Endophytic fungi harbored in the root of Sophora tonkinensis Gapnep: Diversity and biocontrol potential against phytopathogens

Abstract This work, for the first time, investigated the diversity of endophytic fungi harbored in the xylem and phloem of the root of Sophora tonkinensis Gapnep from three geographic localities with emphasis on the influence of the tissue type and geographic locality on endophytic fungal communities and their potential as biocontrol agents against phytopathogens of Panax notoginseng. A total of 655 fungal strains representing 47 taxa were isolated. Forty‐two taxa (89.4%) were identified but not five taxa (10.6%) according to morphology and molecular phylogenetics. Out of identifiable taxa, the majority of endophyte taxa were Ascomycota (76.6%), followed by Basidiomycota (8.5%) and Zygomycota (4.3%). The alpha‐diversity indices indicated that the species diversity of endophytic fungal community harbored in the root of S. tonkinensis was very high. The colonization and species diversity of endophytic fungal communities were significantly influenced by the geographic locality but not tissue type. The geographic locality and tissue type had great effects on the species composition of endophytic fungal communities. Forty‐seven respective strains were challenged by three fungal phytopathogens of P. notoginseng and six strains exhibited significant inhibitory activity. It was noteworthy that endophytic Rhexocercosporidium sp. and F. solani strongly inhibited pathogenic F. solani and other fungal phytopathogens of P. notoginseng.

have provided the impetus for several investigations on endophytic fungi.

Sophora tonkinensis Gapnep, is a well-known medicinal plant of
China that grows in an area of karst topography near the Tropic of Cancer, is mainly distributed in Guangxi province, and oddly is found in Guangdong province as well as Yunnan province (Wang, Xie, Fan, & Liu, 2011). The chemical constituents , including primarily flavonoids, alkaloids, polysaccharides, and saponins, have been isolated from the root of S. tonkinensis, and have pharmacological activities (Commission, 2015b; such as antitumor activity, antimicrobial activity, antiinflammation, antiarrhythmia, antihypertension, hepatoprotection, and immune stimulation. The crude extracts from the root of S. tonkinensis have been effectively applied to control symptoms on Panax notoginseng, a famous traditional Chinese herb with a long history in China as a valuable cardiovascular remedy (Commission, 2015a;Li, Xie, Fan, & Wang, 2011). These symptoms, namely black spots caused by Alternaria panax Whetz (Wei & Chen, 1992), anthracnose by Colletotrichum gloeosporioides (Wei, Chen, & Wu, 1989), and root rot by Fusarium solani (Miao et al., 2006), seriously affect the quality and yield of P. notoginseng in the geo-authenticproducing areas. During the course of plant-endophyte coevolution, it might be possible for endophytes to assist the plant in chemical defense by producing bioactive secondary metabolites according to the theories of "mosaic effect" and "acquired immune systems" (Carroll, 1991;Rodriguez, Redman, & Henson, 2004). Hence, several members of endophytic fungi harbored in the root of S. tonkinensis may have an antagonistic activity against three fungal phytopathogens of P. notoginseng. Therefore, we selected this plant to isolate endophytic fungi.
Most of the studies on endophytic fungi have been carried out in tropical, subtropical, temperate, and boreal regions, but there are only a few studies that have been carried out near the Tropic of Cancer, and overall, no major studies exist on endophytic fungi harbored in medicinal plant from karst topography near the Tropic of Cancer in Guangxi province of China.
The aim of this study was to isolate and identify endophytic fungi harbored in the root of S. tonkinensis, characterize the diversity of endophytic fungal communities, investigate the influence of the tissue type and geographic locality on the colonization, species diversity, and species composition of endophytic fungal communities, and further screen them for potential as biocontrol agents against three phytopathogens of P. notoginseng cultivated in China. The findings will not only enrich the knowledge of endophytic fungi from S. tonkinensis but also benefit the development of organic cultivation techniques for P. notoginseng in China. To the best of our knowledge, this report is the first to describe the diversity, phylogeny, and communities of endophytic fungi harbored in the root of S. tonkinensis, and assess their potential as biocontrol agents against phytopathogens of P. notoginseng.

