Cytoprotective effects of the lipoidic-liquiform pro-vitamin C tetra-isopalmitoyl-ascorbate (VC-IP) against ultraviolet-A ray-induced injuries in human skin cells together with collagen retention, MMP inhibition and p53 gene repression

Preparation of VC-IP

2,3,5,6-O-Tetra-2′-hexyldecanoyl L-ascorbic acid (VC-IP) was prepared as follows: 2-Hexyldecanoyl chloride (128.8 g, 469 mM) was added to a 0–5°C solution of L-ascorbic acid (18.3 g, 104 mM) and pyridine (74 g, 935 mM) in DMF (120 ml) under N₂ and left to react for 1 h. The mixture was extracted with hexane (180 ml), the organic layer was washed with 5% HCl (50 ml), H₂O (3 × 150 ml), dried (MgSO₄) and evaporated. The residual product was purified by silica gel column chromatography using n-hexane/benzene (4:1) as the eluent. The structure of the product was identified by ¹H-NMR (CDCl₃), ¹³C-NMR and IR. VC-IP was obtained as a clear and colorless liquid. The structure was identified by ¹H-NMR (CDCl₃), ¹³C-NMR, and IR as follows. ¹H-NMR (CDCl₃) δ (ppm): 5.4 (m, 1H,C₄-H), 4.27–4.45 (m, 3H, C₅, C₆-H), 0.90–2.40 (m, 124H) ¹³C-NMR δ (ppm): 175.65, 174.56, 171.45, 170.18, 164.92 (C₁), 150.01 (C₃), 122.38 (C₂), 75.10 (C₄), 66.05 (C₅), 61.67(C₆) IRν (cm⁻¹): 724, 1,101, 1,136, 1,341, 1,379, 1,464, 1,708, 1,746, 1,781, 1,800, 2,862, and 2,960.

The Characteristic of Physical Chemistry of VC-IP

VC-IP is colorless clear liquid, almost soluble in oily materials and thermally stable without deterioration in color. Other characters of VC-IP are as follows: little viscous, density 0.93, viscosity 280 cPs (20°C) and surface tension 26.7 dyne/cm.

Solubility of VC-IP

The solubility VC-IP is shown in Table I, VC-IP (5, 10, and 50 wt%) is soluble in various cosmetic oils at temperatures of 25, 50, and 70°C (Table I).

Stability of VC-IP

VC-IP was kept in the thermostat layout at 40 or 60°C for 1 month. It was unchanged on appearance and chrom-phase by sight. The weight was kept at the same value when VC-IP was exposed at 200°C that was measured by Seiko Thermo Gravimetric-Differential Thermal Analysis equipment SSC5200 (Seiko Instruments, Inc., Tokyo) with a speed of increasing temperature 5°C/min, 30–500°C. These findings suggest that VC-IP has excellent thermal stability (Fig. 1).

Cell Culture

Human skin keratinocytes HaCaT were kindly provided from Prof. Norbert E. Fusenig of Deutsches Krebsforschungszentrum (Heidelberg, Germany) [Boukamp et al., 1988], and routinely cultured at 37°C in a humidified atmosphere (5% CO₂ in air) in Dulbecco's modified Eagle minimum essential medium (DMEM) (Nissui Seiyaku, Tokyo) containing 1 mM L-Arg, supplemented with 10% fetal bovine serum (FBS, GIBCO/BRL Life Technologies), 4 mM L-glutamine (Sigma), 100 U/ml penicillin (Sigma), and 100 µg/ml streptomycin (Nissui Seiyaku). Human melanoma cells HMV-II were kindly gifted by Dr. Kasuga of Tokyo Med. Dent. Univ. [Nakamura et al., 1988] and maintained in Ham F12 medium (Nissui Seiyaku) containing 15% FBS, 2 mM L-glutamine (Sigma), and two above-mentioned antibiotics of the same concentrations. Human skin fibroblasts NHDF were purchased from DaiNippon Pharm., Inc. (Osaka) and were maintained in DMEM supplemented with 10% FBS and similarly two antibiotics. Other cell culture reagents were provided from Sigma Chemical Co. (St. Louis, MO). For each study, cells were plated onto 10-cm tissue culture dishes and grew for 4–7 days up to near 80% confluence. The spent culture medium was replaced by fresh medium at 48 h before the experiment.

