Ultraviolet-B exposure causes an inflammatory response, photoaged skin, and degradation of extracellular matrix proteins including collagen and elastin. The regulation of these genes was suggested as an important mechanism to attenuate skin aging. Glycolic acid (GA) is commonly present in fruits and recently used to treat dermatological diseases. We reported that GA slows down cell inflammation and aging caused by UVB. Little is known about GA retarding the skin premature senescence or how to impede these events. To investigate the potential of GA to regulate the expression of MMPs and collagen, GA was topically applied onto human keratinocytes and the C57BL/6J mice dorsal skin. In the present study, we demonstrated that GA reduced UVB-induced type-I procollagen expression and secretory collagen levels. GA reverted and dose-dependently increased the level of aquaporin-3 (AQP3), the expression of which was down-regulated by UVB. The UV-induced MMP-9 level and activity were reduced by GA pre-treatment. Concomitantly, GA reverted mitogen-activated protein kinase (MMP-9) activation and inhibited the extracellular signal-regulated kinase activation (p38, pERK) triggered by UVB. The animal model also presented that GA attenuated the wrinkles caused by UVB on the mouse dorsal skin. Finally, GA triggers the transient receptor potential vanilloid-1 (TRPV-1) channel to initiate the anti-photoaging mechanism in keratinocytes. These findings clearly indicated that the mechanisms of GA promote skin protection against UVB-induced photoaging and wrinkle formation. GA might be an important reagent and more widely used to prevent UVB-induced skin aging.
Glycolic acid (GA) is one of the alpha hydroxy acids and a natural organic acid commonly present in foods . GA is established as a safe and non-toxic substance in fruits and participates in multiple bioactivities such as the induction of antioxidant activity, increase in the efficiency of melasma treatment, and biosynthesis of ceramide [2,3]. GA is often used as a cosmetic ingredient as well as in chemical peels. The early studies reported that topical GA treatment onto the human skin causes the photoprotective and antioxidative effects [4,5].
UV irradiation is one of the most important extrinsic factors that induce skin aging . Chronic exposure to UVB-irradiation results in the activation of pro-inﬂammatory cytokines such as interleukin-1 (IL-1), IL-1β, IL-6, IL-8, tumor necrosis factor-α (TNF-α), and monocyte chemoattractant protein 1 (MCP-1) responsible for skin cell apoptosis, carcinogenesis, collagen degradation, senescence, and skin aging with the appearance of wrinkles [7–9]. UV irradiation causes accumulation of reactive oxygen species (ROS), reduces cellular antioxidant status, and alters signal transduction pathways including the mitogen-activated protein kinase (MAPK), nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB; p65), janus kinase, signal transduction and activation of transcription, and nuclear factor erythroid 2-related factor 2 pathways . Our recent study demonstrated that the GA pre-treatment results in suppressive effects on the UV-induced NF-κB/p65 activation and ROS production [11,12]. We found that GA recruits and induces DNA methyltransferase-3B activity to degrade the inflammasome complex assembly in the HaCaT cell line . Furthermore, we pointed out the optimal concentration (1.5–2.0%) of GA significantly reduces the release of pro-inflammatory molecules and cytokines including cyclooxygenase-2, MCP-1, IL-6, and IL-8 to protect mouse dorsal skin (MDS) against UVB-induced sunburn damage . All these data indicate that the GA pre-treatment might attenuate the UVB-induced inflammatory response, apoptosis, ROS accumulation, and pro-inflammatory cytokine secretion. Therefore, GA might be used as an anti-photoaging agent to protect the skin against UVB radiation. However, the mechanisms through which GA attenuates UVB-induced photoaging and skin damage remain unclear.
Moreover, chronic exposure of the skin to UV irradiation causes the induction of matrix metalloproteinase (MMP) expression, extracellular matrix degradation, and skin dehydration . MMPs are released by keratinocytes and dermal ﬁbroblasts during UVB stress and used as major markers of UVB-induced photoaging [15,16]. ‘Collagenases refer to MMP-1, MMP-8, and MMP-13. MMP-2 and MMP-9 are gelatinases in the skin’ [17–19]. Another major characteristic of photoaging is water loss from the skin. Collagen is used as a functional ingredient in various cosmetic products owing to its efficacy to moisturize and enhance the skin elasticity . UVB irradiation causes the dry, rough, scaly, and aged skin. Water movement across the plasma membrane occurs via the following two pathways: (i) diffusion across the phospholipid bilayer and (ii) water channels such as aquaporins (AQPs) [21,22]. AQP3 is one of the 13 members in the AQP family and speciﬁcally expressed in the basal layer of keratinocytes in the mammalian skin . However, whether GA affects the water channel AQP3 in skin cells and the MDS remains unstudied.
