Designed Mannosylerythritol Lipid Analogues Exhibiting Both Selective Cytotoxicity Against Human Skin Cancer Cells and Recovery Effects on Damaged Skin Cells

Graphical Abstract

We report the synthesis of four kinds of mannosylerythritol lipid (MEL) analogues R-MEL−A ([2R,3S]-erythritol type), S-mannosylthreitol lipid (MTL)-A ([2S,3S]-threitol type), R-MTL−A ([2R,3R]-threitol type), and α-S-MEL−A ([2S,3R]-erythritol type) and their selective cytotoxicity against HSC-5 cancer cells. The results indicate that R-MTL−A has potential as a lead compound for new cosmeceuticals exhibiting both cancer cell-selective toxicity and recovery effects on damaged skin cells.

Abstract

Mannosylerythritol lipids (MELs) are a class of amphipathic molecules bearing a hydrophilic 4-O-β-D-mannopyranosyl-D-erythritol skeleton. Here, we designed and synthesized four kinds of MEL analogues R-MEL−A ([2R,3S]-erythritol type), S-mannosylthreitol lipid (MTL)-A ([2S,3S]-threitol type), R-MTL−A ([2R,3R]-threitol type), and α-S-MEL−A ([2S,3R]-erythritol type) using our previously reported boron-mediated aglycon delivery (BMAD) method and a neighboring group assisted glycosylation method. The selective cytotoxicity of the target compounds against cancer cells was evaluated, with R-MTL−A showing the highest selective cytotoxicity against human skin squamous carcinoma HSC-5 cells. Our findings suggest that R-MTL−A induces necrosis-like cell death against HSC-5 cells by decreasing cell membrane fluidity. R-MTL−A also exhibits an efficient recovery effect on damaged skin cells, indicating that R-MTL−A has potential as a lead compound for new cosmeceuticals with both cancer cell-selective toxicity and recovery effects on damaged skin cells.

1 Introduction

Mannosylerythritol lipids (MELs) are glycolipid biosurfactants that are mainly produced by Pseudozyma sp.¹ Structurally, MELs are amphipathic molecules because they possess a hydrophilic and a hydrophobic moiety. The hydrophilic 4-O-β-D-mannopyranosyl-d-erythritol moiety contains zero, one, or two acetyl groups at the C4’ and C6’ positions of the mannose moiety. MELs are thus classified as MEL−A (4’,6’-di-O-Ac type), MEL−B (6’-O-Ac type), MEL−C (4’-O-Ac type), and MEL−D (non-acetylated type). MELs are also categorized as S-MEL ([2S,3R]-erythritol type) and R-MEL ([2R,3S]-erythritol type) according to the stereochemistry of the hydroxyl groups at the erythritol moiety.² The hydrophobic moiety of MELs contains a mixture of C₆−C₁₈ fatty acyl chains at the C2’ and C3’ positions of the mannose moiety.

