Microbial strains

Four bacterial strains; Staphylococcus aureus ATCC 6538, Escherichia coli K12, Bacillus subtilis ATCC 23,857, and Pseudomonas aeruginosa PAO1 and two fungal strains; Candida albicans ATCC 10,231 and Aspergillus niger ATCC 32,656 were supplied by the Department of Immunology and Microbiology, faculty of pharmacy, Cairo university. The bacterial strains were cultivated on nutrient agar at 37 °C for 24 h., while the fungal strains were cultivated on Sabouraud dextrose agar at 30 °C for 48 h.

Culture media and substrate

Mueller-Hinton broth for bacterial culture was purchased from Titan Biotech LTD, Rajasthan, India. Sabouraud Dextrose broth for fungal culture was purchased from Himedia leading BioSciences company, Mumbai, Maharashtra, India. Standard quercetin was purchased from E. Merck Co., Darmstadt, Germany.

General biotransformation procedure

Biotransformation was executed using a standard two-stage protocol with some modifications (Cichewicz and Kouzi 1998). Stage I: Bacterial cultures were prepared by suspending the cells from agar slant in 50 ml of Mueller-Hinton broth at 250-ml Erlenmeyer flask and incubated for 24 h. on a shaker incubator at 180 rpm and 37 °C. Fungal cultures were prepared similarly using 50 ml of Sabouraud Dextrose broth but incubated in a shaker incubator at 180 rpm and 30 °C. Stage II: Cultures were initiated by inoculating 40 ml of fresh, sterile broth with 10 ml of stage I culture broth and incubated in a shaker for 24 h. Quercetin (25 mg dissolved in 2.5 ml DMSO) was added to each broth culture to produce final concentration of 0.5 mg/ml. Two flasks were prepared for each microbe; one was incubated in a shaker incubator and the other was incubated in a regular incubator (without shaking).

Fermentation was allowed for 24 h. Samples (0.5 ml) were aseptically withdrawn, extracted with ethyl acetate (0.5 ml), mixed for 2 min using a vortex, and centrifuged (13,000 rpm) for 5 min. For each sample, the upper layer was investigated for possible biotransformation using thin layer chromatography (TLC). Pre-coated silica TLC plates 60 F254 (20 × 20 cm, 0.25 mm thickness, Macherey Nagel, Germany) were used as stationary phase, while the mobile phase was chloroform: ethyl acetate: formic acid (60:40:0.01 v/v). Spots were examined in visible and UV light. The disappearance of the parent compound spot and the appearance of new spot/s with different Rf value/s were considered positive sign of biotransformation. Positive/substrate controls prepared from sterile broth were incubated with quercetin without microbes. Negative/culture controls were prepared of broth in which microorganisms were grown under identical conditions without quercetin addition. The flasks which showed positive signs of biotransformation were centrifuged (6000 rpm) for 5 min to separate the bacteria from the broth. Upon lyophilization of 25 ml of the broth, 308 mg dried biotramsformed products (BPs) were obtained. BPs were stored under refrigeration (4°C) until further testing.

LC-ESI-TOF-MS/MS analysis of the biotransformed productsSample preparation

A stock solution of BPs was prepared by dissolving 50 mg in 1 ml of a reconstitution solvent composed of water, methanol, and acetonitrile in a ratio of 2:1:1 v/v. Complete solubilization of the stock solution was achieved by subjecting it to vortex for 2 min and ultrasonication at 30 kHz for 10 min. The stock solution was then centrifuged for another 10 min at 10,000 rpm. An aliquot, 50 µl of stock solution was diluted to 1000 µl by reconstitution solvent to prepare a final concentration of 2.5 µg/µl, and 10 µl was used for injection in negative mode. A blank sample was also injected using 10 µl of reconstitution solvent.

Instruments and acquisition method

Separating small molecules was executed on an Exion LC system (AB Sciex, USA) linked to an autosampler system, a filter disks pre-column (3.0 mm × 0.5 μm, Phenomenex, USA), and an X select HSS T3 (2.1 × 150 mm, 2.5 μm) column (Waters Corporation, Milford, MA, USA) kept at 40 °C. The applied flow rate was 300 µl/min. The mobile phase was composed of 100% acetonitrile (A) and 5 mM ammonium formate in 1% methanol. The pH was adjusted to 3 using formic acid to form solution (B) for positive mode or adjusted to pH 8 using ammonium hydroxide to form solution (C) for negative mode.

The mass spectrometry (MS) was carried out on a Triple TOF 5600 + system provided with a Duo-Spray source working in the ESI mode (AB SCIEX, Concord, ON, Canada). The Triple TOF system was operated using an information-dependent acquisition (IDA) protocol, which was used to assemble full-scan MS and MS/MS information. Thereafter, the top 15 intense ions were chosen for MS/MS fragmentation spectra acquisition after each scan (Hegazy et al. 2021).

LC-MS data processing

Non-targeting, small molecule inclusive analysis of the sample was performed using MS-DIAL 3.70 open-source software (Tsugawa et al. 2015). GNP negative (2351 records) databases were utilized as reference databases. The MS-DIAL output was utilized to run on MasterView which was used for feature (peaks) extraction from total ion chromatogram based on standard criteria (Mohammed et al. 2021).

