Bioconversion of BIOGF1K, a compound-K-rich fraction from ginseng root and its effect on epidermal barrier function

Abstract

BIOGF1K, the ginseng root-based and hydrolyzed ginsenoside-rich fraction, is known to improve skin damage, but there are rare studies on the kinetic of ginsenosides in the epidermis and their effects on epidermal barrier function. The current study investigated the effect of BIOGF1K on epidermal barrier function and its kinetics on epidermal transport. HPLC and LC/MS were used to verify the ginsenosides and the metabolites of BIOGF1K. Human immortalized keratinocytes (HaCaT) and epidermis-dermis artificial skin were treated with BIOGF1K and their metabolites were analyzed by HPLC and LC/MS. The epidermal barrier function was evaluated by transepithelial electrical resistance (TEER). In BIOGF1K, ginsenoside Rg1, Rd, F1, F2, compound Mc, compound Y (CY), and compound K (CK) were detected and CK and CY were the most and second abundant ginsenosides. TEER of HaCaT with 100 and 200 μg/mL BIOGF1K treatment was significantly higher than the control during 600 min of incubation. CK was permeated to the epidermis in a time-dependent manner and its maximum transported rate was observed at 600 min. In the case of artificial skin, CY and CK were permeated to the epidermis-dermis skin as time-dependent. Also, 24 h after treatment of CY, CK was detected as 19.59% of CY. It was proposed that CY was hydrolyzed into CK while permeating the epidermis. Results from the current study suggest that bioconversion of BIOGF1K rich in CK effectively enhances epidermal barrier function and it could be a useful cosmeceutical to exhibit its functionality to the skin.

1. Introduction

Skin is the largest organ of human body with a role in not only protecting against harmful agents and loss of water but also regulating body temperature, blood pressure, and sensation [1]. It consists of three regions including epidermis, dermis, and hypodermis [2]. The epidermis is outermost layer of skin, and it contains keratinocyte in the basal layer and undergoes proliferation and differentiation. There are two main components of the skin barrier including tight junctions (TJs) and outermost layer of epidermis, stratum corneum (SC) [3]. Among them, TJs of epidermis are cell-cell junctions by multiprotein complexes mainly located in the stratum granulosum (SG) underneath SC [3,4]. It was shown that TJs in skin have a role of paracellular barrier as bidirectional contribution to protect pathogen including virus and allergen from body as well as to regulate the passing of ions and molecules [[4], [5], [6], [7]]. Measurement of transepithelial electrical resistance (TEER) based on ohmic resistance is used for measuring skin barrier integrity by tight junction [8,9]. Several studies found that treatment of natural products on skin enhanced the skin barrier function by improving TEER. For instance, saccharomycopsis fermentation product increased cellular integrity and expression of TJ proteins in human keratinocytes [10]. It was observed that Aquaphilus dolomiae extract induced an increase in TEER value from normal human epidermal keratinocyte while recovering cellular integrity, which was reduced by infection of Staphylococcus aureus [11].