These amphiphilic structural features give MELs unique physicochemical properties, such as the ability to self-assemble,³ providing the foundation for their biological activities. MELs have attracted much attention in recent years because of their outstanding properties in cosmetics,⁴ including surface activity,⁵ moisturizing effect,⁶ and recovery effects on damaged skin⁷ and damaged hair,⁵ and also because of their cytotoxicity against cancer cells. For example, S-MEL−A (C₈−C₁₄) isolated from natural products shows cytotoxicity against mouse melanoma B16 cells,^{8, 9} human hepatoma HepG2 cells,¹⁰ and human leukemia K562, HL60, and KU812 cells.¹¹ B16 cell death was due to MEL causing apoptosis by the condensation of chromatin, DNA fragmentation, and sub-G (1) arrest.⁸ A subsequent study independently confirmed these mechanisms and showed that B16 cell apoptosis is associated with endoplasmic reticulum stress.⁹ However, no detailed structure-activity relationship (SAR) studies on the cytotoxicity of structurally well-defined and pure MELs against cancer cells have been reported to date. Our group has described the total synthesis of 20 S-MELs 1--20 with different fatty acid chain lengths (n=5, 7, 9, 11, 13) and different patterns of Ac groups at the C4’ and C6’ positions of the mannose moiety (S-MEL−A, B, C, D) (Figure 1), and we have conducted SAR studies on the antibacterial activity,^{12, 13} self-assembling properties, and recovery effects on damaged skin cells using these chemically synthesized homogeneous MELs.^{7, 13} These results indicated for the first time that slight differences in fatty acid chain length and/or the patterns of Ac groups significantly affect the functional activity of MELs. Specifically, S-MEL−D (n=9) 18 exhibited anti-bacterial activity against a vancomycin-resistant enterococci (VRE) strain whereas the other 19 S-MELs did not. Furthermore, S-MEL−D (n=7) 17 has low skin penetration and shows no recovery effect on damaged skin cells, whereas S-MEL−A-D (n=9) 3, 8, 13, and 18, which have slightly longer fatty chain lengths, have high skin permeability and show remarkable recovery effects on damaged skin cells.^{7, 13} We therefore hypothesized that 3, 8, 13, 18, and their analogues might exhibit both high recovery effects on damaged skin cells and high selective cytotoxicity against skin cancer cells. Compounds demonstrating these two properties are currently attracting attention as next-generation cosmeceuticals.^{14, 15} Here, we evaluated the selective cytotoxicity of 3, 8, 13, and 18 against human skin cancer cells, then we designed, synthesized, and functionally evaluated new synthetic analogues based on the structure of MEL. These analogues showed higher cancer cell-selective toxicity. This led us to create candidate molecules for cosmeceuticals exhibiting both recovery effects on damaged skin cells and cancer cell-selective toxicity.

Figure 1

Chemical structures of S-MELs 1–20 and the objective of this study. Ac=acetyl.

2 Results and Discussion

We chose the human skin squamous carcinoma HSC-5 cell line and the human pulmonary fibroblasts WI-38 cell line as representative cancer and normal cells, respectively, for this study and evaluated the selective cytotoxicity of 3, 8, 13, and 18 using the MTT assay. Each MEL was dissolved in DMSO at 10 mM in a vial as a stock solution and then diluted with culture medium to 100 and 300 μM before being added to 96-well plates.

The results are shown in Figure 2. None of the MELs showed cytotoxicity against normal WI-38 cells (Figure 2a) whereas 3, 8, 13, and 18, which have different patterns of Ac groups at the C4’ and C6’ positions of the mannose moiety, exhibited cytotoxicity against HSC-5 cells at 10 and 30 μM (Figure 2b). S-MEL−A (n=9) 3 exhibited the greatest activity of the MELs tested, indicating that 3 with two Ac groups and two C₁₀ fatty acid chains shows the best selective cytotoxicity against HSC-5 cancer cells. We clarified the SAR of 3 and created compounds with higher activity by designing and synthesizing the analogues R-MEL−A (n=9) 21, S-mannosylthreitol lipid (MTL)-A (n=9) 22 ([2S,3S]-threitol type), R-MTL−A (n=9) 23 ([2R,3R]-threitol type), and α-S-MEL−A (n=9) 24 which has an α-glycosidic linkage (Figure 3), and then evaluated their cytotoxicity against HSC-5 and WI-38 cells.

Figure 2

Effect of 3, 8, 13, and 18 on (a) WI-38 and (b) HSC-5 cell proliferation. Cells were seeded into 96-well plates and incubated at 37 °C in 5 % CO₂ in air for 24 h. After 24 h, the indicated MEL was added to a well, and cells were incubated at 37 °C in 5 % CO₂ in air for 48 h. Cell viability was evaluated using the MTT assay. *p<0.05, **p<0.01.

Figure 3

Chemical structures of R-MEL−A (n=9) 21, S-MTL−A (n=9) 22, R-MTL−A (n=9) 23, and α-S-MEL−A (n=9) 24.

