated under DSE treatment [69], and DSE colonization promoted auxin production in plants [70]. Qiang et al. [71] also found that wheat plants inoculated with an IAA-producing endophytic fungus, Alternaria alternata LQ1230, could improve wheat growth and enhance drought tolerance. Trichoderma strains could also increase IAA secretion in response to drought, showing the ability to protect plants [14,15,58]. Coculture of Arabidopsis seedlings with Trichoderma virens or Trichoderma atroviride showed characteristic auxin-induced phenotypes, including increased biomass production and stimulated lateral root development [72]. In addition, IAA concentration in tobacco plants was enhanced by co-inoculating Glomus versiforme and Bacillus methylotrophicus, and the IAA concentration in co-inoculated plants was higher than that in single-inoculated plants [73]. The interaction between beneficial microorganisms and plants can regulate the hormone metabolism of the associated plants to induce root growth and development and promote nutrient absorption by host plants, thus improving the drought tolerance of plants [72,74].
The results of a variance decomposition showed that DSE and co-inoculation explained more for soil factors than water treatments, so the addition of exogenous microorganisms inevitably affects soil properties. NO 3 − -N and NH 4 + -N are the main inorganic nitrogen sources that plants can absorb and utilize [75]. NR (nitrate reductase) is a key enzyme in the nitrate assimilation pathway of higher plants, which can transform nitric acid to ammonia and improve the absorption of N by plants [74]. Therefore, its activity can be used to evaluate the ability of plants to absorb and assimilate N [76]. Urease can convert urea into ammonia and increase the content of N in soil [77]. In this study, inoculation with DSE and Trichoderma could affect NR, urease, NH 4 + -N, and NO 3 − -N. Upson et al. [78] found that DSE are able to mineralize peptides and amino acids in the rhizosphere, making N more freely available to roots. Wu et al. [79] indicated that colonization by DSE S16 significantly promoted the activity of NR and some genes involved in nitrate regulation in sweet cherry. Some existing research found that DSE hyphal infection and Trichoderma treatment were positively correlated with soil urease activity [80,81]. Thus, DSE and Trichoderma are involved in regulating the uptake of nitrate by plants, and sufficient N supply may contribute to osmoregulation and ROS homeostasis in plants, making plants better adapted to drought [82]. In addition, the content of available P increased, which may be due to DSE and Trichoderma dissolving insoluble phosphate in the soil and using it as a biofertilizer to aid plant growth. Studies have shown that the organic acids produced by DSE cause the dissolution of phosphates, helping to enhance P supply to plants [83]. Trichoderma M14 had the ability to solubilize unavailable P for plant growth [84]. T. Asperellum produced phosphatases and dissolved inorganic or organic phosphates [85]. The content of available P in soils inoculated with M. pseudophaseolina + T. afroharzianum, M. pseudophaseolina + T. longibrachiatum, and P. radicina + T. longibrachiatum was higher than that of single inoculation, so dual inoculation with compatible strains could play a synergistic role in improving plant mineral nutrient absorption. The above results indicate that DSE and Trichoderma inoculation can increase soil enzyme activities to accelerate soil fertility. At the same time, the organic N and insoluble P in the soil were mineralized, the content of available N and available P was increased to improve the utilization rate of macro and trace elements of plants, and it increased the interaction between plants and soil.
The VPA showed that soil nutrients had a greater impact on plant growth under DSE treatment. Correlation analysis showed that DSE inoculation increased the activity of NR in roots and soil enzymes to promote the absorption of nitrate N by plants to promote the growth of plant roots and the content of active ingredients. Deng et al. [86] demonstrated that the biosynthesis of flavonoids is strongly influenced by N status, and N availability alters flavonoid accumulation by affecting the internal carbon/nitrogen balance. Meanwhile, the enlarged plant roots could absorb more soil inorganic matter and form a good cycle. Trichoderma inoculation increased the activity of soil urease and accelerated the production of soil NH 4 + -N, which promoted the growth of plants. Trichoderma inoculation promotes plant uptake of available P to increase aboveground biomass and IAA content. It was found that IAA can upregulate genes encoding enzymes responsible for the synthesis of carboxylates involved in P solubilization to increase soluble P content [87]. P mainly increased root growth by upregulating the positive regulation root growth gene, and then affected root nutrient absorption and plant growth [88]. Hence, P was related to the increased IAA content [89]. Co-inoculation promoted the mineralization of insoluble P and soil available P content, which eventually led to the utilization of available P by plant roots. Co-inoculation increased the NH 4 + -N content in the soil, resulting in an increase in plant root growth and the content of active ingredients. These results illustrate that, in addition to profound physiological changes in plants following DSE and Trichoderma inoculation, DSE and Trichoderma mediated responses to plant-soil feedback. Therefore, functional symbiosis is established by regulating plant drought resistance, improving soil nutrition, and promoting plant growth and other mechanisms.

Conclusions
In this study, we found that DSE can effectively colonize the roots of A. mongholicus, and Trichoderma promoted the colonization of DSE in the roots of A. mongholicus. DSE and Trichoderma inoculation significantly enhanced growth and active ingredient contents of A. mongholicus regardless of the water condition. Interestingly, under drought stress, P. radicina + T. longibrachiatum had larger root length and co-inoculation significantly enhanced the formononetin content of A. mongholicus compared to single inoculation. Additionally, Trichoderma relies on antioxidant enzymes, growth hormones, and a redox system to resist drought, while DSE strains have an additional osmotic regulation system in addition to the drought resistance function possessed by Trichoderma, and the effect of co-inoculation (especially M. pseudophaseolina + T. longibrachiatum and P. radicina + T. afroharzianum) on plant physiology was greater than that of single inoculation. DSE inoculation increased soil enzymes to promote nitrate nitrogen absorption by plants. Trichoderma inoculation increased soil urease activity and plant uptake of available phosphorus. Co-inoculation promoted the mineralization of insoluble phosphorus in soil. Our results supplement previous findings that endophytes can enhance the drought resistance of plants and emphasize the importance of using DSE and Trichoderma in medicinal plant cultivation under drought stress conditions. As A. mongholicus plays an important role in the cultivation of medicinal plants, the DSE-Trichoderma-A. mongholicus association should be further field-tested to determine its ability to suppress drought stress in dryland agriculture.

Data Availability Statement:
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.