Saponins, one of the plant-derived hydrophobic bioactive compounds, have a variety of pharmacological effects including anti-carcinogenic, anti-inflammatory, anti-diabetic effects, and modulation of nuclear receptors [[12], [13], [14]]. Ginsenosides are a major saponins derived from ginseng (Panax ginseng), and they are traditionally used as cosmetic application as well as pharmaceutical agents [[15], [16], [17]]. Based on the structure of ginsenosides, the major pharmacologically effective ginsenosides were classified as dammarane type, mainly divided into protopanaxadiol (PPD) types, including Rb, Rd, Rg3, and Rh2, and protopanaxatriol (PPT) types, such as Re, Rg1, and Rh1 [18,19]. When ginsenosides are hydrolyzed by intestinal bacteria after the oral administration of ginseng, compound K (CK, 20-O-β-D-glucopyranosyl-20(S)-protopanaxadiol) and F1 (20-O- β-D-glucopyranosyl-20 (S)-protopanaxatriol) are produced as metabolites of the PPD and PPT type ginsenosides, respectively [20,21]. These metabolites are also produced by performing physical processes, enzymatic methods, and fermentation [22]. For instance, CK and F1 were produced by reacting enzymes such as pectinase and β-glucosidase, which catalyze the hydrolysis of ginsenosides [23,24]. BIOGF1K, hydrolyzed ginsenoside-rich fraction, was previously isolated from ginseng root [25] and hydrolyzed by glycosidase. Several studies found that BIOGF1K played an important role in inhibiting inflammation, photoaging, and atopic dermatitis of the skin [[26], [27], [28]]. It was reported that ginsenoside F1 and CK provided various pharmacological effects on skin health, including improvement of UVB-induced skin cell damage [26,29]. In the previous in vivo study, it was found that CK enhanced the skin barrier function against harmful exogenous agents [30]. While there is extensive knowledge regarding the effects of ginsenosides on skin health, there is a dearth of studies on the basic transport kinetics of ginsenosides or hydrolyzed ginsenoside-rich fractions across the epidermis and their effect on the epidermal barrier function. To our knowledge, skin permeability of ginsenosides on the epidermis is limited to evaluate ginsenoside-added nanoparticle [31]. Thus, the current study aimed to investigate the effect of BIOGF1K on skin barrier differentiation and integrity and epidermal permeability in human immortalized keratinocytes (HaCaT), and the skin permeability of ginsenosides and bioconversion to its metabolites in artificial skin with profiling ginsenosides in BIOGF1K and its metabolites by using HPLC-UV and HPLC-ESI-MS.

2. Materials and methods

2.1. Chemicals and reagents

Standards of ginsenoside Rg1 and Rd were purchased from Chengdu Biopurify Phytochemicals Ltd (Chengdu, China) and standards of ginsenoside F1, Mc, compound Y (CY) and CK were purchased from Ambo Institute (Daejeon, Korea). Dulbecco’s Modified Eagle Medium (DMEM) was purchased from Biowest (Riverside, MO, USA). Fetal bovine serum (FBS) was purchased from Cytiva (Marlborough, MA, USA). Dulbecco’s phosphate-buffered saline (DPBS) and penicillin/streptomycin (P/S) were purchased from Corning Inc. (Corning, NY, USA). Acetonitrile (ACN), methanol, and water with a grade for HPLC were obtained from J.T.Baker (Phillipsburg, NJ, USA). Dimethyl sulfoxide (DMSO) was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA), respectively. All chemicals were analytical grade and stored in compliance with the shelf life and storage conditions.

2.2. Preparation of hydrolyzed ginsenoside-rich fraction, BIOGF1K

BIOGF1K, hydrolyzed ginsenoside-rich fraction, was gifted by AMOREPACIFIC (Yongsan-gu, Seoul). For treatment of BIOGF1K to cell cultures, BIOGF1K was dissolved in dimethyl sulfoxide (DMSO) at 100 mg/mL as a stock solution and diluted to DMEM.

2.3. HaCaT (human immortalized keratinocytes) cells culture

The growth and differentiation of HaCaT cells were conducted by previous study [32]. The culture medium for HaCaT was composed of 89% DMEM, 10% of FBS, and 1% of P/S with low-calcium growth medium (0.03 mM of calcium). The HaCaT cells were subcultured when the confluency reached 80–90% of confluency for 3–4 days. The cell media was changed every other day for maintenance.

2.4. Cell viability assay using MTT

The cell viability of HaCaT (human immortalized keratinocytes) cells was measured by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay followed by a modified prior method [33]. The 5 × 10³ HaCaT in 0.2 mL of DMEM were plated in 96-well cell culture plate and incubated at 37 °C in a 5% CO₂ incubator for 24 h for cell attachment. And then, the HaCaT cells were treated with BIOGF1K at various concentrations (10, 50, 100, 250, and 500 μg/mL) in DMEM containing less than 0.5% of dimethyl sulfoxide (DMSO) for 24 h. Then, the sample solution was withdrawn and 100 μL of 0.5 mg/mL MTT solution was treated in each well for 2 h. To determine the amount of conversion from MTT to formazan, the MTT solution was withdrawn and 100 μL of 100% DMSO was added to each well. Then, the plate was incubated at 37 °C for 5–10 min until formation of formazan. It was measured at 570 nm by using a microplate reader (Varioskan Flash, Thermo Scientific, CA). The cell viability (%) was calculated as follows:

$$\text{Percent}\text{of}\text{cell}\text{viability}(\%)=\frac{\text{Average}\text{of}\text{sample}\left(\mathrm{O}.\mathrm{D}\right)-\text{Average}\text{of}\text{blank}\left(\mathrm{O}.\mathrm{D}\right)}{\text{Average}\text{of}\text{control}\left(\mathrm{O}.\mathrm{D}\right)-\text{Average}\text{of}\text{blank}(\mathrm{O}.\mathrm{D})}\times 100$$

