Preventive Effects of Anthraquinones Isolated from an Endophytic Fungus, Colletotrichum sp. JS-0367 in Tumor Necrosis Factor-α-Stimulated Damage of Human Dermal Fibroblasts

Reactive oxygen species (ROS) are a major causative factor of inflammatory responses and extracellular matrix degradation. ROS also cause skin aging and diverse cutaneous lesions. Therefore, antioxidants that inhibit the generation of ROS may be beneficial in the relief of skin aging and diseases. We investigated the anti-skin aging effect of anthraquinones from cultures of Colletotrichum sp., an endophytic fungus isolated from Morus alba L. using human dermal fibroblasts (HDFs). We preferentially evaluated the preventive effects of anti-oxidative anthraquinones (1, 4) against the generation of ROS, nitric oxide (NO), and prostaglandins-E2 (PGE2). Among them, 1,3-dihydroxy-2,8-dimethoxy-6-methylanthraquinone (1) suppressed the generation of ROS, NO, and PGE2 in tumor necrosis factor-alpha (TNF-α)-stimulated HDFs. Compound 1 reversed the TNF-induced increase in matrix metalloproteinase (MMP)-1 and a decrease in procollagen I α1 (COLIA1). It also suppressed inducible NO synthase, cyclooxygenase-2, interleukin (IL)-1β, IL-6, and IL-8, which upregulate inflammatory reactions. Mechanistically, compound 1 suppressed nuclear factor-κB, activator protein 1, and mitogen-activated protein kinases in TNF-α-stimulated HDFs. These results suggest that compound 1 may be beneficial for improving skin aging and diverse cutaneous lesions.


Introduction
The skin is the primary protective organ of the human exodermal system and is in direct contact with potentially harmful factors. Because of its direct exposure to the external environment, the skin is the most prominent tissue affected by aging and damage [1,2]. Intrinsic aging involves damage that occurs with time due to decreased skin cell activity by reactive oxygen species (ROS) produced during skin cell metabolism [3]. Extrinsic aging is induced by exposure to external environmental hazards, such as pollution, chemicals, smoking, and ultraviolet (UV) radiation [4][5][6].
UV radiation is a key extrinsic stimulator. It mainly results in cumulative skin damage, referred to as photoaging [2]. UV radiation is classified based on wavelength into three categories UV-A, B, C. UVA and UVB comprise the entire UV spectrum on Earth. These wavelengths pass through the Earth's atmospheric layers to the surface and can damage living things [7]. Exposure of the skin to UV radiation in everyday life produces a variety of physiological effects that include sunburn, photoaging, skin pigmentation, and production of ROS [8]. The foregoing indicates that anthraquinones should prevent oxidative stress related damage to the skin ECM. To explore this, we investigated the prevent effect against skin damages of anthraquinones obtained from Colletotrichum sp. using human dermal fibroblasts (HDFs) to identify potential candidates. Presently, we describe the prevent effect of skin damages of anti-oxidative anthraquinones and verify the mechanism of the active compound for TNF-α-stimulated HDFs.
The cells were seeded in each size well plate at a density of 3 × 10 4 cells/cm 2 and allowed to adhere. The media were subsequently changed with serum-free medium and incubation was continued until analysis. A stock solution (20 µg/mL) of TNF-α (Pepro-Tech, Rocky Hill, NJ, USA) was prepared. The solution also contained 1% bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA) in Dulbecco's phosphate-buffered saline (DPBS; WELGENE Inc. Gyeongsan-si, Gyeongsangbuk-do, Korea) Stock solutions (100 mM) of compounds 1-4 were prepared in dimethyl sulfoxide (DMSO; Sigma-Aldrich).

