Identification of potent anti-Candida metabolites produced by the soft coral associated Streptomyces sp. HC14 using chemoinformatics

Candida albicans is the most common pathogen responsible for both spontaneous and recurrent candidiasis. The available treatment of Candida infections has several adverse effects, and the development of new drugs is critical. The current study looked at the synthesis of anti-Candida metabolites by Streptomyces sp. HC14 recovered from a soft coral. Using the Plackett Burman design, the medium composition was formulated to maximize production. Using GC–MS, the compounds have been identified, and a cheminformatics approach has been used to identify the potential source of activity. The compounds that showed high potential for activity were identified as pyrrolo[1,2-a]pyrazine-1,4-dione, hexahydro-3-(phenylmethyl)-3 and di-n-octyl based on their docking score against the cytochrome monooxygenase (CYP51) enzyme in Candida albicans. As a result of their discovery, fewer molecules need to be chemically synthesized, and fermentation optimization maximizes their synthesis, providing a strong foundation for the development of novel anti-Candida albicans agents.

www.nature.com/scientificreports/ fermentation conditions 16 . External parameters such as inoculation size, medium volume, incubation period, incubation temperature, agitation speed, and starting pH have all had a significant impact on the nature and quantity of antibiotics and/or bioactive chemicals produced [17][18][19][20] . In the field of antibiotics, much effort was invested in directing the product range and enhancing production rates. Changing nutritional or environmental conditions has been demonstrated to result in the development or discovery of novel natural chemicals utilizing various approaches 21 . However, the poor yield of these chemicals limits their application in many circumstances. Furthermore, the nutritional requirements of the microbes used in industrial fermentation processes are as complex and diverse as the bacteria themselves 22,23 . To that end, and thanks to advances in computational biology, drug discovery and repurposing of natural microbial compounds have become easier and faster 24 .
In the current study, the most influential factor for Streptomyces sp. HC14 anti-Candida activity was examined using the Plackett-Burman design. The generated compounds were identified using gas chromatography-mass spectrometry (GC-MS) under near-optimal conditions. The putative active compounds were computationally investigated against a fundamental protein target in Candida albicans utilizing a molecule docking and molecular dynamics approach.

Materials and methods
Bacterial strain and growth condition. The actinobacteria Streptomyces sp. HC14 was previously isolated from the soft coral Sarcophyton glaucum that has been collected from the Red Sea, Hurghada, Egypt 25 . The 16 s rRNA sequence of Streptomyces sp.HC14 was deposited in the NCBI GenBank under the accession numbers JQ929066 25 . For cultivation, a loopful of Streptomyces sp. HC14 was inoculated into 50 mL of ISP2 medium (Difco) in 250 mL Erlenmeyer flasks. Then it was incubated at 37 °C under shaking condition at 200 rpm for 7 days. The cell-free medium filtrate was used to test the antimicrobial activity of the fermented medium. The ISP2 medium has a composition of (g/L) yeast, 4; malt extract, 10; dextrose, 4. The medium pH was adjusted to pH 7.2.
Anti-Candida albicans activity. The cell-free supernatant was used to test the antimicrobial activity of the tested strain. Cells were removed by centrifugation at 5000 rpm for 15 min. Fifty microliters of the Streptomyces sp. HC14 cell-free supernatant were tested against Candida albicans using the well diffusion method. After dispensing of the 50 µL to each the wells, plates were put in the refrigerator for 3 h to allow diffusion, then all the plates were incubated at 37ºC and diameter of inhibition zones were measured after 48 h 26 , and the well diameter was subtracted from the inhibition zone diameter.

Plackett-Burman experimental design (PB).
Streptomyces sp. HC14 were potent in producing antagonistic compounds to Candida albicans. Therefore, it was selected for the optimization experiments in an attempt to maximize the production. Optimization was performed by a two-phase experimental design: the first is screening for critical elements that influence the production of antimicrobial agents in shaken flasks using Plackett-Burman design 27 , and the second was the verification experiment, to confirm the near-optimal conditions for production of anti-Candida compounds.