| Sample collection from selected sites
In 2014, healthy plants of S. tonkinensis were collected in three periods from three different localities of traditional geo-authentic-producing areas  in Guangxi province of south China: Tiandeng county (T), where S. tonkinensis grows as a natural part of an intact shrub forest; Jingxi county (J), where S. tonkinensis grows in the rock crack in limestone mountainous areas; and Guangxi university (G), where S. tonkinensis is cultivated in a medicinal herb garden. Details of three sampling localities and dates were given in Table 1. These plants were carefully up-rooted with the help of a spade, placed in jute bags, labeled, immediately transported to the laboratory, and processed within 24 hr of collection. Import of the plant material from Tiandeng county and Jingxi county was allowed according to the permission of the Department of Forestry of Guangxi province, Guangxi, China, and that from Guangxi university was allowed according to the permission of the College of Agriculture, Guangxi University, Guangxi, China.

| Isolation of endophytic fungi
Root samples (diameter, 1-2 cm) were excised from the plant and cut into segments (length, 5-7 cm). For the surface sterilization and isolation of endophytic fungi, we established the optimum procedures according to previously described methods (Kusari et al., 2013;Tejesvi et al., 2011). The root segments were thoroughly washed in running tap water for 30 min and rinsed with double-distilled water for 10 min. Next, the samples were sterilized with 75% ethanol for 1 min, sodium hypochlorite containing 1% available chlorine for 2 min, and 75% ethanol for 30 s. Finally, these surface-sterilized samples were rinsed three times with sterile, double-distilled water to remove excess surface sterilants, blotted on a sterile filter paper, and dried under aseptic conditions. To ensure the isolation of endophytic fungi, the epidermis and ends of each root segment were removed.
The xylem (X) and phloem (P) were separated from the remaining part of each root segment (R) and transversely cut into 1-cm-long pieces, respectively, which were individually placed in Petri dishes (9 cm in diameter) containing potato dextrose agar (PDA) with chloramphenicol to eliminate any bacterial growth. The dissection of the roots was showed in Figure 1. The Petri dishes with three pieces per dish were incubated at 28°C and checked daily for fungal growth for up to 6 weeks. Each colony which emerged from the segments was transferred to an antibiotic-free PDA medium. Purification was carried out by cutting a small piece of media with mycelia at the edge of a colony and then transplanted on to new medium plates.

| Storage of the purified endophytic fungi
Every purified endophytic fungus sample received a specific code number according to its origin (e.g., TRXY-1 or TRXY-2, from the xylem of the root collected from Tiandeng county, and TRPH-1 TRPH-2, from the phloem of the root collected from Tiandeng county). All endophytic fungi were deposited at the College of Agriculture, Guangxi University, Guangxi, China. For short-term storage, they were cultured on PDA at 28°C for 3-10 days and maintained at 4°C (up to 3 months); for long-term storage, they were preserved with spores or mycelia in 25% (v/v) glycerol at −80°C.

| Fungal identification
The taxonomic identification of endophytic fungi isolated was based on morphology and molecular phylogenetics including phylogenetic position and similarity to reference sequences of the GenBank ( level when the similarity between a query sequence and a phylogenetically related reference sequence was higher than 95%, and the sequence was considered to be conspecific when that was higher than those within same genus. The strain with an ITS sequence showing a divergence greater than 5% with any entry at GenBank was considered as unidentified. These thresholds have been previously employed in other endophyte-related studies to identify fungal taxa (Gonzalez & Luisa, 2011;Kusari et al., 2013;Sanchez, Bills, & Zabalgogeazcoa, 2008).

| Fungal culture and extraction
Every selected endophytic fungus was cultured on a Petri dish containing PDA at 28°C for 5-10 days. The culture materials from each Petri dish were cut into small pieces and transferred to a 2-L Erlenmeyer flask containing sterile solid medium, which included 400 g of potato, 20 g of dextrose, and 20 g of sucrose at 28°C for 30-40 days. The culture materials from the Erlenmeyer flask were successively extracted with methanol to yield crude extracts.

| Preparation of crude extracts from the root of S. tonkinensis
The dried root of S. tonkinensis was pulverized and soaked in 1,000 ml of methanol for 2 weeks at room temperature. The organic solvent was filtered through a filter paper and evaporated to dryness under vacuum to afford crude extracts.

