Korean Red Ginseng Attenuates Particulate Matter-Induced Senescence of Skin Keratinocytes

Skin is a direct target of fine particulate matter (PM2.5), as it is constantly exposed. Herein, we investigate whether Korean red ginseng (KRG) can inhibit PM2.5-induced senescence in skin keratinocytes. PM2.5-treated human keratinocyte cell lines and normal human epidermal keratinocytes showed characteristics of cellular senescence, including flat and enlarged forms; however, KRG suppressed them in both cell types. Moreover, while cells exposed to PM2.5 showed a higher level of p16INK4A expression (a senescence inducer), KRG inhibited its expression. Epigenetically, KRG decreased the expression of the ten-eleven translocation (TET) enzyme, a DNA demethylase induced by PM2.5, and increased the expression of DNA methyltransferases suppressed by PM2.5, resulting in the decreased methylation of the p16INK4A promoter region. Additionally, KRG decreased the expression of mixed-lineage leukemia 1 (MLL1), a histone methyltransferase, and histone acetyltransferase 1 (HAT1) induced by PM2.5. Contrastingly, KRG increased the expression of the enhancer of zeste homolog 2, a histone methyltransferase, and histone deacetyltransferase 1 reduced by PM2.5. Furthermore, KRG decreased TET1, MLL1, and HAT1 binding to the p16INK4A promoter, corresponding with the decreased mRNA expression of p16INK4A. These results suggest that KRG exerts protection against the PM2.5-induced senescence of skin keratinocytes via the epigenetic regulation of p16INK4A.


Introduction
Air pollutants in cities represent a serious health problem, with fine particulate matter (PM 2.5 ) accounting for a large portion, owing to coal combustion and diesel exhaust fumes [1]. Both indoors and outdoors, PM 2.5 damages several human systems, including the cardiovascular, central nervous, and pulmonary immune systems [2][3][4][5]. PM 2.5 mainly penetrates the skin barrier via the appendix pathway and stratum corneum, which can interfere with skin protection activities, resulting in wrinkles and thickening [6,7]. Furthermore, PM 2.5 may induce both oxidative stress and inflammation, leading to skin aging [8]. Previously, we have reported that PM 2.5 contributes to senescence in human keratinocytes through oxidative-stress-dependent epigenetic regulation [9].
Changes in the external environment, such as exposure to PM 2.5 , have been reported to affect gene expression through epigenetic regulation [10,11]; wherein, PM exposure leads to hypo-methylation in the promoters of genes involved in oxidative stress, inflammation, DNA repair, and cell cycle regulation.
In mammals, gene expression is strictly controlled by epigenetic modifications, such as DNA methylation and histone modifications. In general, the methylation of DNA via DNA

PM 2.5 Preparation
Standard diesel PM (SRM 1650b, Sigma-Aldrich Co., Ltd.), with an average diameter of 0.18 µm, was used, mostly comprising polycyclic aromatic hydrocarbons (PAHs) and nitro-PAHs. Information on nitro-PAHs and PAHs in SRM 1650b, such as the certified mass fraction values, can be found in previously published papers [32,33]. PM 2.5 in DMSO was prepared for stock, with DMSO as the control and with concentrations not exceeding 0.01%.

Keratinocyte Culture
The human keratinocytes cell line HaCaT was obtained from the CLS Cell Lines Service (Eppelheim, Germany). Normal human epidermal keratinocytes (NHEK) were obtained from Thermo Fisher Scientific. The HaCaT cells were seeded using a complete DMEM medium with 10% fetal bovine serum and 1% antibiotics. NHEK cells were cultured using EpiLife serum-free medium with added EpiLife undefined growth supplements (Thermo Fisher Scientific). All cells were incubated at 37 • C in a 5% CO 2 incubator.

