Houttuynia cordata Thunb. (HC), a flowering perennial herb from the Saururaceae family native to Southeast Asia, is known for its distinctive fishy smell and heart-shaped leaf. Hence, its common names include fish mint or heartleaf. This plant thrives in shady, moist soil and warm environments at altitudes between 300 m and 2,600 m (Ghosh et al., 2022). HC leaves, roots, and stems are traditionally used as functional foods in salads, soups, and herbal teas. Besides its culinary uses, HC has been used as a medicinal plant to clear heat and detoxicate, induce diuresis, relieve strangury, disperse abscesses, and expel pus. Studies have shown that HC has many beneficial effects, including anti-inflammation, immunomodulation, antioxidation, lung protection, anticancer, antiviral, and antibacterial properties (Rafiq et al., 2022). The pharmacological activities of HC are influenced by the presence of numerous functional compounds, including alkaloids, benzenoids, flavonoids, phenols, and volatile oils (Pradhan et al., 2023).

Fermentation is a metabolic process, wherein enzymes from microorganisms (e.g., bacteria and yeast) chemically break down organic materials. Some studies have proposed that fermenting Chinese herbal medicine (CHM) with probiotics offers several benefits. For example, fermentation can promote the release of active compounds, increase biological activities, reduce toxicities and side effects, produce new functional compounds, and enhance the bioavailability of CHM (Zhang et al., 2023). Lactobacillus is the most commonly used probiotic genus in CHM fermentation (Zhang et al., 2023). Lactiplantibacillus plantarum (formerly known as Lactobacillus plantarum) has been used to ferment Rhizoma Atractylodis macrocephalae and Panax ginseng, and the fermented product can alleviate antibiotic-associated diarrhea and high-fat-diet-induced obesity in rats by regulating intestinal microbiota (Wang et al., 2015; Qu et al., 2021). In addition, L. plantarum has been shown to improve the nutritional quality and flavor properties and increase the antioxidant ability and antibacterial effects in the food industry (Yilmaz et al., 2022).

This study aimed to expand knowledge through the following objectives: first, to compare the flavonoid contents (hyperoside, quercitrin, and quercetin) across different sources, medicinal parts (leaves and stems), and harvesting seasons (summer and winter) of HC raw materials; second, to compare the total polyphenol and flavonoid contents and antioxidant activity in boiling water extracts of HC (BWEHC) leaves and stems harvested in summer and winter; and third, to determine the optimal fermentation conditions using L. plantarum as the candidate strain, with varying solid-liquid ratios and fermentation times as parameters. The post-fermentation bacterial count served as the response factor for optimization. Finally, we compared the content of functional compounds and antioxidant activity before and after fermentation.

Collection and preparation of lyophilized powder of BWEHC leaves and stems

The aerial parts of HC harvested in winter were purchased from four physical herb shops, whereas samples harvested in summer were purchased from only one herb shop (Taichung, Taiwan). All HC samples were authenticated by the testing center of Ko Da Pharmaceutical Co., Ltd. The leaves and stems were separated by hand, dried in an oven at 60°C overnight, ground into powder, sieved by a 40-mesh sifting screen, and stored at room temperature until use. The dried HC leaves and stems were extracted using boiling water at a ratio of 1:15 (w/v) for 1 h. After extraction, the liquid fractions were collected by passing through an 80-mesh sifting screen. Then, the residues were assembled followed by another extraction at a ratio of 1:10 (w/v) for 1 h. Subsequently, all filtrates were collected, subjected to a vacuum concentrator, and lyophilized using a freeze dryer to obtain lyophilized powder.

Determination of quercitrin, hyperoside, and quercetin contents by high-performance liquid chromatography

