Comparison of the Inhibitory Activities of 5,6-Dihydroergosterol Glycoside α- and β-Anomers on Skin Inflammation

Chronic skin inflammatory diseases, such as atopic dermatitis, are associated with a dysfunctional skin barrier due to an increase in various inflammatory stimuli, for instance inflammatory cytokines and chemokines. In particular, CCL17 and CCL22 expression is increased in patients with chronic skin inflammation. In this study, we synthesized several α- and β-anomers of dihydroergosterol (DHE)-glycosides and assessed their effects on CCL17 and CCL22 expression. We confirmed that the β-anomers of DHE-glycosides were superior to α-anomers of DHE-glycosides in inhibiting CCL17 and CCL22 mRNA and protein expression. In addition, we determined that DHE-glycoside β-anomers showed strong inhibitory activity towards pro-inflammatory cytokine mRNA and protein expression, including that of TNF-α, IL-6, and IL-1β- in stimulated HaCaT cells. These results imply that DHE-glycoside α- and β-anomers should be separated during synthesis of drugs for chronic skin inflammation. Our results also suggest that β-anomers of DHE-glycosides may play an important role as new drugs for chronic skin inflammation because of their ability to inhibit the skin inflammatory biomarker proteins CCL17 and CCL22.


Introduction
Natural products from various plants with soap-like properties, such as saponins, have played an important role in medicine and daily life for thousands of years [1]. Saponins are amphipathic glycosides grouped structurally by having one or more hydrophilic glycoside moieties combined with a lipophilic triterpene or steroid derivative, and have well characterized biological properties including hemolytic activity, molluscicidal activity, and anti-inflammatory activity [2].
We previously confirmed that spinasterol-glucose (1) (spinasterol-Glc, Figure 1), extracted and identified from natural herbs, not only exhibited potent anti-inflammatory activity but also inhibited CCL17 mRNA and protein expression in TNF-α/INF-γ-stimulated HaCaT cells [3]. However, spinasterol-Glc is difficult to obtain from either natural plant extracts or using synthetic methods. The synthesis of spinasterol-Glc is very difficult because its steroidal skeleton, α-spinasterol, is not readily available. Recently, synthesis of α-spinasterol from stigmasterol using a double bond migration method has been reported; however, the synthesis was still difficult and inefficient [4,5]. Therefore, we previously designed and synthesized a new sterol, 5,6-dihydroergosterol (DHE or stellarsterol) that is structurally similar to spinasterol and more easily obtained. In previous studies, we demonstrated that DHE and its glycosides have comparable anti-inflammatory activity to spinasterol-Glc. Of the Structures of spinasterol-glucose (spinasterol-Glc) and αand β-anomers of 5,6-dihydroergosterol (DHE)-glycosides.
Generally, glycosides consist of glycone (sugar) and aglycone (non-sugar) components, and can be divided into alpha (α) and beta (β) anomers depending on the linking stereochemistry of the glycone and aglycone portions ( Figure 1). In our synthesis process, we have always obtained both the αand β-anomers of DHE-glycosides, particularly during the glycosylation step. Since there have been few reports describing the relative bioactivity data of αand β-anomers of DHE-glycosides, we were interested in determining their relative biological activities. To do this, we compared the inhibitory activities of αand β-DHE-glycosides on TNF-α/IFN-γ-induced expression of pro-inflammatory cytokines and chemokines, including CCL17 and CCL22, in TNF-α/IFN-γ-stimulated human keratinocytes. Here, we report the synthesis and anti-inflammatory activity of αand β-anomers of DHE-glycosides, and which anomer of DHE-glycosides more effectively inhibits the gene expression of pro-inflammatory cytokines IL1-b, IL-6, and TNF-α, and the chemokines CCL17 and CCL22, in HaCaT cells. Our results indicate that β-anomers of DHE-glycosides inhibit the mRNA and protein expression of pro-inflammatory cytokines and chemokines in TNF-α/IFN-γ-activated HaCaT cells to a greater extent than the α-anomers of DHE-glycosides, and that this is likely because of the instability in solution of the α-anomers of DHE-glycosides. These findings may contribute to the development of useful therapeutic agents for chronic skin inflammation diseases such as atopic dermatitis.