2.5. Differentiation of HaCaT by qRT-PCR

The differentiation of HaCaT culture in the medium with different concentrations of calcium was performed by a previous method [32]. The HaCaT was cultivated in 100 mm cell culture plates and subcultured 3 times. Then, the 5.7 × 10⁵ cells were plated in a 6-well cell culture plate. The cell was grown with low-calcium cell culture medium (0.03 mM of calcium), and when the cell was confluent with 80–90%, the medium was switched with high calcium cell culture medium (2.8 mM of calcium). Then, the HaCaT was cultivated for 5 days with high calcium and when the confluency reached 100%.

After 5 days of culture, the cells were harvested for quantitative real-time polymerase chain reaction (qRT-PCR). To analyze the expression of epidermal differentiation genes, total RNA was isolated from the cells using a commercially available kit (RNeasy Mini Kit, Qiagen, United Kingdom) according to the manufacturer’s instructions. Each RNA sample was synthesized to complementary DNA (cDNA) using PrimeScript™ 1st strand cDNA Synthesis Kit (Takara, Japan). Gene amplification and quantification were performed using a thermocycler (7500 Fast Real-Time PCR Instrument System, Applied Biosystems, USA) and TaqMan™ Universal Master Mix II (Applied Biosystems, USA) under the following conditions: 95 °C for 1 min followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Relative mRNA expression levels were normalized to the housekeeping gene RPLP0 in the same sample. The delta Ct method for quantification of the data was carried out in duplicate.

2.6. Transport of ginsenosides in BIOGF1K by HaCaT cell and artificial epidermis-dermis skin

To determine the transport of ginsenoside by immortal human keratinocyte cell line and artificial skin, 3 × 10⁴ of HaCaT cells were cultured in collagen-coated 12-transwell inserts (3493, Corning Inc.) and the artificial skin (Neoderm®-ED, Tego science, Seoul; Korea) consisted of the epidermis and a dermal matrix were used. The culture medium for HaCaT cells was changed every other day, and HaCaT cells were differentiated by high calcium culture medium (2.8 mM of calcium) after seeding 4 days. Differentiation was confirmed by testing.

To confirm the condition of HaCaT cell appropriate to experiment. The transepithelial electrical resistances (TEER) values were measured when it was confluent with 100% by using Millicell ERS-2 system (Millipore, Bedford, MA, USA), and it was conducted every sampling points. After confirming the stabilization of the TEER value of the cells, several concentrations of BIOGF1K in DMEM were treated in each trans-well insert (apical), and the basolateral of trans-well plate were filled with 1.5 mL of DPBS. During the incubation at 37 °C in a 5% CO₂ incubator for 600 min, the 400 μL of basolateral was collected at predetermined sampling points (0, 30, 60, 240, 360, and 600 min) and 400 μL of DPBS was refilled in basolateral. The TEER values were further analyzed of each time point and calculated by follows:

$$\text{TEER}(\%\text{of}0\min )=\frac{\text{TEER}\text{values}\left(\text{each}\text{timepoint}\right)}{\text{TEER}\text{values}(0\min )}\times 100$$

For the experiment using artificial skin, 500 μL of BIOGF1K and ginsenoside Y dissolved in DMEM (500 μg/mL and 100 μg/mL, respectively) was treated at the apical of artificial skin transwell. To identify the transport of ginsenosides in a time-dependent manner, 500 μL of basal filled with DPBS were collected at sampling points (4 and 10 h after treatment) and refilled with the same amount of DPBS at basal. Cell medium from both apical and basal side were collected after incubation at 24 h after treatment, and then artificial skin tissues were frozen at −80 °C. The frozen tissues were homogenized with cell lysis buffer and sonicated for 30 min followed by centrifugation (4000 rpm, 30 min, 4 °C).

