Near-infrared-emitting nanoparticles activate collagen synthesis via TGFβ signaling

Research efforts towards developing near-infrared (NIR) therapeutics to activate the proliferation of human keratinocytes and collagen synthesis in the skin microenvironment have been minimal, and the subject has not been fully explored. Herein, we describe the novel synthesis Ag2S nanoparticles (NPs) by using a sonochemical method and reveal the effects of NIR irradiation on the enhancement of the production of collagen through NIR-emitting Ag2S NPs. We also synthesized Li-doped Ag2S NPs that exhibited significantly increased emission intensity because of their enhanced absorption ability in the UV–NIR region. Both Ag2S and Li-doped Ag2S NPs activated the proliferation of HaCaT (human keratinocyte) and HDF (human dermal fibroblast) cells with no effect on cell morphology. While Ag2S NPs upregulated TIMP1 by only twofold in HaCaT cells and TGF-β1 by only fourfold in HDF cells, Li-doped Ag2S NPs upregulated TGF-β1 by tenfold, TIMP1 by 26-fold, and COL1A1 by 18-fold in HaCaT cells and upregulated TGF-β1 by fivefold and COL1A1 by fourfold in HDF cells. Furthermore, Ag2S NPs activated TGF-β1 signaling by increasing the phosphorylation of Smad2 and Smad3. The degree of activation was notably higher in cells treated with Li-doped Ag2S NPs, mainly caused by the higher PL intensity from Li-doped Ag2S NPs. Ag2S NPs NIR activates cell proliferation and collagen synthesis in skin keratinocytes and HDF cells, which can be applied to clinical light therapy and the development of anti-wrinkle agents for cosmetics.

Near-infrared (NIR) irradiation has shown great potential for clinical light therapy as well as for cosmetic purposes 1-3 and successfully been applied to photorejuvenation, photoprotection, and treatment for acne and vitiligo [4][5][6][7][8][9] because of its long optical penetration depth of tissue. Moreover, the uniquely high NIR irradiation penetration efficiency has enabled its extensive application in therapeutic approaches to treating hypertrophic scars, including those resulting from skin aging due to skin wrinkling and skin laxity [10][11][12][13][14][15][16][17][18] . Clinical results have shown improvement in skin texture because NIR irradiation activates collagen synthesis and increases the amount of collagen in human dermal fibroblasts (HDF) 19,20 . Changes in collagen have been considered as a leading cause of aging and wrinkle formation because the human dermis is comprised of 90% collagen 21 . Schieke et al. 22 reported that the skin temperature increased because the epidermal layers absorbed most of mid-IR (1.5-5.6 μm) and far-IR (5.6-10,000 μm), whereas NIR (0.8-1.5 μm) penetrated deeper to the subcutaneous tissues without causing an increase in skin temperature.
However, many studies have reported that NIR irradiation using artificial light sources may be deleterious to human skin because it raises the concentration of matrix metalloproteinase 1 (MMP-1) that damage the skin collagen. Although type I procollagen expression increases with a single NIR irradiation, it should be noted that multiple irradiations can reduce its expression 23 . Kim et al. 24 reported that repeated NIR irradiation exposure reduced TGF-β protein expression while a single NIR irradiation increased the transforming growth factor (TGF)-β1, -β2, and -β3 expression in the human skin. TGF-β signaling activates the production of collagen and fibronectin, two important biosynthetic products in normal HDF. Therefore, in NIR-exposed human skin, any modifications in TGF-β signaling can change type I procollagen expression. Despite the benefits of NIR irradiation therapy in wounds and its potential application to the treatment of skin fibrosis, it has been reported to Scientific RepoRtS | (2020) 10:13309 | https://doi.org/10.1038/s41598-020-70415-1 www.nature.com/scientificreports/ have some deleterious effects causing the skin temperature to increase. Research efforts towards developing NIR therapeutics to control the proliferation of human keratinocytes and collagen synthesis in the skin microenvironment have been limited, and the subject has not been fully explored. To address the challenges of the use of NIR-emitting nanoparticle (NP)-based therapeutics and to narrow the gap between current NP-based approaches and their clinical applications, there is a clear need to synthesize effective NIR-emitting NPs and to develop a promising therapeutic modality for a wide range of dermatological and cosmetic applications. Ag 2 S NPs have been extensively studied currently as attractive NIR-emitting NPs for NIR bioimaging due to their high biocompatibility, deep tissue penetration depth, and unique absorption ability in the UV-NIR regions. These characteristics can be used for multispectral absorption applications in the UV, visible, and NIR spectral ranges because of the light absorption properties of these NPs over a broad wavelength range [25][26][27][28][29] . These unique optical properties, including light absorption ability within a broad wavelength range, NIR emission, and the potential application to NIR therapeutics to control the proliferation of human fibroblasts and collagen synthesis in the skin microenvironment, has motivated our research into replacing Ag 2 S NPs with an artificial NIR instrument to enhance the production of collagen. However, the development of simple methods for the preparation of high-quality and monodispersed Ag 2 S NPs is necessary for routine industrial applications, including the manufacturing of dermatological therapies and cosmetic applications.
Herein, we describe the simple preparation of Ag 2 S NPs and the one-pot synthesis of Li-doped Ag 2 S NPs via ultrasonic irradiation, which resulted in a dramatic enhancement of their absorption and emission capabilities in the NIR region (Fig. 1a). The effect of Li + ion doping on the electronic structure of the Ag 2 S system was also investigated by using first-principles calculations, which indicated that the Li-doped Ag 2 S NPs could enhance the photoluminescence of semiconducting NPs. Finally, the effects of NIR irradiation by NIR-emitting Ag 2 S NPs on collagen production were successfully investigated.
Some wavelengths of light are absorbed by Ag 2 S NPs and Li-doped Ag 2 S NPs, which then emit NIR light to induce the activation of TGF-β1 and allows it to bind to its receptors, such as tβR p I and tβR p II. This results in the activation of the transcription factors Smad2, Smad3, and Akt, which regulate target gene expression (type www.nature.com/scientificreports/ I procollagen), thereby increasing the induction of the expression of COL1A1 mRNA expression via the TGF-β signaling pathway (Fig. 1b). As models in this study, we used a keratinocyte cell line and primary HDF cells.

