Area light source‐triggered latent angiogenic molecular mechanisms intensify therapeutic efficacy of adult stem cells

Abstract Light‐based therapy such as photobiomodulation (PBM) reportedly produces beneficial physiological effects in cells and tissues. However, most reports have focused on the immediate and instant effects of light. Considering the physiological effects of natural light exposure in living organisms, the latent reaction period after irradiation should be deliberated. In contrast to previous reports, we examined the latent reaction period after light exposure with optimized irradiating parameters and validated novel therapeutic molecular mechanisms for the first time. we demonstrated an organic light‐emitting diode (OLED)‐based PBM (OPBM) strategy that enhances the angiogenic efficacy of human adipose‐derived stem cells (hADSCs) via direct irradiation with red OLEDs of optimized wavelength, voltage, current, luminance, and duration, and investigated the underlying molecular mechanisms. Our results revealed that the angiogenic paracrine effect, viability, and adhesion of hADSCs were significantly intensified by our OPBM strategy. Following OPBM treatment, significant changes were observed in HIF‐1α expression, intracellular reactive oxygen species levels, activation of the receptor tyrosine kinase, and glycolytic pathways in hADSCs. In addition, transplantation of OLED‐irradiated hADSCs resulted in significantly enhanced limb salvage ratio in a mouse model of hindlimb ischemia. Our OPBM might serve as a new paradigm for stem cell culture systems to develop cell‐based therapies in the future.