Cell Viability Assay

Cell viability was evaluated by photometric assay using the formazan-forming redox indicator dye WST-1 (2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2-disulfophenyl)-2H-tetrazolium, monosodium salt; Dojn Laboratories Co., Kumamoto, Japan). At different time points, cells were rinsed with PR-free DMEM, then incubated for 3 h in PR-free DMEM medium containing 10% WST-1 at 37°C. Cell viability was determined based on mitochondrial conversion of WST-1 to yellowish formazan, being indicative of the number of viable cells [Ishiyama et al., 1996]. The absorbance at 450 nm was read with an absorbance multiplate reader (Bio-Rad^Q4, San Jose, CA, Microplate Photometer).

Measurement of Intracellular ROS Level

In order to investigate the influence of VC-IP on the ROS level in HaCaT cells, we used 6-caboxy-2′, 7′-dichlorodihydrofluorescein diacetate (CDCFH) (Molecular Probes, Eugene, OR) as a redox indicator and used fluorometry as a detection system. The CDCFH-DA dye was taken up by cells, esterolyzed to be membrane-impermeable CDCFH, and oxidized to be highly fluorescent CDCF by reaction with ROS such as hydroperoxide and hydrogen peroxide [Szejda et al. 1984].

The cultures were rinsed with PBS(−) and incubated in phenol red-free DMEM containing 7 µM CDCFH-DA for 30 min. The fluorescence intensity of the oxidative form of CDCFH was measured at 534 nm of an emission wavelength after excitation at 510 nm with a fluorescence plate reader CytoFluor 2350 (Millipore, Bedford, MA). The methanol-killed cells were treated as the blank ones. The images of cellular ROS were observed using a fluorescence microscope (Nikon ECLIPSE E 600, Nikon Co.)

8-Oxodeoxyguanosine (8-OHdG) Assay

Oxidative DNA damage analysis was performed by 8-OHdG assay. Cellular 8-OHdG was measured using a reformative immunofluorescence method based on the previous report [Kimoto et al. 2000]. Briefly, the slides were fixed with 4% paraformaldehyde in PBS (−) (pH 7.4) at room temperature for 25 min and washed again. They were subsequently treated with permeability buffer for 5 min on ice and washed three times with PBS (−). Non-specific binding was blocked with 1.5% normal goat serum, and incubation with the anti-8-OHdG (deoxyguanosine) monoclonal antibody (Nikken SEIL Co. Ltd.) diluted at a rate of 1:50 in goat serum was performed overnight at 4°C. The slides were then incubated with a FITC-conjugated rabbit anti-mouse secondary antibody for 1 h, and observed using a fluorescence microscopy (NIKON ECLIPSE E600).

Flow Cytometric Analysis of Cell Apoptosis and Mitochondrial Membrane Potentials

Apoptosis of HaCaT cells was detected using Annexin V-FITC kit (Beckman Coulter, USA). In brief, HaCaT cells (1 × 10⁶) were collected and washed twice with PBS. Cells were resuspended in 1 ml of FITC-annexin V solution which was prepared freshly for each time. After incubation for 10 min in the dark at room temperature, propidium iodide (PI) solution was added to the cell suspension prior to analysis. Then cells were analyzed with a COULTER® EPICS® XL^TM flow cytometer: display data as two-color dot plot with FITC-Annexin V (green fluorescence, X axis) versus PI (red fluorescence, Y axis).

To detect the reduction of mitochondrial membrane potential in cells, we used MitoCapture™ mitochondrial apoptosis detection kit (BioVision, Inc.). Unfixed, cultured cells were washed in PBS(−), supplemented with MitoCapture™ in incubation buffer. Then cells were incubated for 15 min at 37°C in 5% CO₂, washed, and resuspended in incubation buffer for immediate flow cytometry using a COULTER® EPICS® XL™ flow cytometer. The percentage of MitoCapture™-positive cells was calculated using the EXPO32^TM software (Beckman Coulter).