To determine the quercitrin, hyperoside, and quercetin contents in various HC samples, high-performance liquid chromatography (HPLC) was performed on the Hitachi L-2000 system equipped with an L-2455 diode array detector, an L-2130 pump, and an L-2200 autosampler (Hitachi). The standard solution was prepared by mixing quercitrin (purity >98%, ChemFaces), hyperoside (purity >98%, ChemFaces), and quercetin (purity >98%, Sigma Chemical Co.) with methanol to obtain concentrations of 20, 50, and 50 µg/mL, respectively. The test solution was prepared by mixing 0.25 g of the HC sample powder with 20 mL of 80% (v/v) methanol (Sigma Chemical Co.) under ultrasonication at room temperature for 1 h. After filtration, 80% (v/v) methanol was added to a final volume of 25 mL. Chromatographic separation was performed on a Mightysil RP-18 column (250×4.6 mm, 5 µm) using a gradient solvent system comprising acetonitrile (A) and 0.2% (v/v) acetic acid (B). The gradient profile was set as follows: 20% A from 0 to 30 min and 20%－ 50% A from 30 min to 50 min. The ultraviolet (UV) wavelength, flow rate, and injection volume were 254 nm, 1.0 mL/min, and 10 µL, respectively. The following formula was used to calculate the quercitrin, hyperoside, and quercetin contents. Standard concentration (µg/mL)×(peak area of sample÷peak area of standard)×final volume (mL)÷sample weight÷1,000.

Establishment of L. plantarum growth curve

L. plantarum (BCRC 12251), isolated from sauerkraut, was purchased from the Bioresource Collection and Research Center. L. plantarum was activated by adding 200 µL of bacterial suspension into a glass tube containing 4 mL of MRS broth (Sigma Chemical Co.) and then incubated for 24 h at 37°C. Activated L. plantarum was proliferated by adding 200 µL of the revived bacterial suspension into a glass tube containing 4 mL of MRS broth and incubating for 24 h at 37°C. The optical density value of the proliferated bacterial suspension was adjusted to 1.0 at a wavelength of 600 nm using an ultraviolet/visible (UV/VIS) spectrophotometer (Jasco, V-530). Through two-fold serial dilution, 1 mL of each diluted bacterial suspension was placed into a 9-cm dish and added with 25 mL of MRS agar (Sigma Chemical Co.) using the pour plate method. After incubation for 72 h at 37°C, the number of colonies was counted, and the correlation coefficient between the bacterial counts and the absorbance was calculated.

Fermentation

Table 1 shows the composition of the fermentation substrate. Yeast extract was used as the nitrogen source, and the minimal value for L. plantarum growth was 0.3% (w/v). The content of reducing sugar [glucose equivalent (GE)] in BWEHC leaves from herbal shop C harvested in summer was 0.76±0.08 mg/mg (data not shown). Through calculation, the carbon source in 200 mL of the fermentation substrate containing 1.5% (w/v) BWEHC leaves accounted for 1.14% (2.28 g). Therefore, the carbon sources in all groups, including 0% (w/v), 0.3% (w/v), and 0.9% (w/v), were adjusted to 1.5% (w/v). All prepared fermentation substrates were heated at 80°C for 10 min, and L. plantarum was inoculated at 1×10³ colony-forming units (CFU)/mL and further incubated at 37°C for 24, 48, and 72 h. The number of bacteria was counted using the pour plate method before and after fermentation.

Table 1 . Composition of the fermentation substrate.

Solid (HC)-liquid ratio	Nitrogen source		Carbon source		Final volume (mL)
	Yeast extract (g)		Reducing sugar (g)	Glucose (g)
0% (w/v)	0.6		0.000	2.280	200
0.3% (w/v)	0.6		0.456	1.824	200
0.9% (w/v)	0.6		1.368	0.912	200
1.5% (w/v)	0.6		2.280	0.000	200

HC, Houttuynia cordata Thunb..

Determination of the total polyphenol content

The colorimetric method was used to determine the total polyphenol content in accordance with the method of Singleton and Rossi (1965) with slight modifications. As the standard phenolic compound, gallic acid (Sigma Chemical Co.) was dissolved in distilled water to form 1 mg/mL of the stock solution. Different serial concentrations of gallic acid (25－200 µg/mL) were used to create the standard curve for calculating the total polyphenol content through serial dilution. Briefly, 0.5 mL of gallic acid or sample solution was introduced into glass tubes and then added with 2.5 mL of 10% (v/v) Folin-Ciocalteu reagent (Sigma Chemical Co.) and 2.5 mL of 7.5% (w/v) sodium bicarbonate (Sigma Chemical Co.). The mixture in the glass tubes was homogenized on a vortex mixer, and the absorbance was read at 720 nm against the blank (distilled water) using a UV/VIS spectrophotometer. The total polyphenol content was expressed as mg gallic acid equivalent (GAE)/g.