| INTRODUCTION
Most studies on enhancing the therapeutic effect of light have focused on immediate and drastic changes in cells. However, it is essential to remember that changes in the body induced by light require a latent reaction period. For example, when ultraviolet (UV) rays hit the skin, melanin is formed in melanocytes, located in the base layer of the skin, to afford protection to the skin. This phenomenon occurs a few days after UV light exposure. 1,2 In addition, sunburn caused by excessive light occurs 4-6 h after exposure and peaks in 24 h. 3 Although the effects of light may not always be immediate, to the best of our knowledge, no previous study has undertaken a longterm follow-up of the changes occurring in cells following light induction. Therefore, in this study, we attempted to assess the long-term effects of light on stem cells.
Several studies have reported that the therapeutic effects induced by photobiomodulation (PBM) depend on the differences in wavelength and the total amount of energy. [4][5][6] In general, PBMinduced changes in the cellular microenvironment are known to be related to redox-sensitive transcription factors. Photons from PBM are absorbed by photoacceptors in the mitochondria and induce adenosine triphosphate (ATP) production, which induces the synthesis of reactive oxygen species (ROS) and nitric oxide (NO), resulting in the expression of redox-sensitive transcription factors. 7 Upregulated expression of redox-sensitive transcription factors leads to an increase in growth factor secretion, cell proliferation, and cell migration, which supports tissue regeneration. [8][9][10][11][12][13][14] However, as most PBM strategies are employed for damaged tissues composed of different cell types, identifying specific mechanisms that mediate tissue regeneration becomes challenging. Moreover, previous PBM strategies required repeated exposure to induce an appropriate therapeutic effect owing to the limited transmission of light. [8][9][10][11][12][13]15 In addition, disparate results have been reported with PBM strategies employing identical energy and light sources due to the differences in power density and duration (given that energy and power density are reciprocally related). [16][17][18] Therefore, it is necessary to investigate optimal PBM parameters with a clear understanding of the underlying molecular mechanisms, considering the latent reaction period to improve the therapeutic effect of organic light-emitting diode (OLED)-based PBM (OPBM) in the context of tissue regeneration.
In the present study, we established optimal irradiation parameters to subject human adipose-derived stem cells (hADSCs) to OPBM and investigated the molecular mechanisms that can potentially enhance angiogenesis in ischemic tissues ( Figure 1). We focused on improving the angiogenic efficacy of hADSCs by directly subjecting them to OPBM rather than through the PBM of tissues with complex conformations. We hypothesized that hADSCs subjected to OPBM with optimal irradiation criteria would improve angiogenesis in a mouse model of hindlimb ischemia when compared with conventional hADSCs that have not been subjected to OPBM. We optimized the parameters for the OPBM of hADSCs (wavelength, voltage, current, luminance, and duration), which had not been previously performed.
Next, hADSCs were cultured and irradiated using the optimized OPBM strategy, and the molecular mechanisms related to angiogenesis were examined. Angiogenic paracrine factor expression and various OPBM-regulated aspects, including cell viability, adhesion, apoptotic activity, cell cycle, and glycolysis, were investigated. Unlike previous PBM studies that focused on exploring the molecular F I G U R E 1 Schematic diagram depicting hADSCs subjected to red light. Schematic diagram depicting the OPBM of hADSCs. Bcl-xL, B-cell lymphoma-extra-large; CD31; cluster of differentiation 31; hADSCs, human adipose-derived stem cells; HIF-1α, hypoxia-inducible factor 1-alpha; HGF, hepatocyte growth factor; SM-α-Actin, smooth muscle alpha-Actin; OLED, organic light-emitting diode; OPBM, OLED-based photobiomodulation; VEGF, vascular endothelial growth factor F I G U R E 2 Enhanced expression of angiogenic paracrine factors and adhesion of hADSCs, without cytotoxicity following optimization of OPBM. (a-i) To determine the optimal duration of OPBM for hADSCs, the relative expression of the genes coding for angiogenic paracrine factors (VEGF, FGF2, and HGF) was evaluated by qRT-PCR. Results show the gene expression patterns following the OPBM of hADSCs for 3 (ac), 6 (d-f) or 24 h (g-i). For each OPBM duration, relative gene expression after 0, 24, and 48 h with or without OPBM, was evaluated by normalizing