The arbuscular mycorrhizal fungus Rhizophagus clarus improves physiological tolerance to drought stress in soybean plants

Soybean (Glycine max L.) is an economically important crop, and is cultivated worldwide, although increasingly long periods of drought have reduced the productivity of this plant. Research has shown that inoculation with arbuscular mycorrhizal fungi (AMF) provides a potential alternative strategy for the mitigation of drought stress. In the present study, we measured the physiological and morphological performance of two soybean cultivars in symbiosis with Rhizophagus clarus that were subjected to drought stress (DS). The soybean cultivars Anta82 and Desafio were grown in pots inoculated with R. clarus. Drought stress was imposed at the V3 development stage and maintained for 7 days. A control group, with well-irrigated plants and no AMF, was established simultaneously in the greenhouse. The mycorrhizal colonization rate, and the physiological, morphological, and nutritional traits of the plants were recorded at days 3 and 7 after drought stress conditions were implemented. The Anta82 cultivar presented the highest percentage of AMF colonization, and N and K in the leaves, whereas the DS group of the Desafio cultivar had the highest water potential and water use efficiency, and the DS + AMF group had thermal dissipation that permitted higher values of Fv/Fm, A, and plant height. The results of the principal components analysis demonstrated that both cultivars inoculated with AMF performed similarly under DS to the well-watered plants. These findings indicate that AMF permitted the plant to reduce the impairment of growth and physiological traits caused by drought conditions.

− 0.94 MPa after 3 days of treatment (F 3.15 = 10.51, p = 0.0011), which was significantly lower than all the other treatments (WW, AMF and DS + AMF), which had Ψ w values of no less than − 0.31 MPa (Fig. 3). On day 7, the DS group maintained significantly higher values (Ψ w = − 0.95 MPa), while the DS + AMF plants remained similar to the AMF and WW treatments (Fig. 3).
After 3 days, (F 3.15 = 38.85, p = 0.0001) the DS group of the Desafio cultivar had a significantly lower Ψ w value, than the WW, AMF, and DS + AMF groups, which were, once again all similar to one another (Fig. 3B). At 7 days (F 3.15 = 32.12, p = 0.0001) the DS + AMF had a significantly higher Ψ w , of − 0.54 MPa, than the DS group, which had a value of − 0.79 MPa (Fig. 3B).
Photosynthetic traits. The photosynthetic rate (A) was reduced significantly by drought stress in the Anta82 cultivar on days 3 (F 3.15 = 4.214, p = 0.0298) and 7 (F 3.15 = 4.80, p = 0.0201) (Fig. 4A). The stomatal conductance (g S ) was significantly lower in the DS plants than in the WW plants on day 3 (F 3.15 = 11.825, p = 0.0007) (Fig. 4C). The transpiration rate (E) was altered significantly in the DS group on day 7 (F 3.15 = 18.93, p = 0.0001) (Fig. 4E). The C i /C a ratio was not affected by either treatment (Fig. 4F). Water use efficiency (WUE) increased significantly in the AMF and DS groups on days 3 ( The stomatal conductance (g S ) was higher in the WW and AMF groups on days 3 (F 3.15 = 13.59, p = 0.0004) and 7 (F 3.15 = 14.97, p = 0.0002) (Fig. 4D), as was the transpiration rate (E) on days 3 (F 3.15 = 8.67, p = 0.0025) and 7 (F 3.15 = 12.427, p = 0.0005) (Fig. 4F). The Ci/Ca ratio was lower in the DS plants than in the DS + AMF group on day 3 (Fig. 4H). On day 7, the WUE (F 3.15 = 7.47, p = 0.0044) was significantly higher in the DS + AMF plants than in the WW and DS groups (Fig. 4J).
