Antifungal activity of compounds from Gordonia sp. WA8-44 isolated from the gut of Periplaneta americana and molecular docking studies

Invasive fungal infections are on the rise, leading to a continuous demand for antifungal antibiotics. Rare actinomycetes have been shown to contain a variety of interesting compounds worth exploring. In this study, 15 strains of rare actinobacterium Gordonia were isolated from the gut of Periplaneta americana and screened for their anti-fungal activity against four human pathogenic fungi. Strain WA8-44 was found to exhibit significant anti-fungal activity and was selected for bioactive compound production, separation, purification, and characterization. Three anti-fungal compounds, Collismycin A, Actinomycin D, and Actinomycin X2, were isolated from the fermentation broth of Gordonia strain WA8-44. Of these, Collismycin A was isolated and purified from the secondary metabolites of Gordonia for the first time, and its anti-filamentous fungi activity was firstly identified in this study. Molecular docking was carried out to determine their hypothetical binding affinities against nine target proteins of Candida albicans. Chitin Synthase 2 was found to be the most preferred antimicrobial protein target for Collismycin A, while 1,3-Beta-Glucanase was the most preferred anti-fungal protein target for Actinomycin D and Actinomycin X2. ADMET prediction revealed that Collismycin A has favorable oral bioavailability and little toxicity, making it a potential candidate for development as an orally active medication.


Introduction
The incidence of invasive fungal infections is on the rise due to an increase in the number of immunocompromised patients [1] and the emergence of antibiotic-resistant strains, particularly those resistant to azole drugs [2]. There is an urgent need to develop new anti-fungal antibiotics with novel mechanisms of action to efficiently combat these infections [3]. To address this challenge, researchers have undertaken a diverse range of approaches, including the discovery of new bioactive molecules and the advancement of novel techniques [4]. Microorganisms are an inexhaustible source of new bioactive molecules [5], with actinomycetes having Fractions 2 and 3 exhibited strong anti-fungal activity and were selected for further purification. Fraction 2 was fractionated using open-column chromatography with silica gel and yielded four subfractions (fractions 2-1 to 2-4). Fraction 2-3 was further purified using preparative RP-HPLC with a YMC-Pack ODS-A/S-5 μm/12 nm 250 × 10.0 mm column, a flow rate of 3 mL/min, MeOH/H 2 O (80% at 20 and 22 min), with UV detection at 210 nm. This led to the isolation of two compounds, A (182 mg) and B (20 mg). Fraction 3 was subjected to open-column chromatography using silica gel (gradient of PE/EA) and Sephadex LH-20 column chromatography with MeOH as the eluent. Compound C (29 mg) was then purified using preparative RP-HPLC with a YMC-Pack ODS-A/S-5 μm/12 nm 250 × 10.0 mm column, a flow rate of 3 mL/min, and MeOH/H 2 O (70% in 25 min), with UV detection at 210 nm.

Minimum inhibitory concentration testing
Following the Clinical and Laboratory Standards Institute (CLSI) guidelines, the M27-A3 method was used to determine the minimum inhibitory concentrations (MICs) of C. albicans [23], while the M38-A2 method was used for filamentous fungi [24]. To prepare the samples, C. albicans was diluted with RPMI1640 to a final concentration of 0.5-2.5 × 10 3 spores/ml, A. fumigatus, and A. niger were diluted to 1 × 10 4 spores/ml, and T. rubrum was diluted to 2 × 10 3 spores/ml. The test samples were dissolved in DMSO and then diluted with RPMI1640 to a series of concentrations. Next, 100 μL of the test sample solutions were added to 100 μL of the cell suspension and incubated for 48 h. Each experiment was repeated three times.

Structural prediction of antifungal compounds
The NMR spectra of the chemicals were measured using a 600 MHz spectrometer (Brucker AVANCE III, 600 M, Germany), with tetramethyl silane (TMS) as the internal reference and dimethyl sulfoxide (DMSO) as the solvent. The compounds were injected separately into a Triple-quadrupole mass spectrometer (Thermo Scientific TSQ Endura™, USA) to obtain mass spectra. All solvents used for extraction were of LR grade, whereas HPLC-grade solvents were used for analysis.

