G. Biloba leaf samples were sourced from Guizhou University of Traditional Chinese Medicine in Guiyang, China during the months of September and October 2021. Several fungi were successfully isolated from these leaves. Antagonistic bacteria were collected from rhizosphere soil, with samples taken from a depth of 3–5 cm below the surface after removing the topsoil layer. The culture media utilized in this study included Potato Dextrose Agar (PDA) and Luria-Bertani (LB) medium. All additional biochemical reagents required for direct PCR were procured from Tsingke Biotechnology Co., Ltd.

The collected leaves underwent a series of carefully executed steps to ensure the removal of external contaminants and the isolation of endophytic bacteria. Initially, the leaves were immersed in 75% alcohol for 1 min to effectively sterilize their surfaces. Subsequently, a 2% sodium hypochlorite (NaClO) solution was applied for 3 min to further eliminate any remaining impurities. After this, the leaves were rinsed with sterile distilled water for 1 min to remove residual chemicals. The processed leaves were then dissected into 0.5 cm² fragments using sterilized scissors, maintaining aseptic conditions throughout. These fragments, now devoid of external contaminants, were inoculated onto PDA and incubated at a temperature of 28 C for 2 days.

Single colony cultures that developed on the PDA medium were singled out for further investigation. Each isolated colony was assigned a unique identifier to distinguish and track individual colonies. This process was repeated through several cycles, guaranteeing the successful isolation of individual fungal colonies.

To initiate the gradient dilution, 5 g of soil were meticulously mixed with 45 ml of sterile water in a 250 ml conical flask. This initial mixture served as the basis for the first dilution, referred to as the 10–1 dilution. The flask underwent a rigorous shaking process for 30 min, ensuring the thorough mixing of soil and water, after which it was left undisturbed to allow for settling.

Subsequently, 1 ml supernatant was carefully transferred to a sterile glass test tube and blended with 9 ml of sterile water. This step resulted in the creation of a 10^–2 dilution. The process of iterative dilution was repeated, each time transferring 1 ml of the previous dilution into a new test tube with 9 ml of sterile water. This series of dilutions continued until a 10^–9 dilution was achieved.

Fifty microliters (50 μL) from each dilution were aseptically inoculated onto LB medium in triplicate. These inoculated plates were then incubated at a temperature of 37°C for a duration of 1 day. After the incubation period, individual bacterial colonies that had thrived and developed on the medium were carefully selected based on their distinctive morphological features, coloration, and other observable characteristics.

The plate confrontation method was employed to confirm and identify the purified single-colony bacteria as antagonistic microorganisms. In this procedure, the pathogenic fungus was initially inoculated onto one side of a Petri dish, with a distance of 0.5 cm from the wall, and incubated at 28°C for 24 h. Subsequently, the antagonistic bacteria were inoculated 6 cm away from the fungus on the same Petri dish and placed in a constant temperature incubator at 25°C. The diameter of the inhibition zone was measured using the criss-cross method when the pathogen colony completely covered the plate. This confirmation was essential to ensure that the isolated strains indeed exhibited antagonistic properties against the target pathogen. The experiments were meticulously conducted in triplicate to validate the reproducibility of the antagonistic effects.

Following the identification and confirmation of effective antagonistic strains, these strains were preserved by streaking them on agar slants. The slants were then stored in a refrigerator at a temperature of 4°C, ensuring the long-term viability and availability of these valuable strains for future research and applications.

The T5 Direct PCR Kit (Plus) and Bacterial extraction kit (Invitrogen, the USA) has been specifically tailored for the direct amplification of fungal DNA and bacterial DNA, respectively. The T5 Direct PCR Kit streamlines the process, allowing the use of these fungus samples as templates for PCR amplification. The protocol encompasses the following sequential steps:

Fungal samples were subjected to PCR amplification using the common primers ITS1 and ITS4 as the upper and lower primers. Bacterial samples utilized the common primers 27F and 1492R. The PCR reaction system consisted of a 50 μL reaction volume, comprising 25 μL of 2 × T5 Direct PCR, 2 μL of the forward primer, 2 μL of the reverse primer, 2 μL of DNA template, and 19 μL of ddH₂O.

The resulting PCR products were subsequently sent for sequencing, and the sequenced DNA sequences were compared against the NCBI (National Center for Biotechnology Information) database using Blast, an advanced bioinformatics tool, to determine the species of the isolated strains.

