The skin forms the external surface of the human body, and serves as a barrier protecting the internal organs from ultraviolet radiation, toxins, and bacteria, etc. There are two dependent layers of the skin, the epidermis and dermis, and these consist of many cells such as keratinocytes, melanocytes and fibroblasts [1]. The mechanical strength of the skin is contributed by the dermis, which is composed of the extracellular matrix (ECM) where fibroblasts synthesize ECM components such as collagen and elastin to sustain the skin’s elasticity [2]. However, the structure and function of the dermis can be changed by harmful external factors such as oxidative stress, UV exposure and air pollution. These conditions accelerate aging of the skin by collapsing its dermal structure [3, 4].

Ultraviolet light is a significant cause of exogenous skin damage. The ultraviolet rays that reach the Earth are classified as UVA (320-400 nm) and UVB (280-320 nm) [5]. When exposed to UV on the Earth’s surface, the amount of UV radiation that reaches the human skin is known to be 25 J/cm² under natural sunlight in autumn at 38° N for 4-5 h. It has been reported that this corresponds to ten times the minimum erythema dose in skin [6]. In particular, UVB can penetrate the upper layer of the dermis [7]. UVB exposure increases reactive oxygen species (ROS), such as hydroxyl radicals (•OH), superoxide anion radicals (O•^- ₂), singlet oxygen (¹O₂), and hydrogen peroxide (H₂O₂). To prevent ROS-induced cellular damage, enzymes (superoxide dismutase and catalase) and non-enzymatic antioxidants (L-ascorbic acid and α-tocopherol) are present in skin cells [8, 9]. However, when the balance of this ROS/ antioxidant defense system is upset due to excess ROS generation caused by UVB exposure, skin cells become damaged and skin aging is accelerated [10, 11].

ROS induced by UVB exposure increase the expression of matrix metalloproteinase-1 (MMP-1) in fibroblasts, promoting skin photo-aging [12-14]. MMP-1 degrades collagen type 1, which is an ECM component that provides structural support to the skin. This leads to disintegration of the dermis and acceleration of skin aging [15]. Therefore, the development of anti-aging agents to inhibit UVB- induced ROS generation is an essential strategy for suppressing photo-aging [16].

Pediococcus pentosaceus as a lactobacillus is commonly found in fermented foods such as doenjang and kimchi, and is known to produce lactic acid through anaerobic fermentation [17]. Recently, it was reported that P. pentosaceus enhanced the immune activity of Cordyceps militaris extract, enabling it to boost the phagocytic activity of macrophages in mouse [18]. P. pentosaceus increased the anti-aging activity of Gelidium amansil extract on MMP-1 expression and decrease of collagen by UVB irradiation [19]. We isolated and identified a new lactobacillus, P. pentosaceus DK1, from Diospyros kaki fruit. P. pentosaceus DK1 donated to KCTC (KCTC12963BP) could degrade the tannins in Diospyros kaki fruit and increase the total phenol content by 1.5 fold, and the flavonoid content by 1.4 fold. Diospyros kaki fruit fermented by P. pentosaceus DK1 also enhanced the inhibitory activity of elastase for improvement of skin aging. Thus, P. pentosaceus DK1 was also used to improve the anti-aging activity of L. angustifolia.

L. angustifolia, commonly known as lavender, belongs to the Lamiaceae family [20]. Traditionally, L. angustifolia extract has been used as a remedy for neurological and rheumatic diseases, due to its antibacterial and relaxing properties, and is mainly obtained by extracting volatile components from oils [21, 22]. Another method involves extracting L. angustifolia using water or ethanol, which has been reported to produce extracts that show antioxidant properties, whitening effect, and inhibitory effect on sebum production [23]. The major components of L. angustifoliaextract are linalool, linalyl acetate, ladanein, apigenin, apigenin-7-O-β-glucoside, luteolin, luteolin-7-O-β-glucoside, and 5,4’-dihydroxy flavonoid-7-O-β-pyranglycuronate butyl ester [24, 25]. Recently, Ahn, et al. [26] reported that the antioxidant, tyrosinase, and elastase inhibitory activities of L. angustifolia extract, as well as the ratios of phenolic compounds such as rosmarinic acid, were significantly increased by natural fermentation. However, their fermentation method was very old, and the microorganisms used were also not identified. There have been no studies so far on the anti-aging activity and composition analysis of L. angustifolia extracts fermented by P. pentosaceus DK1.