Determination of the total flavonoid content

The colorimetric method was used to determine the total flavonoid content in accordance with the method of Fattahi et al. (2014) with slight modifications. As the standard flavonoid compound, quercetin was dissolved in dimethyl sulfoxide (DMSO, Sigma Chemical Co.) to form 1 mg/mL of the stock solution. Different serial concentrations of quercetin (25－200 µg/mL) were used to create the standard curve for calculating the total flavonoid content through serial dilution. Briefly, 1.25 mL of distilled water and 0.075 mL of 5% (w/v) sodium nitrite (Sigma Chemical Co.) were placed into glass tubes and then added with 0.25 mL of quercetin or sample solution. The mixture was allowed to react for 10 min under a dark condition. Subsequently, the mixture was added with 0.15 mL of 10% (w/v) aluminum chloride (Sigma Chemical Co.), 0.5 mL of 1 M sodium hydroxide (Sigma Chemical Co.), and 0.225 mL of distilled water. The mixture in the glass tubes was mixed evenly, and the absorbance was measured at 380 nm against the blank (DMSO) using a UV/VIS spectrophotometer. The total flavonoid content was expressed as mg quercetin equivalent (QE)/g.

Determination of the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging ability

The colorimetric method was used to evaluate the free radical scavenging ability against DPPH in accordance with the method of Dibacto et al. (2021) with slight modifications. DPPH (Sigma Chemical Co.) was dissolved in methanol to form a 0.2 mM working solution. Different serial concentrations (1－10 µg/mL) of ascorbic acid (Sigma Chemical Co.), a well-known antioxidant, were used as a positive standard. About 0.5 mL of ascorbic acid, sample solution, or distilled water (control) was introduced into a colorimetric tube and then added with 0.5 mL of 0.2 mM DPPH. The mixture was allowed to react for 30 min under a dark condition. Subsequently, the mixture in the colorimetric tubes was mixed evenly, and the absorbance was measured at 517 nm against the blank (methanol) using a UV/VIS spectrophotometer. The percentage of DPPH radical scavenging ability of samples was calculated as follows: [(OD 517 nm of control−OD 517 nm of the sample)/ OD 517 nm of control]×100.

Trolox equivalent antioxidant capacity assay

The colorimetric method was used to determine the Trolox equivalent antioxidant capacity (TEAC) against 2,2’-azinobis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS⁺) in accordance with the method of Re et al. (1999) with slight modifications. Different serial concentrations (20－100 µM) of Trolox (Sigma Chemical Co.), a water-soluble analog of vitamin E and a well-known antioxidant, were used as a positive standard. About 1.5 mL of distilled water, 0.25 mL of 1 mM ABTS (Sigma Chemical Co.), 0.25 mL of 500 µM hydrogen peroxide (Sigma Chemical Co.), and 0.25 mL of 44 units/mL peroxidase (Sigma Chemical Co.) were placed into a glass tube. The mixture was allowed to stand for 1 h under a dark condition. Then, the mixture was added with 0.25 mL of Trolox, sample solution, or distilled water (control) and allowed to react for 10 min under a dark condition. Subsequently, the mixture in the glass tubes was mixed evenly, and the absorbance was measured at 734 nm against the blank (DMSO) using a UV/VIS spectrophotometer. The percentage of TEAC of samples was calculated as follows: [(OD 734 nm of control−OD 734 nm of the sample)/OD734 nm of control]×100.

Determination of the reducing sugar content

The 3,5-dinitrosalicylic acid (DNS) method was used to measure the reducing sugar content in accordance with the method of Khatri and Chhetri (2020) with slight modifications. The DNS reagent was prepared by dissolving 1 g of DNS (Sigma Chemical Co.), 30 g of potassium sodium tartrate (Sigma Chemical Co.), and 80 mL of 2 N sodium hydroxide through heating in a water bath. After dissolution, the mixture was cooled to room temperature and added with distilled water to a final volume of 100 mL. As the standard compound, glucose (Sigma Chemical Co.) was dissolved in distilled water to form a 10 mg/mL stock solution. Different serial concentrations of glucose (0.2－1.0 mg/mL) were used to establish the standard curve for calculating the reducing sugar content. One milliliter of glucose or sample solution was placed into glass tubes and added with 1 mL of DNS reagent. The mixture was allowed to react under heating in a water bath for 10 min. After cooling, the mixture was added with 1 mL of distilled water and then mixed evenly. Subsequently, the absorbance was read at 540 nm against the blank (distilled water) using a spectrophotometer. The reducing sugar content was expressed as mg GE/g.