Cell Viability
To detect the cytotoxicity effect of KRG, cells were seeded (1 × 10 5 cells/mL) and treated with 2, 4, 6, 8, 10, 20, 40, 80, or 100 µg/mL of KRG. After 48 h, the cells were cocultured with MTT solution. After a four-hour-long incubation period, the transformed formazan crystals were dissolved in DMSO. The formazan product was measured at an absorbance of 540 nm under a scanning multi-well spectrophotometer (Thermo Fisher Scientific).

Colony Formation Detection
To detect the colony-forming ability (CFA), cells were seeded to 5 × 10 2 cells/mL in a 60-mm dish. The colony expansion of single cells was allowed to progress for 3 days, after which they were pre-treated with 20 µg/mL of KRG, and then 50 µg/mL of PM 2.5 was added for 7 days. The colony formation detection was performed using a Diff-Quik kit (Sysmex Corporation, Kobe, Japan). The Diff-Quik staining solution was used to stain the resulting colonies after fixing with Diff-Quik fixation solution.

Immunofluorescence
To measure the expression of epigenetic-related proteins, the cells were seeded to 1.0 × 10 5 cells/mL on microscope slides, were pre-treated with 20 µg/mL of KRG, and then treated with PM 2.5 (50 µg/mL) for 24 h. Subsequently, the cells were fixed with 4% paraformaldehyde for 15 min and permeabilized with a permeable reagent (phosphatebuffered saline, PBS, containing 0.1% Triton X-100) for 15 min. The cells were incubated with a blocking solution (PBS containing 3% BSA) for 1 h and then incubated with the primary antibody diluted in the blocking solution for 2 h. The secondary antibody was incubated for 1 h using Alexa488-binding secondary antibody (Santa Cruz Biotechnology). The cells were observed under a confocal microscope (Zeiss, LSM 510 software) after mounting with a DAPI-containing mounting medium.

β-Galactosidase Activity Detection
To measure β-galactosidase activity, an indicator of cellular senescence, the cells were seeded on microscope slides, pre-treated with 20 µg/mL of KRG, and then treated with PM 2.5 (50 µg/mL) for 24 h. Subsequently, the cells were stained with SPiDER-βGal staining and then incubated at 37 • C for 15 min. After that, the cells were mounted using DAPIcontaining mounting medium and images were captured using a confocal microscope.

Western Blot Analysis
The cells were seeded to 1.0 × 10 5 cells/mL on a 60-mm culture dish. They were pretreated with 20 µg/mL of KRG and then treated with PM 2.5 (50 µg/mL) for 24 h. The cells were collected, and the proteins were extracted by lysis. Subsequently, the lysates of cells (60 µg of protein) were separated via 6, 10, or 12% SDS-polyacrylamide gel electrophoresis and then transferred into the membrane, which was immunoblotted with specific primary antibodies. The following primary antibodies were used; p16 INK4A , DNMT1, DNMT3A, DNMT3B, TET1, TET2, TET3, EZH2, H3K27Me3, MLL1, H3K4Me3, HDAC1, and HAT1. In addition, primary antibodies against actin and TBP were used as a loading control. Membranes bound with primary antibodies were contacted with secondary antibodies (Pierce, Rockland, IL, USA), and bands of protein were assessed using a western blotting detection kit (Amersham, Little Chalfont, Buckinghamshire, UK).

Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)
The cells were seeded to 1.0 × 10 5 cells/mL on a 60-mm culture dish. They were pre-treated with 20 µg/mL of KRG and then treated with PM 2.5 (50 µg/mL) for 24 h. For the real-time qRT-PCT, the qRT-PCR reaction system contained 5.0 µL of 2× SYBR Green Mixture, forward primer (5 µM), and reverse primer (5 µM) each, 1.0 µL of cDNA, and double-distilled water. The qRT-PCR conditions were as follows: pre-denatured for 10 min at 95 • C; 40 cycles at 95 • C for 15 s; and at 60 • C for 1 min on a Bio-Rad iQ5 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA). The qRT-PCR primers used were as follows: forward primer 5 -CTCGTGCTGATGCTACTGAGGA-3 and reverse primer 5 -GGTCGGCGCAGTTGGGCTCC-3 for the amplification of p16 INK4A and forward primer 5 -CACCTTCTACAATGAGCTGCGTGT-3 and reverse primer 5 -CACAGCCTGGATAGCAACGTACA-3 for actin. A prepared 1% agarose gel with ethidium bromide was used to resolve the amplification, and it was photographed under ultraviolet light using Image Quant™ TL analysis software version 10.1 (Amersham Bioscience, Uppsala, Sweden).

