Anhydrous ethanol, ethyl ether, petroleum ether, n-butanol, concentrated hydrochloric acid (37%), sodium hydroxide, concentrated sulfuric acid (98%), Tween 80, Tris (hydroxymethyl) aminomethane, dimethyl sulfoxide, glutaraldehyde, sodium dihydrogen phosphate, disodium hydrogen phosphate, KBr, phosphoric acid, and ethyl acetate were of analytical grade and were purchased from Tianjin Fuyu Fine Chemical Co. Ltd. Analytical grade glucose, peptone, yeast extract, and agar powder were provided by AOBOX. Spectrally pure potassium bromide was purchased from Beijing Inluck Science and Technology Development Co. Ltd., whereas chromatographically pure methanol and acetonitrile were purchased from Honeywell (Burdick & Jackson). D101 and LSA-30 macroporous resins were purchased from Sunresin.

Fresh dandelion whole-grass samples were collected and identified as Taraxacum mongolicum by Professor Xinsheng Li of the Shaanxi University of Technology. The samples were dried to constant weight at 55°C, crushed, passed through a 60-mesh screen, then dried again. Active ingredients of dandelion were extracted using the method illustrated in Fig. 1.

Samples of lyophilized CA (ATCC10231) provided by the China Food and Drug Administration were stored in a low-temperature refrigerator (Thermo Fisher Scientific, Inc.) at −80°C. Prior to use, the samples were rehydrated, passaged, then cultured in Sabouraud agar medium (10 g of peptone, 40 g of glucose, 15 g of agar and 0.1 g of chloramphenicol added to distilled water with final volume 1L). No agar was added to the liquid culture, and other components remained unchanged.

CA (1 µl; OD600=0.4) was inoculated in liquid Sabouraud medium then cultured in a shaker at 37°C and 200 rpm until the optical density (OD)600nm reached 0.60 (Ultraviolet Spectrophotometer, Shimadzu). Next, 100 µl of the CA suspension were pipetted onto six Sabouraud agar plates and spread evenly using an L-shaped glass rod. Later, 10 µl dandelion extract samples I–VI were each pipetted onto sterile filter paper (6 mm in diameter). Then sterile filter papers were each transferred to Sabouraud agar medium (23,24). The plates were incubated at 37°C for 36 h, and the antifungal activity of DAIS was determined based on the inhibition zone of the plate. The same active ingredients were used for subsequent experimental studies. All of the experiments were repeated three times for verification.

An Agilent 6410B HPLC-ESI-MS/MS system (Agilent Technologies, Inc.) was used to analyze DAIS. The ingredients were separated on a DiKMA Diamonsil C18 column (250×4.6 mm, 5 cm) using a mobile phase composed of acetonitrile (eluent A) and 1.0% formic acid (v/v, eluent B). Gradient elution was performed at a flow rate of 1 ml/min, starting with 8% A/92% B at 0 min and passing through 14% A/86% B at 24 min, then 23% A/77% B at 35 min, 24% A/76% B at 44 min, 32% A/68% B at 60 min, 37% A/63% B at 66 min, and finally back to 8% A/92% B at 68 min (maintained for 6 min) (24,25). The detection wavelength of the HPLC system was set to 280 nm and the injection volume to 10 µl. The ESI source was kept at 110°C, and the desolvation temperature was set to 400°C. The nebulizer pressure, fragmentor voltage, and capillary voltage were maintained at 35 psi, 135 V, and 3,000 V, respectively. Full spectra were recorded in the negative ionization mode over the m/z range of 100–1,000 Da.

A single activated CA colony was inoculated in liquid sabourand medium, incubated at 37°C and shaken at 200 rpm for 12 h, then adjusted to a CA concentration of 107 CFU/ml. A 400 mg/ml DAIS mother liquor solution was also prepared. Both solutions were added to test tubes containing liquid Sabouraud medium. The final DAIS concentrations in the tubes were adjusted to 4.0, 8.0, 16.0, 32.0, 48.0, 64.0, or 128 mg/ml, whereas the amount of added CA suspension was set to 5% (v/v) in all of the tubes. A test tube containing an equivalent volume of pure water was used as control. Fluconazole was added to test tubes containing liquid Sabouraud medium and 5% CA suspension at final concentrations of 0.20, 0.40, 0.60, 0.80, 1.00, 1.20, and 1.40 mg/ml, as positive control group. Next, the mixtures were incubated at 37°C and shaken at 200 rpm for 24 h. The MIC was taken to be the lowest sample group concentration beyond which changes in OD600 values compared with the next sample group were ≤5% (26,27).

