Seven strains – Escherichia coli, Streptococcus lutetiensis, Enterococcus gilvus, Clostridium perfringens, Bacteroides fragilis, Lactobacillus acidophilus, and Weissella confuse – were identified and isolated from human feces in our laboratory (11). Two kinds of media, MRS and Brain Heart Infusion (BHI) broths, were purchased from Qingdao Hope Bio-technology Co. and utilized to culture these strains. MRS broth for L. acidophilus and W. confuse; BHI for E. coli, E. gilvus, S. lutetiensis, C. perfringens, and B. fragilis.

Quercetin (Qiyun Biological Science & Technology Limited Co., China) was first dissolved with DMSO, and then 50 mg/mL stock solution was prepared after filtration with a 0.22-µm nylon microporous membrane. The solution was stored at 4°C before use. Before those strains were individually inoculated into their broth media, the quercetin solution was supplemented to a final concentration of 0.2 mg/mL in the broths. After inoculation, the BHI and MRS broths were cultured at 37°C in anaerobic conditions for 24 h.

The above fermented solutions were individually centrifuged at 5,000 rpm for 20 min, and the supernatant solutions were further extracted with an equal volume of ethanol. The extracted solutions were then distilled under reduced pressure at 55°C to remove ethanol and were condensed to five-fold. The remains were filtrated with a 0.45-µm microporous membrane, and the mixture solution of catabolites (MSCs) were individually collected and stored at −20°C before being used.

After MSCs were thawed in a 37°C water bath, 20 µL of sample was pipetted by automatic sampler and analyzed with high-performance liquid chromatography (HPLC). The compounds were separated on an analytical Welch Ultimate XB–C18 column (4.6×250 mm i.d., 5 µm; Welch, Shanghai, China) using a binary gradient delivered by a LC-20AT low-pressure gradient HPLC pump (SHIMADZU, Japan) with acetonitrile as solvent A and 0.1% v/v formic acid water solution as solvent B for the detection of quercetin and its catabolites: 0.01 min, 5% A; 10 min, 10% A; 50 min, 50% A; 65 min, 90% A; 70 min, 90% A; 73 min, 5% A. The flow rate was 0.6 mL/min. Data acquisition was carried out with the software Classic-VP Chromatography Workstation (SHIMADZU, Japan). A diode array detection SPD-M20A plus was applied over the wavelength range of 220–600 nm for peak detection. Compounds were identified by comparison of the UV spectra and the retention times. The following wavelengths were monitored for quantitative analysis: 360 nm for quercetin, 280 nm for metabolites (12, 13).

The fermentation products separated by the HPLC were used for the further experiment. Those components at new peaks were collected and analyzed by mass spectrometry (MS) (Waters, Milford, MA, USA), and the chromatographic scans of selected multiple reaction monitoring ions (MRM) were used for further identification.

Initially, the broths of each bacterium with 10, 20, 30, 40, and 50 mg/mL quercetin were individually fermented for 24 h. BHI and MRS broths with quercetins but without incubating bacteria were used as controls. Then MSCs were prepared following the procedure described above.

The DPPH radical scavenging activities of all the samples of the fermentations were evaluated by the method of Blois (14) with minor modifications. Twenty microliter of each MSC was added into 10 mL colorimetric tubes, made up with 2 mL of distilled water, and then mixed with 2 mL of 0.2 mM DPPH dissolved in methanol. The reaction mixture was incubated for 30 min at 28°C in the dark. The blank sample contained all reagents without MSC. The DPPH radical scavenging activity was determined by measuring the absorbance at 517 nm using a spectrophotometer. The DPPH free radical scavenging activity (%) was calculated using the following formula (1). The test was paralleled three times. 1 Percentage inhibition = [ ( A 0 - A 1 ) / A 0 ] × 100

where A₀ is the absorbance of the control and A₁ is the absorbance in the presence of MSCs.