Quantitative Methylation-Specific PCR (qMSP) and Bisulfite Sequencing
To measure the methylation of the p16 INK4A promoter, a specific methylation kit (Zymo Research, Tustin, CA, USA) was used for DNA (1 µg) conversion, which involves the chemical bisulfite conversion of unmethylated cytosine to uracil and the protection of methylated cytosine. For qMSP detection, qMSP primer pairs were used to assess gene promoter methylation sites, which are located close to the putative transcription start site of the 5 CpG islands. The qMSP was carried out using bisulfite-treated samples, was normalized based on Alu element amplification, and was assessed using the CFX96 TM real-time system (Bio-Rad Laboratories). In addition, for bisulfite sequencing, bisulfitetreated DNA and JumpStart REDTaq DNA polymerase (Sigma-Aldrich Co., Ltd.) were used for the template and amplification, respectively. A gel extraction kit (Qiagen GmbH, Hilden, Germany) and clones using the TOPO TA vector system (Invitrogen, Carlsbad, CA, USA) were used to purify the PCR products in bisulfite sequencing. Then, a NucleoSpin plasmid isolation kit (Macherey-Nagel, Düren, Germany) was used for the isolation and purification of each clone. An M13F primer was used to sequence the randomly selected positive clones. After that, the methylation status of each CpG dinucleotide was examined. qMSP and bisulfite sequencing primers were obtained based on previous reports [9,35,36].

Chromatin Immunoprecipitation (ChIP) Assay
To measure the binding of epigenetic-related proteins to the p16 INK4A locus, the cells were pre-treated with 20 µg/mL of KRG and then treated with PM 2.5 (50 µg/mL) for 24 h. Antibodies against TET1, DNMT1, EZH2, MLL1, HDAC1, HAT1, and normal rabbit immunoglobulin G (IgG) were used for ChIP assays, which were assessed using a SimpleChIP ® enzymatic ChIP kit (Cell Signaling Technology). DNA (200 ng) recovered from the immunoprecipitated complexes was subjected to qPCR. Primers for the p16 INK4A locus have been described previously [9]. The forward primer used was 5 -CCCCTTGCCTGGAAAGATAC-3 and the reverse primer was 5 -AGCCCCTCCTCTTT CTTCCT-3 .

Statistical Analysis
Each experiment was repeated three times, and the results are indicated as the mean ± standard error of the mean (SEM). For the statistical analysis, SigmaStat software v12 (SPSS, Chicago, IL, USA) was used. One-way analysis of variance (ANOVA) and Tukey's post hoc test were used to assess the results. At p < 0.05, differences were considered statistically significant.

KRG Decreased ROS Generation Induced by PM 2.5
Previously, we have demonstrated that 50 µg/mL of PM 2.5 exhibited ROS-induced senescence [9]; therefore, the optimal concentration of PM 2.5 to include cellular senescence was set as 50 µg/mL. We first estimated the cell viability at various concentrations of KRG by using HaCaT and NHEK cells, finding that KRG showed no cytotoxicity up to doses of 20 µg/mL (Figure 1a). The ROS scavenging effect of KRG on ROS generation induced by PM 2.5 in HaCaT and NHEK cells was highest at 20 µg/mL of KRG ( Figure 1b). Therefore, 20 µg/mL of KRG was used as the optimal concentration for the subsequent analyses.