The CA culture and DAIS mother solution, prepared as aforementioned, were added to liquid medium in three test tubes at DAIS concentrations of 0, 1 and 3X MIC, and a CA concentration of 5% (v/v); 0X MIC served as the control group. The mixtures were then incubated at 37°C and 200 rpm followed by obtaining the OD600nm every 3 h for a total of 36 h (Ultraviolet Spectrophotometer, Shimazu). The CA survival and antifungal rates were calculated according to equations (1) and (2) (28–30):

Mixtures of 0, 0.5, 1.0, and 2.0X MIC DAIS and 5% (v/v) CA were incubated at 37°C and shaken at 200 rpm for 5 h. 1 ml aliquots of each sample were centrifuged at 7,104 × g for 6 min (Eppendorf) at 4°C then washed three times with PBS (0.01 M, pH 5.8). Subsequently, the centrifuged aliquots were fixed overnight at room temperature with 2.5% glutaraldehyde (prepared using 25% glutaraldehyde, deionized water, and PBS at a ratio of 1:4:5 in a dark, enclosed environment to prevent evaporation), dehydrated with a graded alcohol solution, and lyophilized for 12 h using a vacuum freeze dryer (Four-Ring). Finally, the samples were coated with gold using an ion coater (31–33) prior to SEM analysis at different magnifications (×2,000, ×5,000 and ×10,000 times (JEOL, Ltd.).

CA cultures prepared as aforementioned were centrifuged at 3,996 × g for 10 min at 4°C then washed three times with PBS (0.01 M, pH 5.8). The cultured cells were subsequently suspended in 0.01% Tween 80/PBS (0.01 M, pH 5.8) and their concentration was adjusted to 109 CFU/ml. DAIS mother solution was added to four test tubes containing suspended CA cells to attain final DAIS concentrations of 0, 0.5, 1.0 and 2.0X MIC. The 0X MIC group served as the control group. The mixtures groups were then incubated at 37°C and 200 rpm. Subsequently, 200 µl aliquots of each sample were collected after 0, 1, 2, 4, and 6 h of incubation at room temperature. The OD260nm values of supernatants were measured after centrifugation at 3,996 × g for 10 min (34) at 4°C.

The 5% (v/v) CA suspensions were added to and 0.5X MIC DAIS in liquid Sabouraud medium. The 0 X MIC group was considered as the control group. After incubation at 37°C and 200 rpm for 48 h, the CA cells were inactivated at 121°C for 20 min then collected by centrifugation at 3,996 × g for 10 min at 4°C. The collected cells were washed three times with 0.01% Tween 80/PBS (0.01 M, pH 5.8) and once with sterile water, followed by lyophilization and weighing.

Cell wall β-(1–3)-D-glucan was extracted according to the method reported in the literature (35), and the extraction rate was determined according to the dextran/cell dry weight. The extracted β-(1–3)-D-glucan was analyzed using infrared spectrometry (Nicolet; Thermo Fisher Scientific, Inc.) in the range of 4,000–500 cm⁻¹.

Data analysis was performed using SPSS 22.0 (IBM, Corp.) and Origin 8.0 (OriginLab) software. ChemDraw v. 17 (CambridgeSoft) software was used to illustrate chemical structures. Experimental data were analyzed using one-way analysis of variance followed by a Tukey's post hoc test to determine the significant differences between the groups. P<0.05 and P<0.01 were considered to indicate statistically significant and highly significant differences, respectively. The results were expressed as the mean ± standard deviation.

In vitro and in vivo assessments have confirm the antifungal activity of dandelion, a plant that has been commonly used in China to treat gynecopathy (20), such as mammitis and vaginitis. Dandelion contains various phytochemicals, including flavonoids, phenolic acids, alkaloids, and terpenes (25). In the present study, 6 samples (I, II, III, IV, V, VI) of DAIS were extracted and purified using hot-water (Fig. 1). The antifungal activity of the dandelion extracts was determined by assessing their effect on the growth of CA using the K-B paper disk method. As presented in Fig. 2, dandelion sample I is the only sample to exhibit an inhibition zone, indicating that all of the other samples (II, III, IV, V and VI) lack notable antifungal activity with respect to CA. The mean diameter of the CA inhibition zone in sample I (concentration=53.2 mg/ml) was found to be 10.38±0.15 mm, signifying marked antifungal effects against CA.