the values to the data collected from the without OPBM group (*p < 0.05, compared with hADSCs without OPBM treatment at each time point, n = 4). We chose 24 h as an optimal irradiation duration for the OPBM of hADSCs. (j) Western blot analysis of relative VEGF expression at 0 and 48 h after the OPBM of hADSCs (*p < 0.05, compared with hADSCs without OPBM, n = 4). (k) Secretion of VEGF and HGF by hADSCs was evaluated after 48 h of OPBM using enzyme-linked immunosorbent assay (ELISA). Relative protein expression is shown as a percentage compared with values obtained from the without OPBM group (*p < 0.05, compared to without OPBM treatment, n = 5). (l) Viability of hADSCs was evaluated using the fluorescein diacetate (FDA)-ethidium bromide (EB) assay (live cells stain green while dead cells stain red). No red cells can be observed in either group (scale bar: 100 μm). Effect of OPBM on hADSC adhesion was analyzed using cell adhesion assays performed at 3, 6, and 12 h after re-attachment of cells. (m) Analysis of cell adhesion ability via cytoplasmic staining of cells with DiI (red). Blue color indicates DAPI staining (nucleus) (scale bar: 50 μm). (n) The adhesion ratio was evaluated by CCK-8 assay using the without OPBM group as a control (*p < 0.05, **p < 0.01, compared with the without OPBM treatment, n = 4). (o) Three hours after re-attachment of the hADSCs, the relative expression of each gene coding for angiogenic paracrine factors (VEGF, FGF2, and HGF) was evaluated using qRT-PCR (*p < 0.05, **p < 0.01 compared with the without OPBM group, n = 4) mechanisms in mitochondria, we focused on receptor tyrosine kinase (RTK)-mediated signaling mechanisms at the cell surface. To the best of our knowledge, this is the first study to evaluate the expression of glycolysis-related enzymes induced by OPBM. Finally, hADSCs subjected to OPBM were transplanted into a mouse model of hindlimb ischemia, and angiogenesis was compared with that observed in response to conventional hADSC treatment. To confirm that OPBM enhanced the angiogenic efficacy of hADSCs, the number of hADSCs was reduced to one-third of that used in conventional hADSC transplantation. A previous study revealed that PBM increases cell adhesion ability following irradiation. 19 To verify the effect of OPBM on cell adhesion, we detached cells subjected to OPBM for 24 h (as well as control cells not subjected to OPBM) from the culture plates using trypsin and then allowed them to re-attach. We investigated cell adhesion 3, 6, and 12 h after the initiation of re-attachment,   cytochrome c oxidase (CCO), which is the complex IV of the mitochondrial electron transport system. 7 Upon stimulation of CCO, NO starts to dissociate from the Cu and Fe centers in CCO, and is replaced by oxygen. 20,21 This results in an increased respiration rate, which in turn, enhances ATP production 22 and ROS levels. 14 However, in our experiments, hADSCs subjected to OPBM exhibited different outcomes. NO production ( Figure 3b) and ATP levels in hADSCs subjected to OPBM (Figure 3c) did not significantly differ from those in hADSCs not subjected to OPBM. In contrast, intracellular ROS staining revealed that ROS levels were significantly different between hADSCs subjected to OPBM and those not subjected to OPBM, as evaluated from the mean fluorescence intensity (MFI) values using 2 0 ,7 0dichlorodihydrofluorescein diacetate (DCF-DA) ( Figure 3d). Furthermore, we analyzed HIF-1α expression by qRT-PCR and observed that HIF-1α expression was significantly increased in hADSCs up to 24 h after OPBM, compared with that in hADSCs without OPBM (Figure 3e). In addition to ROS, which can induce HIF-1α expression, 23 we also investigated the lightinduced mechanism involving RTK signaling. [23][24][25][26] Accordingly, we analyzed the expression of mitogen-activated protein kinase (MAPK), phospho-MAPK (p-MAPK), and phosphoinositide 3-kinase (PI3K) using western blotting. As shown in Figure 3f, MAPK expression was similar in the two treatment groups, but the expression of p-MAPK was markedly increased (>2.5-fold) in hADSCs subjected to OPBM when compared with that in hADSCs not subjected to OPBM. However, the expression of PI3K did not differ between hADSCs with or without OPBM (data not shown). Both factors demonstrate characteristic expression patterns when RTK on the cell surface is stimulated by light or growth factors. [23][24][25][26] Consistent with the findings of previous studies, our results also showed that HGF was stimulated by light ( Figure 2i). 27 Next, we investigated whether increased HIF-1α expression induced changes in the cell cycle in hADSCs subjected to OPBM. HIF-1α is known to cause cell cycle arrest at the G1 phase. 28