Pathogenic fungal gene sequences obtained through sequencing were subjected to thorough analysis. This analysis involved performing a BLAST comparison in the GenBank database, a renowned and comprehensive repository of genetic information. Reference strains were meticulously selected based on a stringent similarity threshold of 98% or higher. This stringent criterion ensures that the reference strains closely resemble the sequences of the pathogenic fungi under investigation.

Following the identification and selection of reference strains, a phylogenetic tree was meticulously constructed. The construction of this tree involved employing the Neighbor-Joining method, a widely recognized approach for inferring evolutionary relationships among biological entities. The process was facilitated using MEGA5.0 software.

The verification of pathogenic fungi adhered to the stringent standards of Koch’s postulates, a well-established set of criteria for establishing the causative relationship between a microorganism and a disease. In this study, Koch’s postulates were applied to healthy G. biloba leaves, following a protocol adapted from Pour et al. (2020).

To prepare the experimental leaves, all petioles were meticulously trimmed using sterilized scissors within a sterile environment to ensure they fit snugly into disposable plates. This process minimized the risk of introducing any contaminants or external factors that could confound the results.

Subsequently, the leaf surfaces underwent thorough cleaning with alcohol pads to achieve a high degree of surface sterilization. This step was crucial to ensure that the leaves were free from any pre-existing microorganisms that could interfere with the pathogenicity testing.

The experimental leaves were then exposed to UV radiation on their undersides for a duration of 30 min. UV radiation is a highly effective method for surface sterilization, eliminating potential contaminants that may be present on the leaf surfaces.

Autoclaved filter paper was thoughtfully placed inside the plates, and 2000 ml ddH₂O were added to the filter paper. This step was essential to maintain suitable humidity levels within the plates, creating an environment conducive to pathogenic activity.

The UV lamp was reactivated to sterilize the upper leaf surfaces for an additional 30 min, ensuring that both sides of the leaves were effectively sterilized.

Finally, pathogenic fungi were inoculated onto the leaves using both punctured and unpunctured methods. This dual approach allowed for the assessment of different modes of infection and disease development. A control group of leaves remained untreated to serve as a baseline for comparison. Each treatment, including the control group, was meticulously replicated three times to ensure the reliability and repeatability of the results.

All disposable plates were hermetically sealed with film to prevent any external contamination during the incubation period. The plates were then placed in an incubator set at a constant temperature of 28 C, creating an environment conducive to disease development.

Leaf changes and disease symptoms were diligently monitored and recorded after a period of 14 days, allowing for the assessment of the pathogenicity of the inoculated microorganisms. Importantly, no additional ddH₂O was introduced during this period to prevent any potential sources of external contamination, thereby upholding the integrity of the experiments.

The leaves, having undergone the pre-treatment procedures detailed in Section 2.5., were subjected to further experimentation. To comprehensively assess the interactions between pathogenic fungi, antagonistic bacteria, and the plant, the leaves were divided into four distinct treatment groups:

The first group (1) involved leaves that were initially inoculated with pathogenic fungi. This inoculation was conducted using a puncture method, wherein the pathogenic fungi were introduced through small punctures made on the leaf surfaces. Subsequently, these leaves received an inoculation of 200 μL antagonistic bacteria.

In the second group (2), leaves were also initially inoculated with pathogenic fungi, but in this case, the inoculation was carried out without puncturing the leaves. After the pathogenic fungi had established themselves, 200 μL antagonistic bacteria were introduced.

The third group (3) comprised leaves that were solely inoculated with pathogenic bacteria, exploring the impact of pathogenic bacteria in isolation.

To serve as a baseline and control for comparison, a blank control group (CK) (4) was established. This control group received no treatment and was maintained in its original state.

Following the allocation of leaves into their respective treatment groups, all plates were carefully placed inside a constant temperature incubator set at 25 C. This controlled environment facilitated the development of interactions between the microorganisms and the plant leaves over a defined period.

The outcomes of these experiments were thoroughly documented through both photography and detailed record-keeping on the 14th day. This comprehensive approach allowed for the observation and assessment of any visible changes, disease symptoms, and the overall impact of the treatments on the plant leaves, providing valuable insights into the interactions and effects of the microorganisms under investigation.

Oneway analysis of variance (ANOVA) with the Least Significant Difference (LSD) test was employed for significance analysis. A significance level of P < 0.05 was considered. Figures were processed using Photoshop CS5.