Eight compounds in dandelion sample I were separated within 78 min and identified at 280 nm using HPLC (Fig. 3). The compounds were further identified by tandem mass spectrometry based on their [M-H]- values and corresponding ion fragments (Table I). Peak with retention times of 21.35 min in Fig. 3 was identified as 4-coumaric acid based on the [M-H]- m/z value of 163. The ion fragment at m/z=119 corresponds to the loss of a CO2 moiety. With an [M-H]- m/z value of 193 and ion fragment at m/z=134, a peak with a retention time of 28.3 min was identified as ferulic acid (36). Similarly, a peak with a retention time of 30.05 min was identified as quercetin pentoside based on the molecular ion peak at m/z=433 and the ion fragment at m/z=301 (formed upon the removal of pentose) (25). Peaks with retention times of 37.29 and 38.45 min both yielded an [M-H]- m/z value of 515, and ion fragments at m/z=353 and 173. However, the former showed a unique fragment at m/z=191, whereas the distinguished fragment in the latter was detected at m/z=179. Based on these values, these two peaks were identified as 3,5-di-O-caffeoylquinic acid and 4,5-di-O-caffeoylquinic acid, respectively (37). The detected fragments at 353, 191, and 173 correspond to the loss of caffeoyl, two caffeoyl, and caffeoyl + quinic groups, respectively (37). The peak with a retention time of 39.56 min was identified as luteolin due to the detection of an ion fragment at m/z=133 (38). Finally, the peaks with retention times of 44.61 and 65.22 min were unidentified. Comparing with reports of Kenny et al (36), 4-coumaric acid, ferulic acid, Quercetin and luteolin had the same [M-H]- values and corresponding ion fragments. 3,5-Di-O-caffeoylquinic acid and 4,5-di-O-caffeoylquinic acid had same [M-H]- values and similar corresponding ion fragments compared with Dias et al (37); quercetin pentoside had the same [M-H]- value and corresponding ion fragment compare with Chen et al (25). Thus, the composition of sample I included 4-coumaric acid, ferulic acid, quercetin pentoside, 3,5-di-O-caffeoylquinic acid, 4,5-di-O-caffeoylquinic acid, luteolin and two unknown compounds. The relative percentages of peaks 1, 2, 3, 4, 5, 6, 7, and 8 were found to be 11.45, 3.96, 10.48, 34.24, 3.91, 11.80, 3.65 and 4.21%, respectively. Therefore, 3,5-di-O-caffeoylquinic acid was recognized as the main ingredient in dandelion extract sample I.

The results summarized in Table II revealed that the CA growth inhibition capacity of dandelion sample I increases with increasing concentration of the extract. The MIC was taken to be the lowest sample group concentration beyond which changes in OD600 values compared with the next sample group are less than or equal to 5%. Therefore, the MIC of sample I was determined as 32.0 mg/ml. The MIC of fluconazole was 0.60 mg/ml which was notably lower than the MIC of sample I because sample I was a mixture and its purity low. The low purity of samples I led to low activity.

Microbial growth curves include four phases corresponding to a delay period, an exponential growth period, a stable period, and a decay period. The CA mechanism of inhibition of dandelion sample I was investigated by analyzing the exponential phase. The kinetic equation of the CA exponential growth period is expressed as (39):

N is the number of CA at time t, and µ is the growth rate of CA.

Integration on both sides yields the following equation:

At constant specific growth rate of microorganisms, the integrated form of kinetic equation given by equation (3), which can be rearranged to produce equation (4), where Δt = t2-t1.

Antimicrobial kinetics refers to the kinetics of antifungal activity as determined by plotting the inhibitory growth curve of CA. As presented in Fig. 4, the OD600nm of CA was significantly reduced upon the addition of dandelion sample I at a concentration of 1.0X MIC. It was further decreased when the concentration of the dandelion extract was augmented to 3.0X MIC. This indicated that the antifungal activity of sample I increases with increasing DAIS concentration. In terms of values, an inhibitory rate of 97.13% was recorded using a DAIS concentration of 3.0X MIC, with a CA growth rate of only 0.00007 (Table III). The obtained results indicate that dandelion sample I inhibits CA proliferation mainly in the exponential phase, resulting in greatly reduced CA growth rates. Such antifungal activity depends on concentration and exposure time (31).