The synthetic schemes of 21–24 are shown in Schemes 1, 2, 3, and 4, respectively, with 21 being synthesized first. Glycosylation of known 1,2-anhydro donor 25 (1.2 equiv.)¹⁶ and racemic alcohol 26 (1.0 equiv.) was conducted using a catalytic amount of borinic acid 27 in MeCN at 0 °C for 1.5 h by the boron-mediated aglycon delivery (BMAD) method developed in our laboratory,^{16, 17} to provide 28 and 29 in 45 % and 43 % yields, respectively, with excellent β-stereoselectivity. Their anomeric configurations were determined by the value of the ¹J_CH (159 Hz for 28 and 155 Hz for 29) coupling constant.¹⁸ Ceric ammonium nitrate (CAN) was used to deprotect the PMB group of the synthesized 28 to give 30. Importing the C₁₀ fatty chain by decanoyl chloride, followed by debenzylation using Pd(OH)₂/C, afforded 31 in 73 % yield. Finally, the synthesis of compound 21 was achieved by acylation of diol 31 with acetic anhydride in pyridine, followed by the deprotection of the TBDPS by treatment of 31 with HF⋅Py and cleavage of the acetonide group by treatment of the resulting alcohol with aqueous 1 M HCl (Scheme 1). Next, 22 and 23 were synthesized as shown in Schemes 2 and 3. After preparation of 33 and 38 from 32 and 37, respectively, BMAD reactions of 33 and 38 with 25 were conducted in MeCN at 0 °C for 1.5 h, affording 34 (¹J_CH=156 Hz) in 75 % yield and 39 (¹J_CH=161 Hz) in 95 % yield.

Scheme 1

Synthesis of R-MEL−A (n=9) 21. Bn=benzyl, PMB=p-methoxybenzyl, TBDPS=tert-butyldiphenylsilyl, CAN=cerium ammonium nitrate, DMAP=4-dimethylaminopyridine, THF=tetrahydrofuran, Py=pyridine.

Scheme 2

Synthesis of S-MTL−A (n=9) 22.

Scheme 3

Synthesis of R-MTL−A (n=9) 23.

Scheme 4

Synthesis of α-S-MEL−A (n=9) 24.

Finally, 22 and 23 were obtained in 5 steps by removing the PMB and Bn groups, followed by acylation, desilylation, and hydrolysis reactions. Compound 24 was synthesized from known compound 43¹³ in 2 steps.

The cytotoxicity of chemically synthesized S-MEL−A (n=9) 3 analogues 21--24 against WI-38 and HSC-5 cells was examined. The results are shown in Figure 4. Comparison of the cytotoxicity of these compounds showed that 3 and 21--23, which have a β-mannosidic linkage, exhibited moderate to highly selective cytotoxicity against HSC-5 cancer cells whereas 24, having an α-mannosidic linkage, showed low selective cytotoxicity, suggesting the importance of the β-mannosidic linkage for selective cytotoxicity against HSC-5 cancer cells. We discovered that the new synthetic R-MTL- A (n=9) 23, which has a [2R,3R]-threitol moiety, exhibits the highest selective cytotoxicity against HSC-5 cells of the compounds tested, higher than that of natural-type S-MEL−A (n=9) 3. These results indicated that slight differences in the steric configuration of the erythritol moiety affect the selective cytotoxicity against HSC-5 human skin cancer cells.

Figure 4

Effects of 3 and its analogues 21–24 on (a) WI-38 and (b) HSC-5 cell proliferation. Cells were seeded into 96-well plates (1.0×10⁴ cells/well in 10 % FBS IMDM or MEM) and incubated at 37 °C in 5 % CO₂ in air for 24 h. After 24 h, the cells were treated with individual compounds (0, 10, and 30 μM) and incubated at 37 °C in 5 % CO₂ in air for 48 h. Cell viability was evaluated using the MTT assay. *p<0.05, **p<0.01.

We next conducted a mechanistic analysis of cancer cell death induced by 3 and 23. We initially hypothesized that 3 and 23 induced apoptosis in HSC-5 cells, based on a report that natural MEL causes the death of mouse melanoma B16 4 A5 and B16 cells by apoptosis.^{8, 9} Apoptosis is generally considered as caspase-mediated programmed cell death,¹⁹ and is accompanied by morphological changes such as cell shrinkage and DNA condensation.²⁰ We therefore used Z-VAD-FMK as a caspase inhibitor and the caspase-dependent apoptosis-inducing agent actinomycin D²¹ was used as a positive control. We found that co-treatment with actinomycin D (AcD) and Z-VAD-FMK significantly increased cell viability compared to when Z-VAD-FMK was not added, indicating the role of caspase-dependent apoptosis. On the other hand, with compounds 3 and 23, Z-VAD-FMK provided no increase in cell viability, suggesting that caspase-dependent apoptosis was not induced (Figure 5a). We next conducted flow cytometry using annexin V-FITC and propidium iodide (PI)²² and found that treating the cells with 3 or 23 for 48 h resulted in more annexin V-negative/PI-positive cells (necrotic) and fewer annexin V-positive/PI-negative cells (apoptotic), as shown in Figures 5b–d. Although the detailed mechanisms of cell death remain to be elucidated, these results suggest that 3 and 23 induce necrosis in a cancer cell-selective manner.