Determination of ROS
Intracellular ROS levels were measured using the dichlorofluorescin diacetate (DCFDA) assay [39]. Briefly, the cells were seeded in a 96-well plate (1 × 10 4 cells/well) and continuously starved in serum-free medium during a 24-h incubation. The serumstarved HDFs were treated with 50 and 100 µM of compounds 1-4 for 1 h, and then with 20 ng/mL TNF-α. After incubation for 12 h, 10 µM DCFDA (Sigma-Aldrich, St. Louis, MO, USA) was exposed for 15 min and were continuously washed with DPBS. Thereafter, the fluorescence intensity of DCFDA was analyzed using a SPARK 10 M microplate reader (Tecan, Männedorf, Switzerland) at a wavelength of 485 nm.

Determination of Nitric Oxide (NO) Production
Nitric oxide (NO) in cell supernatants was measured using Griess reagent [40]. Briefly, the cells were seeded in a 96-well plate (1 × 10 4 cells/well) and incubated for 24 h in serum-free medium. The serum-starved HDFs were exposed to 50 and 100 µM of compounds 1-4 for 1 h, and then with 20 ng/mL TNF-α for 24 h. Thereafter, the supernatant
The cells were seeded in each size well plate at a density of 3 × 10 4 cells/cm 2 and allowed to adhere. The media were subsequently changed with serum-free medium and incubation was continued until analysis. A stock solution (20 µg/mL) of TNF-α (PeproTech, Rocky Hill, NJ, USA) was prepared. The solution also contained 1% bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA) in Dulbecco's phosphate-buffered saline (DPBS; WELGENE Inc. Gyeongsan-si, Gyeongsangbuk-do, Korea) Stock solutions (100 mM) of compounds 1-4 were prepared in dimethyl sulfoxide (DMSO; Sigma-Aldrich).

Determination of ROS
Intracellular ROS levels were measured using the dichlorofluorescin diacetate (DCFDA) assay [39]. Briefly, the cells were seeded in a 96-well plate (1 × 10 4 cells/well) and continuously starved in serum-free medium during a 24-h incubation. The serum-starved HDFs were treated with 50 and 100 µM of compounds 1-4 for 1 h, and then with 20 ng/mL TNF-α. After incubation for 12 h, 10 µM DCFDA (Sigma-Aldrich, St. Louis, MO, USA) was exposed for 15 min and were continuously washed with DPBS. Thereafter, the fluorescence intensity of DCFDA was analyzed using a SPARK 10 M microplate reader (Tecan, Männedorf, Switzerland) at a wavelength of 485 nm.

Determination of Nitric Oxide (NO) Production
Nitric oxide (NO) in cell supernatants was measured using Griess reagent [40]. Briefly, the cells were seeded in a 96-well plate (1 × 10 4 cells/well) and incubated for 24 h in serumfree medium. The serum-starved HDFs were exposed to 50 and 100 µM of compounds 1-4 for 1 h, and then with 20 ng/mL TNF-α for 24 h. Thereafter, the supernatant was collected and homogenized with Griess reagent, and incubated at room temperature. After 10 min, the reaction was analyzed using a SPARK 10 M microplate reader (Tecan, Männedorf, Switzerland) at a wavelength of 540 nm. Sodium nitrite (NaNO 2 ) was used for the detection of nitrite concentration (µM).

Determination of Gene Expression
The mRNA expression of cells was determined using quantitative real-time polymerase chain reaction (qRT-PCR) [41]. Briefly, the cells were seeded in a 6-well plate (3 × 10 5 cells/well) and incubated for 24 h. The cells were continuously starved in serumfree medium for 24 h. Thereafter, the HDFs were exposed to 50 and 100 µM compound 1 for 1 h, and then with 20 ng/mL TNF-α for 4 h (IL-1β, IL-6, and IL-8) and 24 h (PGE 2 , MMP-1, and COLIA1). After incubation for the defined time, the cells were washed with DPBS and continuously harvested with lysis buffer for RNA isolation. Total cellular RNA was isolated using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA). Conversion of the isolated RNA to cDNA was accomplished using the RevertAid First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Waltham, MA, USA). qRT-PCR was carried out using the Quant Studio 3 real-time PCR system (Applied Biosystems, Waltham, MA, USA) and PowerUp SYBR PCR Master Mix (Applied Biosystems) at 95 • C for 10 min, followed by 40 cycles of amplification at 95 • C for 1 s and 60 • C for 30 s. The primers used are shown in Table 1. The mRNA expression was normalized and calculated based on β-actin expression and the ratio to 100% of the non-treated group.