Scanning electron microscopy (SEM) analysis
The morphological changes induced in C. albicans by the active compounds were observed using SEM, as described by Jeyanthi et al. [25]. 4 × MIC of the active compounds were applied to C. albicans suspensions (1-5 × 10 5 CFU/mL) for 48 h at 28 • C. The cells were then fixed in 2.5% glutaraldehyde at 4 • C overnight, washed three times with PBS for 20 min, and subsequently underwent a series of ethanol washes (30-100%). Finally, SEM images were obtained using a Quanta FEG 200 FESEM, which was operated at an accelerating voltage of 2-19 KV under standard operating conditions.

Isolation and screening
15 strains of Gordonia were isolated from the gut of P. americana (as shown in Fig. 1AB). Among these strains, 8 strains demonstrated anti-C. albicans activity, 4 strains showed anti-T. rubrum activity, 3 strains displayed anti-A. fumigatus activity, and 3 strains exhibited anti-A. niger activity (as listed in Table 1). Notably, strain WA8-44 demonstrated significant activity against all the tested fungi. Therefore, strain WA8-44 was chosen for further investigation (as shown in Fig. 3ABCD and Table 1).

Identification of potent isolate WA8-44
Gordonia sp. WA8-44 exhibited good growth on Gauze's Agar No. 1 medium, with irregularly shaped and raised colonies that were either reddish-brown or milky yellow, with a moist and opaque surface ( Fig. 2A). The strain was observed to be a gram-positive bacterium based on Gram staining results (Fig. 2B). Short, smooth, rod-shaped cell was observed using SEM (Fig. 2CD).
The 16 S rRNA sequence of strain WA8-44 was submitted to NCBI, and an accession number was obtained in GeneBank (MH605445). A phylogenetic tree of strain WA8-44, based on the Neighbor-Joining method, formed a single clade involving Gordonia terrae (NR_118,598) with a similarity of 99.55% (Fig. 2E). Molecular characterization confirmed that the strain belonged to Gordonia terrae.

Extraction, purification, and characterization
A bioassay-guided investigation was performed to isolate the active constituents from an ethyl acetate extract of strain WA8-44. Gradient elution was performed using different organic solvents and the activity of different fractions was determined using C. albicans ATCC 10231 as an indicator. The active components were further purified and the final purification of the sub-fractions was achieved by preparative RP-HPLC. In total, three active compounds were isolated and identified ( Fig. 1S-9S). Compound A appeared as

Anti-fungal activity of isolated bioactive compounds
The minimum inhibitory concentrations (MICs) of the isolated compounds are presented in Table 2. Actinomycin D, Actinomycin X 2 , and Collismycin A showed potent activity against C. albicans ATCC 10231, T. rubrum ATCC 60836, A. fumigatus ATCC 96918, and A. niger ATCC 16404, with MICs ranging from 3.19-101.96 μmol/L. This suggests that Actinomycin D, Actinomycin X 2 , and Collismycin A have the potential to be potent antifungal agents with broad spectra. The anti-C. albicans effect of Actinomycin D (Fig. 4CG), Actinomycin X 2 (Fig. 4BF), and Collismycin A (Fig. 4DH) was observed through SEM analysis. The cell membrane structure of C. albicans ATCC 10231 was damaged after treatment with these compounds, which may have resulted in cell lysis.

Molecular docking studies
In this study, molecular docking analysis was conducted to evaluate the potential binding affinity of Actinomycin D, Actinomycin X 2 , and Collismycin A to selected receptors. The receptors selected were N-myristoyltransferase, dihydrofolate reductase, Secreted aspartic proteinase 5, 1,3-Beta-Glucanase, lanosterol 14-alpha demethylase, chitin Synthase 2, fructose-1,6-bisphosphate aldolase, thymidylate synthase, and squalene epoxidase ( Table 3). The results showed that Actinomycin D and Actinomycin X 2 exhibited the   These results suggest that Actinomycin D, Actinomycin X 2 , and Collismycin A have the potential to target these receptors and may be useful as antifungal agents. We also provided visual representations of the interaction between these compounds and their respective receptors in Fig. 5AB, 6AB, and 7AB, as well as detailed information about the intermolecular interactions in Table S1. The stability of the interactions was mainly due to the formation of H-bonds and hydrophobic interactions between the compounds and the receptor active sites. At 1,3-Beta-glucanase, Actinomycin D formed nine conventional H-bonds, five carbon H-bonds, and two hydrophobic interactions, while Actinomycin X 2 formed four conventional H-bonds, four carbon H-bonds, and seven hydrophobic interactions (Fig. 5C). The researchers also noted that although Actinomycin D and Actinomycin X 2 have a slight difference in chemical structure, the two compounds only have six same interactions (Fig. 6C). At Chitin Synthase 2, Collismycin A formed two carbon H-bonds with Tyr302 and Val303, and four hydrophobic interactions with Arg300, Arg300, Val303, and LysA306 (Fig. 7C). Overall, these findings suggest potential new targets for these compounds and provide insights for further drug design and development.