Results
Preparation of NIR-emitting Ag 2 S and Li-doped Ag 2 S nps. A simple synthesis of 10 nm Ag 2 S NPs was performed by using a sonochemical method in which the decomposition of raw materials was induced by ultrasound under ambient conditions to enhance collagen production using NIR-emitting NPs. Silver nitrate (AgNO 3 ) in 1-dodecanethiol was sonicated to generate localized hot spots within the acoustic cavitation of collapsing bubbles during ultrasonic irradiation (reaction time: 10 min, power: 50%, temperature: ~ 160 ℃).
Li-doped Ag 2 S NPs were synthesized by adding a suitable amount of Li + to the reaction bottle, a method to fabricate undoped Ag 2 S NPs, which improves NIR emission intensity, enhancing absorption properties in the broad wavelength range to increase its general photoluminescence (PL) performance. The structure and morphology of Ag 2 S NPs and Li-doped Ag 2 S NPs were analyzed using a transmission electron microscopy (TEM). Ag 2 S NPs TEM images verified its monodispersity and narrow particle size distribution (Fig. 2a); thus, it can be inferred that phase separation by nucleation and growth processes during ultrasonic irradiation at 160 ℃ is effective. Similar results were published in a previous study 26 . The peaks in the XRD patterns that displayed varying amounts Li + doped Ag 2 S NPs corresponded to the monoclinic Ag 2 S phase (JCPDS No. 014-0072) ( Supplementary Fig. S1), and its TEM images displayed monodispersed spherical NPs. Li + concentration did not significantly affect particle morphology, size, and fundamental properties such as the FT-IR spectra and the NPs charges after and before Li + doping except for some differences in the hydrodynamic diameter (Fig. 2a, Supplementary Figs. S2, S4, S5, S6). Figure 2b shows Ag 2 S NPs and Li-doped Ag 2 S NPs PL excitation and emission spectra. Within the UV-NIR range, the NPs can be effectively excited in contrast to www.nature.com/scientificreports/ PbSe and PbS quantum dots. It has been reported that Ag 2 S NPs, at various excitation ranges, emit efficiently, making them promising candidates in research requiring particular absorption properties in various wavelength regions ranging from UV to NIR. The Li-doped Ag 2 S NPs demonstrated a remarkable augmentation of emission intensity of up to two orders in magnitude when compared with undoped Ag 2 S NPs, exhibiting an emission peak at 1,250 nm. It is well known that Li + ions, even at minimal concentrations, play an essential role as co-dopants in increasing the lumious efficiency of phosphors 30 .