Statistical analysis

Data were expressed as the mean±standard deviation from three independent experiments. The independent sample t-test was used to compare two groups, and two-tailed significance at P<0.05 was considered significant. One-way analysis of variance followed by Duncan’s multiple range test was conducted for pairwise comparisons of group means using SPSS Statistics software (version 22.0.0, IBM Corp.). Statistical significance was considered at P<0.05.

Effects of different sources and medicinal parts on the specific flavonoid content in HC raw materials

Consumers can purchase HC samples through retail channels, including physical herb stores and e-commerce platforms. However, these channels need to be certified by competent authorities as the quality of CHM is often questionable. The present study evaluated the functional compounds of HC from various herbal shops, reflecting real purchasing practices in Taiwan, where herb quality can be uncertain. Nguyen et al. (2023) collected 32 HC leaf samples from 32 diverse provinces in Vietnam and found differences in the quercitrin, hyperoside, and rutin contents across distinct places of origin.

The quantitative results from the four physical herbal shops revealed that hyperoside and quercitrin were the predominant flavonoids in the raw material of HC leaves and stems (Fig. 1). The leaves and stems from herbal shop C had the highest hyperoside, quercitrin, and quercetin contents. The hyperoside and quercitrin contents in the leaves were much higher than those in the stems, with a difference of about 14- to 40-fold. Quercetin could only be detected in the leaves collected from various sources, not the stems. A significant variation was observed in the specific flavonoid content from the four herbal shops, possibly because of their places of origin, environmental factors, and storage conditions. To the best of our knowledge, this study is the first to report that the specific flavonoid content varied in the raw materials of HC leaves and stems.

Figure 1. Hyperoside, quercitrin, and quercetin contents in the raw materials of Houttuynia cordata Thunb. leaves and stems harvested in winter from four herbal shops. Values, expressed as the mean±SD from three independent experiments with different lowercase letters in the specific compound, are significantly different (P<0.05).

The quercitrin content is one of the indices for quality control proposed by the Taiwan Herbal Pharmacopeia (THP), which should not be less than 0.2% in HC. The proportion of leaves and stems in the dried aerial part accounted for 58% and 42% of HC, respectively (data not shown). Using this proportion, the calculated quercitrin content in the dried aerial parts of HC samples was 0.35%, 0.31%, 0.48%, and 0.21% for the four herbal shops, A, B, C, and D, respectively (data not shown), which met the specifications set by THP. Herein, we selected HC from herbal shop C for the following experiments.

Effects of different harvesting seasons and medicinal parts on the specific flavonoid content in HC raw materials

According to a monograph on HC in the Pharmacopeia of the People’s Republic of China, summer is the optimal season for harvesting dried HC when the stems and leaves are full of flowers and spikes. Moreover, the levels of specific volatile compounds, including (E)-3,7-dimethyl-2,6-octadien-1-ol and β-phellandrene, in the dried aerial parts and fresh whole grass of HC harvested in summer are significantly higher than those harvested in autumn (Pan et al., 2021).

The quantitative results revealed for the first time that the raw materials of HC leaves or stems harvested in summer had a significantly higher specific flavonoid content than those harvested in winter (Fig. 2), with a difference of about 7.5- to 11.6-fold. Similarly, a higher specific flavonoid content could be observed in the raw materials of HC leaves harvested in summer than those in the raw materials of HC stems, with a difference of about 1.7- to 3.6-fold.

Figure 2. Hyperoside, quercitrin, and quercetin contents in the raw materials of Houttuynia cordata Thunb. leaves and stems harvested in winter and summer from herbal shop C. Values are expressed as the mean±SD from three independent experiments. P<0.05 indicates a significant difference.