In-silico drug-likeness and ADMET prediction of bioactive compounds
In this study, Lipinski's rule of five was used to predict the drug-like properties of the bioactive compounds. According to Lipinski's rule of five, a molecule that is most likely to be developed as a candidate for an orally active medication should meet certain criteria, including a molecular weight less than 500 Da, no more than 10 hydrogen bonds, no more than 5 hydrogen H-Bond Donors, and MLogP<4.15. As shown in Table 4, the results indicate that Collismycin A complies with Lipinski's rule of five and can be strongly recommended as an oral drug.
Pharmacokinetics refers to how a drug behaves after it is introduced into the body, with ADMET parameters being the factors influencing its pharmacokinetics. The ADMET parameters of the bioactive compounds in silico were presented in Table 5. Caco2 permeability of all compounds was above 0.9, which indicates that the compounds can be administered orally. The blood-brain barrier (BBB) plays a crucial role in maintaining homeostasis in the central nervous system (CNS) by preventing unrestricted movement of poisonous or harmful compounds, transit of nutrients, and removal of metabolites from the brain. However, both compounds cannot cross into the brain as their BBB permeability was above 0.3. Actinomycin D and Actinomycin X 2 have a class II acute oral toxicity, which means they are fatal if swallowed, with LD 50 values between 5 and 50 mg/kg, while Collismycin A has a class IV acute oral toxicity, which means it is harmful if swallowed, with LD 50 values between 300 and 2000 mg/kg.

