Endophytic Bacillus velezensis strain B‐36 is a potential biocontrol agent against lotus rot caused by Fusarium oxysporum

Abstract Aim The aim of this study was to screen potential lotus plant endophytic bacterial isolate for effective inhibition against lotus rot causing fungal pathogen Fusarium oxysporum. Methods and Results In this study, endophytic bacteria were isolated from lotus tissues and tested for antagonistic activities against the pathogenic fungus F. oxysporum. Among the putative endophytic Bacillus strains identified, suspensions of the strain B‐36 showed the highest inhibition rate against F. oxysporum growth. Pot assays indicated that B‐36 was effective in controlling F. oxysporum‐inducing lotus rot. However, the control efficiency varied with the inoculation method and concentration, where injection of 800 μl B‐36 suspension per plant (2 × 108 CFU per ml) into stems showed the highest control efficiencies of 77·1 and 60·0% for pre‐inoculation and post‐inoculation. In addition, the colonizing population levels (CPLs) of B‐36 on lotus also varied with the inoculation method and concentration, with the highest CPLs, that is, 3·05 and 2·83 log(CFU per gram), being observed on lotus leaves and stems respectively for stem injection of 200 μl per plant. Moreover B‐36 showed no noticeable effects on lotus seed germination rate or seedling growth. Finally, B‐36 was characterized as Bacillus velezensis based on its morphology, Gram‐positive characteristics, as well as its 16S rDNA and gyrB sequences. Conclusion The isolate B‐36 can be applied as a biocontrol agent against F. oxysporum‐inducing lotus rot. Significance of Impact of the Study The soil‐borne fungus F. oxysporum causes lotus rot and severe yield loss, and currently available control methods are very limited. Here we identify a new promising biocontrol agent against lotus rot caused by F. oxysporum.


Introduction
Lotus (Nelumbo nucifera) is an aquatic plant widely cropped worldwide owing to its applications in food and drugs (Zhang et al. 2015;Sharma et al. 2017). The soilborne fungus Fusarium oxysporum is distributed worldwide with a great diversity of host-specific strains, causing considerable damage to various crops (Gordon 2017). Lotus rot induced by F. oxysporum can result in the wilt of lotus plants causing significant yield loss (Rana et al. 2017) and postharvest damage to lotus roots (Tang et al. 2017).
In recent years, researchers have been paying increasing attention to fungi and bacteria, especially plant endophytes, as biocontrol agents of plant pests (Miliute et al. 2015;Zheng et al. 2016). Currently, several biological control products are available for F. oxysporum infecting ornamentals and crops (Lecomte et al. 2016;Raza et al. 2017), but few are available for controlling F. oxysporum infections in ornamentals. Additionally, a number of endophytic fungi and bacteria have been isolated from lotus plants, including the genera Colletotrichum, Diaporthe, Rhizobium and Pseudomonas (Pawlik et al. 2017;Chen and Kirschner, 2018). These endophytic bacteria have also been shown to possess plant growth-promoting traits (Pawlik et al. 2017). For instance, Bacillus sp. are considered bacteria with the widest potential application in plant pest control (Wu et al. 2015), and some of them are endophytic bacteria with plant growth-promoting activity (El-Deeb et al. 2013;Radhakrishnan et al. 2017;Zhang et al. 2019). The production of antifungal and antibacterial compounds has been reported as the main mechanism by which Bacillus sp. control plant pests (Santoyo et al. 2012;Mora et al. 2015).
Bacillus amyloliquefaciens that has been recently classified as being B. velezensis is reported to have antagonistic activities against fungal and bacterial pests on plants (Ji et al. 2013;Dunlap et al. 2016;Masum et al. 2018;Nikolic et al. 2019). Its antagonistic activities were identified by genome analysis of B. velezensis FZB42 that was previously named as B. amyloliquefaciens FZB42, and four gene clusters were found to be involved in the synthesis of antibacterial compounds (Chen et al. 2009;Fan et al. 2018). However, the control efficiency of B. velezensis against lotus rot caused by F. oxysporum remains unknown.
The main objective of our work was to evaluate the potential of the endophytic Bacillus strain B-36 isolated from lotus plants for the biocontrol of F. oxysporum-inducing lotus rot through analysing its antagonistic activity against F. oxysporum-inducing lotus rot and its colonizing ability on lotus plants.

