For the isolation and culture of hMSCs, we used a sub-fractionation culturing method (SCM) as described previously (10). Human BM aspiration for MSC isolation was approved by the Seoul St. Mary’s Hospital Institutional Review Board (IRB number #KC15CSSE0336) and written informed consent was obtained from healthy donors. For culturing human dermal fibroblast (HDF) and keratinocytes, cells were obtained from American Type Cell Collection (ATCC PCS-201-012 for HDF and ATCC PCS-200-011 for keratinocyte, respectively; ATCC, Virginia, US). Cells were grown in DMEM (Thermo Fisher Scientific, 11885-084; Massachusetts, US) supplement with 10% FBS (Thermo Fisher Scientific) and 1% penicillin-strepto-mycin (Thermo Fisher Scientific).

hMSCs were cultured in phenol red-free-DMEM containing 10% FBS or 4% hPL. When the cell confluency reached 80 to 90%, the culture flask was rinsed three times with Dulbecco’s Phosphate Buffered Saline (DPBS). Subsequently, 20 ml of phenol red-free, choline chloride-free, serum free medium was added. After 72 h, the conditioned medium (CM) obtained after culturing in FBS-containing medium and in hPL-containing medium hereafter defined as FBS-MSC-CM and hPL-MSC-CM, respectively, were harvested. They were centrifuged at 1500 rpm for 5 min then the supernatant was filtered through a 0.22 μm filter and stored at 4℃ until further analysis.

The concentration of major growth factors (VEGF, TGF-β1, and HGF) in MSC-CM was quantitatively analyzed by ELISA. The Human VEGF Quantikine (R&D SYSTEMS, SVE00; Minnesota, US), Human TGF-β1 Quantikine (R&D SYSTEMS, SB100B), Human HGF Quantikine Kits (R&D SYSTEMS, SHG00) were used for analyses according to the manufacturer’s instructions. The optical density of the samples was measured using a microplate reader (Biotek, Synergy H1; Vermont, US).

The expression of 105 cytokines was recorded using the Human XL cytokine Array kit (R&D Systems, ARY022B) according to the manufacturer’s instructions. The membrane was detected using a chemiluminescence imager (GE Healthcare, Image Quant LAS 4000 Min; Illinois, US) and the mean pixel density of secreted proteins from hMSCs was analyzed using image J (National Institutes of Health, Maryland, US).

To determine the effects of FBS-MSC-CM and hPL- MSC-CM on human dermal fibroblast (HDF) and keratinocyte proliferation, HDF and keratinocytes were treated with 50% and 100% FBS-MSC-CM and hPL-MSC-CM for 24 h. HDF and keratinocyte proliferation were measured using a CCK-8 kit (Cell Counting Kit-8, Dojindo; Maryland, US) according to manufacturer’s instructions.

Collagen and elastin production of HDF was measured to compare the effects of MSC-CM on skin elasticity. The Sircol soluble collagen assay (Biocolor, S1000) and Fastin Elastin assay kits (Biocolor, F2000) were used. HDF were cultured in a 6-well culture plate and treated with MSC- CM for 72 h. After performing the assays according to the manufacturer’s instructions, collagen and elastin concentrations were measured using a microplate reader (Biotek, Synergy H1).

MMP-1 and MMP-2 expression were analyzed in HDF subjected to UVA irradiation with 168 mJ/cm² by a UV lamp (model TN-4LC, K-ACE, Korea) by real-time reverse transcription-polymerase chain reaction (RT-PCR). During irradiation, control cells were prepared equally without exposure of UVA irradiation in the presence or absence of MSC-CM for 72 h. The administration of ascorbic acid (50 μg/ml) for 72 h was applied to positive control for reduction of MMP-1 and MMP-2 expression. Isolated mRNA from cells with RNeasy Mini Kit were used for amplification with TaqMan reverse transcription reagents (Applied Biosystems, California, US) and GAPDH were used for normalized control.

Scratch assays were performed to evaluate the effects of MSC-CM on the migration ability of HDF. In each well of the 24-well culture plate containing wound inserts, 1×10⁵ HDF were seeded. After the cells formed a monolayer with the wound, FBS-MSC-CM or hPL-MSC-CM were treated. After 6 h, the wells were fixed and stained with 10% formaldehyde and 1% crystal violet solution, respectively, and images of the wound healing process were captured. For quantification, the width (cm) of the wound was measured to compare the wound healing levels of cells cultured in different conditioned medium.

All results are presented as mean±standard deviation (SD). One-way ANOVA was used to compare the treatment groups and a p-value <0.05 was considered statistically significant. Statistical tests were conducted using GraphPad Prism 6.0 software (Graphpad; California, US).