A total of 19 fungal strains and 5 bacterial strains were isolated from infected leaves and rhizosphere soil of G. biloba. The identical strain isolates were distinguished based on their morphological characteristics using both a light microscope and visual observation and identified through sequencing and genetic alignments. Among these isolated fungi, Botryosphaeria dothodea and Neofusicoccum parvum were identified as potential pathogen through checking on the reference literatures. Colonies of these two strains on PDA medium were initially white, gradually transitioned to lime green, with some eventually turning black in the later stages (Figure 1).

A phylogenetic tree was constructed using Mega 5.0 (Figure 2) to elucidate the evolutionary relationships among isolated B. dothidea SHY0821 and other related strains. The tree’s topology highlighted an intriguing insight into their genetic divergence. Notably, B. dothidea SHY0821 shares a common branch with MN856391.1 and MN856396.1, suggesting a common ancestry, however, further analysis indicated separate evolutionary trajectories. This observation strongly suggests significant genetic distinctions between SHY0821 and its counterparts within the species.

The resulting phylogenetic tree offered valuable insights into the evolutionary relatedness of the pathogenic fungi and their potential genetic relationships with the selected reference strains. This in-depth analysis is instrumental in shedding light on the genetic diversity and evolutionary context of the pathogenic fungi, aiding in the understanding of their pathogenicity and genetic characteristics.

The experimental results detailed in Figure 3 demonstrated the successful validation of Koch’s postulates for the pathogenic fungi, B. dothidea and N. parvum, associated with Ginkgo leaf spot disease. The inoculation of G. biloba leaves with these fungi, whether through punctured or unpunctured methods, led to visible necrosis. Notably, necrosis appeared sooner in the puncture-inoculated leaves compared to the unpunctured group, suggesting a more rapid onset of symptoms due to the method of inoculation.

The disease progression was characterized by necrotic lesions, initially observed at leaf tips or margins, gradually developing into yellow to brown lesions spreading toward the mid rib. Control leaves from the same branch remained healthy, reaffirming the association of these specific fungi with the observed symptoms. Re-isolation and genetic identification of pure cultures from these lesions further confirmed the presence of B. dothidea and N. parvum, affirming their pathogenic nature in causing Ginkgo leaf spot disease.

Notably, N. parvum demonstrated a higher impact on leaf spot disease compared to B. dothidea. The punctured group displayed more severe symptoms, indicating a correlation between method of inoculation and disease severity (Table 1). This observation aligns with existing literature that describes N. parvum as a more aggressive pathogen in various plant hosts.

From the inter-rhizosphere soil samples of G. biloba, fourteen bacterial strains were isolated and purified. Among these strains, two displayed a significant inhibitory effect on the growth of the pathogenic fungus B. dothidea. Molecular analysis based on 16S rRNA identified these strains as Bacillus velezensis (B. velezensis ST2-1) and Bacillus amyloliquefaciens (B. amyloliquefaciens ST2-1-12). The average inhibitory zones measured for these strains against B. dothidea were 4.78 cm and 3.46 cm (Figure 4), respectively, with B. velezensis exhibiting a more pronounced inhibitory effect. The inhibition zone of B. velezensis against N. parvum was 4 cm (Figure 4).

Co-inoculation of antagonist microorganisms, B. velezensis and B. amyloliquefaciens, along with the pathogenic microorganism B. dothidea on Ginkgo leaves was conducted. Initially, leaves were inoculated with B. dothidea, followed by the application of live bacterial suspensions of B. velezensis and B. amyloliquefaciens. After 14 days, it was observed that the leaf lesions inoculated with the antagonist microorganisms exhibited significantly smaller lesions compared to the control group solely inoculated with the pathogen (Figure 5). Additionally, the results indicated that B. velezensis showed a stronger inhibitory effect compared to B. amyloliquefaciens on the development of lesions caused by B. dothidea (Table 2).

B. subtilis, a widely distributed Gram-positive bacterium, has been previously reported to exhibit effective antibacterial properties against various pathogenic bacteria. This antibacterial effect is attributed to the secretion of various secondary metabolites, including polyketides, bacillomycin, and 3-hydroxypropionaldehyde. However, the precise mechanism by which B. velezensis and B. amyloliquefaciens inhibit B. dothidea requires further investigation.