Western Blots for P53 Proteins

One-quarter- to one-half-confluent cultures were collected, rinsed with PBS, pelleted, lysed in 20 mM Tris buffer (pH 7.3) plus 2% sodium dodecyl sulfate (SDS)- plus a protease inhibitor cocktail (Wako Pure Chemical Industries, Ltd, Osaka), and sonically disrupted. Then 40 µg of protein (assayed with the Bio-Rad protein reagent) from each extract was separated by SDS–polyacrylamide gel electrophoresis with precasted 4–20% gradient polyacrylamide gels (Bio-Rad). Gels were electrotransferred to nitrocellulose membranes, where proteins were detected with antibodies specific for p53, followed by peroxidase-labeled secondary antibody (Santa Cruz) and chemiluminescence reagent (Santa Cruz).

Collagen Synthesis

Human skin fibroblasts NHDF were administered with L-2-[³H]proline (Amersham Biotech) in the presence or absence of VC-IP or ascorbic acid (Asc) of 20–50 µM for 24 h, and underwent cytolysis by NaOH treatment and subsequent neutralization. The cell (1 × 10⁶) extract obtained was thoroughly hydrolyzed with collagenase from Clostridium perfringens (Wako Pure Chemical Industries, Ltd), and separated by TCA/tannin treatment and centrifugation to generate the supernatant and precipitate as the collagen fraction and the non-collagen protein fraction, respectively, both of which were evaluated with an Aloka liquid scintillation counter LC-3600 sci.

Enzymatic Activity of MMPs

After culturing, the conditioned media were collected and clarified by centrifugation. MMP activity was analyzed by zymography, using a procedure described in reference (Sudel et al. 2003). Briefly, 50 µl of conditioned media were mixed with SDS sample buffer, without β-mercaptoethanol, and heated for 37°C/30 min. Normalized samples (30 µg/lane; Bio-Rad protein assay) underwent the standardized electrophoresis in 10% SDS–PAGE containing 0.1% gelatin (Bio-Rad). Molecular weight markers (6–210 kDa), recombinant human latent MMP-9 (92 kDa), and active MMP-9 (83 kDa; Calbiochem) were loaded as positive controls. The gel was washed in 2.5% Triton X-100 (TX)-PBS to remove SDS and renature the proteins, incubated at 37°C for 48 h in an activation buffer (50 mM Tris–HCl, 200 mM NaCl, 10 mM CaCl₂; pH 7.5), rinsed in double distilled water, and finally stained for 60 min with 0.25% Coomassie brilliant blue R250 in 40% isopropanol. Gelatinolytic activity was identified as transparent bands on a uniform blue background following destaining in 7% acetic acid, indicating the area where gelatin was digested.

Measurement of Intracellular L-Ascorbic Acid (Asc)

Cell samples were subjected to homogenization and extraction in the presence of 3% metaphosphoric acid and 0.2% ethanol, followed by HPLC separation on an ODS column and UV/coulometric electrochemical detector (ECD) detection with 0.1 M KH₂PO₄-H₃PO₄ buffer (pH 2.35) containing 0.02% ethanol and 0.1 mM EDTA-2Na as the mobile phase.

Statistical Analysis

All data, expressed as mean ± standard deviation, were processed statistically by the software of SPSS 11.5 for Windows. The differences of the data were considered significant when P < 0.05.

In Vitro Cytotoxicity

Before performing the series tests of bio-functions of VC-IP, we evaluated the cell viability of the keratinocytes, to which VC-IP was administered for 48 h. As shown in Figure 2, after cultured in medium containing VC-IP of the varying concentrations, it was found that the survival rates of HaCaT cells were not decreased below a concentration of 5,000 ppm (4,500 µM). This finding suggests that VC-IP has no or scarce cytotoxicity to human skin keratinocytes even at higher concentrations. According to our other experimental results for toxicity in vivo, VC-IP showed LD₅₀ > 2,000 mg/kg to rats. VC-IP occurred no irritating reaction to human skin and exerted no significant cytotoxicity to reconstructed three-dimensional skin model (data are not shown). Therefore, as a cosmetic material, VC-IP is very safe (Fig. 2).