Determination of Protein Expression
The protein expression of cells was measured using Western blot analysis [42]. Briefly, the cells were seeded in a 6-well plate (3 × 10 5 cells/well) and continuously starved in serum-free medium during a 24-h incubation. The serum-starved HDFs were exposed to 50 and 100 µM of compound 1 for 1 h and then to 20 ng/mL TNF-α for 15 min (phos-phoextracellular signal-regulated kinase [p-ERK], ERK, phospho-C-Jun, T-terminal ki-nase [p-JNK], JNK, p-p38, p38, and glyceraldehyde 3-phosphate dehydrogenase [GAPDH]), 4 h (nuclear factor-kappa B [NF-κB], activator protein-1 [AP-1] and GAPDH), and 6 h (iNOS, COX-2 and GAPDH). Among the proteins described above, the proteins except for GAPDH are activated by TNFα, and occur increasing proinflammatory cytokines and collagenase. GAPDH is housekeeping gene, that was utilized for the normalization of data.
After incubation for each time, the cells were washed with DPBS and continuously harvested with 1× radioimmunoprecipitation assay buffer (RIPA buffer; Tech & Innovation, Gangwon, Korea). Protein concentration was determined using the Pierce™ BCA Protein Assay Kit (Pierce, Rockford, IL, USA).
Equal amounts of protein samples were separated by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride membrane (Merck Millipore, Darmstadt, Germany). The membranes were blocked in 5% non-fat milk and then the primary antibodies individually added to a membrane and incubated for 4 h at room temperature. The antibodies were directed to iNOS, COX-2, NF-κB (p65), AP-1, ERK1/2, phospho-ERK1/2, p38, phospho-p38, JNK, phospho-JNK, and GAPDH (all from Cell Signaling Technology, Danvers, MA, USA). The antibodies were applied in immunoreaction enhancer solution (Can get signal, Toyobo, Osaka, Japan). After washing, the membranes were incubated with appropriate secondary antibodies (Cell Signaling Technology) for 1 h at room temperature. Subsequently, protein signals were visualized by enhanced chemiluminescence using the Fusion Solo Chemiluminescence System (PEQLAB Biotechnologie GmbH, Erlangen, Germany) and SuperSignal ® West Femto Maximum Sensitivity Chemiluminescent Substrate (Pierce). The relative expression of proteins was quantified using ImageJ software (version 1.8.0) from the National Institutes of Health (Bethesda, MD, USA).

Statistical Analyses
The data from the experiments performed in triplicate are expressed as mean ± standard error of the mean (SEM). Statistical analyses were performed using one-way analysis of variance (ANOVA) complemented by the Tukey's honest significance test. The results were considered statistically significant at p < 0.05, p < 0.01, and p < 0.001.