Evaluation of the cytotoxicity of the biotransformed productsCell lines, cell culture, and substrate preparation

Adeno-carcinomic human alveolar basal epithelial cells: A549 cell line and adeno-carcinomic human colorectal epithelial cells: Caco-2 cell line were purchased from the American Type Cell Culture (ATCC, USA), while normal human primary skin fibroblast: hFB cell line was purchased from Rio de Janeiro Cell Bank (BCRJ, Brazil). All cells were cultured in Dulbecco’s Modified Eagle (DMEM) high glucose medium which comprised 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1% antibiotic-antimycotic cocktail. All were purchased from Biowest (Nuaillé, France). Cells were kept in humidified air having 5% CO2 at sub-confluency at 37 °C. For sub-culturing step, monolayer cells were treated with trypsin/EDTA then harvested at 37 °C. Cells were used upon reaching 75% confluency. The substrate (S) was prepared by mixing 0.5 mg of quercetin with 21.5 mg of Mueller Hinton broth.

Cytotoxic assay (MTT)

The MTT colorimetric assay was conducted based on the original procedure proposed by Hansen et al. (Hansen et al. 1989). The MTT (3-[4,5-dimethylthiazole-2-yl]-2,5-diphenyltetrazolium bromide) was purchased from Merck KGaA (Darmstadt, Germany). It was utilized to evaluate the examined samples’ cytotoxicity. The assay is based on the ability of the living cells’ active mitochondrial dehydrogenase enzyme to cleave the yellow MTT tetrazolium rings forming insoluble dark blue formazan crystals. The crystals’ solubilization leads to the dark blue color formation directly proportional to the live cells’ number. In brief, cells (1 × 104 cells/well) in serum-free media were seeded in a 96-well flat bottom microplate and treated with 20 µl of different concentrations of the examined samples for 72 h. in a humidified air with 5% CO2 at 37ºC. The samples’ final concentration range was 3.125 to 1000 µg/ml. At the end of the incubation period, media were discarded. MTT solution (40 µl) was added for each well and incubated for 4 hr. The MTT crystals solubilization was executed by adding acidified isopropanol 180 µl/ well. The plates were shaken at room temperature. The photometric absorbance was determined at 570 nm using a microplate ELISA reader (FLUOstar OPTIMA, BMG LABTECH GmbH, Ortenberg, Germany). Each concentration was repeated twice and the average was calculated. The formula used to calculate the cytotoxicity was as follows:

$$\begin}=\\ \frac}} \times 100\end$$

Docking of quercetin and A7P to the binding sites of PI3Kδ and Cat B

The crystal structure of PI3Kδ (PDB ID: 8BCY, PDB DOI: https://doi.org/10.2210/pdb8B4T/pdb) and CatB (PDB ID: 8B4T, PDB DOI: https://doi.org/10.2210/pdb8BCY/pdb) were downloaded from Protein Data Bank (https://www.rcsb.org). 8BCY consists of two chains: chain A (Phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit delta isoform, Gene Names: PIK3CD) and chain B (Phosphatidylinositol 3-kinase regulatory subunit alpha, Gene Names: PIK3R1) (Mazzucato et al. 2023). 8B4T consists of one chain (Cathepsin B, Gene Names: CTSB, CPSB) (Rubesova et al. 2023). All the molecular modeling studies were carried out by Molecular Operating Environment (MOE, 2019.0102) software. The proteins were firstly prepared by eliminating the crystallographic water molecules. Only the chain that is co-crystallized with the ligand was kept. The selected protein chain was refined to the root-mean-square (RMS) gradient of 0.001 kcal/mol /A2 and energy minimized to 10− 5 kcal/mol /A. The active sites of the target proteins were identified for docking study using pre-co-crystallized ligands; 9-[2-(3,4-dichlorophenyl) ethyl]-2-(3-hydroxyphenyl)-8-oxidanylidene-7~-purine-6-carboxamide C20H15Cl2N5O3 (DHOPC) for PI3Kδ and (2S)-2-[[(2S)-2-[(4-chloranylphenoxy) carbonyl amino]-3-cyclohexyl-propanoyl] amino]-3-phenyl-propanoic acid C25H29ClN2O5 (PPA) for CatB. The grid co-ordinates of the interacting amino acids with the pre-co-crystallized ligands were 13.197 (X), -1.856 (Y), and 17.052 (Z) for 8BCY, and 12.335 (X), -0.512 (Y), and − 3.466 (Z) for 8B4T.

The 2D structures of the parent compound (quercetin) and the main biotransformed product (A7P) were drawn using ChemDraw professional 17.0 software. The structures were copied as smiles and pasted on MOE interface as 3D structures. The proposed compounds were protonated. Their energies were minimized using RMS gradient 0.1 Kcal/mol /A2 to obtain the most stable conformers. For each docking run, 10 poses were generated using the Triangle Matcher placement method and London dG scoring function. The docking results were visualized using Biovia Discovery Studio Visualizer 2024. The pose with ideal binding mode for each compound was used in portending the ligand–enzyme interactions at the active site. Analysis of the docking results was performed by comparing the interactions and docking scores of the docked compounds with that of the pre-co-crystallized ligand.

Statistics

Data is expressed as mean ± SD. Non-linear regression analysis was used in calculating the dose-response curve with 5 parametric logistic curve equations using GraphPad Prism 10.2.2 (341) (San Diego, California USA, www.graphpad.com).