In the case of the Desafio cultivar, the minimum fluorescence did not change significantly on either day 3 (F 3.15 = 0.120; p = 0.9469) or day 7 (F 3.15 = 1.107; p = 0.3845) (Fig. 5B). The maximum quantum yield of PSII increased slightly in the inoculated plants on day 3, although no differences were found among the treatments on day 7 (Fig. 5D). Similarly, neither the effective quantum yield of PSII (F 3.15 = 2.071, p = 0.1577) nor the electron transport rate (F 3.15 = 2.234, p = 0.1368) varied among the treatments after 3 days, although a significant increase was observed in the DS + AMF plants in comparison with the DS group after 7 days (YII: F 3.15 = 6.077, p = 0.009;   Principal components analyses. The first two components explained 43.4% of the variance in the Anta82 data and 46.9% of that in the Desafio data. The plot of the first two PCA axes clearly differentiated a DS cluster from the other groups in the Anta82 data (Fig. 8A,B).
The first PCA axis of the Desafio data, which explains 24.6% of the variance, is not systematically related to the different treatments (Fig. 8C,D), while the second axis clustered the DS records to the negative side of the plot, the DS + AMF and WW in the middle, and the AMF records toward the positive side. In the case of the Anta82 cultivar, most of the variables (LDM, E, Ψw, Chl, Carot, S root , g S , A, ETR, RDM, Y II , F v /F m , Mg leaf , Nleaf, MCR, C i /C a , SD, PH, Ca leaf , and Ca root ) were significantly correlated (p < 0.05) with the first PCA axis, which means that www.nature.com/scientificreports/ they decreased with drought stress. In contrast, WUE, Mg root , K root , N root and NPQ correlated negatively with this axis, which means that these variables increased with drought stress in the Anta82 cultivar.
In the case of the Desafio cultivar, Ψw, Carot, LDM, E, F v /Fm, Ci/Ca, g S , PH, LA, Nroot, ETR, Caleaf, WUE, Mgleaf, and RDM were all correlated positively with PC2, while S root and NPQ were correlated negatively with this axis. This means that the former group of variables tends to increase with AMF and decrease with the DS treatments, whereas S root and NPQ increase with DS and decrease with AMF in the Desafio cultivar.
The third axis explained 10.3% of the variance in the Anta82 data and 9.3% of that in the Desafio data. PC3 was thus able to differentiate WW from the other treatments in both cultivars. In the case of the Anta82 cultivar, PC3 also clustered the DS + AMF data. The K leaf , Y II , ETR, P leaf , WUE, Carot, and N root variables were positively correlated with the third PCA axis in the Anta82 data, which means that these variables increase with DS + AMF and decrease with WW, while F 0 , E and PH were negatively correlated with this axis.. These variables thus presented the opposite pattern, increasing with WW and decreasing with DS + AMF. In the case of the Desafio cultivar, F v /F m , Carot, NPQ, M groot , WUE, and M gleaf were all positively correlated with PC3, while K leaf , RDM, and N leaf were negativelycorrelated . As the WW data were located toward the negative side of the third axis, the former set of variables decreased and the latter increased with this treatment.
Overall, the results of the PCA demonstrated that plants under drought stress colonized with AMF clustered together with the well-watered plants of both cultivars, which clearly indicates that inoculation with AMF contributed to a reduction in the physiological impairment of the plant provoked by drought conditions.
The negative effects of drought also appear to be more pronounced in the Anta82 cultivar than in Desafio. In particular, the third axis of the Anta82 dataset and the second axis of the Desafio data provide insights into the physiological impact of AMF on the plants, independent of their watering regime.

Discussion
The Anta82 and Desafio soybean cultivars presented differential responses to the inoculation and water treatments. The results of the present study support the hypothesis that inoculation with the arbuscular mycorrhizal fungus R. clarus leads to the colonization of the roots of the soybean plant and that, under drought stress, it favors the water status and metabolic activities of the plant, particularly in the case of the more drought-tolerant Desafio cultivar. Arbuscular mycorrhizal fungi are abundant and widely distributed in an ample range of environments www.nature.com/scientificreports/ and may significantly enhance the stress tolerance of host plants 15 . It is also important to note that the presence of different AMF species is related to the type of host plant and edaphoclimatic conditions 43 . Colonization by mycorrhizae is the principal route of symbiosis with the host plant 44 , which favors both the organisms involved in the relationship. The application of AMF is a potential strategy for the enhancement of the capacity of a plant to tolerate drought in arid ecosystems 45,46 .