Seven independent variables based on the ISP2 medium's composition were chosen for screening the most influential component. Namely, dextrose, yeast extract, malt extract, seawater, pH, culture volume, and inoculum size. For each variable, high ( +) and low (-) levels were tested (Tables 1, 2). Trial number 9 in Table 2 represents the basal control. All experiments were carried out in duplicate, and the arithmetic means of antibacterial activity were assessed by measuring the inhibition diameter as a response. The main effect of each variable was determined by the following equation: where E xi is the variable main effect, M i+ and M i− are inhibition zone diameters in high and low levels respectively, where the independent variable (xi) was present in high and low concentrations, respectively, and N is the number of trials divided by 2. Statistical t-values for equal unpaired samples were calculated using Microsoft Excel www.nature.com/scientificreports/ to determine the variable significance. From the main effect calculations, the composition of the near optimized medium was predicated, which gives a maximum inhibition zone as a response.
Extraction of secondary metabolites. At the end of the fermentation period, the bioactive compounds were extracted from the culture medium using the method described by El-Naggar et al. 28 . Briefly, the cell-free extract was obtained by centrifugation for 15 min at 5000 rpm, then, it was mixed with chloroform in a 1:1 ratio and agitated for one hour and repeated 3 times. The organic phase was separated from the aqueous phase using a separation funnel and dried under a vacuum using a rotary evaporator at a temperature of no more than 50 °C. The recovered crude extract was then dissolved in methanol for further characterization.
Structural and molecular docking. Analysis by UV absorption and GC-MS. The crude extract was analyzed by UV absorption using a Perkin Elmer-lambada 4B-UV/VIS Spectrophotometer. Identification of the metabolites was done by GC-MS analysis; briefly, 1 μL of the crude extract was injected into RTX-5 column (7 m × 0.32 mm) (model GC-MS-QP-2010 plus from Shimadzu, Japan), and Helium (3 mL/min) was used as a carrier gas. The following temperature gradient program was used: At 75 °C for 2 min, followed by an increase from 75 to 175 °C at a rate of 50 °C per min, and finally 7 min at 175 °C. The m/z peaks, representing mass to charge ratios characteristic of the metabolites, were compared with those in the mass spectrum library of the corresponding organic compounds 29 .
Ligand and protein model and docking. Candida target was retrieved from the PDB with the accession number 5V5Z. MarvinSketch (v22.11.0) was used to draw all the metabolites (ligands) 3D structures of the identified compounds. The ligands and protein model preparation and docking were conducted using Schrödinger software, released 2018-4 30,31 . Briefly, the drawn 3D structures were prepared using LigPrep tool. OPLS-2005 force field module was applied to minimize the energy of the ligands. The energy of each molecule was minimized using LigPrep. The docking was done using extra-precision docking default settings and the docking score results were displayed using XP visualizer.
Ligand-target complex free energy. To calculate the free energy of complexes, Maestro's Prime integrated tool was used. Molecular mechanics of the Generalized Born and surface area solvation (MM-GBSA) method was used to determine the binding free energies 32 . MM-GBSA was used to improve the accuracy of the docking score 33 . Thus, the free energy of optimized receptors, ligands, and ligand-receptor complexes was calculated. These calculations were performed under the solvation condition of the VSGB 2.0 model and forcefield OPLS_2005 34 . The strain energy for the ligands as well as the relative binding free energies of the complexes were also estimated 34,35 . The visualization of energy was generated by the primary energy visualizer 36 .
Ethics approval. This article does not contain any studies on humans or on animals.

Plackett-Burman experimental design (PB). Plackett-Burman design was used in a few controlled
experiments to determine the factors affecting the production of bioactive chemicals and figure out their likely ideal amounts. The responses in Table 3 demonstrate a significant variation in the diameter of the inhibition zones, highlighting the significance of parameter tuning to achieve high activity. This illustrates that the combined influence of the nine components in their inhibition zone diameter range from 0 to 15 mm against Candida albicans.