In this study, the anti-aging effects of L. angustifolia extract fermented by P. pentosaceus DK1 were evaluated on MMP-1 expression, collagen production, and antioxidant activity, and analyzed on the converted components during fermentation.

Reagents and Chemicals

Human skin fibroblasts (HS68 cells) were purchased from Lonza (Basel, Switzerland). Dulbecco’s modified Eagle’s medium (DMEM), foetal bovine serum (FBS), trypsin, and penicillin-streptomycin were obtained from Capricorn Scientific (Ebsdorfergrund, Germany).

Fermentation of L. angustifolia Extract

Fermented and non-fermented L. angustifolia extract were obtained from GFC Life Science (Korea). The dried L. angustifolia flowers (2 kg) were homogenized in 70% ethanol (40 L) for 20 days at room temperature. The extract powder (410. 93 g) was obtained by filtration and vacuum evaporation drying

Fermentation of L. angustifolia Extract

In order to cultivate the P. pentosaceus DK1 strain, lactobacilli MRS broth (BD 288130) purchased from Difco (USA) was selected as the culture medium for the optimal growth conditions of the strain. MRS medium was developed to favor the growth of lactobacilli in 1960 by De Man, Rogosa and Sharpe and is known to support the growth of lactobacilli including Pediococcus. MRS broth was dissolved in distilled water, and L. angustifolia extract was added (Table 1). P. pentosaceus DK1 was pre-cultured in MRS broth and cultured at 37°C for 24 h. Then, 10% of the strains pre- cultured in the broth containing the L. angustifolia extract were cultured at 37°C for 14 d. The fermented broth was then centrifuged at 9,010 g for 15 min and the supernatant was filtered using a 0.22 µm cellulose filter. Finally, the fermented broth was fractioned with ethyl acetate. The ethyl acetate fraction was evaporated to obtain fermented L. angustifolia extract. All the processes were indicated in Scheme 1.

Table 1 . Fermentation conditions of L. angustifolia extract..

Sample Group	Fermentation condition
	L. angustifolia extract Concentration (%)	MRS medium Ratio (%)
A1	0.1%	0.69%
A2	0.1%	1.38%
A3	0.1%	2.75%
A4	0.1%	5.50%
B1	0.5%	0.69%
B2	0.5%	1.38%
B3	0.5%	2.75%
B4	0.5%	5.50%

Figure 7. Preparation of fermented and non-fermented L. angustifolia extracts.

Cell Culture

Human skin fibroblasts (HS68 cells) were incubated in DMEM medium supplemented with 10% FBS, 100 U/ml, of penicillin and 100 µg/ml of streptomycin. The cells were incubated in medium at 37°C in a humid incubator under 5% CO₂ atmosphere.

UVB Radiation

For supply of UVB radiation, a CL-1000 Ultraviolet Crosslinker (UVP, USA), with an emission spectrum of 280–370 nm and a peak at 312 nm, was used. HS68 cells with 70–80% confluence were treated with fermented and non-fermented L. angustifolia extract at various concentrations in a plastic, 60-mm Petri dish in FBS-free medium for 24 h. The medium was then removed, and the cells were rinsed twice with phosphate-buffered saline (PBS). The cells were irradiated through the cover of the dish by UVB after addition of 1 mL PBS in the wells (height; 3 mm). After UV exposure, the medium was replaced with FBS-free medium.

Cell Viability

A 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, USA) assay was used to analyze cell viability. HS68 cells were seeded in 96-well plates at 1 × 10⁴ cells per well and incubated for 24 h at 37°C. After treatment with various doses of fermented or non-fermented L. angustifolia extract for 24 h or exposure to UVB radiation, the medium was replaced with FBS-free medium and the cells were further incubated for up to 72 h. Subsequently, 0.5 mg/ml MTT solution was added and the cells were further incubated for 30 min at 37°C. Then, 100 µl dimethyl sulfoxide (DMSO) was added to the wells respectively, and the absorbance was determined at 570 nm using a microplate reader (TECAN, Austria).