All collected samples was diluted at a 1 to 1 ratio (v/v) with methanol, and the epidermal transport of ginsenoside was calculated by follows:

$$\mathrm{E}\mathrm{p}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m}\mathrm{H}\mathrm{a}\mathrm{C}\mathrm{a}\mathrm{T}(\%)=\frac{\mathrm{G}\mathrm{i}\mathrm{s}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{b}\mathrm{a}\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}{\mathrm{T}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}}\times 100$$

2.7. Identification of ginsenosides and its metabolite in BIOGF1K by HPLC-ESI-MS

Identification of ginsenosides and its metabolite in BIOGF1K by high performance liquid chromatography (HPLC)-UV and HPLC- electrospray ionization (ESI)-mass spectrometry (MS) equipped with Ultimate 3000 and LCQ Fleet Ion Trap Mass Spectrometer (Thermo Fisher Scientific, Waltham, USA) was followed by a previous method with modification [34]. The reverse phase column, Mightysil RP-18 GP (4.6 × 250 mm, 5 μm; Kanto chemical, Tokyo, Japan) column, was used, and the column oven was set to 40 °C. The stock solution of ginsenoside Rg1, Rd, F1, Mc, CY and CK standard was prepared by using 100% methanol solvent. The gradients of mobile phase were 10% Acetonitrile (A) and 90% Acetonitrile (B) as follows: 0–10 min, 15% B; 10–50 min, 15–60% B; 50–70 min, 60–85% B; 70–71 min, 85-15% B; 71–80 min, 15% B. UV detector was set at a wavelength of 203 nm. The sample injection volume was 10 μL. MS was performed in a positive ESI mode, and the MS parameters were as follows: sheath gas flow of 70 arb, auxiliary gas flow of 20 arb, heater temperature of 300 °C, capillary voltage of - 35 V, and capillary temperature of 275 °C. As sample preparations for HPLC-UV and HPLC-ESI-MS, all standard and BIOGF1K was finally diluted to 50% methanol, and samples including collected media and homogenized artificial skin tissue were diluted at a 1 to 1 ratio (v/v) with methanol. Calibration curves of ginsenosides were evaluated by measuring each analytical standard of ginsenosides with 5 concentrations (5, 25, 50, 75, and 100 μg/mL) as triplicate. The correlation coefficient (r²) was calculated by the regression equation of calibration curve using Microsoft Excel 365. For quantification of ginsenosides in BIOGF1K and collected samples, all samples and analytical standards are measured triplicate by HPLC-UV.

2.8. Statistical analysis

One-way and two-way repeated-measures analysis of the variance (ANOVA), t-test, and Tukey’s post hoc test were performed to measure the significant differences among the groups by using Graphpad Prism 6.0 software (Graphpad, CA, USA).

3. Results & discussion

3.1. Identification and quantification of ginsenosides in BIOGF1K

By HPLC-UV and HPLC-ESI-MS analysis, ginsenosides in BIOGF1K were identified and quantified. As shown in Fig. 1 and Fig. S1, seven peaks detected in UV chromatogram of BIOGF1K (Fig. 1A) were identified as corresponding to both retention time of each ginsenoside standards and mass spectra (Fig. 1B). Identified and quantified ginsenosides in BIOGF1K, including patterns of MS/MS fragmentation for targeted ginsenosides, are described in Table 1: Rg1 at m/z 823.9167 [M + Na]+ → 643.2972 (RT = 21.94 min); Rd at m/z 969.7500 [M + Na]⁺ → 789.4223 (RT = 38.10 min); F1 at m/z 662.4167 [M + Na]⁺ → 203.0767, 481.1351 (RT = 38.53 min); F2 at m/z 808.5833 [M + Na]⁺ → 627.2935 (RT = 47.33 min); Mc at m/z 778.1667 [M + Na]⁺ → 335.2395 (RT = 54.59 min); CY at m/z 777.9167 [M + Na]⁺ → 335.2124 (RT = 55.71 min); CK at m/z 645.9167 [M + Na]⁺ → 465.3109, 203.1558 (RT = 62.81 min). Due to a lack of commercial standard, F2 was temporarily identified with production based on previous findings [35,36]. The contents of six ginsenosides in BIOGF1K were 1.16 ± 0.03, 1.37 ± 0.11, 1.72 ± 0.04, 27.86 ± 1.65, 93.49 ± 1.60, and 343.38 ± 4.76 for Rg1, Rd, F1, Mc, CY, and CK, respectively (Table 1). To our knowledge, there is no study on the quantification of ginsenosides in BIOGF1K comparable with our results. Since BIOGF1K is known as a hydrolyzed ginsenoside fraction of ginseng root by glycosidase, CK, attached the smallest number of glycoside among PPD type of ginsenosides, was the most abundant in BIOGF1K, and CY was second highest content in BIOGF1K, while other ginsenosides of the PPD family were detected in low content.