Effect of Anthraquinones on Production of Intracellular ROS, and Proinflammatory Mediators NO and PGE2 in TNF-α-Stimulated HDFs
In our previous study, four anthraquinones (1-4) were isolated from an endophytic fungus, Colletotrichum sp. JS-0367. Among them, 1,3-dihydroxy-2,8-dimethoxy-6methylanthraquinone (1 in Figure 1) and evariquinone (4 in Figure 1) displayed potent radical scavenging activities of 1,1-diphenyl-2-picrylhydrazyl (DPPH). Compound 4 also displayed neuroprotective effects in glutamate-stimulated murine HT22 hippocampal neuronal cells. The compound was the most potent of the four. It prevented the generation of ROS, intracellular calcium ion levels, and phosphorylation of MAPKs in glutamatemediated apoptosis [38]. Moreover, many studies have reported that anthraquinones, as antioxidants, have the potential to suppress oxidative stress within cells [43][44][45][46]. Thus, we focused on suppressing the oxidative potential of these anthraquinones in HDFs. In preliminary experiments, the four anthraquinones were not cytotoxic to HDFs at 100 µM (data not shown).
Therefore, compounds 1-4 were expected to have oxidative stress-induced anti-skin aging effects without damage to HDFs. As mentioned above, UV exposure induces intercellular ROS generation and proinflammatory cytokines, such as TNF-α. Furthermore, mitochondrial-derived ROS act as signaling molecules that upregulate inflammatory cytokines, including TNF-α. Excessively increased TNF and ROS regulate increased levels of each other and are activated, triggering diverse inflammatory responses and collagen cleavage. Thus, ROS and TNF-α can be used to test mechanisms similar to those of processes that are induced by UV-induced skin inflammation and aging. We researched the inhibitory effects of anthraquinones 1-4 on the production of TNF-α-induced ROS, NO, and PGE 2 .
The change in intracellular ROS levels were measured using DCFDA fluorogenic dye. Serum-starved HDFs were treated with compounds 1-4 and subsequently with TNF-α for 12 h. The cells were then exposed to 10 µM DCFDA for 15 min and their fluorescence intensities were measured. TNF-α stimulation increased 2.30 ± 0.03-fold (p < 0.01, Figure 2A). It was substantially but non-significantly reduced 1.79 ± 0.02-fold and 1.83 ± 0.01-fold (p < 0.05) after treatment with 100 µM of 1 and 4, respectively. These results showed that ROS generation was suppressed in TNF-α-stimulated HDFs by compounds 1 and 4. This might be the mechanism involved in ameliorating skin damage induced by oxidative stress.
Antioxidants 2021, 10, x FOR PEER REVIEW 7 of 17 (3) and evariquinone (4) on production of intracellular ROS, proinflammatory mediators NO and PGE2 in TNF-α-stimulated HDFs. HDFs were untreated or exposed to TNF-α, followed by treatment with 1-4 for 24 h. The levels of (A) ROS and (B) NO and (C) PGE2 were determined using DCFDA dye, Griess reaction assay, and ELISA. The data are presented as mean ± SEM of at least three independent experiments. ## p < 0.01 and ### p< 0.001 difference compared to untreated cells. * p < 0.05 and ** p < 0.01 difference compared to TNF-α-stimulated cells.
The four compounds are presented typical anthraquinone structure. In moieties of anthraquinone, they equally attached with the one hydroxyl group, the one methyl group and the one methoxy group to carbon positions 1, 6 and 8 (C-1, 6, 8), respectively. They are substituted the hydroxy groups or the methoxy groups at C-2, 3, but their numbers and positions are different. In detail, compound 1 was attached with the one hydroxyl groups at C-3 and the one methoxy groups at C-2. Differently, compound 3 was attached at changed positions, that with the one hydroxyl group at C-2, and the one methoxy groups at C-3. Compounds 2 and 4 were bound to only one functional group, the two methoxy groups or the two hydroxyl groups at C-2 and 3. In ROS generation, compounds 1 and 4 were presented with EC 20 (concentration of compound that produces 20% biological effect) of 51.1 and 71.2 µM, respectively, whereas compounds 2 and 3 were not shown below 100 µM (Tables S1-S6). These results indicate that the mechanism of ROS scavenging by anthraquinone, may depend on the correlation between the hydroxyl group at C-3 and the ketone group at C-10).
TNF-α treatment remarkably increased the levels of PGE 2 from 20.3 ± 0.10 pg/mL to 47.4 ± 1.33 pg/mL ( Figure 2C). Compound 1 treatment dramatically reduced the increased PGE 2 levels in a dose-dependent manner (50 µM; 41.8 ± 0.98 pg/mL, 100 µM; 31.8 ± 2.93 pg/mL, p < 0.05). Compound 3 treatment also suppressed the increased PGE 2 levels in a dose-dependent manner (50 µM; 45.5.8 ± 1.64 pg/mL, 100 µM; 39.4 ± 3.05 pg/mL). Compounds 2 and 4 did not inhibit PGE 2 production. In PGE 2 generation, compounds 1 and 3 were showed with EC 20 (concentration of compound that produces 20% biological effect) of 49.3 and 75.3 µM, respectively, whereas compounds 2 and 4 were not significant EC 50 below 100 µM (Table S1). The mechanism of PGE 2 production was also expected to be related to the hydroxyl group of C-3 and the ketone group of C-10, but it was not clear because other differences were also discovered.
Taken together, these results indicate that compound 1 potently scavenges excess ROS and suppresses the production of NO and PGE 2 , compared with compounds 2-4. For this reason, we subsequently focused on compound 1.