The potential for colonization depends on the plant and the type of fungus, in addition to the cultivation conditions and exposure time [47][48][49][50] . The Anta82 cultivar is more sensitive to drought stress (DS) and presented an increase in mycorrhizal colonization, especially under drought stress on days 3 and 7. However, the Desafio cultivar presented a pronounced increase in mycorrhizal colonization after 7 days of drought stress. Moreira et al. 51 found that coffee plants inoculated with R. clarus had a higher mycorrhizal colonization (39%) and root dry matter than noninoculated plants, in soils at 71% of field capacity. Inoculation with AMF of different species has also been shown to potentially mitigate the effects of drought stress and increase resistance in bean 52 , rice 53 , tomato 54 , wheat 55 , and in orange 15 and apple 56 trees.
The mechanisms associated with the maintenance of the water status of a plant are triggered rapidly when hydrological conditions become limited, which means that drought stress is one of the principal factors affecting plant growth and production 6,57 . These mechanisms include stomatal regulation, which responds rapidly to water stimuli 10,58 . As the root-bound hyphae of mycorrhizal fungi improve water uptake, their association with plants is a potential ally for the maintenance of their water status 47 . Under conditions of drought stress, there is an increase in the space and accumulation of air between the soil particles and the roots, which can be compensated for the presence of AMF, guaranteeing the transport of water 15 . Our results show that AMF mitigated the adverse effects of drought stress in both cultivars. Similarly, the water potential of Poncirus trifoliata plants was also increased by 20% under conditions of water deficit when they were inoculated with mycorrhizal fungi 59 . Under drought conditions, AMF helps the host plants to absorb more water, which means that inoculation with these fungi may provide an important strategy for the improvement of the productivity of plants grown in semi-arid regions 60 .
A mechanism that maintains plant turgor, the closure of the stomata may limit photosynthetic processes and compromise crop yields [61][62][63] . Low photosynthetic yields related to stomatal limitations have been observed in plants exposed to drought stress, although inoculation with AMF increases stomatal conductance under drought stress in comparison with noninoculated plants. Photosynthesis and transpiration were also higher in the presence of mycorrhizae under drought stress conditions in both cultivars in the present study, which indicates that plants associated with AMF can improve their water status under drought stress through their ability to use water more efficiently. Quiroga et al. 64 found that maize plants in symbiosis with AMF R. irregularis under drought conditions had increased stomatal conductance and photosynthesis parameters. Inoculation with AMF may have beneficial effect, even in the absence of drought stress. Coffee plants of three different cultivars, which were well-watered and inoculated with the spores of R. clarus and/or Acaulospora colombiana presented an increase in photosynthetic rates, stomatal conductance, transpiration, water use efficiency, and the percentage of mycorrhizal colonization 65 . These findings indicate that the AMF had an active role in this process, permitting greater stomatal opening and higher photosynthetic rates, associated with higher turgor, as recently observed in eggplant 27 , tomato 54 , and Olea europaea 66 . This was possible because the AMF enhanced water uptake, even under limiting conditions 67,68 , thereby inducing changes in the critical substrate water potential and increasing the water transport in the colonized substrates 26 , which would permit greater water-use efficiency. The maintenance of water status also favors the fixation of atmospheric CO 2 and increases the movement of photoassimilates (the "sink effect") from the aerial parts of the plant to its roots 48 .