Figure 5

A mechanistic analysis of HSC-5 cell death caused by 3 and 23. (a) Cells were seeded into 96 well plates (1.0×10⁴ cells/well in 10 % FBS IMDM) and incubated at 37 °C in 5 % CO₂ in air for 24 h. Compounds were added individually to the wells, and cells were incubated at 37 °C in 5 % CO₂ in air for 48 h. Cell viability was evaluated using the MTT assay. **p<0.01. (b–d) Cells were seeded into small flasks (6.5×10⁵ cells/flask). After 24 h, 3 or 23 was added to each flask and the cells were incubated for 48 h at 37 °C in 5 % CO₂ in air. The cells were double-stained with annexin V-FITC and PI and analyzed for apoptosis by flow cytometry. The data show the percentage of PI-positive/FITC-negative cells (necrotic), PI-negative/FITC-positive cells (apoptotic), and PI-negative/FITC-negative cells (viable).

We next focused on the mechanism of selective cytotoxicity against cancer cells. Regardless of the cancer cell type, the membranes of cancer cells are typically softer than those of their normal counterparts,²³ suggesting that a reduction in membrane fluidity would be more detrimental to cancer cells than to normal cells.²⁴ Several studies show that cell membrane rigidity is induced by compounds such as flavones,²⁵ quercetin, epigallocatechin, and apigenin, resulting in selective cytotoxicity against cancer cells^24-26 and suggesting that increased membrane rigidity is a key factor for selective cytotoxicity against cancer cells. We investigated whether the selective cytotoxicity of 3 and 23 is related to cell membrane fluidity by using DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine)²⁷ liposomes as a cell membrane model and Nile red as a fluorescent probe, and evaluated the change in membrane fluidity upon the addition of 3 or 23. As shown in Figures 6a and b, the treatment of DOPC liposomes with 3 or 23 shifted the maximum emission wavelength of Nile red toward lower wavelengths in a concentration-dependent manner, indicating that 3 and 23 decreased membrane fluidity.²⁸ Based on these preliminary results, we next investigated the change in membrane fluidity upon the addition of 3 or 23 to HSC-5 and WI-38 cells using a commercial membrane fluidity kit.^{29, 30} Treating WI-38 cells with 3 or 23 for 24 h and 36 h had no effect on cell membrane fluidity, as shown in Figure 7a. In sharp contrast, the treatment of HSC-5 cells with 3 or 23 decreased cell membrane fluidity. Incubation time dependencies were clearly observed between 24 and 36 h. Furthermore, consistent with the trend in cytotoxicity to HSC-5 cells, 23 induced a slightly higher change than did 3 (Figure 7b), suggesting that a change in cell membrane fluidity is important for the selective cytotoxicity of MEL and its analogue against cancer cells.³¹

Figure 6

Fluorescence spectra of DOPC-dispersions containing Nile red and different concentrations (0, 0.3 1, 3 mM) of (a) 3 or (b) 23. The maximum wavelength of the dispersions was measured using a fluorescence spectrophotometer (λ_ex=553 nm, λ_em=637 nm).

Figure 7

Effects of 3 and 23 on (a) WI-38 and (b) HSC-5 cells. Cells were seeded into 96-well plates (1.0×10⁴ cells/well in 10 % FBS DMEM and MEM) and incubated at 37 °C in 5 % CO₂ in air for 24 h. The cells were then treated with 3 or 23 (10 and 30 μM) for 24 and 36 h. Membrane fluidity was measured using a membrane fluidity kit.