Effect of Compound 1 on COX-2 and iNOS Expression in TNF-α-Stimulated HDFs
The activation of COX-2 and iNOS is important role in NO and PGE 2 production. We investigated the effect of compound 1 on the protein expression of COX-2 and iNOS by serum-starved TNF-α-stimulated HDFs.
dose-dependent manner ( Figure 3B). The protein expression of iNOS was significantly reduced to 3.78 ± 0.34 (p < 0.05) and 3.10 ± 0.17-fold (p < 0.01) by 50 and 100 µM of compound 1, respectively. The protein expression of COX-2 was also reduced to 13.0 ± 1.06 (not significant) and 8.96 ± 1.33-fold (p < 0.05) by 50 and 100 µM of compound 1, respectively. These results indicate that compound 1 may inhibit inflammation in TNF-α-stimulated HDFs. Previous studies have demonstrated that anthraquinones inhibit inflammatory mediators (NO and COX-2) and suppress inflammatory responses via the NF-κB pathway [47,48]. Consistent with the prior data, compound 1 also suppressed the inflammatory Previous studies have demonstrated that anthraquinones inhibit inflammatory mediators (NO and COX-2) and suppress inflammatory responses via the NF-κB pathway [47,48]. Consistent with the prior data, compound 1 also suppressed the inflammatory response to iNOS and COX-2 in TNF-α-stimulated HDFs. Therefore, compound 1 can ameliorate inflammation induced by the generation of ROS.

Effect of Compound 1 on MMP-1 and COLIA1 mRNA in TNF-α-Stimulated HDFs
The ECM of skin is a complex collection of collagen and non-collagen components. The generation of ROS is induced by external stimuli that include UV radiation. ROS alters the gene and protein structure, including collagen and collagen-degrading enzymes. Ultimately, the changes damage the skin ECM, leading to aging related features like wrinkles [5,49]. MMP-1 is a collagenase that plays a critical role in the degradation of collagen in the skin. Therefore, inhibitors of MMP-1 activity may be a potential candidate for anti-skin aging, such as wrinkle formation [50]. We investigated MMP-1 expression in TNF-α-stimulated HDFs.
Collagen was synthesized from procollagen, which is a precursor molecule containing additional peptide sequences. Because these sequences are cleaved during collagen secretion, several sequences have indirect information about collagen synthesis levels. Thus, we determined COLIA1 to investigate collagen synthesis. As shown in Figure 4A, TNF-α treatment significantly decreased the mRNA expression of COLIA1 to 0.37 ± 0.00-fold compared with the untreated group. It was significantly increased to 0.56 ± 0.03-fold (p < 0.05) and 0.69 ± 0.06-fold (p < 0.01) by 50 and 100 µM of compound 1, respectively. Analogously, TNF-α treatment also reduced the protein secretion of COLIA1 to 6.25 ± 0.70 ng/mL compared with untreated group (15.8 ± 0.20 ng/mL). COLIA1 was decreased to 7.09 ± 0.22 (not significant) and 8.95 ± 0.48 ng/mL (p < 0.01) by 50 and 100 µM of compound 1, respectively. These results demonstrate that compound 1 increased both gene expression and protein of procollagen in TNF-α-stimulated HDFs. Therefore, compound 1 might potentially enhance of skin ECM degradation by oxidative stress.