Effect of NIR emission by Ag 2 S and Li-doped Ag 2 S nps on human skin cells. The NIR-emitting
properties of Ag 2 S NPs may affect the skin, including wound-healing and anti-wrinkle effects; therefore, we examined their cytotoxicity in HaCaT cells and HDF cells. After 24 h of incubation with NPs, the cell viability was not decreased. Rather, cell proliferation was increased in a dose-dependent manner (Fig. 3a,b). To confirm whether NIR emission was associated with TGF-β signaling and collagen biosynthesis, we exposed HaCaT cells and HDF cells to NIR for 20 min, and the expression of collagen biosynthesis genes was measured by realtime PCR. The expressions of TGF-β1, tissue inhibitor of metalloproteinase 1 (TIMP1), and type 1 collagen (COL1A1) were upregulated by NIR exposure in HaCaT cells and HDF cells when compared with that in cells that were not exposed to light or that had normal light exposure ( Supplementary Fig. S3).
To investigate the role of Ag 2 S NPs and Li-doped Ag 2 S NPs, the expression of genes involved in collagen synthesis was examined in HaCaT cells and HDF cells. HaCaT cells were treated with various concentrations of Ag 2 S NPs and Li-doped Ag 2 S NPs. Only TIMP1 was upregulated when HaCaT cells were treated with Ag 2 S NPs, but in the presence of Li-doped Ag 2 S NPs, there was a dramatic upregulation of the expression of TGF-β1, TIMP1, and COL1A1 (Fig. 4a). In HDF cells, only TGF-β1 was upregulated when cells were treated with Ag 2 S NPs, but both TGF-β1 and COL1A1 were upregulated when cells were treated with Li-doped Ag 2 S NPs (Fig. 4b).
Then, collagen synthesis in HaCaT cells was analyzed by immunoblot analysis to identify the proteins involved in the TGF-β signaling pathway that were upregulated by Ag 2 S NPs or Li-doped Ag 2 S NPs. TGF-β1 binds to the receptors tβRpI and tβRpII, which are located on the surface of the plasma membrane and phosphorylates downstream transcription factors that regulate the expression of target genes such as TGF-β1, TIMP1, and COL1A1. In addition, the Akt signaling pathway, independently of the TGF-β signaling pathway, activates collagen synthesis. When HaCaT cells were treated with Ag 2 S NPs or Li-doped Ag 2 S NPs, the phosphorylation of www.nature.com/scientificreports/ Smad2 was increased by both types of NPs. In contrast, phosphorylation of Smad3 was slightly increased by Ag 2 S but decreased by Li-doped Ag 2 S (Fig. 5a,c). Additionally, the Akt pathway was determined by immunoblot analysis. Although Akt phosphorylation was not altered by Ag 2 S or Li-doped Ag 2 S NPs, collagen production was increased in the immunoblot analysis (Fig. 5a,c). Fibroblasts are another skin cell type associated with collagen production. Therefore, we examined the effects of NPs on HDF cells. We found that the phosphorylation of Smad2 and Smad3 was decreased by Ag 2 S or Li-doped Ag 2 S NPs. In contrast, the phosphorylation of Akt was increased by Ag 2 S or Li-doped Ag 2 S NPs, and the degree of Akt phosphorylation was higher during Li-doped Ag 2 S NPs treatment (Fig. 5b,d). As a result, collagen synthesis was activated only by Li-doped Ag 2 S NPs. These results suggest that both types of Ag 2 S NPs activate the TGF-β signaling pathway in HaCaT cells and HDF cells and activate collagen synthesis in skin.