| Enhanced expression of glycolytic enzymes in hADSCs subjected to OPBM
Based on the increased expression of paracrine factors associated with angiogenesis and cell cycle transition, we investigated the effects of OPBM on hADSC metabolism. To identify the changes in cellular metabolism following the OPBM of hADSCs, we analyzed the expression of enzymes involved in glycolysis as well as the levels of the metabolic product, lactate, and related factors ( Figure 3h). 32 Glycolysis is a metabolic pathway during which glucose is converted to pyruvate in the cytoplasm. [33][34][35] Glucose transport across the cell membrane is facilitated by the glucose transporter 1 (GLUT1). 36 In general, glycolysis proceeds through 10 different enzyme-catalyzed processes. In the present study, we analyzed the following representative processes: HK1, PFKP, GAPDH, and PKM1/2 ( Figure 3h). When glucose enters the cytosol, it is phosphorylated to glucose-6-phosphate (G6P). The enzymatic reaction associated with this step involves the phosphorylation of hexokinase1 (HK1). 37 to adenosine diphosphate (ADP) to generate ATP and pyruvate. 44,45 Following glycolysis, pyruvate is reduced to lactate by lactate dehydrogenase A (LDHA). 35

| DISCUSSION
In the present study, we aimed to verify the intensified therapeutic efficacy of hADSCs through latent angiogenic molecular mechanisms triggered by area light sources. We demonstrated that red light at a wavelength of 610 nm from OLEDs resulted in the marked enhancement of angiogenic paracrine factor expression, cell adhesion ability, cell cycle, and glycolysis in hADSCs after exposure to optimized light irradiation. We also revealed that the viability and angiogenesisinducing ability of transplanted hADSCs were enhanced following OPBM treatment. Using a mouse model of hindlimb ischemia, we demonstrated that when using hADSCs subjected to OPBM for treating ischemic disease, the number of transplanted stem cells can be reduced to one-third of the number used in the conventional method. We directly irradiated hADSCs with red light from the OLED and confirmed a significant enhancement in their angiogenic efficacy, compared with that of hADSCs without OPBM, after the latent reaction period. Importantly, due to the use of OLEDs, the properties of the light source in this study differed from those from the previous studies. Unlike lasers or LEDs, which are point light sources, OLEDs are surface light sources that can distribute light evenly over a wider area rather than focusing on a specific area. In addition, the brightness per area of OLEDs is lower than that of point light sources (laser or LED); thus, they can be used for long-term treatments without a temperature increase that can induce cellular damage. In other words, OPBM irradiation used in our study can stimulate stem cells for an extended period and with a wider range when compared with conventional PBM, without causing cellular damage.
We optimized the duration of hADSCs exposure to OLEDs to increase angiogenic, and adhesion efficacy while decreasing cellular damage after the latent reaction period. In the present study, hADSCs showed increased expression of angiogenesis-related genes for paracrine factors (VEGF and HGF) after 24 h of OPBM, which persisted for 48 h after irradiation. Moreover, the expression of VEGF and HGF was higher in hADSCs subjected to OPBM than in hADSCs not subjected to OPBM. In addition, we evaluated the expression of paracrine factor genes associated with angiogenesis after OPBM and  Figure S4). Interestingly, we observed that the ratio of cells in the G0/G1, S, and G2/M phases was also altered following the OPBM of hADSCs. These results indicate that OPBM of hADSCs affects the cell cycle, which may be associated with the upregulation of HIF-1α.
We analyzed the cellular metabolism in hADSCs after OPBM to elucidate the precise molecular mechanisms of action. Analysis of the expression of enzymes (HK1, PFKP, GAPDH, and PKM1/2) involved in the various steps of the glycolysis pathway 48 h after the OPBM of hADSCs revealed that levels of all enzymes were increased. Therefore, we concluded that the glycolysis pathway was activated in response to OPBM. Lactate not only affects cell proliferation, migration, and collagen synthesis, but also plays a critical role in angiogenesis and wound healing, [56][57][58][59][60][61] and is essential for the increased production of angiogenic F I G U R E 5 Enhanced limb salvage, angiogenesis, and decreased muscle degeneration in mice injected with hADSCs subjected to OPBM.

| CONCLUSION
In the present study, we optimized the parameters of OPBM to enhance the expression of angiogenic paracrine factors in hADSCs and verified the underlying molecular mechanism, which differed from previous reports. Importantly, we established the criteria for PBM using a commercially available light source and optimized the custom-

| OLED-based PBM
The experiments were conducted using a red OLED (Kaneka, Osaka, Japan) that emitted 610 nm near-infrared radiation. hADSCs were allowed to adhere to the culture dishes for 1 day. The next day, the nonadherent cells were washed off using phosphate-buffered saline (PBS; Gibco BRL), and fresh culture medium was added. The cells were treated with light for 3-24 h at varying energy densities (ranging from 40 to 300 J/cm 2 ).

| qRT-PCR
Total RNA was extracted from samples using 1 ml TRIzol (Ambion, GAPDH served as an internal control. For the in vivo assays, qRT-PCR was used to quantify the relative expression of CD31 and SM-α-actin.
β-actin served as an internal control. The sequences of primers used for qRT-PCR are listed in Table 1.

| In vitro cell adhesion assay
In brief, cells subjected to or not subjected to OPBM (24 h) were detached using trypsin (Gibco BRL) and immediately re-attached in 6-or 24-well plates. After 3, 6, or 12 h of incubation, the unat-

| NO assay
The NO concentration in the culture supernatants was measured using the Griess reagent. Briefly, hADSCs were seeded in 24-well plates (2 Â 10 4 cells/well) and incubated for 24 h. Culture supernatants were collected after 24 h of OPBM (n = 4). NO production was measured using the Griess Reagent System, according to the manufacturer's instructions (Promega Corp., Madison, WI, USA).
The OD of each well was recorded at 560 nm using a microplate reader (Tecan).

| Western blotting
For the in vitro assay, 1 Â 10 6 hADSCs were subjected to OPBM for 24 h.
For the in vivo assay, mouse limb muscle samples were lysed in RIPA buffer using an electric homogenizer. After centrifugation at 10,000 g for 10 min,

| Histology and IHC
Twenty-eight days after the treatment of mouse models with hindlimb ischemia, muscle tissue samples were obtained from the limbs and embedded in optimal cutting temperature compound (Scigen Scientific, Gardean, CA, USA). After the freezing step, the samples were cut into 10-μm sections at À20 C. Sections containing ischemic regions were stained with H&E to assess muscle degeneration and tissue inflammation. In addition, the sections were subjected to immunofluorescence with anti-CD31 (Abcam, ab28364) and anti-smooth muscle

| Statistical analysis
All quantitative data are expressed as mean ± SD. Statistical analysis was performed using Student's t-test or one-way ANOVA using a Bonferroni test. Statistical significance was set at p < 0.05.

CONFLICT OF INTEREST
Authors declare that they have no competing interests.