Fig. 1.

Identification of ginsenosides in BIOGF1K. A: UV chromatogram of BIOGF1K 500 ppm; B: MS/MS spectra.

Table 1.

Identification and quantification of ginsenosides in BIOGF1K.

Peak No.	RT (min)	Ginsenosides	Class	Molecular weight	Molecular formula	Calibration curve	Correlation coefficient (r2)	[M+Na]+ (m/z)	Product ion (m/z)	Content in BIOGF1K (mg/g)
1	21.65	Rg11)	PPT	801.0	C42H72O14	y = 2759.4x + 250.32	0.9992	823.9167	643.2972	1.16 ± 0.03
2	37.43	Rd1)	PPD	947.2	C36H62O9	y = 2221.6x + 377.96	0.9993	969.7500	789.4223	1.37 ± 0.11
2	37.88	F11)	PPT	638.9	C36H62O9	y = 6914.5x – 6759	0.9998	662.4167	203.0767, 481.1351	1.72 ± 0.04
3	46.58	F22)	PPD	785.0	C42H72O13	NA	NA	808.5833	627.2935	NA
4	53.31	Mc1)	PPD	755.0	C41H70O12	y = 5425.2x – 1695.8	0.9994	778.1667	335.2395	27.86 ± 1.65
5	54.34	Compound Y1)	PPD	755.0	C41H70O12	y = 5578.7x – 6934.3	0.9996	777.9167	335.2124	93.49 ± 1.60
6	61.55	Compound K1)	PPD	622.9	C36H62O8	y = 6738x – 16394	0.9998	645.9167	465.3109, 203.1558	343.38 ± 4.76

1) Identified by standard; 2) temporarily identified by product ion; NA: Not available.

3.2. Effect of cell BIOGF1K on cell viability in HaCaT cells

Cell viability after treating various concentrations from 10 to 500 μg/mL of BIOGF1K into HaCaT cells is shown in Fig. 2. Cell viability in HaCaT cell treated with 10 and 50 μg/mL of BIOGF1K was significantly higher than the control (p < 0.05). Cell viability from BIOGF1K in HaCaT cells was significantly reduced at 250 and 500 μg/mL. The IC₅₀ of BIOGF1K to HaCaT cell was determined to be 235.94 μg/mL, and a further study was conducted with BIOGF1K concentration which was lower than IC₅₀.

Fig. 2.

Effect of BIOGF1K treatment on cell viability of HaCaT for 24 h. Control: HaCaT treated with only DMEM. An asterisk (*) indicates significant differences between control and each treatment group (*p < 0.05, **p < 0.01). Different letters on the bar indicate significant differences among concentrations (p < 0.05).

3.3. Differentiation of epidermal barrier in HaCaT

Expression of representative epidermal differentiation markers, including keratin 1 (KRT1), keratin 10 (KRT10), and involucrin (IVL), is presented in Fig. 3. As shown in Fig. 3A–B, the gene expression levels of KRT1 and KRT10 were remarkably higher in HaCaT cells cultured with high-calcium media (2.8 mM) than HaCaT cells cultured with low-calcium media (0.03 mM) without a significant difference (p > 0.05). However, as shown in Fig. 3C, the gene expression level of IVL was significantly higher in HaCaT cells cultured with high-calcium media than low-calcium media (p < 0.05). KRT1 and KRT10 are differentiation-specific cytokeratins in keratinocytes observed at the initial and late phases of differentiation [37]. IVL is one of the important structural proteins for the formation of SC [6,38]. In accordance with findings from the current study, treatment of extracellular Ca²⁺ in human epidermal keratinocytes induced differentiation and formation of adherens junctions [10]. This finding confirmed that HaCaT cells cultured in high calcium concentration conditions were properly differentiated, and that was suitable for subsequent experiment.