Recent studies indicated that UV irradiation, low-pH cell environment, and heat induce transient receptor potential vanilloid-1 (TRPV1) activation in human keratinocytes . TRPV1 is a member of the nonselective cationic channel family. The photoaged skin exhibited an increase in the TRPV1 expression . Moreover, GA is known to induce keratinocyte proliferation in a skin equivalent model via the TRPV1 activation . However, little is known regarding the TRPV1 expression after the GA pre-treatment in UVB-irradiated skin cells.
As UV irradiation induces the skin inflammation, dehydration, and wrinkles, as well as TRPV1, it plays an important role as an aged skin sensor; in this study, we aimed to investigate the following questions in HSCs and an animal model: (i) whether GA can revert UVB-induced wrinkle formation, collagen degradation, and elastin outflow in the MDS; (ii) whether GA can restore the reduction in AQP3 expression by UVB in HSCs and the MDS; (iii) whether GA regulates the UV-induced MAPK pathway; and (iv) the mechanism through which GA regulates the expression of MMPs and TRPV1 receptors.
Materials and methods
C57BL/6J mice (age: 8–10 weeks) were purchased from the National Laboratory Animal Center (R.O.C.) and housed in the Laboratory Animal Center, Tzu-Chi University. Our animal research methods were approved by the Animal Care, and they comply with the commonly accepted three R's of Tzu-Chi University (approval ID 103-29-1). The mice were maintained under the standard conditions of 12 h light/dark cycles at 24 ± 2°C and relative humidity of 50 ± 10% in the Animal Resource Facility for at least 1 week prior to their use in experiments. The mice were anesthetized using an intraperitoneal injection of ketamine:xylazine (80 : 10 mg/kg body weight) prior to being subjected to experimental procedures. The mice were categorized into four groups (n = 5). The hair on the dorsal side of the mice was removed using commercially available hair removal creams containing thioglycolate trihydrate (∼250 μl/mouse) 1 or 2 days prior to performing the experiments. In this study, we determined the effects of topical application of an aqueous solution of GA (2%, Sigma Chemical Co., U.S.A.). The GA concentration was established based on our recent study.
Cell culture and drug treatments
HaCaT and Hs 68 fibroblast cell lines were obtained from Cell Lines Service, Germany. Cells were cultured in Dulbecco's modified Eagle's medium (Gibco–Invitrogen, Carlsbad, CA, U.S.A.) supplemented with 10% fetal bovine serum (Gibco–Invitrogen), and incubated in a humidified atmosphere containing 5% CO2 at 37°C . Cells at 70–80% confluency were pre-treated with various concentrations of GA (0.1 (pH 7.4), 1 (pH 7.5), 2 (pH 7.3), or 5 mM (pH 7.1) for 24 h.
UVB irradiation was supplied by a closely spaced array of KLBiotech STS-1 sunlamps. The energy output of UVB (290–320 nm) was measured using a UVB photometer (LT Lutron, UV-340A photometer, International Light, Taiwan) and determined as 1.5 mW/cm2. Immediately prior to UVB irradiation, the culture medium was replaced with phosphate-buffered saline (PBS, pH 7.3).
Reverse transcription polymerase chain reaction (RT-PCR) and quantitative real-time PCR (RT-qPCR)
Total RNA was isolated using ReliaPrep™ Cell RNA Miniprep System (Promega) according to the manufacturer's instructions. RT-PCR and RT-qPCR were performed by following the recently described protocols . The primers to perform RT-qPCR ampliﬁcation are shown in Table 1. The relative gene expression was evaluated using the 2[(-Delta Delta) C(T)] method. The results were expressed in the linear form using the formula 2-ΔΔCT and the relative expression values >2-fold variation were considered as significant .