Turnover rate of hyperoside and quercitrin from the raw material to the BWEHC leaves

The decoction extracts the active compounds from CHM by boiling them in water for a specific time. CHM decoction is increasingly popular because of its health benefits and fewer regulatory restrictions (Zhang et al., 2023). Calculating the turnover rate from raw materials to boiling water extracts for a specific ingredient is crucial for determining the transfer capacity of the ingredient during decoction. The hyperoside and quercitrin contents in the raw materials of HC leaves and BWEHC leaves harvested in summer from herbal shop C were 5.90±0.06 and 13.5±0.2 mg/g and 13.6±0.8 and 28.8±1.9 mg/g, respectively (Table 2). The yield rate from the raw material to the boiling water extracts was 22%±2% (Table 2). Through calculation, the turnover rates for hyperoside and quercitrin were 50.7%±2.4% and 47.1%±2.8%, respectively. The turnover rate for a specific ingredient could provide a clue for Chinese medicine manufacturers to ensure quality consistency when manufacturing herbal products.

Table 2 . Hyperoside and quercitrin contents in raw materials and boiling water extracts of Houttuynia cordata Thunb. leaves and the turnover rate for specific compounds.

	Raw materials	Boiling water extracts	Turnover rate (%)1)
Hyperoside (mg/g)	5.90±0.06	13.6±0.8	50.7±2.4
Quercitrin (mg/g)	13.5±0.2	28.8±1.9	47.1±2.8

Values are presented as the mean±SD from three independent experiments..

¹⁾The turnover rate percentage was calculated as follows: [hyperoside or quercitrin content in boiling water extracts (mg/g)/hyperoside or quercitrin content in raw materials (mg/g)]×yield rate (%)×100, where the yield rate from raw material to boiling water extracts was 22%±2%..

Effects of harvesting seasons and medicinal parts on the total polyphenol and flavonoid contents and antioxidant activity in BWEHC

To determine the total polyphenol and flavonoid contents in the BWEHC leaves and stems harvested in summer and winter, a standard curve was constructed using different concentrations (25－200 µg/mL) of gallic acid and quercetin, along with their corresponding absorbance values. The linear regression coefficients (R²) for the total polyphenol and flavonoid contents were 0.9992 and 0.9981, respectively (data not shown), indicating the reliability of these analytical approaches.

To validate the efficacy of the DPPH and TEAC assays, we used different concentrations of vitamin C (1－10 µg/mL) and Trolox (20－100 µM). The results showed that the linear regression coefficients (R²) for the DPPH and TEAC assays were 0.9986 and 0.9984, respectively (data not shown). Through calculation, the IC₅₀ values of vitamin C and Trolox in the DPPH and TEAC assays were 6.11 µg/mL and 67.2 µM (16.8 µg/mL), respectively, indicating the efficacy of these assays.

The results showed that the BWEHC leaves exhibited a significantly higher total polyphenol and flavonoid contents and a considerably more potent antioxidant activity than the stems (Table 3). These results are consistent with those obtained from previous studies, indicating that the 95% (v/v) ethanol extracts of HC leaves contain the highest total polyphenol and flavonoid contents and exhibit the most potent antioxidant activity among the different medicinal parts of HC (Song et al., 2022). A possible explanation for the difference in the content of functional compounds, such as quercitrin, hyperoside, total polyphenol, and total flavonoid, between the leaves and stems is that flavonoids accumulate in the leaves and exhibit a protective effect against damage caused by UV radiation in plants (Tattini et al., 2000). By contrast, the stem serves as the organ for transporting water and nutrients and supporting plants’ overall structure. In addition, the BWEHC leaves and stems harvested in summer contained higher total polyphenol and flavonoid contents and exhibited more potent antioxidant activity than those harvested in winter.

Table 3 . Comparison of the total polyphenol and flavonoid contents and antioxidant activity in boiling water extracts of Houttuynia cordata Thunb. leaves and stems harvested in winter and summer from herbal shop C.

	Leaves			Stems
	Summer	Winter		Summer	Winter
Total polyphenol (mg GAE/g)	81±7a	41±6b		31±2c	14±2d
Total flavonoid (mg QE/g)	120±5a	46±3b		39±3b	17±1c
DPPH (IC50, µg/mL)	67±3d	82±3c		164±11b	268±20a
TEAC (IC50, µg/mL)	123±13d	197±17c		362±16b	643±25a

Values are presented as the mean±SD from three independent experiments with different lowercase letters in the same row, are significantly different (P<0.05)..

GAE, gallic acid equivalent; QE, quercetin equivalent; DPPH, 2,2-diphenyl-1-picrylhydrazyl; TEAC, Trolox equivalent antioxidant capacity..