Presently, SEM has become necessary for the study of cell morphology. The SEM micrographs of CA cells in the control groups were shown in Fig. 5A-C. CA cells exhibited an almost ellipsoid shape, with complete structure, smooth surface, and notable refraction. The morphology of CA cells was notably unchanged upon treatment with 0.5X MIC of sample I for 5 h; however, the surface became rough, and the refraction was weakened (Fig. 5D-F). After treatment with 1.0X MIC of sample I for 5 h, the CA cell structure remained intact, but most cells were no longer ellipsoidal in shape. In fact, the surfaces of many cells were found to be wrinkled, and some of them appeared to be depressed (Fig. 5G-I). When the concentration of DAIS was further increased to 2.0X MIC, all CA cells showed non-ellipsoidal shape with relatively rough surfaces. The cells also began to shrink, and their morphology was irregular. Some cells showed marked compression, and damaged cell membranes of CA appeared (Fig. 5J-L). These results suggest that dandelion sample I is capable of destroying the membrane of CA cells, changing their normal morphology, shrinking their size, wrinkling their surface, and giving them an asymmetrical structure (40). The reason behind such changes may be attributed to significant leakage in cell content.

Under normal circumstances, intracellular macromolecules cannot pass through the cell membrane. However, when cells are adversely affected by unfavorable growth conditions or antibacterial agents, their membranes become damaged, thereby allowing UV-absorbing macromolecules, such as DNA and RNA, to leak into the culture medium, leading to an increase in cell absorbance at 260 nm (28). Thus, the integrity of the cell membranes can be estimated based on cellular absorbance at 260 nm.

The UV absorbance values of the culture medium at a wavelength of 260 nm were relatively low in the control group (0X MIC) compared with the sample groups (0.5, 1.0, and 2.0X MIC sample I-treated CA) (Fig. 6). After incubation for 1 h, the OD260nm values of the 0, 0.5, 1.0 and 2.0 X MIC supernatants were found to be 0.0926±0.0252, 0.1027±0.0183, 0.1286±0.0214, and 0.1579±0.0152, respectively. These values increased to 0.2982±0.0333, 0.3594±0.0165, 0.3942±0.0231, and 0.4756±0.0216, respectively, after 6 h of incubation. With increases in treatment time, the absorbance of the culture medium at 260 nm was gradually increased. These results indicated that dandelion sample I could be capable of destroying the cell membranes of CA, resulting in extensive nucleic acid leakage and ultimately, the inhibition of growth (41).

β-(1–3)-D glucan is the main component of the CA cell wall (42), and structural abnormalities in this component may lead to cell wall rupture. Therefore, the integrity of the CA cell wall was investigated by assessing the cellular content of β-(1–3)-D-glucan using infrared spectroscopy. The results indicated β-(1–3)-D-glucan contents of 22.13 and 20.86% for the 0 and 0.5X MIC groups, respectively (data not shown). The difference between the two groups is insignificant, which implies that sample I has no significant effect on the concentration of β-(1–3)-D-glucan in the CA wall.

The infrared spectrum of β-(1–3)-D-glucan comprises a peak between 3,000 and 2,800 cm⁻¹ corresponding to the ‘C-H’ stretching vibration (Fig. 7). A wide peak between 800 and 400 cm⁻¹ corresponds to the ‘-CH2-’ moieties. The peaks at 2,920, 1,370, 1,260, and 1,250 cm⁻¹ characterize the vibrations of polysaccharides, whereas the peak at 1650 cm⁻¹ peak is related to the ‘C=O’ vibration. Finally, the absorption peak at 890 cm⁻¹ is indicative of the presence of a β-configuration polysaccharide glycosidic bond.

Compared with the control group, the ‘C-H’, ‘C=O’, and ‘-CH2-’ absorption peaks were all increased in the 0.5X MIC group. Meanwhile, the absorption peak of the β-configuration polysaccharide glycosidic bond is decreased. However, no significant change was noted in the other absorption peaks. These results indicate that dandelion sample I altered the structure of β-(1–3)-D glucan in CA cell walls, possibly via the partial breaking of the glycosidic bond to form more ‘C=O’ and ‘-CH2-’ groups (Fig. 7). Another plausible explanation could be that after treatment with sample I, glycosidic bond synthesis was inhibited (42), resulting in the formation of dextran molecules with an incomplete structure. The reduced number of glycosidic bonds inhibits the cross-linking of dextran, ultimately affecting the synthesis of CA cell walls. Abnormal cell wall structure leads to fragile cells that are easily ruptured (43,44).