Effect of Compound 1 on Proinflammatory Cytokines in TNF-α-Stimulated HDFs
Cellular oxidative stress produces proinflammatory cytokines, such as TNF-α, IL-1β, IL-6, and IL-8, and is involved in the upregulation of the inflammatory response [51,52]. These inflammatory responses cause skin aging and diverse cutaneous lesions [53,54]. To assess the inhibitory effect of compound 1 on the inflammatory response in skin dermal cells, we verified the effect of compound 1 on IL-1β, IL-6, and IL-8 mRNA expression in TNF-α-stimulated HDFs. To investigate whether compound 1 inhibited the inflammatory response in skin cells, we directly determined mRNA gene expression of IL-1β, IL-6, and IL-8 in TNF-α-stimulated HDFs.

Effect of Compound 1 on NF-κB and AP-1 Expression in TNF-α-Stimulated HDFs
MMP-1 and proinflammatory cytokine levels are upregulated by the AP-1 and NF-κB. Thus, compound 1 might promote collagen synthesis by inhibiting MMP-1 and proinflammatory cytokine expression levels. Western blot analysis was conducted to further investigate the role of the AP-1 and NF-κB in the action of compound 1.

Effect of Compound 1 on TNF-α-Stimulated Phosphorylation of MAPKs in HDFs
AP-1 and NF-κB pathways regulate MMP-1 and proinflammatory cytokines. The pathways are regulated by signaling of MAPKs. To determine whether compound 1 could inhibit MAPK phosphorylation in TNF-α stimulation, we investigated the effects of compound 1 on TNF-α-stimulated phosphorylation of MAPKs in HDFs. Serum-starved HDFs were challenged with compound 1 followed by TNF-α for 15 min. Expression of protein was measured by Western blotting.

Effect of Compound 1 on TNF-α-Stimulated Phosphorylation of MAPKs in HDFs
AP-1 and NF-κB pathways regulate MMP-1 and proinflammatory cytokines. The pathways are regulated by signaling of MAPKs. To determine whether compound 1 could inhibit MAPK phosphorylation in TNF-α stimulation, we investigated the effects of compound 1 on TNF-α-stimulated phosphorylation of MAPKs in HDFs. Serum-starved HDFs were challenged with compound 1 followed by TNF-α for 15 min. Expression of protein was measured by Western blotting.

Conclusions
ROS are major causative factors of inflammatory responses and ECM degradation. ROS cause skin aging and diverse cutaneous lesions. Thus, ROS inhibitors may lessen skin aging and diseases. The present data demonstrate that 1,3-dihydroxy-2,8-dimethoxy-6methylanthraquinone (1), a novel anthraquinone isolated from Colletotrichum sp. JS-0367 reduces TNF-α-stimulated ROS, NO, and PGE 2 , attenuated MMP-1 expression, and enhances collagen synthesis. Furthermore, compound 1 inhibits the expression of TNF-α-stimulated proinflammatory cytokine mediators, including iNOS and COX-2, and proinflammatory cytokines IL-1, IL-6, and IL-8. The mechanism by which compound 1 inhibits TNF-αstimulated skin aging in HDFs involves the inhibition of NF-κB, AP-1, and MAPKs activation. The present data provide the potent evidence that compound 1 may be beneficial in improving skin damage. Although more extensive studies are needed for a thorough understanding of the protective effects of compound 1 on skin aging, the compound is a potential candidate for improving skin aging and diverse cutaneous lesions.