KRG Decreased ROS Generation Induced by PM2.5
Previously, we have demonstrated that 50 µg/mL of PM2.5 exhibited ROS-induced senescence [9]; therefore, the optimal concentration of PM2.5 to include cellular senescence was set as 50 µg/mL. We first estimated the cell viability at various concentrations of KRG by using HaCaT and NHEK cells, finding that KRG showed no cytotoxicity up to doses of 20 µg/mL ( Figure 1a). The ROS scavenging effect of KRG on ROS generation induced by PM2.5 in HaCaT and NHEK cells was highest at 20 µg/mL of KRG ( Figure 1b). Therefore, 20 µg/mL of KRG was used as the optimal concentration for the subsequent analyses.

KRG Attenuated Cellular Senescence Induced by PM2.5
Oxidative stress induced by PM2.5 significantly showed senescence phenotypes in cells, such as an irregular size, a flattened and enlarged cell shape (Figure 2a), reduced CFA (Figure 2b), increased SAHF in the nucleus (Figure 2c), and higher cytoplasmic βgalactosidase activity (Figure 2d). However, the KRG treatment nearly restored the normal cell shape (Figure 2a), increased the CFA (Figure 2b), decreased the SAHF (Figure 2c), and reduced the β-galactosidase activity (Figure 2d). The expression of the p16 INK4A protein, a well-known CDK inhibitor and senescence inducer, was strongly induced in the PM2.5-treated cells, compared with that in the control cells; however, KRG treatment reduced its expression in both of the cell types (Figure 2e). In agreement with the western blotting results, p16 INK4A mRNA in the PM2.5-treated cells was also induced, compared with that in the control cells; however, KRG application reduced p16 INK4A mRNA ( Figure  2f).