|Gene||Accession number||Primer sequences 5′ → 3′||Length (bp)|
|Gene||Accession number||Primer sequences 5′ → 3′||Length (bp)|
Western blotting analysis
Total proteins were extracted using RIPA buffer and quantified by reagent (Bio-Rad). Equal amounts of protein were resolved on 12% gradient gel by performing sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and electrophoretically transferred onto PVDF membranes (Amersham, Buckinghamshire, U.K.). The membranes were incubated with specific antibodies against mouse and human TRPV1 (VR1(E-8), Santa Cruz Biotechnology, INC), human AQP3 (Bioss Inc, Boston, MA), MMP-9 (Neomarker, Fremont, CA), antibodies against ERK, JNK, p38, c-Jun, and c-Fos (Cell Signaling Technology, Beverly, MA), and β-actin (Sigma Chemical Co.). The blots were washed with Tris-buffered saline (Tween-20) and incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies (1 : 5000 dilution; Dako Cytomation) for 1 h. The immunoreactivity bands were visualized using the enhanced chemiluminescence (Amersham) and exposure to UVP Biospectrum Imaging System (Upland, CA, U.S.A.).
Immunohistochemical stain and Masson trichrome stain
Samples were ﬁxed in 10% formalin for 24 h and embedded in parafﬁn. Serial sections (3.5 mm) were mounted on to slides and stained with Bouin's Solution (Picric acid (saturated) 75 ml; Formaldehyde (37–40%) 25 ml and glacial acetic acid 5 ml). This solution will improve Masson Trichrome staining quality. The amounts of collagen ﬁbers in MassonTrichrome-stained sections and elastic ﬁbers in Gomori's aldehyde fuchsin-stained section were measured using ImageJ software. IHC stain was defined through the pathologist's specialty interpretation.
The culture medium was suspended in a loading buffer (10% SDS, 25% glycerol, 0.25 M Tris (pH 6.8), and 0.1% bromophenol blue) and without prior denaturation run on 10% SDS–PAGE gels containing 0.5 mg/ml gelatine. After electrophoresis, gels were washed to remove SDS and incubated for 30 min at room temperature in a renaturing buffer (50 mM Tris, 5 mM CaCl2, 0.02% NaN3, 1% Triton X-100). In the next steps, gels were incubated in a developing buffer [50 mM Tris (pH7.8), 5 mM CaCl2, 0.1 of 5 M NaCl, and 1% Triton X-100]. Gels were subsequently stained with 0.5% Coomassie Brilliant Blue G-250 and destained in 30% methanol, 10% acetic acid to detect gelatinolytic activity.
Quantification of total collagen synthesis
The total collagen (TC) synthesis was measured using a Sircol soluble collagen assay kit (Biocolor Ltd., U.K.). The cell culture medium was mixed with Sircol dye reagent and incubated at room temperature for 30 min. After centrifugation, ice-cold acid-salt washing reagent was added to the precipitate and the mixture was centrifuged. The precipitate was dissolved in alka reagent and the absorption was determined at 555 nm using an enzyme-linked immunosorbent assay reader (Tecan, Grodig, Austria).
‘The SPSS 18.0 (SPSS Inc., IL, U.S.A.) statistical software packages were used for analyzing the data. Data are presented as the mean ± standard deviation. The parametric independent samples t-test and the nonparametric Mann–Whitney U-test were used to compare differences between the two groups.’ Statistical significance was set at P < 0.05 and P < 0.001.
GA attenuated the reduction in TC synthesis caused by UVB exposure in human skin keratinocytes
To determine whether GA regulates the TC expression in skin cells, cultured keratinocytes and fibroblasts were treated with various GA doses for 24 h and the TC mRNA expression was analyzed by performing RT-qPCR. GA signiﬁcantly increased the TC synthesis at 0.1 and 0.5 mM concentrations in both the skin cell lines (P < 0.05) (Figure 1A). GA notably stimulated the TC production at the 24 h time point (Figure 1B). We established the optimal GA concentration to be 0.5 mM to perform further experiments. Cells were treated with or without GA prior to UVB irradiation. The UVB group exhibited a significant reduction in the TC mRNA and protein production compared with the control group; however, TC levels were significantly restored in the GA pre-treatment group (P < 0.01) (Figure 1C,D). We determined elastin mRNA expression in the aforementioned treatment groups of skin cells, which decreased in a dose-dependent manner in the UVB-irradiated group (Figure 1E). GA drastically reverted the UVB-inhibited elastin mRNA expression in the GA pre-treated and UVB-exposed group compared with the UVB-irradiated group (Figure 1F).