Synthesis of α-and β-Anomers of 5,6-Dihydroergosterol-Glycosides (DHE-Glycosides)
As shown in Scheme 1, βand α-anomers (2-5) of DHE-glycosides can be prepared by acid-catalyzed Schmidt glycosylation of DHE with OH-protected sugars followed by deprotection. Previously, we determined which of various glycosyl donors, including glycosyl halide, glycosyl sulfide, and glycosyl trichloroacetimidate, in the presence of several catalysts, resulted in the most efficient glycosylation of DHE. We found that when O-tetrabenzoyl-protected glycosyl trichloroacetimidates (7 and 8) were used as glycosyl donors and Cu(OTf) 2 was used as a catalyst, an effective stereoselective glycosylation reaction occurred, especially for β-anomers [6]. Therefore, glycosyl trichloroacetimidates (7 or 8), glycosyl donors for Cu(OTf) 2 -catalyzed glycosylation, were used in this study. The two glycosyl donors, glucopyranosyl and galactopyranosyl trichloroacetimidate, and DHE were synthesized according to previous methods [6].
The anomeric effect in sugar chemistry predicts that α-anomers will be the major product of glycosylation reactions. However, β-anomers have been reported to be produced when acyl groups capable of neighboring group participation (NGP) are present at adjacent carbon positions. Unlike β-anomers, which are the major products in our reaction with a benzoyl group in an adjacent position, α-anomers are not readily available in sufficient quantities to carry out biological activity studies. According to the literature, α-anomers are primarily produced when a benzyl (Bn)-protected sugar is used as a glycosyl donor [19]. We attempted to glycosylate with a Bn-protected sugar, but there was an issue during the debenzylation step. Therefore, we examined the various reaction conditions under which α-anomers are obtained with a benzoyl (Bz)-protected sugar. Experimental results for reactions of glucosyl imidate (7) with DHE (6) under various conditions are shown in Table 1. As expected, β-anomers were primarily synthesized in the four acid catalytic conditions (Entry 1-4), and α-anomers were obtained at a 10% yield under the SiO 2 -H 2 SO 4 acid condition (Entry 2).

Effect of α-and β-Anomers of 5,6-Dihydroergosterol(DHE)-Glycosides on Cell Viability
The viability of HaCaT cells treated with αand β-anomers was determined with MTT assays to determine experimental DHE-Glc and DHE-Gal concentrations. We have previously confirmed the effects of β-anomer DHE-sugar derivatives on cell viability [15]. As shown in Figure 2, none of the tested αand β-anomers of DHE-Glc and DHE-Gal had an effect on cell viability up to a concentration of 20 µM. In addition, we confirmed that cell survival remained at about 80% when the concentration of the compounds was 20 µM, and cell viability at these concentrations was similar to that obtained with a 0.1% DMSO working solution alone. Therefore, all experiments were performed using concentrations up to 20 µM, and results can be interpreted independently of cell survival.

Effects of DHE-Glycoside α-and β-Anomers on mRNA Expression Levels of Pro-Inflammatory Cytokines and Chemokines in TNF-α/IFN-γ Induced HaCaT Cells
We examined the effects of the αand β-anomers of DHE-Glc and DHE-Gal on pro-inflammatory cytokine and chemokine gene expression in HaCaT cells. We evaluated the mRNA expression levels of pro-inflammatory cytokines TNF-α, IL-1β, and IL-6 using an RT-PCR assay in TNF-α/IFN-γ-treated HaCaT cells. In addition, we also measured mRNA expression levels of chemokines CCL17 and CCL22, which are known to have a role in chronic skin inflammation. We confirmed that mRNA expression levels of the pro-inflammatory cytokines TNF-α, IL-6, and IL-1β were dramatically increased in TNF-α/IFN-γ-treated HaCaT cells. CCL17 and CCL22 mRNA expression levels had a similar expression pattern to TNF-α, IL-6, and IL-1β. Neither DHE (6), nor the DHE-glycoside α-anomers (3 and 5), had an effect on TNF-α, IL-6, or IL-1β mRNA levels in TNF-α/IFN-γ-treated HaCaT cells. In contrast, we found that DHE-glycoside β-anomers (2 and 3) reduced pro-inflammatory cytokine and chemokine mRNA expression in activated HaCaT cells ( Figure 3). These results indicate that the β-anomers of DHE-glycosides were more effective than the α-anomers of DHE-glycosides at inhibiting gene expression of pro-inflammatory cytokines including IL1-β, IL-6, TNF-α and chemokines CCL17 and CCL22 in TNF-α/IFN-γ-treated HaCaT cells.