Scavenging Effect of VC-IP on UVA-Enhanced Intracellular Oxidative Stress in Human Melanocytes

It has been reported that, VC-IP could suppress UVB, hydrogen peroxide and tert-butyl hydroperoxide induced intracellular oxidative stress [Ochiai et al., 2006]. To examine whether VC-IP could influence UVA-induced elevation of intracellular ROS level in human keratinocytes HaCaT, we quantified the intracellular ROS by CDCFH method. As shown in Figure 3A, after irradiation with UVA (100 J/cm²), significant accumulation of intracellular ROS in non-pretreated cells was observed. In the cells pretreated with VC-IP at 88 ppm, the ROS level was markedly reduced to 60%, 48%, and 35% versus the sham-irradiated level at different time points, respectively. Figure 3B shows that at 5 and 30 min after UVA irradiation, the high fluorescence intensity in the non-pretreated HaCaT cells were observed, whereas VC-IP-pretreated cells showed much lower fluorescence intensity. The UVA-induced intracellular ROS enhancement was prevented by addition of 88 ppm VC-IP.

Cytoprotective Effects of VC-IP on UVA-Irradiation in Human Keratinocytes HaCaT and Human Melanocytes HMV-II

To examine the cytoprotective effect of VC-IP, at 48 h after UVA irradiation, we determined cell viabilities of HaCaT and HMV-II cells which were pre-administered with VC-IP at different time points. As shown in Figure 4A, UVA-irradiation decreased the cell viability in a dose-dependent manner. After irradiated with UVA at the doses of 20, 60, and 100 J/cm², the cell viability was reduced to 77%, 47%, and 31% of that of sham-irradiated control, respectively. When the cells were pre-treated with VC-IP of 72 ppm (65 µM) for 2 h, the cell viability was significantly increased to 71% at 60 J/cm² and 47% at 100 J/cm² of a UVA irradiance. When the pre-treatment time of VC-IP was further extended to 24 h, the cell viability was markedly increased to 100% at 20 J/cm², 69.5% at 60 J/cm², and 63.6% at 100 J/cm² of UVA irradiance, respectively. When the pre-treatment time of VC-IP was prolonged to 18 h, the cell viability was significantly increased to 99.8% at 20 J/cm² and 49.5% at 60 J/cm² of UVA. Why the cell viability was lower upon pretreatment with VC-IP at 18 h than those at 2 and 24 h before UVA-irradiation at 60 and 100 J/cm²? The reason is considered to be that because a doubling time of HaCaT cells is about 20 h, cellular Asc that was esterolytically supplied from VC-IP, at 18 h after administration with VC-IP, is largely expended to achieve continued mitosis, and at this time there may be therefore no enough Asc to scavenge ROS when cells are exposed to strong UVA-irradiation. These results suggested that VC-IP could protect HaCaT keratinocytes from UVA injures, especially upon pretreatment at 24 h before UVA irradiation, suggesting a opportune period necessary for esterolytic conversion of VC-IP into a really active form of the ROS-scavenger such as Asc.

Melanocyte is another very important component cell of skin. Melanocytes protect skin from UV-induced harmful effect through melanogenesis. Here we use human melanoma cells HMV-II which are considered as a good substitute of human normal melanocyte [Nakamura et al., 1988] to observe the cytoprotective effect of VC-IP against UVA injury as shown in Figure 4B. At 48 h after UVA irradiation, HMV-II cell viability was reduced to 43% of that of sham-irradiated control. When the cells were pretreated with VC-IP at concentrations of 2.5, 11, and 22 ppm for 3 h, the cell viability was increased to 29%, 57%, and 65%, respectively. VC-IP significantly enhanced the cell viability at concentrations of 11 and 22 ppm, being marked cytoprotection for human melanocytes.