Establishment of the optimal fermentation conditions

The basic procedure for liquid fermentation included inoculating live microorganisms into a fermented substrate containing CHM extracts, providing suitable nitrogen and carbon sources, and incubating under proper temperature and pH values (Zhang et al., 2023). Liquid fermentation requires highly controlled conditions to increase the conversion rate of the active compounds and decrease the pollution risk. Therefore, optimizing the liquid fermentation process is crucial for improving the stability of fermented products (Li et al., 2020).

Herein, we compared different solid-liquid ratios (0.3－ 1.5%, w/v) and fermentation times (24－72 h) to determine the optimal fermentation conditions for L. plantarum growth. First, we established the L. plantarum growth curve to assure the correctness of the inoculation amount using the absorbance at OD_{600 nm} and their corresponding bacterial count. The correlation coefficient (R²) was 0.9993 (data not shown). The bacterial count before fermentation in different solid-liquid ratios and fermentation times was 3.08±2.12 and 3.14±2.21 log CFU/mL, suggesting that the growth curve is predictable for L. plantarum inoculation (Table 4).

Table 4 . Effects of different solid-liquid ratios and fermentation times on Lactiplantibacillus plantarum count.

Solid (HC)-liquid ratios	Bacterial count (log CFU/mL)	Fermentation times	Bacterial count (log CFU/mL)
Initial	3.08±2.12a	Initial	3.14±2.21a
0.3% (w/v)	7.51±6.42b	24 h	7.88±6.66b
0.9% (w/v)	7.92±6.82d	48 h	8.21±6.78c
1.5% (w/v)	7.82±6.66c	72 h	7.84±6.52b

Values, expressed as the mean±SD from three independent experiments with different lowercase letters in the same column, are significantly different (P<0.05)..

HC, Houttuynia cordata Thunb.; CFU, colony-forming unit..

Specifically, the solid-liquid ratio at 0.9% (w/v) for 24 h of fermentation resulted in a 2.57- and 1.26-fold higher bacterial count than that at 0.3% (w/v) and 1.5% (w/v), respectively. Then, the solid-liquid ratio of 0.9% (w/v) was selected to determine the time effects on the bacterial count. The results showed that the bacterial count at 48 h of fermentation was 2.19- and 2.37-fold higher than that at 24 and 72 h of fermentation, respectively.

To further compare the growth-promoting effects of BWEHC leaves, the bacteria were counted in an optimal fermentation condition [solid-liquid ratio of 0.9% (w/v) and 48 h of fermentation] and compared with a control containing 0.3% (w/v) yeast extract and 1.14% (w/v) glucose. The results showed that the BWEHC leaves significantly increased the growth of L. plantarum by 16.2-fold (bacterial count: 8.23±6.58 log CFU/mL) compared with the control group (bacterial count: 7.02±6.45 log CFU/mL) (Fig. 3), indicating that BWEHC leaves could be a potential prebiotic for L. plantarum.

Figure 3. Effects of boiling water extracts of Houttuynia cordata Thunb. leaf on Lactiplantibacillus plantarum growth. The control group, comprising 0.3% (w/v) yeast extract, 1.14% (w/v) glucose, and 10³ colony-forming units (CFU)/mL L. plantarum inoculation, was incubated at 37°C for 48 h. The fermented group, comprising 0.3% (w/v) yeast extract, 1.14% (w/v) glucose (1.368 g of reducing sugar for boiling water extracts of H. cordata Thunb. leaf and 0.912 g of glucose in 200 mL of fermented substrate), and 10³ CFU/mL L. plantarum inoculation, was fermented at 37°C for 48 h. Values are expressed as the mean±SD from three independent experiments. P<0.01 indicates a significant difference.

Improvement of the total polyphenol and flavonoid contents and antioxidant activity after fermentation

L. plantarum is a probiotic specializing in food fermentation but has not been previously used to ferment HC. Zhang et al. (2023) proposed that the probiotic fermentation of CHM has the advantages of increasing the release of active compounds and improving the biological activities of CHM. Herein, we found a significant increase in the total polyphenol and flavonoid contents and antioxidant activity in the fermented products compared with the nonfermented substrates (Table 5). Similarly, Lee et al. (2018) demonstrated higher total polyphenol content and radical scavenging ability in HC fermented with Lactobacillus brevis B84 than in nonfermented HC. In addition, Li et al. (2020) found that certain CHMs, including Semen vaccariae and Leonurus artemisia, fermented with probiotics, such as Lactobacillus casei, Enterococcus faecalis, and Candida utilis, exhibited increased soluble total flavonoid, total alkaloid, crude polysaccharide, and total saponin contents in the fermented product than in nonfermented herbs. A possible explanation for these findings is that probiotics can generate hydrolytic enzymes to degrade plant cell walls and increase the release of active compounds, including flavonoids, glycosides, anthraquinones, terpenoids, alkaloids, and organic acids (Woldemariam Yohannes et al., 2020).