KRG Attenuated Cellular Senescence Induced by PM 2.5
Oxidative stress induced by PM 2.5 significantly showed senescence phenotypes in cells, such as an irregular size, a flattened and enlarged cell shape (Figure 2a), reduced CFA (Figure 2b), increased SAHF in the nucleus (Figure 2c), and higher cytoplasmic β-galactosidase activity (Figure 2d). However, the KRG treatment nearly restored the normal cell shape (Figure 2a), increased the CFA (Figure 2b), decreased the SAHF (Figure 2c), and reduced the β-galactosidase activity (Figure 2d). The expression of the p16 INK4A protein, a well-known CDK inhibitor and senescence inducer, was strongly induced in the PM 2.5 -treated cells, compared with that in the control cells; however, KRG treatment reduced its expression in both of the cell types (Figure 2e). In agreement with the western blotting results, p16 INK4A mRNA in the PM 2.5 -treated cells was also induced, compared with that in the control cells; however, KRG application reduced p16 INK4A mRNA (Figure 2f).   Our previous study demonstrated that p16 INK4A is epigenetically controlled via the methylation of DNA during cellular senescence induced by PM 2.5 [9]. To determine whether KRG is involved in the regulation of PM 2.5 -induced p16 INK4A transcription via epigenetic DNA methylation, we measured the p16 INK4A promoter methylation status using qMSP and bisulfite sequencing analysis. The DNA methylation level of the p16 INK4A promoter region (from -150 to +200 bp, including 35 CpG sites in the promoter of p16 INK4A ) significantly declined in the PM 2.5 -treated cells compared with that in the control cells; however, KRG increased the DNA methylation level in both of the cell types ( Figure 3a). Subsequently, the DNA methylation status of the p16 INK4A locus was examined with bisulfite sequencing, revealing that the PM 2.5 -treated cells had lower methylation levels than those of the control cells. However, KRG increased methylation in both of the cell types ( Figure 3b). Therefore, KRG reduced the transcription of p16 INK4A in the PM 2.5 -treated cells by attenuating epigenetic promoter methylation. In addition, western blot analyses showed that DNMT1, DNMT3A, and DNMT3B expression decreased in the PM 2.5 -treated cells but was increased by KRG (Figure 3c). Conversely, TET1, TET2, and TET3 expression increased in the PM 2.5 -treated cells but was reduced after applying the KRG treatment ( Figure 3c). The expression of DNMTs and TETs by western blot analysis was confirmed using immunofluorescence (Figure 3d,e). We then assessed whether DNMT and TET could directly bind to the p16 INK4A promoter region in both cell types by employing ChIP-qPCR analysis. The binding of DNMT1 to the p16 INK4A promoter region decreased in the PM 2.5treated cells; however, KRG increased its binding in both cell types, which showed a similar pattern to that of protein expression (Figure 3f). The binding of TET1 to the p16 INK4A locus increased in the PM 2.5 -treated cells but was decreased in both cell types by KRG (Figure 3g). Furthermore, TET1 siRNA decreased the expression of p16 INK4A mRNA in the PM 2.5 -treated cells, while KRG decreased PM 2.5 -induced p16 INK4A expression (Figure 3h). These results indicate that PM 2.5 -induced p16 INK4A expression involves the release of DNMT and the recruitment of TET participating in senescence induction. However, KRG attenuates these effects, leading to the inhibition of PM 2.5 -induced senescence-related gene transcription. Figure 2. Inhibitory effect of KRG against cellular senescence induced by PM2.5. (a) Microscopic assessment reveals that PM2.5 induces morphological alterations typical of cellular senescence. Red square area means enlarged cell morphology. (b) Colony-forming ability (CFA) was enumerated after Diff-Quik staining. (c) Senescence-associated heterochromatin foci (SAHF) structure after cell staining with 4′,6-diamidino-2-phenylindole (DAPI) fluorescent dye was assessed. The arrows indicate SAHF in senescent cells induced by PM2.5. * p < 0.05 and ** p < 0.05 indicate significant differences with control cells and PM2.5-exposed cells, respectively. (d) β-galactosidase activity was revealed by staining with DAPI (blue) and the SPiDER-βGal (green) working solution. (e) Senescence marker p16 INK4A was detected by western blotting using the corresponding antibodies after cell lysates were electrophoresed. Actin represents a loading control. (f) Expression of p16 INK4A mRNA was assessed by qRT-PCR. * p < 0.05 and ** p < 0.05 indicate significant differences with control cells and PM2.5-exposed cells, respectively.