Quantitative Analysis of MMP-1 Expression

HS-68 fibroblasts were incubated up to 70-80% confluence on 60-mm plates in a humid incubator maintained at 37 C. After treatment with different concentrations of the fermented or non-fermented L. angustifolia extract, the cells were exposed to 80 mJ/cm² UVB under PBS and then incubated further for 48 h in fresh medium without FBS. Secreted MMP-1 proteins within the culture medium were measured by an ELISA kit (R&D Systems, USA) according to the manufacturer’s instruction.

HPLC and LC/ESI-MS Analysis

HPLC analysis of the lavender was carried out using a Shimadzu LC-20A HPLC system (Shimadzu, Japan) equipped with a UVD 170s DIONEX detector and Shim-pack VP-ODS C18 column (L: 250 mm, LD: 4.6 mm, 5 µm). The mobile phase was composed of A (2% acetic acid in H2O) and B (0.5% acetic acid in 50% acetonitrile aqueous solution). The working conditions were as follows: 0–25 min, 0% (v/v) of B; 25–40 min, 0–25% (v/v) of B; 40-80 min, 25–35% (v/v) of B; 80–110 min, 35–40% (v/v) of B; 110-120 min, 40–20% (v/v) of B; 120–130 min, 20–10% (v/v) of B; 130-140 min, 10– 0% (v/v) of B; 140–150 min, 0% (v/v) of B. The flow rate was 1.0 ml/min and the samples were observed at 365 nm. The samples were passed through a 0.2-um filter and then 20 µl of the samples with 10,000 µg/ml were injected into the HPLC.

The mass spectrometric analysis was performed using an LCQ Ion Trap Mass Spectrometer (Thermo Finnigan, USA) with an ESI interface and detection was done in positive ion mode by the National Instrumentation Center for Environmental Management College of Seoul National University (Seoul, Korea). The operating conditions were as follows: capillary voltage, 33 V; capillary temperature, 400°C; nebuliser pressure, 10 psi; and drying gas, N₂. Compounds within STE were identified by comparing the UV spectra, retention times, and fragment ions of standard materials.

Intracellular ROS Evaluation

The intracellular ROS levels were evaluated using the fluorescence dye 2’,7’-dichlorodihydrofluorescein diacetate (H₂DCF-DA, Sigma- Aldrich, USA), which is generated by the conversion of non- fluorescent H₂DCF-DA to highly fluorescent 2’,7’-dichlorofluorescein (DCF) by intracellular ROS. HS-68 cells were incubated with 20 µM H₂DCFH-DA for 30 min at 37°C. The cells were then washed twice with PBS and irradiated by 80 mJ/cm UVB. Fluorescence signals were detected using a fluorescence ELISA reader (Perkin Elmer, USA; excitation, 490 nm; emission, 530 nm).

Statistical Analysis

All statistical analyses were performed using SPSS 17.0 (SPSS, USA) software. The results were presented as mean ± standard deviation (SD). Statistically significant differences were calculated by one-way ANOVA, where p < 0.01 was considered statistically significant.

Effect of Fermented L. angustifolia Extract on UVB- Increased MMP-1 Expression in Human Skin Fibroblasts

UVB-treated fibroblasts secrete a greater amount of matrix metalloproteinase-1 (MMP-1) than untreated fibroblasts. MMP-1 is a metalloproteinase with a zinc ion in its center, and promotes skin aging through degradation of collagen type I and III, which act as support fixtures in the dermis.

We studied the effects of fermented and non-fermented L. angustifolia extracts on UVB-induced MMP-1 expression in human fibroblasts. To design an optimized system for measurement of anti-aging activity, the fibroblasts were irradiated by UVB at 20-100 mJ/cm². As the UVB irradiation dose was increased, the amount of MMP-1 expression in the UVB-exposed fibroblasts also increased (Fig. 1A). However, at 100 mJ/cm², there was a slight decrease in cell viability (Fig. 1B). We then set the ultraviolet radiation (UVB) at 80 mJ/cm², which produced a high MMP-1 expression level without cytotoxicity in the fibroblasts.

Figure 1. Effects of UVB irradiation on human skin fibroblasts. The cells were irradiated with various doses of UVB and incubated 72 h. Then, the levels of MMP-1 proteins (A) and cell viability (B) were measured. Data were presented as mean ± SD of three independent experiments. *p < 0.01 compared with the negative control.