Discussion
Rare actinomycetes have recently gained popularity in research, and several valuable compounds have been discovered from them [36]. Examples include vancomycin from Amycolatopsis [37], macrolides from Micromonospora [38], and erythromycin from Saccharopolyspora [39]. Gordonia is a member of the rare actinomycetes family that can be found in various environments such as the human body, soil, sewage, and oil wells [40]. However, it is rarely found in insects. In this study, 15 strains of Gordonia were isolated from the gut of P. americana for the first time, which provides new sources of Gordonia.
Insects are the largest group in nature and they harbor a large number of symbiotic bacteria in their bodies. It has been suggested by many researchers that gut-inhabiting bacteria protect their insect hosts against pathogens by synthesizing specific antimicrobial compounds [41][42][43]. For instance, Nurdjannah et al. found that Bacillus cereus from Apis dorsata gut produced surfactin, fengycin, and iturin A that exhibited potent anti-Neisseria gonorrhoeae activity [44]. Pleosporales sp. BYCDW4 from Odontotermes formosanus produced 5-hydroxyramulosin and biatriosporin M, which showed strong antibacterial effects against E. coli, C. albicans, B. subtilis, and S. aureus [45]. It is widely acknowledged that microorganisms isolated from insects are crucial sources for discovering and producing bioactive compounds.
The microbiomes of insects hold great promise for developing new antibiotics to treat fungal infections. In a study by MG Chevrette et al., insect-associated actinomycetes showed significantly greater activity against fungi and a wider range of biosynthetic capabilities compared to soil-associated actinomycetes, as determined through ecologically optimized bioassays, genomic, and metabolomic analyses [46]. While previous studies have not reported the anti-fungal activity of Gordonia isolated from different environments, our research found that 53.3% of Gordonia strains isolated from the gut of P. americana exhibited anti-fungal effects. This discovery highlights a new source of microorganisms for exploring novel natural anti-fungal products.   Out of all the isolates, WA8-44 showed robust anti-fungal activity against common human pathogens. Therefore, WA8-44 was selected for further investigation. Through morphological characteristics, molecular identification, and 16 S rRNA phylogenetic tree analysis, WA8-44 was identified as Gordonia terrae (GenBank Accession Number: MH605445). Using bioassay-guided isolation methods, three active anti-fungal compounds were separated from the fermentation broth of WA8-44. The purified compounds were chemically characterized through MS, 1 H NMR, and 13 C NMR. The three active anti-fungal compounds were identified as Collismycin A, Actinomycin D, and Actinomycin X 2 .
Collismycin A is a natural product that is known to have cytotoxic activity against cancer cells and is typically produced only by Streptomyces [47]. However, our study discovered that Gordonia also produces Collismycin A, which showed remarkable activity against fungi. While previous studies have not reported the anti-filamentous fungi activity of Collismycin A, our findings reveal its potential as an effective anti-filamentous fungi agent. Additionally, Actinomycin D and Actinomycin X 2 , two chromopeptide lactone antibiotics with various biological activities, including anti-cancer [48], anti-bacterial, anti-fungal [49], and anti-leishmanial properties [50], were also identified in Gordonia. These compounds are typically produced by Streptomyces [51] and, in some cases, Micromonospora [52], but this study demonstrated that they can also be produced by Gordonia. The MIC assay showed that all three active compounds had significant anti-fungal activity against C. albicans, T. rubrum, A. fumigatus, and A. niger, with MICs ranging from 3.19-101.96 μmol/L. Molecular docking is a powerful computational method used to predict the binding affinities and conformations of drug candidates [53,54]. In this study, in-silico virtual screening was employed to determine the mechanism of action of three antifungal bioactive compounds derived from Gordonia sp. WA8-44 against C. albicans. Nine modeled protein targets from C. albicans were docked with the bioactive compounds, and the results showed that Actinomycin D and Actinomycin X 2 were the most potent against 1,3Beta-Glucanase. This protein is a potential therapeutic target for treating Candida infections because it is responsible for most of the glucanase activity in the Candida cell wall [55]. Although Actinomycin X 2 and Actinomycin D are similar in chemical structure, Actinomycin D exhibited stronger anti-C. albicans activity due to its six additional H-bonding interactions with 1,3Beta-Glucanase [56]. Likewise, in the Hamdoon's study, methylated taxifolin demonstrated enhanced binding affinity towards various pro-and antiapoptotic proteins compared to taxifolin, thereby displaying heightened anticancer efficacy [57]. Although previous studies have reported the anti-Candida albicans effect of Collismycin A, its mechanism of action is still unclear [58]. In Candida albicans, Chitin Synthase II is primarily responsible for synthesizing the primary septum and is considered a potential target for antifungal drugs [59]. These results provide insights into the possible specific mechanisms of action of these compounds and could guide future experimental investigations.
Drug development is a time-consuming and costly process, and inadequate pharmacokinetics is a significant cause of failure in  Table 5 In silico ADMET properties of each compound.
clinical trials [60]. Therefore, ADME data, which provide information on a drug's absorption, distribution, metabolism, and excretion, are becoming increasingly crucial in drug development. In-silico techniques offer a cost-effective way to filter and optimize drug candidates in the early stages of drug development since experimental acquisition of ADME data can be expensive. In this study, an in-silico assessment of the ADME properties of the three bioactive compounds was conducted. The Caco2 permeability values of all compounds were greater than 0.9, suggesting their suitability for oral administration. The BBB permeability of all three substances was low, indicating that none of them would have any significant negative impact on the central nervous system. In the metabolic process of drugs, CYP enzymes, including CYP450 inhibitors, play a central role. None of the three drugs were predicted to inhibit P450, indicating that they would not present drug-drug interaction problems with medications targeting those CYP enzymes. Consistent with previous reports in the literature, Actinomycin D and Actinomycin X 2 were predicted to have moderate toxicity [48]. The results obtained from the in-silico ADMET studies predicts Collismycin A as a potent oral drug candidate due to its low toxicity and compliance with the Lipinski rule of five. This study represents the first attempt to isolate anti-fungal secondary metabolites from Gordonia associated with P. americana. A total of 15 strains of Gordonia were isolated from the gut of P. americana, and 53.3% of them exhibited anti-fungal activity. Among these strains, three anti-fungal compounds (Collismycin A, actinomycin D, and actinomycin X 2 ) were identified and purified from the secondary metabolites of Gordonia sp. WA8-44. Notably, Collismycin A was isolated from Gordonia for the first time, and its antifilamentous fungi activity was firstly identified in this study. Based on the predicted ADMET properties, Collismycin A demonstrated good oral bioavailability and low toxicity, making it a promising candidate for the development of orally active drugs.

Data availability statement
Data included in article/supp. Material/referenced in article.

Additional information
Supplementary content related to this article has been published online at [URL].

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.