Fig. 3.

Effect of calcium concentration on differentiation of HaCaT by measuring expression of related genes, including KRT1 (A), KRT 10 (B), and IVL (C) using qPCR. Low Ca: HaCaT cultured with 0.03 mM of calciumin culture medium; High Ca: HaCaT cultured with 2.8 mM of calcium in culture medium.

3.4. Effect of BIOGF1K on the epidermal barrier function in HaCaT

In order to measure the epidermal barrier function, cell integrity was measured by TEER from HaCaT cultured in transwell at 0, 30, 60, 240, 360, 600 min after incubation with various concentrations (20, 50, 100, 200 μg/mL in DMEM media) of BIOGF1K (Fig. 4). Except for 20 μg/mL of BIOGF1K treatment, the TEER value from other concentrations of BIOGF1K appeared to increase for 600 min of incubation. Comparing the TEER of the BIOGF1K treatment group with that of the control group, 100 μg/mL of BIOGF1K treatment was significantly higher than that from the control group measured at 60 min (p < 0.05) and 240 min (p < 0.01). In addition, the TEER from the group with 200 μg/mL of BIOGF1K treatment was significantly higher than that of the control group measured at 600 min (p < 0.01). The current study found that 100 and 200 μg/mL of BIOGF1K is effective in improving the cell barrier function of keratinocytes by measurement of TEER. TJs are important for the barrier function of epidermis as well as SC, and TEER can easily be measured in the principle of Ohm’s law, reflecting the skin barrier function of TJs [39,40]. It was observed that a significant reduction in TEER from patients with atopic dermatitis induced expression of claudin-1 and claudin-23, which are TJ proteins determining TJ resistance and permeability [41]. It is plausible that BIOGF1K could be effective in increasing skin barrier function from TJs. The current study further investigated the epidermal transport of BIOGF1K.

Fig. 4.

Effect of BIOGF1K on TEER values (% of 0 min) of HaCaT cells at each time point during incubation in trans-well. TEER value (%) was expressed by comparing to own TEER (Ω × cm2) at 0 min of each concentration. * and ** indicate a significant difference compared with the control group at p < 0.05 and p < 0.01, respectively.

3.5. Transport and bioconversion of ginsenosides through HaCaT and artificial epidermis-dermis skin

In order to characterize the epidermal transport of ginsenosides in BIOGF1K treated in HaCaT cell cultured in a transwell, the basal media was sampled at 30, 60, 240, 360, 600 min after treatment of 100 μg/mL of BIOGF1K. The basal media was analyzed using HPLC-UV with same methods used for the identification and quantification of ginsenosides in BIOGF1K (Fig. 5A). Among the four major ginsenosides from BIOGF1K which can be quantified, the amount of epidermal transport for ginsenoside CK time-dependently increased, but that of other ginsenosides including Rg1, Rd, F1 and Mc except CY were not detectable. CY detected in basal media increased until 240 min after treatment, but the content was decreased from 360 min after treatment (Fig. 5B). In contrast to CK and CY, other ginsenosides in BIOF1K seemed to be accumulated in the HaCaT cell rather than transported. As a matter of fact, CK not only contains a relatively higher amount (343.38 ± 4.76 mg/g) than other ginsenosides in BIOGF1K but also has a relatively lower molecular weight (MW: 622.9). Our findings confirmed that CK, which is more abundant in BIOGF1K than other ginsenosides by enzymatic processing, more easily penetrated the epidermal barrier. Interestingly, the detected amount of CK at 600 min of incubation was beyond the quantified amount from BIOGF1K, implying that other PPD ginsenoside families in BIOGF1K, particularly CY having the second-highest content, could be metabolized into CK in the epidermis.

Fig. 5.