First, we explored whether the expression of secreted proteins varies in the two different culture media (FBS- supplement vs. hPL-supplement). We identified a higher number of proteins in hPL-MSC-CM (n=39) than in FBS-MSC-CM (n=25) (Fig 1A). More specifically, 25 proteins were identified in both MSC-CMs, whereas 14 proteins were identified only in the hPL-MSC-CM (Fig. 1B). We found that the concentrations of 11 proteins, Angiopoietin-1, Chitinase 3-like 1, Cystatin C, IL-17A, IL-22, IGFBP-3, MCP-1, MIF, Pentraxin3, Serpin E1, VCAM-1, were significantly higher in both FBS-MSC-CM and hPL-MSC-CM than in the control medium, but there was no significant difference in proteins between FBS- MSC-CM and hPL-MSC-CM (Fig. 1C). All 9 proteins, except Cystatin C and IL-17A, are reported to be involved in skin cell proliferation, skin-repairing angiogenesis, and wound healing (11-17). Cystatin C and IL-17A were reported to regulate antibacterial activity in skin (18, 19).

The 14 proteins, angiogenin, complement factor D, Dkk-1, endoglin, emmprin, FGF-7, FGF-19, HGF, IGFBP-2, osteopontin, SDF-α1, Thrombospondin-1, uPAR, and VEGF, were significantly upregulated in hPL-MSC- CM comparing with FBS-MSC-CM. Among these, 13 proteins except complement factor D are involved in skin rejuvenation processes such as wound healing, skin cell growth, skin regeneration, and ECM remodeling (20-29). However, to the best of our knowledge, there is no report on the direct effect of complement factor D on the skin, known to involved in the alternative pathway of the complement system (Fig. 1D). Importantly, 14 proteins, adiponectin, apolipoprotein A-1, BDNF, C5/C5a, C reactive protein, DPPIV, GDF-15, IL-6, IL-8, LIF, PF4, RANTES, RBP-4, and Vitamin D BP, were detected only in hPL- MSC-CM, but not in FBS-MSC-CM (Fig. 1E). Among them, 6 proteins, Adiponectin, BDNF, DPPIV, PF4, RANTES, and RBP-4, are involved in skin regeneration (30-35), whereas 3 proteins, IL-6, IL-8, and Vitamin D BP, play a role in reducing skin inflammation (36, 37). IL-6 is also known to stimulate collagen production in HDF (38). However, the rest 5 proteins, apolipoprotein A-1, C5/C5a, C-reactive protein, GDF-15, and LIF, are not associated with skin rejuvenation. The protein array map and full name of the cytokine arrays from the array kit used in this experiment are summarized in the supplementary data (Supplementary Table S1).

In addition, we quantified representative growth HGF, VEGF, and TGF-β1 that enhanced skin regeneration and reduced skin inflammation. We found that all the three growth factors were significantly higher in hPL-MSC-CM than in FBS-MSC-CM (Fig. 1F) indicating that hPL culture conditions are more advantageous than FBS in increasing the concentrations and numbers of skin-related proteins secreted from hMSCs.

Next, we conducted experiments to determine whether hPL-MSC-CM, which is rich in skin-related proteins, has better efficacy in skin improvement. Increased cell proliferation of HDF improves the regeneration of damaged skin and increased cell proliferation of keratinocytes strengthens the skin barrier by increasing intercellular adhesion (1). We compared the effects of FBS-MSC-CM and hPL-MSC-CM on the proliferation of skin cells, HDF and keratinocytes.

We found that HDF proliferation was significantly higher in hPL-MSC-CM than in FBS-MSC-CM (Fig. 2A). Similarly, hPL-MSC-CM demonstrated superior efficacy on keratinocyte proliferation compared with FBS-MSC- CM (Fig. 2B). Thus, hPL-MSC-CM stimulates prolifera-tion of skin cells more effectively than FBS-MSC-CM expecting better efficacy in skin regeneration and rejuve-nation.

A previous study showed that increased proliferation and migration of HDF suggests enhanced skin regeneration (1). Accordingly, to determine the effect of MSC-CM on migration of HDFs, a wound scratch assay was performed. Both FBS-MSC-CM and hPL-MSC-CM increased the migration of HDF compared to control, without any treatment of conditioned medium, however, The hPL-MSC-CM showed the superior efficacy compared to FBS-MSC-CM (Fig. 3A). In consistent, the wound width (cm) was significantly reduced in both FBS-MSC- CM and hPL-MSC-CM compared with control and the superior efficacy of HDF migration was observed in hPL-MSC-CM (Fig. 3B).

To determine the effect of MSC-CM on ECM production and remodeling by HDF, we firstly measured the concentrations of collagen and elastin. We found that the synthesis of collagen and elastin increased in both MSC- CMs, however, the increased extent was higher in hPL- MSC-CM than FBS-MSC-CM (Fig. 4A). Additionally, the expression levels of MMP-1 and MMP-2 were examined in HDF subjected to UVA irradiation to determine the efficacy in ECM remodeling. In response to UVA, MMP-1 and MMP-2 were induced, which completely blocked by ascorbic acid (AA, 50 μg/ml) treatment in HDF (Fig. 4B). In the presence of both MSC-CMs, the blockade of MMP-1 induction was observed in similar with AA, moreover, hPL-MSC-CM was better than FBS-MSC-CM. However, inhibition of MMP-2 induction was observed with similar extent in both conditions. This suggests that hPL-MSC-CM may be more effective at skin regeneration for ECM production and ECM remodeling via MMP-1 than FBS-MSC-CM.