p16 INK4A Expression in KRG-Treated Cells Was Attenuated through Decreased DNA Demethylation in Cellular Senescence Induced by PM2.5
Our previous study demonstrated that p16 INK4A is epigenetically controlled via the methylation of DNA during cellular senescence induced by PM2.5 [9]. To determine whether KRG is involved in the regulation of PM2.5-induced p16 INK4A transcription via epigenetic DNA methylation, we measured the p16 INK4A promoter methylation status using qMSP and bisulfite sequencing analysis. The DNA methylation level of the p16 INK4A promoter region (from -150 to +200 bp, including 35 CpG sites in the promoter of p16 INK4A ) significantly declined in the PM2.5-treated cells compared with that in the control cells; however, KRG increased the DNA methylation level in both of the cell types ( Figure 3a). Subsequently, the DNA methylation status of the p16 INK4A locus was examined with bisulfite sequencing, revealing that the PM2.5-treated cells had lower methylation levels than those of the control cells. However, KRG increased methylation in both of the cell types ( Figure 3b). Therefore, KRG reduced the transcription of p16 INK4A in the PM2.5-treated cells by attenuating epigenetic promoter methylation. In addition, western blot analyses showed that DNMT1, DNMT3A, and DNMT3B expression decreased in the PM2.5-treated cells but was increased by KRG (Figure 3c). Conversely, TET1, TET2, and TET3 expression increased in the PM2.5-treated cells but was reduced after applying the KRG treatment ( Figure 3c). The expression of DNMTs and TETs by western blot analysis was confirmed using immunofluorescence (Figure 3d,e). We then assessed whether DNMT and TET could directly bind to the p16 INK4A promoter region in both cell types by employing ChIP-qPCR analysis. The binding of DNMT1 to the p16 INK4A promoter region decreased in the PM2.5-treated cells; however, KRG increased its binding in both cell types, which showed a similar pattern to that of protein expression (Figure 3f). The binding of TET1 to the p16 INK4A locus increased in the PM2.5-treated cells but was decreased in both cell types by KRG (Figure 3g). Furthermore, TET1 siRNA decreased the expression of p16 INK4A mRNA in the PM2.5-treated cells, while KRG decreased PM2.5-induced p16 INK4A expression ( Figure  3h). These results indicate that PM2.5-induced p16 INK4A expression involves the release of DNMT and the recruitment of TET participating in senescence induction. However, KRG attenuates these effects, leading to the inhibition of PM2.5-induced senescence-related gene transcription.
(a)  Our previous study demonstrated that histone methylation plays an important role in p16 INK4A epigenetic regulation [9]. To determine whether KRG is involved in epigenetic histone methylation of p16 INK4A expression induced by PM 2.5 , we detected the expression of HMT proteins using western blot analysis. EZH2, a transcriptional suppressor and a polycomb complex component with H3K27 methyltransferase activity, and its target protein, H3K27Me3, decreased after PM 2.5 treatment; however, they were increased by KRG (Figure 4a). The expression of MLL1, a transcriptional activator with H3K4 methyltransferase activity, and its target protein (H3K4Me3) increased after PM 2.5 treatment; however, they were decreased by KRG (Figure 4a). The expressions of EZH2 and MLL1 were confirmed by immunofluorescence analysis (Figure 4b,c). We assessed whether histone methyltransferase-related proteins could directly bind to the p16 INK4A locus in HaCaT and NHEK cells using ChIP-qPCR analysis. The binding of EZH2 to the p16 INK4A locus was decreased in PM 2.5 -treated cells; however, KRG increased their binding, which showed a similar pattern to that of protein expression (Figure 4d). The binding of MLL1 to the p16 INK4A locus increased in the PM 2.5 -treated cells but was decreased by KRG in both of the cell types (Figure 4e). In addition, MLL1 siRNA decreased the expression of p16 INK4A mRNA in the PM 2.5 -treated cells, and KRG decreased the PM 2.5 -induced p16 INK4A expression ( Figure 4f). Collectively, these results indicate that the PM 2.5 -induced expression of p16 INK4A involves the recruitment of MLL1, whereas the release of EZH2 is linked to senescence induction. However, KRG attenuates these effects, leading to the inhibition of p16 INK4A transcription induced by PM 2.5 .