Isolation of lotus endophytic Bacillus strains
Endophytic Bacillus strains were isolated from healthy lotus plants in fields around Wuhan, China, according to the protocols described by Lai et al. (2012) with some modifications. Briefly, lotus tissues, including leaves, lotus pods, stems, rhizomes and seeds, were washed with tap water and incubated in an ultrasonic bath for 10 min, surface sterilized with 75% ethanol and 2% NaClO 3 for 5 min each, followed by washing three times with sterilized water. The tissues were then cut into small pieces and treated with 80°C sterilized water for 10 min to enrich for Bacillus sp., followed by culture on nutrient agar (NA) medium at 28°C for 16 h to check surface sterilization. Finally, the surface sterilized lotus tissues were ground in 5 ml of sterilized water and sterilized quartz sand using a mortar and pestle. After incubation on NA medium at 28°C for 24 h, individual bacterial colonies were isolated and purified.
Screening of the putative Bacillus strains with antifungal activities against F. oxysporum The dual culture technique was applied to detect the antagonistic activities of the putative Bacillus strains against F. oxysporum that we had previously isolated from lotus with rot disease. A 3-mm-diameter disc from a 7-day-old mycelial culture of F. oxysporum was placed in the centre of fresh potato dextrose agar (PDA) plates (90 mm). A fresh bacterium colony was then inoculated longitudinally on the right and left sides of the fungal disc at 2 cm distance using a sterile inoculating needle. The PDA plate inoculated only with F. oxysporum was used as a negative control. After incubation at 25°C for 3 days, the antagonistic effect was evaluated by measuring the inhibition zones and colony diameters. The percentage of growth inhibition was calculated by the following equation: n = [(AÀB)/A] 9 100, where A is the colony area of uninhibited fungi and B is the colony area of treated fungi (Etebarian et al. 2005). The values were recorded as the means of four replicates, and each experiment was repeated twice.
The antifungal activity of sterilized broth filtrates from B-36 against F. oxysporum was assessed as follows. Specifically, B-36 was cultured in 4 ml nutrient broth (NB) medium at 28°C and 280 rev min À1 for 24 h. One millilitre of culture was transferred to 100 ml of potato dextrose broth (PDB) medium and further cultured under the same conditions for 72 h. The cell-free supernatant was collected by centrifugation at 6000 g for 10 min, followed by filtration through 0Á45 lm cellulose nitrate filters. Different broth filtrate concentrations (V B-36 /V PDA : 1, 5, 10 and 20%) were added into the PDA medium, and F. oxysporum was placed in the centre of fresh PDA plates as described above and cultured at 25°C. Meanwhile, sterilized water was used as a negative control. After 7 days of incubation, the colony diameters and percentage of growth inhibition of F. oxysporum were estimated as described above. The values were recorded as the means of three replicates, and each experiment was repeated twice.
The antifungal effect of B-36 volatiles against F. oxysporum was analysed as described by Fernando et al. (2005) with minor modifications. Briefly, after incubation in NB medium for 24 h, 200 ll of B-36 suspension was plated on one half of a divided plate containing PDA medium, and F. oxysporum was inoculated on the other half, followed by resealing the plates and culturing at 25°C. Meanwhile, plates without B-36 volatiles were used as a negative control. After 7 days of incubation, the growth inhibition percentage of F. oxysporum was estimated as described above, and its mycelial morphology was observed through an optical microscope. There were three replicates for each treatment, and each experiment was repeated twice.