Functional validation of RNA-seq results.
To identify various genes that may be involved in the development of dermal fibrosis mediated by NPs in human skin, we performed high-throughput RNA sequencing (RNA-seq) using in vitro HaCaT cells treated with Ag 2 S or Li-doped Ag 2 S NPs at two concentrations (160 and 320 μg/mL) for 24 h. Gene expression profiling identified several novel transcripts in human keratinocyte cells that were significantly up-and down-regulated including the selected targer genes (TGF-β1, TIMP1 and COL1A1), and others ( Fig. 6a; Supplementary Tables S1, S2). In a previous study, Gliga et al. 31 used RNA-seq to measure the effect of Ag-based NPs in human bronchial epithelial cells, and reported that Ag-based NPs are pro-fibrotic and induce epithelial-mesenchymal transition and cell tranformation. We also found a group of 19 genes that make up the extracellular matrix component through gene expression profiling, all of which showed increased expression ( Fig. 6b; Supplementary Table S3). However, unlike the real-time PCR results, no dramatic increase depending on NPs concentration was found (data not shown). This is thought to be a discrepancy arising from the sensitivity of the method for measuring gene expression. Among the differentially-regulated genes, Transglutaminase 2 (TGM2) was the most highly up-regulated (log 2 ratio = 6.93 and 6.33, in Ag 2 S and Li-doped Ag 2 S treated HaCaT cells). TGM2 knockout mice developed significantly reduced pulmonary fibrosis compared www.nature.com/scientificreports/ with wild-type mice, and overexpression of TGM2 led to increased fibronectin deposition in vitro 32 . The various genes that have been altered by NPs will provide important targets for mechanisms and diseases associated with fibrosis of the skin through future studies. The top 100 most significantly differentially expressed genes regulated by NPs is shown in Supplementary data (Tables S1, S2).

Discussion
The successful simple preparation of Ag 2 S NPs and Li-doped Ag 2 S NPs can aid the development of NIR www.nature.com/scientificreports/ therapeutics that can trigger human keratinocytes and collagen to proliferate in the skin microenvironment. Upon ultrasonic radiation, Ag 2 S NPs and Li-doped Ag 2 S NPs enhanced absorption and emission in the NIR region via a one-step process. Further study of the electronic structure of Ag 2 S doped with Li + ion using the firstprinciples calculations indicated that Li-doped Ag 2 S NPs exhibited increased signalling of NIR photoluminescence and absorption and upregulated collagen expression in skin keratinocytes. Using next-generation sequencing analysis, the increase in expression of various genes constituting the extracellular matrix was observed, and these results are expected to contribute to the study of skin fibrosis in the future. Also, NIR radiation may enhance the wound healing process and increase collagen, and elastin contents form the stimulated fibroblasts, despite the poorly understood biologic effects 2,33 . Ag-based NPs have inherent antibacterial and anti-inflammatory traits, due to its single metallic nanoparticles, that can be altered to develop augmented would and burn dressings 34,35 .
Ag-based NPs aid in early wound-healing stages in diabetic patients, although leaving minor scars 36 . Considering applications of Ag-based NPs in wound therapy due to their effective and enhanced antibacterial characteristics, their biocompatibility and safety need to be thoroughly analyzed 37 .
In this study, the effects of infrared radiated NPs on collagen expression in human keratinocytes were researched, and results revealed some beneficial properties for skin aging and wound healing. Taken together, we have proposed a photoluminescence enhancement mechanism observed in Fig. 2b shown in Fig. 7. Additional carriers in Ag 2 S NPs interstitially doped with Li + accumulate within the NPs, which further enhances the metallic properties. According to the results of Kang et al. 38 , the electronic structure changes from a semiconducting structure to a metallic band gap structure only in the presence of Li + interstitial doping. Additional electrons begin to migrate to the semiconducting NPs near metallic Ag 2 S NPs as the doping concentration of Li + increases. Accordingly, positively charged metallic NPs can improve the photoluminescence efficiency of the semiconducting NPs. However, this photoluminescence enhancement is only observed in optimum Li + doping conditions. At certain doping conditions, there is less absolute number of semiconducting NPs participating in the photoluminescence process and a decline in their contribution; thus, the total photoluminescence of the sample is reduced. This mechanism of photoluminescence enhancement and quenching has already been proposed in a