Photosynthetic limitation is directly related to the photochemical responses of the plant, as demonstrated by the chlorophyll a fluorescence variables 69 . Drought stress may compromise the functionality of the chloroplast electron chain 9,70,71 , limit the production of energy, and reduce the energy required for the completion of the photosynthetic process. In the present study, the inoculation of soybean plants with AMF under drought stress permitted the maintenance of the maximum quantum yield of PSII (F v /F m ) in comparison with well-watered plants, as observed previously in rice 72 , watermelon 73 , wheat 74 and maize 14 . By day 3 of the present study, the plants under drought stress used a thermal dissipation mechanism (NPQ) to avoid excess energy expenditure. Thermal dissipation by the xanthophyll cycle, which is activated in response to the pH gradient formed by the cyclic electron cycle 75,76 , is known to be an initial response to abiotic stress and permits the regulation of the amount of excitation energy directed to the reaction centers of the photosystems 77,78 . Higher NPQ rates are important early protection mechanisms for the photosystem, as they avoid the effects of photo-oxidative stress on the photosynthetic photochemical protein complex. The increase in NPQ was efficient in the presence of AMF, especially in plants of the Desafio cultivar, even after 7 days of exposure to drought stress.
The exposure of plants to drought stress can cause an excess of energy expenditure that is not devoted to the photosynthetic process. This results in the increased production of reactive oxygen species (ROS), which promotes the peroxidation of lipids, proteins, and chloroplast pigments 79,80 . There was also a reduction in the concentration of photosynthetic pigments, in particular chlorophyll a, which also acts on the photosystem reaction centers. In the present study, drought stress contributed to a reduction in the photosynthetic pigments of the soybean plants, although this damage was mitigated in the plants inoculated with AMF. These results are consistent with the findings of Baslam and Goicoechea 81 , who observed that the association of the AMF Rhizophagus intraradices (Glomus intraradices) and Funneliformis mosseae associated with Lactuca sativa plants resulted in an increase in the chlorophyll and carotenoid content, even after exposure to drought stress. In addition, R. clarus and a mixture of AMF spores (including those of R. clarus) induced an increase in the activity of antioxidant enzymes and the concentration of malondialdehyde in strawberry 82 and soybean plants 83 under drought conditions, which may also contribute to the avoidance of oxidative stress and the degradation of pigments.
The impairment of the photochemical stage, associated with the oxidative stress caused by drought stress, may limit plant growth. Under unfavorable water conditions, the plant tends to initially expands its root system www.nature.com/scientificreports/ to increase the area of contact with the soil 12,84,85 . This compromises aerial growth, including the development of new leaves and shoots 86 . In the present study, even after a short period of stress, it was possible to observe a reduction in the vegetative growth of the soybean plants exposed to drought stress, which was mediated in part in the plants inoculated with R. clarus. Inoculation with Funneliformis geosporus and Funneliformis mosseae contributed to the growth of Fragaria ananassa plants under drought stress, as observed by Boyer et al. 87 . Oliveira et al. 88 found an increase in the yield of Cicer arietinum following inoculation with the AMF Rhizophagus irregularis and the bacterium Mesorhizobium mediterraneum in the absence of drought conditions. Under drought stress, however, these authors observed an increase in plant biomass and the crude protein content of the grains in comparison with the control plants. Bernardo et al. 89 found that wheat plants (Triticum spp.) inoculated with Funneliformis mosseae (Glomus mosseae) accumulated more dry matter in the aerial parts under drought stress than the noninoculated plants. Plant growth occurs through the accumulation of nutrients, which is enhanced by the interaction between the soybean plants and the AMF. Under drought stress, nutrients may be accumulated in or adsorbed by the roots, due to the reduction in the flow of mass that is normally promoted by transpiration. When associated with AMF, plants increase the efficiency of their water use, which permits the transfer of nutrients to the shoots. The symbiosis between AMF and legumes may be relevant to the rhizobia nodulation of N 2 -fixing bacteria. This interaction may be responsible for nutrient recycling and favor nutrient uptake 67 , as well as improving the tolerance of abiotic stress 90 . Rhizophagus clarus is also known to increase the effectiveness of chemical fertilizers in soybean plants under field conditions, and to increase the P and N contents in inoculated plants 91 . This is an economically important finding, given that it would favor a reduction in the application of fertilizers to farm crops. Overall, inoculation with R. clarus increased the capacity of soybean plants to tolerate drought stress, by modifying their metabolism to permit the maintenance of or even an increase in their development under moderate drought conditions. The potential mitigation of the effects of drought stress by R. clarus provides important insights for the development of further research, including field experiments to verify the response of the plants under natural conditions, considering the potential occurrence of multiple abiotic stressors, to support the development of more efficient agricultural practices in regions subject to moderate water stress.