Effects of DHE-Glycoside α-and β-Anomers on mRNA Expression Levels of Pro-Inflammatory Cytokines and Chemokines in TNF-α/IFN-γ Induced HaCaT Cells
We examined the effects of the α-and β-anomers of DHE-Glc and DHE-Gal on proinflammatory cytokine and chemokine gene expression in HaCaT cells. We evaluated the mRNA expression levels of pro-inflammatory cytokines TNF-α, IL-1β, and IL-6 using an RT-PCR assay in TNF-α/IFN-γ-treated HaCaT cells. In addition, we also measured mRNA expression levels of chemokines CCL17 and CCL22, which are known to have a role in chronic skin inflammation. We confirmed that mRNA expression levels of the pro-inflammatory cytokines TNF-α, IL-6, and IL-1β were dramatically increased in TNF-α/IFN-γ-treated HaCaT cells. CCL17 and CCL22 mRNA expression levels had a similar expression pattern to TNF-α, IL-6, and IL-1β. Neither DHE (6), nor the DHE-glycoside α-anomers (3 and 5), had an effect on TNF-α, IL-6, or IL-1βmRNA levels in TNFα/IFN-γ-treated HaCaT cells. In contrast, we found that DHE-glycoside β-anomers (2 and 3) reduced pro-inflammatory cytokine and chemokine mRNA expression in activated HaCaT cells ( Figure 3). These results indicate that the β-anomers of DHE-glycosides were more effective than the α-anomers of DHE-glycosides at inhibiting gene expression of pro-inflammatory cytokines including IL1-β, IL-6, TNF-α and chemokines CCL17 and CCL22 in TNF-α/IFN-γ-treated HaCaT cells. To determine the effects of synthetic α-and β-anomers of DHE-Glc and DHE-Gal on pro-inflammatory cytokine IL-1b, IL-6, and TNF-α expression, mRNA levels were determined using an RT-PCR assay. mRNA expression levels of the chemokines CCL17 and CCL22 were compared under similar conditions. RT-PCR reactions were performed in triplicate and repeated at least three times.

Effects of DHE-Glycoside α-and β-Anomers on Pro-Inflammatory Cytokine and Chemokine Protein Expression Levels in TNF-α/IFN-γ-Induced HaCaT Cells
Next, we used ELISA assays to examine whether DHE-glycoside α-and β-anomers have an effect on pro-inflammatory cytokine and chemokine protein expression in TNF-α/IFN-γ-induced HaCaT cells. In TNF-α/IFN-γ-induced HaCaT cells, TNF-α, IL-6, IL-1β, CCL17, and CCL22 protein To determine the effects of synthetic αand β-anomers of DHE-Glc and DHE-Gal on pro-inflammatory cytokine IL-1b, IL-6, and TNF-α expression, mRNA levels were determined using an RT-PCR assay. mRNA expression levels of the chemokines CCL17 and CCL22 were compared under similar conditions. RT-PCR reactions were performed in triplicate and repeated at least three times.