p16 INK4A Expression in KRG-Treated Cells Was Attenuated via Changes in Histone Acetylation and Deacetylation in Cellular Senescence Induced by PM 2.5
Many studies have shown that histone acetylation plays an important role in p16 INK4A epigenetic regulation [25]. To determine whether KRG is involved in the epigenetic histone acetylation of p16 INK4A expression induced by PM 2.5 , we detected the expression of histone acetylation and deacetylation proteins using western blot analyses. The expression of the transcriptional repressor HDAC1 decreased in the PM 2.5 -treated cells with deacetylase activity; however, it was increased by KRG (Figure 5a). The transcriptional activator HAT1 increased in the PM 2.5 -treated cells with acetyltransferase activity; however, treatment with KRG decreased it (Figure 5a). The expressions of HDAC1 and HAT1 were confirmed using immunofluorescence (Figure 5b,c). Furthermore, we assessed whether histone acetylrelated proteins could directly bind to the p16 INK4A locus using ChIP-qPCR analysis. The binding of HDAC1 to the p16 INK4A locus was decreased in the PM 2.5 -treated cells but was increased by KRG in both cell types, with a pattern comparable to that of protein expression (Figure 5d). The binding of HAT1 to the p16 INK4A locus increased in the PM 2.5treated cells but was decreased by KRG in both cell types (Figure 5e). In addition, HAT1 siRNA decreased the expression of p16 INK4A mRNA in the PM 2.5 -treated cells, while KRG decreased the PM 2.5 -induced p16 INK4A expression (Figure 5f). Collectively, these results indicate that p16 INK4A expression induced by PM 2.5 involves the recruitment of HAT1, while the release of HDAC1 is associated with cellular senescence induction. However, KRG has been shown to weaken this effect and induce the inhibition of p16 INK4A transcription through increased histone deacetylation.
firmed by immunofluorescence analysis (Figure 4b,c). We assessed whether histone methyltransferase-related proteins could directly bind to the p16 INK4A locus in HaCaT and NHEK cells using ChIP-qPCR analysis. The binding of EZH2 to the p16 INK4A locus was decreased in PM2.5-treated cells; however, KRG increased their binding, which showed a similar pattern to that of protein expression (Figure 4d). The binding of MLL1 to the p16 INK4A locus increased in the PM2.5-treated cells but was decreased by KRG in both of the cell types (Figure 4e). In addition, MLL1 siRNA decreased the expression of p16 INK4A mRNA in the PM2.5-treated cells, and KRG decreased the PM2.5-induced p16 INK4A expression (Figure 4f). Collectively, these results indicate that the PM2.5-induced expression of p16 INK4A involves the recruitment of MLL1, whereas the release of EZH2 is linked to senescence induction. However, KRG attenuates these effects, leading to the inhibition of p16 INK4A transcription induced by PM2.5.   Many studies have shown that histone acetylation plays an important role in epigenetic regulation [25]. To determine whether KRG is involved in the epigen PM2.5-treated cells but was decreased by KRG in both cell types (Figure 5e). In addition, HAT1 siRNA decreased the expression of p16 INK4A mRNA in the PM2.5-treated cells, while KRG decreased the PM2.5-induced p16 INK4A expression (Figure 5f). Collectively, these results indicate that p16 INK4A expression induced by PM2.5 involves the recruitment of HAT1, while the release of HDAC1 is associated with cellular senescence induction. However, KRG has been shown to weaken this effect and induce the inhibition of p16 INK4A transcription through increased histone deacetylation.

Discussion
PM2.5 has been reported to cause skin diseases such as allergies, inflammatory dermatitis, and skin senescence, resulting in harmful effects on the skin. We have previously reported that PM2.5 induced an increase in oxidative stress through the aryl hydrocarbon receptor-ROS pathway and promoted the binding of TET1 and MLL1, instead of DNMT1 and EZH2, to the p16 INK4A promoter region. This binding induced the expression of p16 INK4A , ultimately inducing human epidermal keratinocytes senescence [9].
KRG has been reported to have more pharmacological efficacy than that of fresh and white ginseng [37]. About 40 types of ginsenosides, including Rb1, Rb2, Rc, Rd, Re, and Rg1 as main bioactive ingredients, have been identified in KRG [37,38]. All compounds exhibit their own biological activity, such as anticancer, anti-inflammatory, antioxidant, antibacterial, antiviral, and antifungal effects, but when mixed or combined, they can produce additive, synergistic, or increased effects [38][39][40][41][42]. (a) Nuclear fractions were electrophoresed, and histone deacetyltransferase 1 (HDAC1) and histone acetyltransferase 1 (HAT1) were detected using western blotting with the corresponding antibodies. As indicated, TBP serves as the loading control. (b,c) The nuclear locations of (b) HDAC1 and (c) HAT1 were determined by confocal microscopy after Alexa488-labeling (green) with the corresponding antibodies and staining with DAPI (blue). (d,e) ChIP assays using antibodies against (d) HDAC1 and (e) HAT1 were performed and analyzed by qPCR. HAT1 siRNA was transfected into cells and incubated for 24 h. (f) Nuclear fractions were electrophoresed, and HAT1 was detected using western blotting with the corresponding antibodies. As indicated, TBP serves as the loading control. The mRNA levels of HAT1 and p16 INK4A were detected by qRT-PCR. * p < 0.05 and ** p < 0.05 indicate significant differences with control cells and PM 2.5 -treated cells, respectively.