Inhibitory Effects of VC-IP on Production of 8-OHdG in the Nuclei of HaCaT Cells

The formation of 8-OHdG-adducted bases is an oxidative modification of DNA and is considered as a deteriorative consequence of oxidative stress. After UVA irradiation (100 J/cm²), formation of 8-OHdG markedly became detectable in the nuclei and cytoplasma of HaCaT cells at post-irradiational 1 h, as shown in Figure 5B, most of formed 8-OHdG were accumulated in the nucleus of UVA-irradiated and non-preadministered cell, and then became fadeout or bleached as shown by immunofluorescence stain using anti-8-OHdG antibody. When VC-IP was administered to HaCaT cells at 72 ppm at 24 h before UVA irradiation, the formation of 8-OHdG was significant reduced (Fig. 5A). The result suggested that VC-IP could protect keratinocytes from UVA-caused DNA oxidative damage assumedly through scavenging of UVA-derived ROS by VC-IP or its esterolytic active form.

Inhibitory Effects of VC-IP on UVA-Induced Apoptosis of HaCaT Cells

At 12 h after UVA (100 J/cm²) irradiation, apoptotic HaCaT cells were identified by double supravital staining with recombinant FITC-conjugated annexin-V and propidium iodide and analyzed by flow cytometry. A middle panel of Figure 6A shows that after UVA irradiation, about 18.9% of cells showed annexin V-positive staining which means those cells underwent early apoptotic change and 45.7% of cells showed annexin V-PI-positive staining, which indicates those cells were suffering from late stage apoptosis and necrosis. In contrast, VC-IP (88 ppm, 24 h)-preadministered cells showed early stage apoptosis for 10.8% out of the total cell population and late stage apoptosis/necrosis for 1.7% (Fig. 6A, the right panel), both being markedly diminished as compared with UVA-irradiated and non-preadministered cells. These results suggested that VC-IP could largely reduce UVA-caused apoptotic and necrotic cell death in keratinocytes.

Mitochondrial alterations, including oxy-radical production and membrane permeability changes, play a pivotal role in UVA-caused cell death and the concomitant oxidative insults [Suschek et al., 2001]. We therefore determined whether VC-IP may modify mitochondrial changes involved in the death of HaCaT cells subjected to UVA irradiation. The green fluorescent dye MitoCapture™ mitochondrial apoptosis detection kit was used to assess mitochondrial membrane potential and then analyzed by flow cytometry. Exposure of cell cultures to UVA irradiation resulted in a significant increase in the amount of FITC-fluorescence in HaCaT cells within 12 h (Fig. 6B, a middle panel). Pre-treatment with VC-IP of 88 ppm at 24 h before UVA irradiation that significantly attenuated the number of apoptotic HaCaT cells (Fig. 6B, a right panel). These findings suggest that the cytoprotective effect of VC-IP is associated with preservation of mitochondrial membrane potential. VC-IP can stabilize mitochondrial function and thereby protect the cells against apoptotic death induced by UVA insults.

Effects of VC-IP on UVA-Modulated Apoptosis-Related Protein p53 Expression

The wild-type p53 protein acts as the molecular policeman of the genome by halting the cell cycle to allow repair, or if repair is not possible, promoting cell death by apoptosis. DNA-damaging agents such as UV and ionizing radiation induce high levels of wild-type p53 [Saied and Shamsuddin, 1998]. To determine whether VC-IP prevents UVA-induced above-mentioned response in HaCaT cells, we investigated the levels of wild-type p53 in 40 µg of total protein from each cell extract at 3 h after UVA-irradiation at 100 J/cm² by Western blot. The protein bands were scanned and analyzed by an ACT-II software. Figure 7 shows that UVA-modulated expression of p53 proteins was significantly suppressed in HaCaT cell 24 h pretreated by VC-IP in a concentration-dependent manner. The expression of p53 proteins in VC-IP (44 and 88 ppm)-pretreated cells was as low as 49% and 12% of the control, respectively. This result suggests that VC-IP can stabilize UVA-activated wild-type p53 proteins to prevent apoptosis.