Effects of DHE-Glycoside α-and β-Anomers on Pro-Inflammatory Cytokine and Chemokine Protein Expression Levels in TNF-α/IFN-γ-Induced HaCaT Cells
Next, we used ELISA assays to examine whether DHE-glycoside αand β-anomers have an effect on pro-inflammatory cytokine and chemokine protein expression in TNF-α/IFN-γ-induced HaCaT cells. In TNF-α/IFN-γ-induced HaCaT cells, TNF-α, IL-6, IL-1β, CCL17, and CCL22 protein expression increased by approximately 4-to 9-fold compared to controls. DHE treatment had no effect on pro-inflammatory cytokine protein expression in HaCaT cells relative to control cells. Moreover, TNF-α protein expression was slightly reduced compared to the positive control. HaCaT cells treated with α-anomers of DHE-Glc and DHE-Gal (4 and 5) had similar levels of IL-1β, IL-6, TNF-α, CCL17, and CCL22 protein expression as DHE-treated cells. These results were consistent with the mRNA gene expression results. However, TNF-α/IFN-γ-induced HaCaT cells treated with β-anomers of DHE-Glc and DHE-Gal (2 and 3) showed reduced pro-inflammatory cytokine and chemokine protein expression. We confirmed that the β-anomer of DHE-Glc was more effective than the β-anomer of DHE-Gal at reducing cytokine and chemokine protein expression ( Figure 4). Our results indicate that the β-anomer of DHE-Glc (2) inhibited TNF-α/IFN-γ-activated expression of IL-1β, IL-6, TNF-α, CCL17, and CCL22 proteins, which resulted in the reduction of cytokine and chemokine mRNA levels in HaCaT cells. To determine the effects of synthetic DHE-glycoside αand β-anomers on IL-1β, IL-6, and TNF-α protein expression, protein expression was determined using ELISA assays. CCL17 and CCL22 protein expression was determined under same conditions. The values for protein expression intensity are the mean ± S.D. of three independent experiments (n = 3). * p < 0.01 compared to mock medium alone, ** p < 0.01 compared to treatment with TNF-α/IFN-γ.