Conclusions
The arbuscular mycorrhizal fungus Rhizophagus clarus supported the maintenance of the water status of Anta82, a drought-sensitive soybean cultivar, mitigating the negative effects of drought stress on the photosynthetic apparatus of this plant. In the case of the Desafio cultivar, which is moderately drought-tolerant, greater colonization by Rhizophagus clarus increased the concentration of photosynthetic pigments and improved the physiological performance of the plant and its growth. These data indicate that inoculation with Rhizophagus clarus is a potentially important tool for the improvement of soybean yields, especially in regions with low precipitation that are subject to drought.

Materials and methods
Plant material and experimental conditions. The experiment was conducted in a climated-controlled greenhouse (~ 27 °C and relative humidity of ~ 75%) at the Laboratory of Ecophysiology and Plant Productivity at the Rio Verde campus of the Goiano Federal Institute of Science and Technology in Rio Verde, Goiás, Brazil. In the greenhouse, soybean plants (Glycine max (L.) Merrill) of two cultivars-Anta82 RR (Anta82; Geneze Seeds, São Paulo, Brazil), which is drought-sensitive, and the moderately drought-tolerant Desafio 8473 RSF (Desafio; Brasmax Sementes, Cambé, PR, Brazil)-were grown in 3-dm 3 pots. Each pot contained a mixture (2:1) of Red Latosol (LVdf), which is the typical soil of the Brazilian Cerrado savanna, and sand, which had been corrected to 60% of base saturation with PNRT 100 dolomitic limestone. This substrate was fertilized based on the results of the chemical analysis (Supplementary Table S1) and the recommendation for Cerrado soils 92 .
The arbuscular mycorrhizal fungus (AMF) Rhizophagus clarus was donated by the germplasm bank of the Soil Microbiology Laboratory at the Ilha Solteira campus of São Paulo State University (UNESP) in Ilha Solteira, São Paulo, Brazil. To multiply the inoculum, the sand/soil (2:1) substrate was first placed in cotton bags, which were sterilized in an autoclave at 1.5 atm at a temperature of 121 °C for two hours. The substrate was then dried in an oven at 105 °C for 24 h, and rehydrated with distilled and sterilized water for 24 h 93,94 . Each plastic pot (1 dm 3 ) containing sterilized substrate received approximately 10 g of the AMF inoculum together with Sorghum bicolor seeds, which acted as host plants. After 3 months, the soil was collected and stored in an ultrafreezer prior to use in the experiment. The density of spores was determined by placing a sample of the substrate on an acrylic plate with concentric rings, and examined under a SteREO Discovery.V8 stereomiscroscope (Zeiss, Göttingen, Germany) to count the spores, following Gerdemann and Nicolson 95 and Jenkins 96 . The spores were stored in an ultrafreezer. For the experiment, each seeding hole in the 3-dm 3 pots was inoculated with 10 g of the inoculum containing 3.5 g of R. clarus spores during the sowing of the soybean 97 .
The experimental was based on a random-block design, with four replicates. Each pot containing four plants was considered to be an experimental unit. The experimental water treatment (drought stress) was applied when the plants reached vegetative developmental stage V3, approximately 30 days after germination, when the waterholding capacity (WHC) was maintained at 60%, based on the gravimetric method. The plants in the control treatment (well-watered) were maintained at 100% WHC. The treatments were as follows: well-watered (WW); well-watered and inoculated with R. clarus (AMF); drought stress (DS); and DS plants inoculated with R. clarus (DS + AMF). The plants in all four groups were evaluated on days 3 and 7 after the initiation of the treatment.