Epidermal transport of ginsenoside CK (A) and CY (B) from HaCaT seeded in a transwell at 30, 60, 240, 360 and 600 min after incubation with BIOGF1K.

To confirm transport of ginsenosides through dermis as well as epidermis, further study on transport of ginsenosides and bioconversion of CY to CK was performed by using artificial epidermis-dermis skin. When BIOGF1K was treated into artificial epidermis-dermis skin, only CY and CK was transported into basal media with increasing as a time-dependent manner (Fig. 6A). Compared with CY, CK was detected from apical media, skin and basal media as higher ratio from initial treated concentration of CK in BIOGF1K (Fig. 6B). This result implies that CK was possibly bio-converted from CY during incubating in artificial epidermis-dermis. In order to confirm the bioconversion from CY to CK, CY was treated into that artificial epidermis-dermis skin for 24 h. Although only CY was treated, CK was detected from artificial epidermis-dermis skin (Fig. 6C). Approximately 5.38% of CY was remained and 19.59% of CY was converted to CK (Fig. 6D). The result indicates bioconversion from CY into CK was occurred from epidermis-dermis skin.

Fig. 6.

Transport of BIOGF1K and ginsenoside CY in Neoderm®. Transported ginsenoside CY and CK as time-dependent (A), and distribution of ginsenoside CY and CK (B) 24 h after treatment of BIOGF1K in Neoderm®. MS/MS chromatogram of remained ginsenoside CY and bioconverted ginsenoside CK from CY (C) and cumulative percentage of ginsenoside CY from initial value (D) 24 h after treatment of ginsenoside CY in Neoderm® tissue.

Similar to our findings, PPD ginsenosides were generally hydrolyzed by intestinal microorganism or hydrolases such as β-glycosidase and pectinase [42]. CY also can be hydrolyzed into CK by recombinant β-glycosidase from Microbacterium esteraromaticum and β-Glycosidase from Sulfolobus solfataricus [24,43]. It was reported that the activity of endoglycosidase such as glucocerebrosidase and heparanase-1 was detected in the epidermis [44]. However, studies on effect of endoglycosidases activity from epidermis on hydrolysis of ginsenosides was limited yet, and the elucidation of bioconversion requires further study. A previous study found that CK increased not only hyaluronan synthesis by enhancing the expression of hyaluronan synthase 2 via in vitro study, but also hyaluronan content in the epidermis via in vivo study [45]. In addition to the epidermis, CK effectively prevented UVA-irradiated inflammation and photoaging in dermal fibroblasts [46]. Based on previous findings and our current results, it is plausible that CK can play a pharma-cosmetical role in both epidermis and dermis by permeating through epidermis.

Taken together, the current study proposed a transport mechanism of ginsenosides on the epidermal transport of BIOGF1K (Fig. 7). Among the various ginsenosides from BIOGF1K, CK, which has relatively lower molecular weight, readily penetrated into the epidermal barrier, contributing to the skin barrier function. Other PPD-class ginsenoside such as CY in BIOGF1K could be metabolized to CK in the epidermis by endoglycosidase activity, more readily permeable form to skin barrier. However, further study is required to investigate the bioconversion of ginsenoside by endoglycosidases in the epidermis and dermis. Based on our results, it is plausible that ginsenoside CK, which is known to be effective in preventing skin aging as well as penetrating into skin epidermis, exhibits its functionality to the skin. Results from the current study suggest that BIOGF1K rich in CK effectively enhances the barrier function of the epidermis and could be a useful cosmeceutical to exhibit its functionality to the skin by epidermal permeated CK.

Fig. 7.

Proposed schematic pathway of ginsenosides through the epidermis.

Author contribution statement

Woo-Hyun Kim: Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Jeong-Eun Kim; Sehyun Kim: Performed the experiments.

Yongjoo Na: Conceived and designed the experiments.

Yong-Deok Hong: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data.

Joonho Choi: Conceived and designed the experiments; Analyzed and interpreted the data.

Won-Seok Park: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Soon-Mi Shim: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This study did not receive any specific grant from funding agencies.

Data availability statement

Data will be made available on request.

Declaration of interest's statement

The authors declare no conflict of interest.

Appendix ASupplementary data

The following is the Supplementary data to this article.