The effects of GA on collagen and elastin expression in human skin cells.
GA inhibited the UVB-induced collagen degradation and wrinkle formation in the MDS
To examine the effects of GA on the UVB-induced collagen degradation and wrinkle formation in in vivo condition, GA was applied on the dorsal skin of mice for 24 h and exposed to UVB radiation. UVB exposure resulted in macroscopic wrinkle formation and prominent skin folds in the MDS of vehicle-treated and UVB-exposed group compared with that in the unexposed control group. However, GA drastically inhibited wrinkle formation of UVB-exposed skin, consistent with macroscopic observations (Figure 2A). Next, the tissue sections were subjected to MDS to evaluate the changes in collagen fiber formation in the dermal areas of UVB-exposed MDS. The normal collagens were present in the bundle formation and abundant in the dermis. However, after the UVB was exposed, the skin collagen exhibited a greater extent of damage, including fibers broken, less extent of bundle formation, and loosely scattered in the tissues compared with the vehicle-treated group. In the pretreat GA group, GA restored the collagen fiber formation and attenuated the UVB-induced necrosis in the upper layer of the dermis of GA-treated and UVB-exposed mice (Figure 2B). The collagen content of UVB-irradiated cells was signiﬁcantly lower than that in the control group (P < 0.01); however, the UVB-induced collagen fiber loss and necrosis in the upper layer of dermis were significantly attenuated in the GA-treated group (Figure 2C).
GA protect C57BL/6J mouse skin from UVB-induced wrinkle formation and collagen degradation in the dorsal skin of hairless mice.
GA restored the UVB-inhibited AQP3 expression in HSCs and the MDS
A significant up-regulation of AQP3 mRNA was observed in all the GA-treated groups of keratinocytes and fibroblast cells, especially in HaCaT cells in a dose-dependent manner (Figure 3A). The protein levels were similar to the gene expression levels (Figure 3B). To confirm the UVB-induced AQP3 level degradation status, cells were exposed to various doses of UVB radiation and AQP3 mRNA levels were determined by performing RT-qPCR. AQP3 mRNA expression was decreased in a dose-dependent manner in UVB-irradiated cells compared with that in the control group (Figure 3C). AQP3 was degraded by UVB irradiation; however, this was restored in the GA pre-treated group (Figure 3D) and this result was consistent even at the protein level (Figure 3E). Furthermore, we pre-treated the MDS with GA (2%) for 24 h prior to UVB irradiation and determined the AQP3 expression by performing IHC assay. GA promoted the AQP3 protein concentration and increased the protein expression in mouse skin. However, UVB irradiation damaged keratinocytes, broke (black triangle) the MDS, and induced the release of inflammatory cells 2.25-folds compared to the control group (gray triangle). The dispersion and shallow distribution of AQP3 level were presented as a white triangle in the image. GA attenuated the damage of keratinocytes in the MDS and restored the AQP3 expression level at least 1.6-fold (Figure 3F). All these results indicated that GA restored UVB-inhibited AQP3 expression in human keratinocytes and the MDS.
GA inhibits UVB-induced down-regulation of AQP3 expression in cultured human skin cells and mice dorsal skin cells.
GA inhibited the UVB-mediated induction of MMP expression in HaCaT cells
UVB irradiation caused a marked increase in the MMP-9 mRNA level; however, this was inhibited by GA pre-treatment (Figure 4A). This result was consistent even in the case of protein levels (Figure 4B). To study the effect of GA on the UVB-mediated alteration in MMP inhibitor, the tissue inhibitor of metalloproteinases (TIMP-1) expression was determined by performing RT-qPCR. UVB irradiation caused a decrease in TIMP-1 level; however, GA pre-treatment restored the TIMP-1 expression (gray column) (Figure 4C). Furthermore, we analyzed the MMP-9 activity by performing a gelatin zymography assay. The MMP-9 activity was significantly higher by 1.42-fold in the UVB-irradiated group than that in the control group; however, this activity was decreased in the GA pre-treated and UVB-irradiated group (Figure 4D). However, GA did not affect the MMP-2 expression in the UVB-treated cells and the result was consistent in the case of MMP-2 activity (Figure 4A).
GA decreased the MMP-9 enzyme activities in UVB-treated keratinocytes.