Mycorrhizal colonization rate. Mycorrhizal  www.nature.com/scientificreports/ initially immersed in 2% KOH, heated in a stove at 90 °C for 120 min, and then washed in distilled water. The samples were then immersed in 1% HCl for 30 min, washed in water, and stained with 0.05% trypan blue in a lactoglycerol solution (1:1:1 lactic acid, glycerol, water) at 90 °C for 10 min 99 . The root fragments were mounted on microscope slides and observed at a magnification of 200 × under an Olympus BX61 microscope (Tokyo, Japan) to determine the percentage of root volume colonized by the fungus 100 .
Physiological traits. Gas exchange was measured in fully expanded leaves to determine the net photosynthetic assimilation rate (A, µmol CO 2 m −2 s −1 ), stomatal conductance (gs, mol H 2 O m −2 s −1 ), transpiration rate (E, mmol H 2 O m −2 s −1 ), and the ratio between internal and external CO 2 concentrations (C i /C a ). The measurements were obtained using an Infrared Gas Analyzer (IRGA; LI-6400xt, Licor, Lincoln, NE, USA). The instantaneous water-use efficiency (WUE, in µmol CO 2 mmol −1 H 2 O) was calculated as the ratio between A and E. All measurements were obtained under a constant photosynthetic photon flux density (PFFD, 1000 µmol photons m -2 s −1 ) and at the ambient atmospheric CO 2 concentration (Ca, ~ 400 µmol mol −1 ), temperature (~ 25 °C), and relative humidity (~ 50%). Morphological traits and nutrient content. The plants were measured to determine the plant height (PH, cm) and stem diameter (SD, mm). The leaves were separated out to obtain the leaf area (LA, cm 2 ). The leaves and roots were dried to a constant weight at 65 °C in a forced-air circulating oven to obtain the leaf dry matter (LDM, g) and root dry matter (RDM, g).
To quantify the nutrient content of the leaves and roots, the content of the dry material (~ 500 mg) was extracted by nitric-perchloric (3:1) digestion and analysed following Embrapa 102 . Nitrogen (N) was measured by the Kjeldahl titration method using a nitrogen distiller (TE-0364, Tecnal, Piracicaba, Brazil). Phosphorus (P) was determined by the molybdenum blue method, and sulfur (S) was determined by the barium chloride turbidity approach, using molecular absorption spectrophotometry (SP1105, Tecnal, Piracicaba, Brazil). Potassium (K) was analyzed using flame photometry (B462, Tecnal, Piracicaba, Brazil) and calcium (Ca) and magnesium (Mg) were determined using atomic absorption spectrophotometry (SavantAA, GBC Scientific Equipment, Braeside, Australia).

Statistical analysis.
The variation in the data were evaluated using analysis of variance (ANOVA) and pairs of means were compared using Tukey's post hoc test (p < 0.05). These analyses were run in SISVAR software (v. 5.6, Lavras, MG, Brazil).
Principal components analyses (PCA) were run on the whole dataset, and for each cultivar separately. These analyses were run in the FactoMineR 103 and factoextra 104 packages in R software 105 . The data were first scaled using the scale function and then analyzed using the PCA function. The eigenvalues were evaluated to determine the number of axes that should be evaluated. We used the screen plot to visualize the data, based on Cattell's rule, which states that the components that correspond to the eigenvalues to the left of the straight line (eigenvalues lying on the straight line correspond to random variation) should be retained 106 . To better understand the variables represented in each component, we used the fviz_contrib and dimdesc functions, which determine the contribution of the variable to each component and its correlation with the component, respectively.
Ethics approval and consent to participate. Manuscripts reporting on studies do not involve any human participants, human data, or the analysis of human tissue are exempt from these considerations.

Complies with international, national and/or institutional guidelines. The experimental research
reported here complies with all the relevant institutional, national, and international guidelines and legislation. www.nature.com/scientificreports/ Correspondence and requests for materials should be addressed to G.C.M.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.