Discussion
PM 2.5 has been reported to cause skin diseases such as allergies, inflammatory dermatitis, and skin senescence, resulting in harmful effects on the skin. We have previously reported that PM 2.5 induced an increase in oxidative stress through the aryl hydrocarbon receptor-ROS pathway and promoted the binding of TET1 and MLL1, instead of DNMT1 and EZH2, to the p16 INK4A promoter region. This binding induced the expression of p16 INK4A , ultimately inducing human epidermal keratinocytes senescence [9].
However, total extracts of ginseng are more beneficial and effective than single ginsenosides or combinations of specific ginsenosides [42]. Here, we determined the anti-senescent efficacy of KRG and the underlying molecular mechanism involving the epigenetic regulation of p16 INK4A , a senescence sensor, in PM 2.5 -induced senescence. Our results have shown that KRG decreases the cellular senescence phenotypes associated with PM 2.5 -induced oxidative stress, including flattened and enlarged cell shapes, SAHF-like chromatin foci, and β-galactosidase activity in skin keratinocytes.
As a CDK4/CDK6 inhibitor, p16 INK4A plays a vital role in senescence formation. The p16 INK4A protein is relatively stable, and its expression is mainly controlled at the transcriptional level. Abnormal p16 INK4A hypermethylation is found in most tumors and reduces gene expression [51][52][53]. In mammalian cells, DNMTs maintain global and gene-specific de novo DNA methylation [24]. TETs have the ability to reverse this methylation process. When cells were exposed to PM 2.5 in this study, TET1 replaced DNMT1 in the p16 INK4A promoter region, promoting p16 INK4A transcription. KRG pre-treatment in the cells that were treated with PM 2.5 reversed the altered expressions of DNMTs and TETs. In addition, KRG decreased TET1 binding to the p16 INK4A locus and increased DNMT1 binding, thereby reversing the increase in p16 INK4A expression and preventing cellular senescence.
Histone methylation via EZH2 and MLL1 is also vital for gene transcription. A methyltransferase EZH2 binds to the p16 INK4A locus, inhibiting p16 INK4A transcription. Conversely, the p16 INK4A locus binding of MLL1 induces transcription during replication and premature senescence [54,55]. In the presence of PM 2.5 , MLL1 replaced EZH2 in the skin keratinocytes, thereby promoting the transcription of p16 INK4A . However, KRG suppressed this effect. In addition to EZH2 and MLL1, the modification of histone acetylation via HAT or HDAC plays an important role in gene transcription [25,56]. In our study, PM 2.5 induced p16 INK4A expression via histone acetylation with HAT1. However, KRG reversed p16 INK4A expression via histone deacetylation with HDAC1.
Considering these results, we conclude that p16 INK4A expression induced by PM 2.5 is suppressed by KRG via the epigenetic regulation of DNA and histones, which results in the inhibition of cellular senescence. As KRG inhibited the p16 INK4A expression induced by PM 2.5 through epigenetic regulation, it can be inferred that KRG protects the skin against PM 2.5 exposure and reduces skin senescence. However, it remains unclear how other epigenetic regulation processes, such as those involving microRNA, are regulated during cellular senescence induced by PM 2.5 , and how KRG executes epigenetic regulation process of microRNA modulated by PM 2.5 . Hence, further studies are needed in order to determine other epigenetic regulation processes affecting cellular senescence.