| In vitro antagonistic assays of endophytes against fungal phytopathogens
In order to screen antagonistic fungi against phytopathogens of P. notoginseng, we took three fungal phytopathogens including . Every selected strain with antagonistic activity was cultured and extracted to yield crude extracts as mentioned above. All crude extracts were dissolved in 1% (v/v) dimethyl sulfoxide (DMSO).
Appropriate volumes of the solutions of each crude extract were incorporated into the PDA medium and poured into the Petri dishes to obtain final concentrations ranging from 2 to 8 mg/ml according to the concentration of carbendazim wettable powders used in field, and each concentration was tested in triplicate. The positive control with carbendazim wettable powders was treated in the same way.
The growth control without drug was maintained with 1% DMSO mixed with PDA medium. Every mycelial plug (6 mm diameter) from each 3-day-old fungal phytopathogen was, respectively, placed at the center of the Petri dishes. The cultures were incubated at 28°C, and the colony growth diameter of each fungal phytopathogen was measured when the fungal growth in the growth control had completely covered the Petri dishes. The radial growth of each fungal phytopathogen in the treatment measured by removing 6 mm from the growth diameter was recorded as D 1, and that in the growth control was recorded as D 2 . The inhibition percentage of mycelial growth was calculated with the help of the modified formula as follows:

| Statistical analysis
The colonization frequency (CF) was expressed in percentages and calculated as the number of segments colonized by a single endophyte divided by the total number of segments examined ×100 (Mishra et al., 2012). The percentage of species composition (S i ) was calculated as the number of taxa that belong to a specific phylum, class, or order divided by the total number of taxa in the sample (Botella and Diez, 2011). The relative species frequency (P i ) was calculated as the number of isolates that belong to taxon i divided by the total number of isolates in the sample (Kusari et al., 2013).    Table 3). The species richness consisted of frequent species (29 species, 61.7%) and rare species (18

| Phylogeny and fungal diversity analysis
T A B L E 2 Summary of the endophytic fungi isolated from the root of S. tonkinensis with their respective strain numbers, GenBank accession numbers, and closest affiliations of the representative isolates in the GenBank according to ITS analysis

| The influences of geographic locality and tissue type on endophytic fungal communities
The influences of geographic locality and tissue type on the coloni-  Table 4.
These results indicated that the colonization and species diversity of these communities were significantly influenced by geographic locality but not tissue type.
F I G U R E 2 Phylogenetic tree of identifiable endophytes harbored in the root of S. tonkinensis based on ITS sequences using the software of MEGA version 6.0 T A B L E 3 Summary of the endophytic fungi isolated from the xylem and phloem of the root of S. tonkinensis from three sampling localities with their taxa and the number of isolates from each taxon  Sorenson's and Jaccard's similarity indices of the endophytic fungal communities between two tissues or two localities were rather low, as exhibited in  b The number of isolates obtained from the dominant species is followed by the relative species frequency (Pi>1/s, 1/s = 0.0213) in brackets.

T A B L E 3 (Continued)
from the xylem of that, providing clear evidence for tissue specificity (Table 3). These results revealed that geographic locality and tissue type had great effects on the species composition of these communities.

| In vitro antagonistic assays of endophytic fungi against fungal phytopathogens of P. notoginseng
All respective strains from 47 taxa isolated from the root of S. tonkinensis were screened for antagonistic activity against three fungal phytopathogens of P. notoginseng using the coculture method.
Six endophytic strains with more than 60% inhibition against all of three fungal phytopathogens were selected to further test the antifungal activity of their crude extracts using the mycelial growth method (Table 7, Figure 3, Figure 4). In the test plates, mycelial growth inhibition, including no growth, only growth on the mycelial plug and growth on the medium, was observed. The colonial T A B L E 4 The influences of geographic locality on CF, Shannon-Wiener index, and Simpson's diversity index T A B L E 5 Sorenson's and Jaccard's similarity indices of endophytic fungi communities between two tissues or two localities F I G U R E 3 The coculture interactions between endophytic fungi strains and three fungal phytopathogens on PDA morphology was changed in the plates with crude extracts from different strains. Crude extracts from strains TRPH-73, TRPH-105, and TRPH-87 exhibited more than 90% inhibition against all of three fungal phytopathogens even at the low concentration of 2 mg/ml.
The most susceptible phytopathogen was C. gloeosporioides whose mycelial growth was completely inhibited by the crude extracts of strains TRPH-73, TRPH-87, and TRPH-105, even at the low concentration of 2 mg/ml, and by the crude extracts of strains TRXY-34-1, TRXY-69, and TRXY-63 at the concentration of 8 mg/ml. The six strains showed significant antifungal activity against three fungal phytopathogens, based on that of carbendazim wettable powders, which were widely applied to control fungal phytopathogens of P. notoginseng using the concentration range of 2-8 mg/ml.
Particularly, the antifungal activity of the crude extracts from strains TRPH-73, TRPH-105, TRPH-87, and TRXY-34-1 was more than that of carbendazim wettable powders against A. panax and almost equal to that of carbendazim wettable powders against C. gloeosporioides.
The inhibitory activity of the crude extracts from strains TRPH-73 and TRPH-105 was also equal to that of carbendazim wettable powders against F. solani.