Discussion
The genesis of this study was evaluating the anti-inflammatory activity of extract of the leaves of Stewartia koreana (SKE), which is a natural herb native to Korea. We found that SKE had potent anti-inflammatory activity. SKE inhibited the LPS-induced NF-κB signaling mechanism in RAW264.7 mouse macrophages, and strongly inhibited the production of NO and PGE 2 [20]. Next, we investigated the active ingredient of SKE, and identified spinasterol-Glc (1) as the main active ingredient of SKE extract. Spinasterol-Glc (1) inhibited the production of LPS-induced NO in RAW264.7 mouse macrophage cells, in addition to inhibiting the expression of the pro-inflammatory cytokines TNF-α, IL-6, and IL-1β. This inhibition proved to be the result of inhibition of the IκB-α/IKK phosphorylation process [21]. We also found that spinasterol-Glc (1) strongly inhibited the expression of TARC/CCL17 stimulated by TNF-α/IFN-γ in HaCaT cells. This phenomenon was caused by suppression of phosphorylation of c-Raf, p38 MAPK, and JAK2. Furthermore, spinasterol-Glc (1) inhibited NF-κB and STAT1 promoter activation, and we confirmed that spinasterol-Glc (1) was a potential treatment option for chronic skin inflammation due to its ability to suppress CCL17 expression [3]. However, spinasterol-Glc (1) is difficult to isolate from natural extracts or synthesize. Therefore, we designed and synthesized an analogue of spinasterol, 5,6-dihydroergosterol (DHE). DHE has the same steroidal backbone as spinasterol, but different side chains. We have examined the biological activity of four DHE-glycosides, including ergosterol, in previous studies. We showed that DHE-glycosides strongly inhibited the production of NO induced by LPS in RAW264.7 mouse macrophages [6]. Next, we investigated the inhibitory effect of DHE-Glc on chronic skin inflammation in DNCB-induced animal models. We found that DHE-Glc inhibited the infiltration of epidermal eosinophil and mast cells in our DNCB-induced skin inflammation animal model and that DHE-Glc reduced the concentration of IgE and histamine and mRNA expression and protein levels of CCL17/22 in the plasma of DNCB-treated animals. We also confirmed that inhibition of CCL17/22 expression was due to suppression of NF-κB and STAT1 signaling [7]. Therefore, we reasoned that DHE-Glc (2) would be useful as a therapeutic agent for chronic skin inflammation.
In our synthesis process, however, the αand β-anomers of DHE-glycosides have always been obtained together, particularly during the glycosylation step. To determine the relative biological activity of αand β-anomers, we prepared αand β-anomers of DHE-glycosides and evaluated their anti-inflammatory activity. In more detail, we synthesized αand β-anomers of DHE-Glc and DHE-Gal (2, 3, 4, and 5) via acid-catalyzed glycosylation of DHE (6) with OH-protected glucosyl trichloroimidate (7) or OH-protected galactosyl trichloroimidate (8), followed by deprotection (Scheme 2). Based on the results of glycosylation reactions in various acids (Table 1), β-anomers were synthesized by Cu(OTf) 2 -catalyzed glycosylation followed by deprotection, and α-anomers were obtained by SiO 2 -H 2 SO 4 -catalyzed glycosylation followed by deprotection. We determined the anti-chronic inflammatory activities of these compounds by measuring mRNA and protein expression levels of pro-inflammatory cytokines TNF-α, IL-6, and IL-1β and chemokines CCL17 and CCL22 in TNF-α/IFN-γ-stimulated HaCaT cells after treatment of these cells with these compounds. We found that the β-anomers of DHE glycosides had a greater inhibitory effect than the α-anomers of DHE-glycosides on the gene expression of the pro-inflammatory cytokines IL1-β, IL-6, TNF-α and the chemokines CCL17 and CCL22 in HaCaT cells. In addition, the β-anomers of DHE-glycosides decreased protein expression levels of these pro-inflammatory cytokines and chemokines. This inhibitory effect is likely due to inhibition of the phosphorylation of IKK/IκBα and activation of NF-κB and STAT1 by the β-anomers of DHE-glycosides [7]. In chronic skin inflammatory sites, keratinocytes and various immune cells are activated by inflammatory stimuli, such as TNF-α and IFN-γ [22,23]. The secretion of chemokines and pro-inflammatory cytokines from keratinocytes is responsible for chronic skin disorder diseases as well as abnormal immune responses [24][25][26]. TNF-α and IFN-γ induce strong inflammation in vivo [26,27]. According to previous reports, TNF-α and IFN-γ increased the expression and secretion of pro-inflammatory cytokines IL-1β and IL-6 and chemokines CCL17 and CCL22 in chronic skin inflammatory lesions [10,[28][29][30]. Therefore, inhibiting the secretion of pro-inflammatory cytokines and chemokines is important for developing drug candidates that can effectively treat chronic skin inflammation.

Cell Viability Assay
Cultured cells were removed from cell culture plates, and HaCaT cells at a concentration of 5 × 10 4 cells/well were re-cultured in 96-well plates for 18 h. After the growth medium was discarded, HaCaT cells were treated with the indicated concentration of each compound in serum-free medium for 24 h. After removing the culture medium, cells were treated with 100 µg/mL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-thetazolium bromide (MTT) for 1 h for formazan formation in live cells. Crystallized formazan was dissolved in 200 µL DMSO and the wavelength of each sample was analyzed by spectrophotometer using an absorbance measurement at 560 nm. Analyses were repeated three times, and results are expressed as means of three independent experiments.

Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) Analysis
The RNA-Bee isolation kit (Tel-Test, Friendswood, TX, USA) was used according to the manufacturer's recommendations for isolating total RNA from HaCaT cells. cDNA was prepared using reverse-transcriptase M-MuLV (Fermentas Life Science, Pittsburgh, PA, USA). PCR amplification was performed using specific primers (Bioneer, Daejeon, Korea).

Enzyme-Linked Immunosorbent Assay
HaCaT cells were cultured in 6-well plates and then the growth medium was removed. Cells were treated with TNF-α/IFN-γ (10 ng/mL) in the presence or absence of each compound for 18 h. Supernatants of cultured cells were collected and analyzed using an enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions.

Statistical Analysis
Unless otherwise stated, all experiments were performed with triplicate samples and repeated at least three times. Data are presented as means ± S.D. and statistical comparisons between groups were performed using 1-way ANOVA followed by a Student's t-test.