Fermented and non-fermented L. angustifolia extracts (A1-B4) were prepared under various medium conditions, using the concentrations (Table 1) of L. angustifolia extract (0.1% or 0.5%) and the concentrations of MRS medium (0.69%, 1.38%, 2.75%, or 5.50%) as described in Materials and Methods.

In order to select the optimal fermentation conditions for the L. angustifolia extract in terms of anti-aging activity, 10 μg/ml of L. angustifolia extract fermented product and non-fermented product were used to treat fibroblasts for 24 h and these were then irradiated with 80 mJ/cm²UVB. As shown in Fig. 2A, all fermented and non-fermented L. angustifolia extracts reduced the increased MMP-1 protein expression under UVB irradiation. Fermented L. angustifolia extract showed a greater inhibitory effect on MMP-1 expression than non-fermented L. angustifolia extract. In particular, the inhibitory effect of MMP-1 expression of the L. angustifolia extract fermented at B2 condition (0.5% of L. angustifolia extract and 1.38% of MRS medium) was the largest of all extracts. All extracts fermented at all conditions showed no cytotoxicity (data not shown). We selected this condition as optimal for fermentation and used these same conditions in subsequent experiments.

Figure 2. Effect of fermented and non-fermented L. angustifolia extracts on UVB-induced MMP-1 expression in human skin fibroblasts. (A) MMP-1 expression was determined after pre-treatment with 10 μg/ml of fermented and non-fermented L. angustifolia extracts and irradiation of 80 mJ/cm2 UVB. The cell viability (B) and MMP-1 expression (C) were determined after pre-treatment of fermented and non-fermented L. angustifolia extracts at indicated concentration and irradiation of 80 mJ/cm2 UVB. Data are presented as the mean ± SD of three independent experiments. *p < 0.01 compared with the UVB-treated control. §p < 0.01 compared with cells treated with non-fermented L. angustifolia extract.

As shown in Fig. 2B, non-fermented L. angustifolia extracts indicated cytotoxicity at more than 50 μg/ml but fermented L. angustifolia extracts decreased the cytotoxicity at the same concentrations. In particular, fermented L. angustifolia extract showed no cytotoxicity up to 50 μg/ml. We selected 25 μg/ml as the maximal concentration for further experiments.

Fermented and non-fermented L. angustifolia extracts decreased UVB irradiation-increased MMP-1 expression in a dose-dependent manner (Fig. 2C). The MMP-1 expression levels of fermented L. angustifolia extracts were 8.2-14.0% lower than those of non-fermented L. angustifolia extracts at all concentrations (3.1-25 μg/ml). These results suggest that L. angustifolia extracts fermented by P. pentosaceus DK1 are effective in the suppression of MMP-1 protein expression increased by UVB irradiation.

Effect of Fermented L. angustifolia Extract on Collagen Production

When the skin is exposed to UVB, the levels of type 1 procollagen decrease. Anti-aging agents can protect against the degradation of collagen in UVB-exposed fibroblasts. The effects of fermented and non-fermented L. angustifolia extract were evaluated on UVB irradiation-induced decrease in collagen production (Fig. 3).

Figure 3. Effect of fermented and non-fermented L. angustifolia extracts on UVB-mediated production of procollagen type I in human skin fibroblasts. Data are presented as the mean ± SD of three independent experiments. *p < 0.01 compared with the UVB-treated control. §p < 0.01 compared with cells treated with non-fermented L. angustifolia extract.

In UVB-irradiated fibroblasts, the level of procollagen type 1 was 42% lower than the levels of collagen in normal cells without UVB irradiation. Fermented and non- fermented L. angustifolia extracts increased the levels of collagen reduced by UVB irradiation in a dose-dependent manner. Fermented L. angustifolia extracts presented a 20% higher level of collagen production than non-fermented L. angustifolia extracts at 12.5 and 25 μg/ml. However, treatment with fermented and non-fermented L. angustifolia extracts without UVB irradiation showed no increase in collagen production in fibroblasts (data not shown). These results suggest that the fermentation of L. angustifolia extracts recover collagen production after reduction from UVB irradiation.