Hydroxyl radical, generated from the Fe^{3 +}–ascorbate–H₂O₂ (Fenton reaction), was evaluated by degradation of deoxyribose that produced thiobarbituric acid reactive species (TBARS) (15). The reaction mixture contained 25 mmol/L deoxyribose, 10 mmol/L ferric chloride, 100 mmol/L ascorbic acid, and 2.8 mmol/L H₂O₂ in 10 mmol/L KH₂PO₄ (pH 7.4), and it was then mixed with of 0.4 mL water solutions, including 20 µL of each MSC. The blank sample contained all reagents without MSC. The reaction mixture was incubated at 37°C for 1 h. Subsequently, 1 mL of 1% thiobarbituric acid and 1 mL of 3% trichloroacetic acid were added, and the mixture was heated at 100°C for 20 min. TBARS was measured spectrophotometrically by taking absorbance at 532 nm. The results were expressed as percentage inhibition of deoxyribose oxidation, as determined by the following formula (2). The test was paralleled three times. 2 Percentage inhibition = [ ( A - B ) / A ] × 100

where A is the malondialdehyde produced by Fenton reaction treated alone, and B is the malondialdehyde produced in the presence of MSCs.

Each sample was analyzed in triplicate, and data are presented as mean values±standard error (n=3). Multivariate correlation analysis of appropriate data was performed according to Pearson's correlation.

It is difficult to separate all the metabolites of bacteria by HPLC because of numerous substances in fermented broths. In this experiment, the metabolites of seven strains transforming quercetin were separated with HPLC. Only one compound was discovered, and it presented among the metabolites of both C. perfringens and B. fragilis at the retention time of 25.024 min in the charts (Fig. 1A and B). The retention time of quercetin was about 49.013 min. Its peak appeared in the fermented solution of E. gilvus, E. coli, S. lutetiensis, W. confuse, and L. acidophilus, but not of C. perfringens and B. fragilis.

The component at the retention time of 25.024 min was collected for MS. Two significant molecular ion peaks of 167 and 123 m/z were observed (Fig. 2) and deduced as 3,4-dihydroxyphenylacetic acid, one main metabolite of gut bacterial fermenting quercetin, according to its possible fractures (16).

The quercetin in broths removed DPPH free radicals depending on its concentration (Fig. 3). About 70% DPPH was removed by 1 mg/mL quercetin. Compared to quercetin, MSCs of seven bacterial species scavenged less DPPH free radicals (Fig. 3). Among MSCs of all strains, that of L. acidophilus can remove the least amount of DPPH free radicals in a series of doses of quercetin added in broths. The MSCs of W. confuse and E. gilvus, scavenged more DPPH free radicals than that of L. acidophilus, but less than the others when 0.4 mg/mL quercetin or more was added in broths. The MSCs of the other two strains, that is, W. confuse and E. gilvus, scavenged more DPPH free radicals than that of L. acidophilus but less than the others when more than 0.4 mg/mL quercetin was added in broths. Their removal rates of DPPH by MSCs reached the highest value when the concentration of quercetin was 1 mg/mL, the highest concentration of quercetin in broth. The abilities of scavenging free radicals were improved along with more quercetin added in broths.

The hydroxyl group free radical scavenging activity of quercetin was also related to its concentration (Fig. 4). The removal rate of OH free radicals reached almost 100% when the concentration of quercetin was 1 mg/mL. The MSCs of B. fragilis presented the strongest activity in scavenging OH free radicals in all concentrations of quercetin added in broths. When 1 mg/mL quercetin was added in both, the MSCs of B. fragilis almost removed the same amount of OH free radicals with quercetin (100%) and 80% in the case of C. perfringens. The OH free radicals scavenging activities of MSCs from S. lutetiensis and B. gilvus presented the weakest with only about 30% DPPH removed when more than 0.6 mg/mL quercetin was added in broth. The activities of all strains increased along with more quercetin being added.