Table 5 . Comparison of the hyperoside, quercitrin, quercetin, and total polyphenol and flavonoid contents and antioxidant activity before and after fermentation.

	Before fermentation	After fermentation	P-value
Hyperoside (mg/g)	13.63±0.82	10.32±0.14	<0.05
Quercitrin (mg/g)	28.83±1.94	22.42±0.12	<0.05
Quercetin (mg/g)	1.22±0.04	1.38±0.08	<0.05
Total polyphenol (mg GAE/g)	81.42±7.02	111.22±15.32	<0.05
Total flavonoid (mg QE/g)	120.22±5.01	139.41±4.42	<0.05
DPPH (IC50, µg/mL)	67.42±3.02	53.22±7.41	<0.05
TEAC (IC50, µg/mL)	123.13±13.22	101.08±7.21	<0.05

Values are expressed as the mean±SD from three independent experiments..

GAE, gallic acid equivalent; QE, quercetin equivalent; DPPH, 2,2-diphenyl-1-picrylhydrazyl; TEAC, Trolox equivalent antioxidant capacity..

An interesting finding in the present study is that the hyperoside and quercitrin contents decreased by approximately 25%, whereas the quercetin content increased by approximately 13% in the fermented products than in the nonfermented substrates. In contrast to the findings of Lee et al. (2018), these authors found that the quercitrin, quercetin, rutin, and kaempferol contents were not significantly different between HC fermented with L. brevis B84 and nonfermented HC, whereas the vanillic acid and caffeic acid contents were significantly increased after fermentation.

In addition, Kwon and Ha (2012) analyzed BWEHC whole herb fermented with six Bacillus strains and found that the quercitrin, rutin, and quercetin contents increased by 1.4-, 1.9-, and 2.5-fold, respectively, in the fermented products than in the nonfermented products. The different inoculated probiotics and medicinal parts of HC may contribute to the variation of the specific flavonoid content before and after fermentation. One limitation of this study is that only the predominant flavonoids were determined by HPLC. Therefore, more studies are needed to compare the differences between nonfermented and fermented HC using nontargeted metabolomics.

Hyperoside and quercitrin are the glycoside derivatives of quercetin. Quercitrin has been shown to hydrolyze into quercetin by fecal flora, and glycoside hydrolases may be involved in the degradation (Bokkenheuser et al., 1987). Mao et al. (2021) compared the genomics of 133 L. plantarum strains isolated from various sources and found that 114 strains contained glycoside hydrolases, which can degrade the glycosidic linkage to release glycan and glycoside. The glycoside hydrolases in L. plantarum may be essential in changing the hyperoside, quercitrin, and quercetin contents before and after fermentation. However, further research is needed to demonstrate this relationship.

The present study showed that different sources, medicinal parts, and harvesting seasons could affect the functional compounds and antioxidant activity of HC. The HC leaves harvested in summer had a higher hyperoside, quercitrin, quercetin, and total polyphenol and flavonoid contents and antioxidant activity than the HC stems harvested in winter. The optimal fermentation conditions were a solid-liquid ratio of 0.9% (w/v) and a fermentation time of 48 h. The BWEHC leaves could be a potential prebiotic for L. plantarum. The quercetin content, total polyphenol and flavonoid content, and antioxidant activity significantly increased after fermentation than before fermentation. By contrast, the lower hyperoside and quercitrin contents in the fermented products than in nonfermented products may be attributed to glycoside hydrolases in L. plantarum.

None.

The authors declare no conflict of interest.

Concept and design: CMY. Analysis and interpretation: CCC, SAK. Data collection: CCC, SAK. Writing the article: CMY, SAK. Critical revision of the article: CMY, CHC. Final approval of the article: all authors. Statistical analysis: CCC. Obtained funding: none. Overall responsibility: CMY.