Effect of Fermented L. angustifolia Extract on UVB- Induced ROS Generation

UVB irradiation increases ROS generation in cells and ROS-stimulated fibroblasts increases expression of MMP-1 protein and inhibits collagen production [27]. Therefore, it is important to identify the active compounds that are able to inhibit ROS generation stimulated by UVB for anti-aging activity [28]. H₂DCF-DA [29] was used to determine the effect of fermented and non-fermented L. angustifolia extracts on UVB-induced ROS generationin fibroblasts.

As shown in Fig. 4A, fibroblasts irradiated with UVB showed a 421.2% ± 20.4% higher ROS generation than cells without UVB irradiation. At all concentrations (3.1-25 μg/ml), fermented L. angustifolia extracts reduced the amount of ROS generation by more than 16% compared to those of non-fermented L. angustifolia extracts. Non-fermented and fermented L. angustifolia extracts at 6.3 μg/ml indicated 335.4% ± 11.3% and 283.2% ± 14.8% of ROS levels on UVB-irradiated fibroblasts.

Figure 4. Effect of fermented and non-fermented L. angustifolia extracts on UVB-induced ROS generation in human skin fibroblasts. The cells were treated with different concentrations of fermented and non-fermented L. angustifolia extracts for 24 h and then treated with 20 μM H2DCFH-DA for 30 min. Subsequently, the cells were irradiated with 80 mJ/cm₂ UVB. ROS generation was measured using a fluorescence reader (A) and fluorescence microscope (B). In Fig. 4B, the cells were treated with (a) non-treatment; (b) 80 mJ/cm₂ UVB irradiation; treatment of 6.3 (c) and 25 μg/ml (d) non-fermented L. angustifolia extract and UVB irradiation; treatment of 6.3 (e) and 25 μg/ml (f) fermented L. angustifolia extract and UVB irradiation. Scale bar, 50 μm. Data are presented as the mean ± SD of three independent experiments. *p < 0.01 compared with the UVB- treated control. §p < 0.01 compared with cells treated with non-fermented L. angustifolia extract.

ROS levels of fermented L. angustifolia extract at 25 μg/ml were 116.3% ± 12.5% lower than that in the non-fermented L. angustifolia extract (70.6%) (Figs. 4A and 4B). These results suggest that fermentation using P. pentosaceus DK1 enhances the inhibitory effect of L. angustifolia extract on UVB-induced ROS generation.

Component Analysis of L. angustifolia Extract before and after Fermentation

Changes in the composition of natural products fermented by microorganisms can affect their anti-aging activity [31]. In this study, the anti-aging activity of L. angustifolia extract was increased after fermentation, with reduced ROS generation, decreased MMP-1 expression, and increased collagen production.

We observed the changes in the components of the L. angustifolia extract after fermentation by P. pentosaceus DK1. The components were identified by HPLC, UV- spectrophotometry, and LC/ESI-MS (Table 2), and quantified using standard materials to confirm the change in the amounts of components. As shown in Fig. 5, luteolin-7-O- glucoside (peak 3) and apigenin-7-O-glucoside (peak 5) in fermented L. angustifolia extract were decreased compared with the non-fermented L. angustifolia extract and luteolin (peak 6) and apigenin (peak 7), as their aglycone components, were increased, respectively. In addition, chlorogenic acid (peak 1) decreased after fermentation. Interestingly, there was no significant change in luteolin 7-O-glucuronide (peak 2) after fermentation, despite having a similar structure to luteolin-7-O-glucoside.

Table 2 . Mass and UV spectrum of identified compounds in L. angustifolia extract fermented or non-fermented by P. pentosaceus DK1.

HPLC Peak No.	Name of the compound	Molecular formula	Retention time (min)	Measurement
				λmax (nm)	Negative ions (m/z) [M+H]-	Positive ions (m/z) [M+H]+
1	Chlorogenic acid	C16H18O9	49.754	244, 322	353.4	355.1
2	Luteoln-7-O-glucuronide	C21H18O12	69.408	253, 347	-	463.8
3	Luteolin-7-O-glucoside	C21H20O11	80.261	254, 346	447.9	449.4
4	Rosmarinic acid	C18H15O8	82.956	290, 331	359.2	-
5	Apigenin-7-O-glucoside	C21H20O10	92.844	267, 334	431.8	433.8
6	Luteolin	C15H10O6	98.003	254.349	285.2	287.4
7	Apigenin	C15H10O5	106.566	270, 334	269.1	271.2

Figure 5. HPLC chromatograms of L. angustifolia extracts fermented (A) and non-fermented (B). The peaks were detected at 365 nm of wavelength for 150 min.