| DISCUSSION
Morphological characteristics and ITS sequences analysis have been employed for the identification of endophytic fungi in this work.
However, this work still failed to identify some taxa based on >5% divergences of ITS sequences and no spores. These unidentifiable taxa require the analysis of other gene markers to provide better taxonomic resolution. Many other markers which have been used for fungal identification are 28S rDNA gene, cytochrome c oxidase subunit I, and beta-tubulin 2 gene (Liu, Xu, & Guo, 2007;Rivera-Orduna et al., 2011;Robideau et al., 2011). Some endophytic fungi belonging to unidentifiable taxa may represent novel species. The taxonomic novelty of endophytic fungi may also correspond to chemical novelty of their secondary metabolites (Kumar et al., 2013). Furthermore, these endophytic fungi have not been explored for their natural products. Thus, these organisms will be given priority to isolate and characterize novel molecules from their secondary metabolites.
The total CF of endophytic fungi with 17% in the root of S. tonkinensis was much lower than that with the range of 33% to 53% in the roots of other medicinal plants (Jin et al., 2013;Kharwar, Verma, Strobel, & Ezra, 2008;Mishra et al., 2016). Two reasons may account for the high CF in the roots of medicinal plants in previous reports.
One likely reason is that the soil fungi and rhizospheric fungi are so prevalent and diversified to easily establish an endophytic relationship with the roots (Ghimire, Charlton, Bell, Krishnamurthy, & Craven, 2011). The other reason is that roots as important sources of easily accessible substrate may provide a relatively stable environment favoring many fungal survival and coexistence (Angelini et al., 2012;Garbeva, Veen, & Elsas, 2004). However, the low CF in the root of S. tonkinensis may be associated with the presence of antimicrobial chemicals as mentioned above that may have suppressed the growth of some endophytic fungi.
In the root of S. tonkinensis, the majority of endophyte taxa were Ascomycota, a finding that was in agreement with that of previous reports (Qadri et al., 2014;Rivera-Orduna et al., 2011;Vieira et al., 2012). The low proportion of the phylum Basidiomycota and Zygomycota were consistent with that reported in other studies (Gonzalez & Luisa, 2011;Sánchez, Bills, Acuña, & Zabalgogeazcoa, 2010;Tejesvi et al., 2011). However, recent papers have suggested that Basidiomycota constitute an important component of certain endophytic communities (Pinruan et al., 2010;Rungjindamai, Pinruan, Hattori, & Choeyklin, 2008). The abundant classes Sordariomycetes, Eurotiomycetes, and Dothideomycetes, were similar to that of endophytic fungal community associated with ferns in Costa Rica (Del Olmo-Ruiz & Arnold, 2014) and Huperzia serrata in China (Xiong et al., 2015). The abundant orders Eurotiales, Hypocreales, and Pleosporales, were in line with that of endophytic fungal community in Ficus tree (Solis, Edison Dela Cruz, Schnittler, & Unterseher, 2016), Annona squamosa (Lin et al., 2010), and Stellera chamaejasme L. (Jin et al., 2013), respectively. The species F. solani, F. oxysporum, C. perangustum, Cladosporium sp., T. pinophilus, P. coffeae, and M. verrucaria were dominant in this work, possibly due to the high spore production of these fungi and their cosmopolitan nature, which statistically increases their chance to become established as endophytes, as indicated in previous studies (Mishra et al., 2012;Raviraja, 2005;Schulthess & Faeth, 1998). In addition, based on the "balanced antagonisms" hypothesis (Schulz, Haas, Junker, Andree, & Schobert, 2015;Schulz, Rommert, Dammann, Aust, & Strack, 1999), they as dominant species might not only secrete toxic metabolites to inhibit microbial competitors (Breinholt et al., 1997;Lee & Lee, 2012;Zhai et al., 2015) but also possess the ability to resist the attack of the host alkaloids T A B L E 7 Percent of inhibitory activity on mycelial growth of fungal phytopathogens produced by the crude extracts of six endophytic strains from the root of S. tonkinensis on PDA However, 38.3% rare species in this work imply that some members of these fungi are host-specific and occupy specific ecological niche in this community (Yuan et al., 2011).