Promotive Effects of VC-IP on Collagen Synthesis in Human Skin Fibroblasts NHDF

The loss of collagen is considered as the characteristic histological finding in aged skin. To investigate whether VC-IP could influence type-IV collagen synthesis in human skin fibroblasts NHDF, we quantified the collagen level of NHDF after VC-IP or Asc treatment. Figure 8 shows that after NHDF cells were treated with VC-IP at the concentrations of 10, 20, and 50 µM for 24 h, the collagen synthesis in 1 × 10⁶ cells was markedly increased to 144.6%, 174.1%, and 193.3% of the control, respectively. The collagen synthesis levels of Asc (20 and 50 µM)-administrated cells (1 × 10⁶ cells) were enhanced to 110% and 127%, respectively. These results indicate that VC-IP could largely promote collagen synthesis in human skin fibroblasts NHDF in a concentration-dependent manner. VC-IP showed a collagen-increasing activity more potent than Asc at the same concentrations.

Inhibitory Effects of VC-IP on MMPs in Human Skin Fibroblasts NHDF

The expression and functional activities of MMP-2 and MMP-9 were known to drive specifically collagen type-IV cleavage [Sancéau et al., 2003; Marin-Castaño et al., 2005]. To examine whether VC-IP could influence MMPs activity in human skin fibroblasts, we quantified the activity of MMP-2 and MMP-9 of NHDF cells after incubation with VC-IP or Asc. As shown in Figure 9, the activities of MMP-2 and MMP-9 were obviously decreased by VC-IP at 10, 20, and 50 µM and Asc at 20 and 50 µM. Both MMP-2 and MMP-9 levels of VC-IP-treated cells are significantly lower than Asc at the same concentrations. The result indicates that VC-IP not only promotes the collagen synthesis but also reduce the activity of MMPs to prevent collagen degradation.

Conversion Ability of VC-IP to Intracellular AsA in NHDF Fibroblasts

To examine whether VC-IP permeates through the plasma membrane to be transformed to Asc in NHDF fibroblasts, we observed the concentrations of intracellular Asc in VC-IP- or Asc-treated NHDF cells by HPLC separation and UV/coulometric ECD detection. Figure 10 shows that the intracellular Asc concentrations of VC-IP (5, 10, 20, and 50 µM)-pre-treated cells were 2.18, 5.48, 10.15, and 9.46 nmol/10⁶ cells, respectively. The intracellular Asc levels of NHDF cells which treated with Asc at higher concentrations of 20, 50, and 100 µM, were 0.86, 1.42, and 1.29 nmol/10⁶ cells, respectively. VC-IP enhanced intracellular Asc through its persistent conversion to Asc, than administration with Asc. This finding indicates a possibility that VC-IP could cross the hydrophobic plasma membrane and is transformed to the major antioxidant Asc in cells.

In conclusion, the new lipophilic Asc derivative, 2,3,5,6-O-tetra-2′- hexyldecanoyl-L-ascorbic acid (VC-IP) that is a colorless and transparent liquid has the extremely high thermal stability in cosmetic formulations without losing the favorable Asc-based activities. VC-IP did not reduce the survival rates of HaCaT cells even at a concentration as high as 5,000 ppm (4,500 µM). VC-IP significantly protected human skin cells such as HaCaT keratinocytes and HMV-II pigment cells against UVA-induced cell death in vitro as shown by WST-1 method. The mitochondrion and the plasma membrane, which are considered to be two main sites of ROS production [Agarwal et al., 2005], may be dually protected by lipophilic Asc derivatives. It has been reported that there was a relationship between structure and bio-activities of Asc. Lipophilicity is an important determinant of antioxidant activity as it can regulate membrane permeability and component distribution in lipid bilayer membrane. According to our previous study, VC-IP could decrease UVB-induced photo-oxidation on human sebum such as squalene (data are not shown). In the present study, VC-IP was demonstrated to enhance intracellular accumulation of Asc, assumedly through esterolytic conversion to Asc, more markedly than administration with Asc itself being susceptible to oxidative degradation. Flow cytometric analysis showed that VC-IP could stabilize the plasma membranes and mitochondria membrane potential to prevent UVA-induced apoptosis. These effects may ensue from the abilities of VC-IP to repress production of 8-OHdG and down-regulate the expression of apoptosis-related gene p53. VC-IP also showed a promotive effect on collagen synthesis and a repressive effect on MMP activities.

Thus application with VC-IP may alleviate skin damage following physiologic doses of UVA irradiation, and should be adopted in the preventative and protective strategies against UVA injuries.