GA regulated the UV-induced MAPK pathway
In the UVB-irradiated group, the procollagen expression was signiﬁcantly decreased by 0.7-fold compared with the control group. The procollagen expression reduced by UVB irradiation was increased by 1.9-fold in the GA-treated and UVB-irradiated group compared with that in the control group. Similarly, UVB irradiation caused a 2.1-fold, 1.8-fold, and 1.1-fold increase in p-MAPK, phospho extracellular signal-regulated kinase (p-ERK1/2), and p-c-Jun levels, respectively, relative to their levels in the control group. However, GA significantly reverted the UV-induced p-MAPK phosphorylation level (Figure 5).
Effect of GA on the regulation of type-I procollagen and MAPK signal pathway.
Synergistic regulation of TRPV1 expression in GA-treated and UVB-irradiated keratinocytes
UVB irradiation increased the TRPV1 expression in a dose-dependent manner (Figure 6A). Moreover, we validated the effect of acidic environment on the TRPV1 expression in keratinocytes. HaCaT cells were treated with various GA concentrations for 24 h and RT-qPCR was performed. TRPV1 level was significantly increased by 4-fold in GA (0.1 mM)-treated group which is higher than that in the control group (Figure 6B) and the results were consistent in the case of protein levels (Figure 6C). GA treatment prior to UVB irradiation strongly increased the TRPV1 expression (Figure 6D). To estimate the optimal GA concentration that triggered the TRPV1 expression in in vivo condition, the MDS was moistened with various GA concentrations and stained using the specific antibody against TRPV1 by performing IHC analysis. TRPV1 level was significantly increased and distributed in the stratum corneum increased folds at least 2, 10, 18, and 17 folds, respectively. Stimulated with more than 2% GA treatment significantly increased TRPV1 expression (Figure 6E).
TRPV1 expression profile in response to GA with and without UVB irradiation.
We observed that in HaCaT cells and the hairless dorsal skin of C57BL/6J mice, UVB inhibited AQP3 expression and decreased the moisture content. UVB also induced wrinkle formation and collagen degradation in the MDS and induction of MMP-9 expression via TIMP-1 regulation via activated the p-MAPK pathway in HaCaT cells. In this article, we demonstrated that topical application of GA prior to the UVB irradiation could signiﬁcantly attenuate these manifestations. GA also has the synergistic regulation to promote the TRPV1 expression in UVB-irradiated keratinocytes. To the best of our knowledge, this is the first in vivo study to evaluate the efficiency of GA to prevent the UVB-induced collagen degradation and that indicates the mechanism of action of GA by which it attenuates the inhibition and induction of AQP3 and MMP-9 expression, respectively, by UVB irradiation via MAPK pathway in keratinocytes and the MDS.
The photoaging process involving dermal collagen is dissimilar to the chronological aging process in the human skin . The chronological aging in the skin was proved to be associated with the telomerase activity and telomeres that are the specialized structures present at the ends of chromosomes and believed to be an integral part of natural cellular aging as well as in the carcinogenesis . However, UVB-induced skin aging has been elucidated in different pathways, such as NF-kB signaling pathway that leads to the increase in proinflammatory cytokines including IL-1β, IL-6, and the MAPK/activator protein-1 (AP-1) signaling process that results in the elevation of MMP levels that degrade collagen fibers [30,31], and free radicals, etc. (An increased expression of MMPs is associated with the collagen degradation in photoaged and chronologically aged human skin that leads to the mild wrinkle formation and aged appearance ). Therefore, the balance between the levels of MMPs and procollagen expression plays an important role in chronologically aged and photoaged skin. Dissimilar to the chronological aging process, the photoaging of skin is clinically characterized by the occurrence of dryness, pigmentation, laxity, and deep wrinkling owing to the accumulation of environmental damage, particularly exposure to UVB irradiation . In this study, we proved that GA is an efficient photoprotective agent against the UVB-induced skin wrinkles, collagen degradation, and loss of moisture and MMP-9 expression in in vitro as well as in vivo conditions. We demonstrated that GA at 0.1 and 0.5 mM concentrations stimulated the procollagen synthesis (Figure 1A) as well as inhibited the UVB-induced MMP-9 mRNA expression, protein level, and activity (Figure 4). ‘MMP-9 mRNA is massively increased following UVB radiation in HaCaT cells, and that may be resulted from MAPK-mediated AP-1 activation.’ MMPs are regulated by endogenous inhibitors named TIMPs, especially TIMP-1; however, UVB radiation disrupts the balance between MMPs and TIMPs [34,35]. GA induced the TIMP-1 expression in UVB-irradiated skin cells (Figure 4C). All these data indicated that GA plays an important role in the repression of MMP-9 level and improved the procollagen synthesis.