Geographic locality significantly affected the colonization, species diversity, and species composition of endophytic fungal communities harboring the root of S. tonkinensis, possibly because ecological environment primarily including temperature, rainfall, altitude, and geographic coordinates are diverse in three geographic localities as mentioned above. In different ecosystems, the fungi are subjected to different selection pressures (Goere & Bucak, 2007;Petrini, Sieber, Toti, & Viret, 1993). Furthermore, in order to adapt to the ecological environment, a plant may produce several toxic metabolites toward which biotransformation abilities of many endophytic fungi to a certain extent decide the colonization range of their hosts (Saunders & Kohn, 2009;Wang & Dai, 2011;Zikmundova, Drandarov, Bigler, Hesse, & Werner, 2002). These lead to the establishment of a quite specific endophytic fungal community at each geographic locality, as reported previously (Goere & Bucak, 2007;Hoffman & Arnold, 2008).
However, there were few works about endophytic fungi communities in the xylem and the phloem of the root tissue in previous studies.
In this work, results showed that the xylem and phloem of the root influenced the species composition of the endophytic fungi communities but not the colonization and species diversity of that. The striking difference in the species composition of fungal communities between the xylem and phloem may be due to tissue specificity as reported in other tissues (Mishra et al., 2012;Raviraja, 2005). These tissues may represent two distinct microenvironments including toxic metabolites, oxygen, nutrition, anatomy, and endophytic bacteria consequently shaping their difference in species composition (Huang, Cai, Hyde, Corke, & Sun, 2008;Qadri et al., 2014;Schulz et al., 2015). Further work is needed to investigate the reasons for similarity in the colonization and species diversity of fungal communities between the xylem and the phloem.
This work also demonstrated that geographic locality affected the endophytic fungi communities harbored in the root of S. tonkinensis more strongly than the tissue type, a finding that was not in agreement with a previous report (Mishra et al., 2012).
Some endophytic fungi with strongly antimicrobial activities as biological agents are of increasing public interest (Bailey et al., 2008;Rubini et al., 2005). Because the crude extracts from the roots of S. tonkinensis were effectively used to control symptoms on P. notoginseng cultivated in Guangxi province, we attempted to screen antagonistic fungi from endophytic fungi isolated from them against three fungal phytopathogens of P. notoginseng.
The results that 24 strains showed 50% or more inhibition against three fungal phytopathogens of P. notoginseng, suggested that it is possible to effectively screen potential biocontrol agents against fungal phytopathogens of P. notoginseng from the root of S. tonkinensis.
The antifungal activity of the crude extracts from six strains was more than or almost equal to that of carbendazim wettable powders against three fungal phytopathogens of P. notoginseng in vitro, therefore, future investigations will be conducted to study their potential as biocontrol agents on an agronomic scale.
It was noteworthy that there was a few works in the antagonistic activity and compounds of Rhexocercosporidium species in previous study. Therefore, the strains Rhexocercosporidium sp TRPH-87 and Rhexocercosporidium sp TRPH-105 probably produce new natural compounds with antifungal activity, and the isolation and characterization of the active substance from them are in progress.
The result that the endophytic strain F. solani TRX-34-1 strongly inhibited pathogenic F. solani compelled reconsidering whether F. solani TRX-34-1 was capable of producing associated plant secondary metabolites as a result of horizontal gene transfer (Gogarten & Townsend, 2005) from host plant to endophytic fungus during the course of evolution. In this work, F. solani TRX-34-1 is likely a nonpathogenic strain based on its antagonistic activity against three fungal phytopathogens of P. notoginseng. Thus, key research on the mode of action of F. solani TRX-34-1 against phytopathogens of P. notoginseng by several methods is progress.
In conclusion, endophytic fungal communities harbored in the roots of S. tonkinensis with high diversity were affected by geographic locality more strongly than tissue type, and they have great promise not only as potential sources of bioactive secondary metabolites, but also as biocontrol agents against fungal phytopathogens of P. notoginseng and possibly other pathogens.