The main effect of each variable was estimated as the difference between the average of the measurements made at the high (+) and low (−) levels of the factors. It was observed for antagonistic compounds produced by Streptomyces sp. HC14 for C. albicans. High levels of dextrose, sea water, and yeast extract enhanced anti yeast activity. On the contrary, low levels of malt extract, pH, volume, and inoculum size increased the zone of inhibition (Fig. 1).  www.nature.com/scientificreports/ Statistical analysis of the Plackett-Burman experiment is shown in Table 4 for Streptomyces sp. HC14. From a closer look, malt extract was the most significant variable for Streptomyces sp. HC14, as it had significant levels of 95%.
Verification experiment. To validate the obtained data and to evaluate the accuracy of the applied Plackett-Burman statistical design, a verification experiment was carried out in triplicate for each strain to predict the near optimum combination levels of independent variables. The Plackett-Burman design predicted that the higher inhibition zone formed by Streptomyces sp. HC14 was grown in a medium containing (g/L): yeast extract 6, Dextrose 6, Malt extract 5, Sea Water + 5 g NaCl, pH 6.5 with an inoculum size of 0.5 mL in 25 mL culture volume. Three independent repeated experiments were performed to verify the validity of the near optimum settings. The base condition was compared to the average inhibition zone of anticipated near-optimal levels of independent variables. Using the optimal conditions established from the Plackett-Burman experiment, the inhibition zone of Streptomyces sp. HC14 rose 2.37 folds over the basal conditions (Fig. 2).  www.nature.com/scientificreports/ Structural and molecular docking. Identification of secondary metabolites. For the identification of secondary metabolites, UV absorption followed by GC-MS were applied to the active crude extract. The UV absorption spectrum of the extract with antimicrobial activity has recorded a maximum absorption at 437.4 nm for Streptomyces sp. HC14. The National Institute of Standard and Technology (NIST) database, which contains more than 62,000 patterns, was used to interpret the mass spectrum of the GC-MS. A comparison was made between the mass spectra of the unknown and known components recorded in the NIST collection. The components of the test materials' names, molecular weights, and structures were determined. The GC-MS chromatograms of Streptomyces sp. HC14 metabolites is given Fig. 3. On comparison of the mass spectra of the constituents with the NIST library, eighteen peaks were obtained for Streptomyces sp. HC14, ( Table 5). As clearly shown in the Fig. 3 the highest peak was displayed by the 18th compounds which has been identified to be di-n-octyl phthalate. The second highest peak was compound ordered the 4th, named Cyclohexane, 1, 5-diisopropyl-2, 3-dimethyl.    Table 6. The main purpose of this step was to explore and predict the source of activity in the extract and the potential target. From the results listed in Table 6, it was observed that pyrrolo[1, 2-a]pyrazine-1, 4-dione, hexahydro-3-(phenylmethyl) showed the best scores for docking, GScore, lipophilic score and H-Bond energy with values of − 10.428 kcal/ mol, − 10.428 kcal/mol, − 3.747 kcal/mol, and − 1.725 kcal/mol respectively. Figure 4 depicts the chemical structures of the top metabolites based on docking score.  www.nature.com/scientificreports/ The graphical representations of the interactions in ligand-target complex are shown in Fig. 5a,b. The figures depict that there are three hydrogen bonds formed between the ligand and the SER-378 and HIE-377 amino acid residues of the target protein. Moreover, this metabolite showed one π-π with TYR-118 residue.
Prime MM-GBSA results in Table 7 revealed that the calculated binding energy of di-n-octyl phthalate-1 was − 80.92 kcal/mol which is highest among the tested compounds as well as it is higher compared to the best docking compound (pyrrolo[1, 2-a]pyrazine-1, 4-dione, hexahydro-3-(phenylmethyl)-3), which has a numerical values calculated to be − 54.51 kcal/mol. This indicates that di-n-octyl phthalate has a higher affinity to CYP51 enzyme.