Assessment of biocontrol activity of B-36 against lotus rot caused by F. oxysporum
Similar-sized healthy rhizomes of lotus (N. nucifera) cultivar Taikong 6 were transferred into plastic pots (30 9 15 cm, diameter 9 height) filled with sterilized soil (one rhizome per pot). The lotus plants were cultured at approximately 25°C in a greenhouse and plants showing the appearance of the first leaf were used in the following experiments. Briefly, B-36 and the pathogen F. oxysporum were cultured on PDB and PDA medium respectively. B-36 cultures and F. oxysporum spores were collected by centrifugation at 8000 g for 5 min, followed by three washes with sterilized water. Finally, 2 9 10 8 CFU per ml B-36 and 1 9 10 4 spores per ml F. oxysporum spore suspension were prepared as described by Lin et al. (2018).
To analyse control efficiency with different inoculation methods and concentrations, B-36 was inoculated on lotus plants by foliar spraying or injecting 200, 400 and 800 ll B-36 suspension per plant into stems. F. oxysporum was inoculated by injecting 100 ll spore suspension per plant into stems. B-36 control efficiencies on F. oxysporum-inducing lotus rot were analysed in terms of preand post-inoculation with different inoculation methods and concentrations. In this study, B-36 pre-inoculation refers to the inoculation of B-36 on lotus plants 15 days before the inoculation of F. oxysporum, whereas B-36 post-inoculation refers to the inoculation of B-36 7 days post-inoculation (dpi) of F. oxysporum on lotus plants. Corresponding positive and negative controls were set up by spraying the fungicide carbendazim on the leaves and by injecting 800 ll of sterilized water into stems per plant respectively. There were six replicates and two biological replicates for each treatment.
The disease indices of each plant were measured at 30, 60 and 90 dpi of the second microbe on lotus plants. Disease severity was classified as 0, 1, 3, 5, 7 and 9 depending on the diseased leaf area (0, <10, 10-25, 25-50, 50-75 and >75% of the total respectively). The disease index (DI) and control effects (Ce) for the pathogen were calculated by the following equation: Analysis of endophytic colonization of B-36 in lotus B-36 was acclimated to the different concentrations of the antibiotic rifampicin (Sigma, R3501, St. Louis, MO) by continuous culture at 28°C for 24 h on NA medium, that is, 1, 5, 25, 50, 100, 200 and 400 lg ml À1 rifampicin. The acclimated strain B-36 R that could grow on NA medium containing 400 lg ml À1 rifampicin was collected for further culture on PDB medium containing 400 lg ml À1 rifampicin at 28°C and 160 rev min À1 for 16 h. Finally, a 2 9 10 8 CFU per ml B-36 R suspension was obtained as described above.
To test B-36 R colonization efficiencies on lotus tissues, lotus plants were inoculated with B-36 R suspension by three different methods: (i) soaking seeds for 10 min, (ii) foliar spraying of B-36 R suspension on lotus plants with one leaf, and (iii) injecting 100 and 200 ll B-36 R suspension per plant into stems of the lotus plants with one leaf. For each experiment, sterilized water was used as a negative control. After co-cultivation at 25°C in a greenhouse for different periods of time (1, 7, 15, 30 and 45 days), the colonization efficiencies of B-36 R on lotus leaves and stems were analyzed for each time point. B-36 R colonies in inoculated lotus leaves and stems were isolated as follows. Briefly, approximately 3 g of fresh leaves and stems were weighed and ground separately in 5 ml sterilized water. Thereafter, 100 ll of each tissue suspension was diluted, plated on NA medium containing 400 ll ml À1 rifampicin, and cultured at 28°C for 24 h. Finally, the number of B-36 R colonies on each plate was counted, and B-36 R population densities were evaluated in lotus stems and leaves. There were three technical replicates and two biological replicates for each treatment.

Evaluation of B-36 on lotus seed germination and growth
Similar-sized healthy lotus seeds were washed with tap water, followed by two washes with sterilized water and soaking the seeds in B-36 suspension (2 9 10 8 CFU per ml) for 1 h, using seeds soaked in sterilized water as a negative control. All the seeds were sown in sterilized sandy soils and cultured at 25°C in a 16 h light/8 h dark cycle. After 15 days of co-cultivation, the seed germination rate, average height and above-ground biomass were evaluated. There were seven individual plants and two biological replicates for each test group.

Statistical analysis
All data were expressed as mean AE SE. The experimental data were analysed using standard analysis of variance (ANOVA) followed by Duncan's multi-range test. Values with P < 0Á05 were considered statistically significant. ANOVA was conducted using SPSS software, ver. 30.

Results
Endophytic bacterial strain B-36 isolated from lotus showed high antagonistic activity against F. oxysporum Of all the endophytic bacterial strains isolated from lotus, 10 strains exhibited antagonistic activities against F. oxysporum (Table 1). Among these 10 strains, strain B-36 showed the highest antifungal activity against F. oxysporum, with a mycelial growth inhibition rate of 74Á1% (Table 1). In addition, B-36 broth filtrates displayed a strong antifungal activity against F. oxysporum in a dosedependent manner, with mycelial growth inhibition rates of 42Á7 and 90Á5% for 1 and 20% B-36 broth filtrates respectively (Fig. 1). However, for B-36 volatiles, no significant antifungal activities against F. oxysporum were observed (data not shown).

B-36 has high biocontrol efficiency against lotus rot caused by F. oxysporum
Pot experiments were designed to evaluate the biocontrol efficiency of strain B-36 against lotus rot caused by F. oxysporum. At 90 dpi, pre-inoculation B-36 biocontrol efficiencies against lotus rot were 71Á8 and 77Á1% after injecting 400 and 800 ll B-36 per plant into stems respectively compared to no significant biocontrol effect by foliar spraying of B-36 (Fig. 2). In addition, significant post-inoculation biocontrol efficiencies of B-36 against lotus rot were also recorded for stem injection of 400 and 800 ll per plant at 60 dpi and 800 ll per plant at 90 dpi,   that is, 50Á7, 58Á8 and 60Á0% respectively (Fig. 3). No apparent biocontrol against lotus rot was observed for any inoculation method at 30 dpi for B-36 pre-inoculation and post-inoculation (Figs 2 and 3).