Water movement across the plasma membrane occurs via diffusion through the lipid bilayer or water channels that traverse the membranes. Multiple studies indicated that the skin of AQP3-null mice is relatively dry, rough, and aged. Moreover, the AQP3 expression is involved in skin diseases including wound healing, atopic dermatitis, and psoriasis. [36–38]. The activation or up-regulation of AQP3 level might increase the water content, improve the barrier function and skin appearance possibly by reducing the skin aging and wrinkling . We observed that GA up-regulated the AQP3 mRNA level in a dose-dependent manner (Figure 3A) and restored the UVB-inhibited AQP3 expression in human keratinocytes and the MDS (Figure 3). We suggest that GA has moisture maintenance functions and prevents the UVB-induced dryness and aging of the skin.
TRPV1 is an acid-sensitive ion channel linked to heat and pain reception and is expressed in human epidermal keratinocytes . We indicated that the minimum concentration of GA (2%) stimulated the TRPV1 expression in the MDS. The GA concentration (5 mM) used in in vitro experiments was equivalent to 0.038% (W/V). Additionally, we demonstrated that either UVB-induced photoaging (Figure 6A) or GA-treated group (Figure 6B) induces the TRPV1 expression in keratinocytes. The GA pre-treatment along with UVB irradiation synergistically induced the TRPV1 expression in human keratinocytes (Figure 6D). Interestingly, GA attenuated and restored the UVB-induced collagen degradation and AQP3 reduction, respectively. Morris et al. indicated that TRPV1 is a calcium-permeable channel and the alteration of calcium level in the epidermis induces various changes in the epidermal physiology such as cellular differentiation, permeability barrier homeostasis, and epidermal cell proliferation [40,41]. We indicated that co-treatment with GA and UVB radiation enhanced the release of cytoplasmic calcium level and endoplasmic reticulum stress . Oh et al. proposed that TRPV1 plays a signiﬁcant role in the pain transduction pathway and exhibits a well-deﬁned pro-inﬂammatory role in various diseases and injury stresses [43,44]. GA has been established as a non-toxic fruit substance. However, our previous studies have demonstrated that up to 3% of GA resulted in prominent wrinkled appearances and skinfolds or more than 5 mM of GA caused cell survival rate only 20%. Those data indicated that GA still has the limited concentration. We still do not know the half-life of GA and how to regulate the performance of TRPV1. These need more experiments in the future to prove.
We clearly demonstrated that GA attenuates the UVB-induced MMP-9 expression, AQP3 level reduction, and collagen degradation in human keratinocytes and the MDS. Additionally, we demonstrated that MAPK acts as a signal transducer in this aforementioned process. Notably, we observed the synergistic regulation of TRPV1 expression in GA-treated and UVB-irradiated keratinocytes. All these findings presented in the graphical representation (Figure 7). As GA regulates the AQP3, procollagen, and MMP-9 expressions, we suggest GA as a useful agent to inhibit the UVB-induced skin aging.
Scheme of GA inhibition of UVB-induced AQP3, MMP-9 expression, and collagen degradation in Keratinocytes.
mitogen-activated protein kinase
monocyte chemoattractant protein 1
mouse dorsal skin
nuclear factor kappa B
normal human epidermal keratinocyte
polyacrylamide gel electrophoresis
reactive oxygen species
tissue inhibitor of metalloproteinases
tumor necrosis factor-alpha
transient receptor potential vanilloid-1
S.-C.T. conceived and designed the experiments. S.-C.T., J.-C.L., and P.-Y.L. performed the experiments. L.-C.T. contributed the data statistics. S.-C.T. and J.-H.Y. discussed the results and prepared the manuscript. C.-H.L. provided accurate statistical methods and recalculated the raw data. Chin-Hung Liu also provided the results of the IHC interpretation. C.-H.L. and S.-C.T. rewrote and revised the manuscript. All authors read and approved the final manuscript.
This study was supported by the Buddhist Tzu Chi Medical Foundation, Hualien, R.O.C. [TCMMP 10401-01].
The Authors declare that there are no competing interests associated with the manuscript.