Discussion
In the current study, the metabolites of Streptomyces sp. HC14 were computationally evaluated to predict likely active compounds from a mixture of metabolites that demonstrated activity against Candida albicans. The Plackett-Burman screening stage revealed that the culture conditions had a significant impact on the production of active chemicals. Similar observation was found during the optimization of anti-Klebsiella compounds production by Streptomyces sp. 2A 37 . Similar trend was also observed by Streptomyces albus AN1 and J1074 38 . The coherence occurred between Streptomyces sp. 2A and Streptomyces sp. HC14 in such a way that certain Plackett-Burman trials resulted in the entire absence of activity, indicating that the growing conditions are extremely important. From main effect calculation and analysis, malt extract was crucial component in the synthesis of anti-Candida metabolites, this was agreed with results the production of anti-Agrobacterium tumefaciens from Streptomyces sp. TN71 39 . With respect to concentration, the best concentration of malt extract was 5 g/L in the current study. The same concentration was reported for Streptomyces sp. TN71 39 . Smaoui et al., on the other hand, reported that a low concentration of glucose (2 g/L) was advantageous to the antimicrobial activity, whereas the highest concentration of glucose was helpful in the current study. Glucose and malt extract, together with other medium components, were crucial for enhancing the anti-fouling agent produced by Streptomyces sampsonii PM33 40 . Furthermore, malt extract was particularly beneficial in boosting up to 10.5-fold the Ilamycin-E1/E2 from Streptomyces atratus SCSIO ZH16-ilaR mutant, which is active against Mycobacterium TB 41 . Subsequently, the GC-MS analysis identified the composition of the metabolite, which, when combined with molecular docking, offered some new insight on the possible function of such molecules. Among the identified compounds, the concentration of di-n-octyl phthalate was the highest. It also resulted in high docking and MM-GBSA scores, which support the argument to be the main source of the activity of Streptomyces sp. HC14 extracts against Candida albicans. The activity is supported by the results reported by Shafeian et al. 42 , the compound di-n-octyl phthalate was purified from the marine sponge Haliclona (Soestella) caerulea and proved to be effective against the Candida albicans. Additionally, it was also isolated from sources for instance, Streptomyces melanosporofaciens and Streptomyces albidoflavus, Penicillium skrjabinii, and Penicillium olsonii 43 . It has been also identified www.nature.com/scientificreports/ in plant extract of Sisymbrium irio and exhibited antifungal activity 44 . This demonstrates the distribution of this molecule and demonstrates its promise as an antifungal drug, as well as supporting its activity and explaining the high affinity scores provided by cheminformatics. The second potential active compound with high docking results was [1,2-a]pyrazine-1,4-dione, hexahydro-3-(phenylmethyl) which might have a role in the anti-Candida activity. The strong docking score supports it as a source of action and is consistent with prior studies, as it was isolated from Streptomyces sp. VITPK9 demonstrated anti-Candida activity with low MIC values. The MIC values ranged from 0.78 to 1.6 µg/mL against Candida krusei MTCC9215, Candida tropicalis MTCC184, and C. albicans MTCC227 45 . This backs with previous research on its antimicrobial/anti-Candida capabilities 45,46 .
The molecular backbone is made up of diketopiperazine and pyrrole. The diketopiperazine and its derivatives were found to have antifungal properties, such as cyclo(L-Ile-L-Pro) and cyclo(Gly-Leu), which were isolated from Gram-negative bacteria such as Pseudomonas aeruginosa and Lactobacillus plantarum, respectively 47 . Furthermore, pyrrole has been discovered to be a high potential chemical with intriguing applications that is efficacious against multidrug resistant bacteria 46 . In other words, it has maintained the antifungal activity as its precursors.  www.nature.com/scientificreports/ In conclusion, the production of anti-Candida metabolites from Streptomyces HC14 is mainly associated with high concentrations of malt extract and glucose. GC-MS study combined with computational biology techniques has identified two putative metabolites as source of activity against Candida albicans: pyrrolo[1,2-a] pyrazine-1,4-dione, hexahydro-3-(phenylmethyl), and di-n-octyl phthalate. This expands our understanding of the range of antifungal compounds derived from Streptomyces and can serve as a foundation for further research into ligand-target interactions in Candida albicans.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/