B-36 can successfully colonize lotus
To evaluate B-36 colonization in lotus, the rifampicin-resistant strain B-36 R was developed. Results indicated that B-36 R could grow well in NA medium containing 400 lg ml À1 rifampicin, which showed that B-36 R obtained high resistance to rifampicin (Fig. 4a).
Therefore, the strain B-36 R was used to study colonization dynamics in lotus plants. Results showed that B-36 R could successfully colonize lotus leaves and stems via three different inoculation methods, that is, seed soaking, foliar spraying and stem injection (Fig. 4b-d). For each inoculation method, B-36 R CPL was significantly decreased at 30 dpi; however, it was higher in stems than in leaves except for foliar spraying (Fig. 4b-d). In addition, among three inoculation methods, the highest B-36 R CPL was observed in stems and leaves at approximately 3Á05 and 2Á83 log(CFU per gram) respectively at 30 dpi of treatment with 200 ll suspension per plant. In contrast, the lowest CPLs in the stems and leaves were around 2Á04 and 1Á99 log(CFU per gram) after soaking seeds with B-36 R (Fig. 4b-d). No B-36 R colonies were isolated in any of the negative controls.
B-36 showed no apparent effects on lotus seed germination and seedling growth Compared to the control seeds, soaking seeds with B-36 suspension showed no significant change (P > 0Á05) in seed germination rate (approximately 86Á4%) (Fig. S1). It also had no apparent effects on the height and aboveground biomass of 15-day-old seedlings, which were approximately 18Á61 cm and 4Á73 g respectively (Fig. S1).

B-36 was characterized as Bacillus velezensis
The colony of strain B-36 on NA medium was opaque, white and dry (Fig. S2a). In addition, B-36 was characterized as a Gram-positive bacterium by Gram staining with the typical characteristics of the genus Bacillus (Fig. S2b). This was further confirmed by 16S rDNA (GenBank accession no. MK182731) and gyrB-encoding gene (Gen-Bank accession no. MK185102) sequence analyses, which were 99Á9 and 99Á4% identical respectively to the known sequences of B. amyloliquefaciens (AB244285 in RDP database and JN412504 in GenBank database) (data not shown). Based on these data, B-36 belongs to B. amyloliquefaciens that has recently been classified as B. velezensis.

Discussion
Fusarium oxysporum is a soil-borne pathogen that causes severe damage in a number of crops (Dean et al. 2012). Recent data indicate an expanded host range of F. oxysporum (Matic et al. 2018), implying the urgent need of effective management strategies. Several biological agents against plant wilt diseases caused by F. oxysporum have been identified, including plant endophytic fungi and bacteria (Lecomte et al. 2016;Raza et al. 2017;Hajji-Hedfi et al. 2018).
Among bacteria, Bacillus sp. are believed to have the highest potential for pest biocontrol (Shafi et al. 2017). As a Bacillus sp. strain, B. velezensis has been reported to be effective in controlling F. oxysporum-caused wilt disease in tomato (Elanchezhiyan et al. 2018), indicating its potential as a pest biocontrol agent on other crops. In this study, an endophytic bacterium, B. velezensis strain B-36, was isolated from lotus plants, which showed high antifungal activity against F. oxysporum. B-36 appears to inhibit F. oxysporum growth mainly through the production of antifungal agents, as its broth filtrates showed relatively high antagonistic activity against F. oxysporum. Previous studies have also identified antifungal compounds synthesized by B. velezensis, such as peptides and lipopeptides (Romano et al. 2011;Alvarez et al. 2012;Kim et al. 2015;Luna-Bulbarela et al. 2018).
Pot assays showed that B-36 had significant control efficiency against lotus rot caused by F. oxysporum. The control efficiency was closely related to B-36 concentration and inoculation method. Injecting over 400 ll B-36 per plant into the stems could result in a significant control efficiency against lotus rot, probably owing to a higher B-36 CPL in lotus plants. In contrast, a lower B-36 CPL from foliar spraying resulted in no obvious control efficiency against lotus rot. Moreover B-36 controls lotus rot in a dose-dependent manner, implying that its significant control efficiency against lotus rot could be mainly achieved by antifungal compounds rather than by priming lotus defenses, despite several reports on the induction of the plant defense response by B. velezensis (Yamamoto et al. 2015a;Yamamoto et al. 2015b). Meanwhile, the detection of remarkable B-36 control efficiencies against lotus rot at later stages post-B-36 inoculation rather than at the early stage might be caused by persistent inhibition of F. oxysporum growth by B-36, as well as higher disease tolerance of lotus plants at late developmental stages. It is worth noting that injecting 800 ll B-36 per plant into stems showed significantly higher control efficiency against lotus rot than the fungicide carbendazim, indicating the promising potential of B-36 as a biological agent against F. oxysporum on lotus.
In conclusion, B. velezensis strain B-36 could serve as a potential biocontrol agent against F. oxysporum in lotus plants due to its significant biocontrol properties and colonizing ability on lotus plants, as well as the lack of apparent negative effects on lotus seed germination and seedling growth.