Materials.
A similar synthesis method and characterization of NPs has been already proposed in a previous study 38 . Silver nitrate [Ag(NO) 3 , 99%], Li 2 CO 3 (98%), and 1-dodecanethiol (DDT) were purchased from Sigma Aldrich. Chloroform and ethyl acetate were used to disperse and to isolate the NPs. All chemicals were used without further purification.

Characterization of Ag 2 S NPs and Li-doped Ag 2 S nps.
The absorption spectra of the solutions containing 0.01 g of Ag 2 S and Li-doped Ag 2 S NPs (in 10 mL of chloroform) were measured using a SolidSpec-3700 UV-Vis-NIR spectrophotometer from Shimadzu. The photoluminescence (PL) spectra were measured by a Fluorolog-3 in TCSPC mode (HORIBA Scientific). All samples were excited by a CW 450 W xenon source, and directed to a single-grating spectrometer. The PL spectra were obtained using an InP/InGaAs detector equipped with an LN cooler. All TEM images were acquired on a JOEL JEM-2100F transmission electron microscope operating at 200 kV. The TEM samples were prepared by drop casting very thin nanoparticle solutions onto a 200 mesh copper grid with a carbon film (Ted Pella). Then X-ray diffraction spectroscopy (XRD) was conducted using a Rigaku D/MAX-220 V X-ray diffractometer coupled with a Cu K-alpha (1.540598 Å) source. The FT-IR was measured by EQUINOX 55 (Bruker). The Zeta potential and DLS were measured by ELS-77 (Otsukael). www.nature.com/scientificreports/ niR irradiation. The medium was replaced with phosphate-buffered saline (pH 7.2) and using an UIM-250 (Unix, Korea) the cells were exposed to NIR at a distance of 20 cm at room temperature before irradiation. The NIR device emitted NIR spectra between 1,100 and 1,800 nm. The irradiation conditions were 20 J/cm 2 and 40 J/cm 2 in this study. It has been reported that under these conditions after immediate exposure to NIR, no temperature rise occurs in the phosphate-buffered saline 40 . As a control, cells were left on a clean bench with natural light during the day time. The cells were maintained in a serum-free medium for 6 h after treatment, and total RNA was extracted using an Easy-spin Total RNA Extraction Kit (Intron Biotechnology, Korea) for mRNA measurement by real-time PCR.

Surface modification of Ag 2 S NPs and Li-doped Ag 2 S NPs with 3-mercaptopropionic acid (MPA
RNA-sequencing and data analysis. RNA-seq analyses were performed at Theragen Bio Institute (Gyeonggi-do, Korea). The libraries were prepared for 150 bp paired-end sequencing using TruSeq RNA Sample Prep Kit (Illumina, CA, USA). A total of 1 μg of RNA molecules was purified and fragmented, then synthesized as single-stranded cDNAs via random hexamer priming. Using this as a template to synthesize the second strand, a double-stranded cDNA was prepared. cDNA libraries were amplified with PCR after a sequential process of end repair, A-tailing, and adapter ligation. The quality of these cDNA libraries was evaluated with the Agilent 2100 BioAnalyzer (Agilent, CA, USA) and were quantified with the KAPA library quantification kit (Kapa Biosystems, MA, USA) in accordance with the manufacturer's library quantification protocol. Cluster amplification of denatured templates was followed by paired-end (2 × 150 bp) sequencing using Illumina Novaseq6000 (Illumina, CA, USA). Differential expression analysis was performed by Cuffdiff 41 . With most options set at default values, only multi-read-correction and frag-bias-correct options were applied for better analysis accuracy. DEGs were identified based on the q value threshold of less than 0.05 for correcting errors caused by multiple-testing 42 . The raw datasets of the RNA-seq analysis will be available to the researchers of interest upon request via correspondence.