Anti-Aging Effect of the Identified Compounds in UVB-Irradiated Human Skin Fibroblasts

Identified components of the L. angustifolia extract converted by P. pentosaceus DK1 were evaluated on UVB- induced ROS generation, MMP-1 expression and collagen degradation. UVB was irradiated to human fibroblasts after treatment with luteolin, luteolin-7-O-gluconisde, apigenin and apigenin-7-O-glucoside.

Luteolin and apigenin reduced 80% and 85.8% of ROS generation increased by UVB irradiation compared to their glycosides (Fig. 6A). Luteolin and apigenin also reduced 26.5% and 22.4% of UVB-enhanced MMP-1 levels less than their glycosides, respectively (Fig. 6B). In UVB-reduced collagen production, luteolin and apigenin increased 9.5% and 12.5% more than their glycosides (Fig. 6C).

Figure 6. Effect of identified compounds on ROS generation, MMP-1 expression and collagen production in UVB-irradiated human skin fibroblasts. Data are presented as mean ± SD of three independent experiments. *p < 0.01 compared with the negative control. §p < 0.01 compared with cells treated with non-fermented L. angustifolia extract.

Taken together, these results suggest that aglyconation of the flavonoids in L. angustifolia extract through fermentation by P. pentosaceus DK1 is a major factor in promoting anti- aging activity through a reduction in UVB-induced ROS production.

Recently, fermentation using microorganisms and enzymes has been shown to enhance both the safety and activity of natural products for cosmetics agents [31]. The mechanism of efficacy enhancement of materials by microbial fermentation is as follows: 1) the microorganism breaks down the substance to a small size to increase its absorption into the skin; 2) the microorganism enhances the product’s physiological activities by removing glycoside; 3) favorable nutrients including vitamins and organic acids are produced during microbial cultivation [32, 33]. In addition, heavy metals and pesticide residues in the extracts of raw plants are harmful to the skin, microorganisms can enhance the safety of the extracts through decomposition and/or adsorption [34].

In this study, L. angustifolia extract was fermented using P. pentosaceus DK1, a new microorganism isolated from Diospyros kaki fruit, commonly known as persimmon, to develop a novel material for the inhibition of skin-aging caused by UVB irradiation. We optimized the fermentation conditions using P. pentosaceus DK1 in 1.38% MRS medium supplemented with 0.5% L. angustifolia extract. Fermentation of L. angustifolia extract by P. pentosaceus DK1 increased the anti-aging activities while decreasing MMP-1 expression and ROS generation, and increased collagen production in UVB-irradiated human skin fibroblasts. We identified that the flavonoid glycosides, luteolin-7-O-glucoside and apigenin-7-O-glucoside, contained in L. angustifolia extract were decreased and the contents of sugar-removed luteolin and apigenin after fermentation were increased. These results suggest that P. pentosaceus DK1 may remove glucosides from flavonoid glucosides as well as degradation of tannins. Luteolin and apigenin indicated higher anti- aging activities than the glucosides of them on MMP-1 expression, collagen production and ROS generation in UVB-irradiated fibroblasts. Sung, et al. reported that luteolin had a higher antioxidant activity than the glycoside in intracellular systems [35]. Luteolin and apigenin were also reported to suppress UV-induced MMP-1 expression through anti-oxidative activity on human skin fibroblasts and human keratinocytes (HaCaT cells) [37]. Apigenin increases collagen synthesis in fibroblasts [38].

In conclusion, P. pentosaceus DK1 could increase the anti-oxidative and aging activities of the L. angustifolia extract through bioconversion which increases the levels of apigenin and luteolin by aglyconation of their glucosides. Thus, L. angustifolia extract fermented by P. pentosaceus DK1 could have applicability as an anti-aging agent.

This study was supported by a grant of the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (Grant No. HN15C0104).

The authors have no financial conflicts of interest to declare.