We evaluated the recovery effect of 23 on damaged skin cells using 3D-cultured human epidermis models. The results are summarized in Figure 8. Damaging the cells with 1 wt % sodium dodecyl sulfate (SDS) decreased cell viability to 10 %. In contrast, treatment of the damaged skin cells with 6 mM 3 or 23 increased cell viability to 50 % and 75 %, respectively. These results indicate that 23 is a promising candidate for cosmeceuticals exhibiting both a recovery effect on damaged skin cells and cancer cell-selective toxicity.

Figure 8

The recovery effect of 3 and 23 on damaged skin cells. Cultured skin cells were treated with 1 % SDS, the SDS solution was removed, and then the cells were treated with 3 or 23 dissolved in olive oil (6 mM). The viability of the cells was determined with an MTT assay at 570 nm. Vertical bars show the standard error of the mean based on three independent measurements. **p<0.01.

3 Conclusions

In summary, we evaluated the selective cytotoxicity of S-MELs 3, 8, 13, and 18 using normal WI-38 cells and human skin cancer HSC-5 cells and found that 3 exhibited the greatest selective cytotoxicity against HSC-5 cells. We then designed and synthesized four new analogues, R-MEL−A 21, S-MTL−A 22, R-MTL−A 23, and α-S-MEL−A 24 to conduct a detailed SAR study and found that 23 exhibits better selective cytotoxicity against HSC-5 cells than does 3. This demonstrated for the first time that the anomeric configuration and slight differences in the steric configuration of the erythritol moiety in MEL can affect the selective cytotoxicity of the compound. A mechanistic analysis of cancer cell death induced by 3 and 23 suggested that 3 and 23 induce necrosis-like cell death in a cancer cell-selective manner, and that 3 and 23 selectively decrease cancer cell membrane fluidity, with 23 exhibiting slightly higher activity than 3, consistent with the cytotoxicity of the compounds towards HSC-5 cells. These results suggested that the change in cell membrane fluidity is important for the selective cytotoxicity of MEL and its analogues. Finally, we evaluated the recovery effect of 3 and 23 on damaged skin cells. Synthetic analogue 23 exhibited higher activity than natural type MEL 3. These results indicate that 23 is a promising candidate for cosmeceuticals, with both cancer cell-selective toxicity and recovery effects on damaged skin cells. Further investigations are in progress in our laboratories.

Experimental Section

General Methods for Chemical Synthesis

All chemicals and reagents were commercially available with no further purification unless otherwise stated. NMR spectra were recorded on a JEOL ECA-500 (500 MHz for ¹H, 125 MHz for ¹³C) spectrometer or a JEOL ECZ-400S (400 MHz for ¹H, 100 MHz for ¹³C) spectrometer. ¹H-NMR data are reported as follows; chemical shift in parts per million (ppm) downfield or upfield from CDCl₃ (δ 7.26) integration, multiplicity (br=broad, s=singlet, d=doublet, t=triplet, q=quartet, and m=multiplet), and coupling constants (Hz). ¹³C-NMR chemical shifts are reported in ppm downfield or upfield from CDCl₃ (δ 77.0). ESI-TOF MS spectra were measured on a Waters LCT premier XE. Melting points were determined on a micro hot-stage (Yanako MP−S3) and were uncorrected. Optical rotations were measured on a JASCO P-2200 polarimeter. Silica gel TLC was performed on a Merck TLC 60F-254 (0.25 mm) or a Merck PLC 60F-254 (0.5 mm). Column chromatography separation was performed on a Silica Gel 60 N (spherical, neutral, 63–210 μm or 40–50 μm) (Kanto Chemical Co., Inc.). Air- and/or moisture-sensitive reactions were carried out under an argon atmosphere using oven-dried glassware.

Synthesis

The synthesis of compounds R-MEL−A 21, S-MTL−A 22, R-MTL−A 23, and α-S-MEL−A 24 is provided in the Supporting Information.

Materials and Experimental Methods for Biochemical Assays

Cell Culture

The HSC-5 and WI-38 cell lines were routinely grown in Iscove's modified Dulbecco's medium (IMDM) or Dulbec-co's modified Eagle's medium (DMEM) supplemented with 10 % (v/v) Fetal bovine serum, 0.5 % (v/v) penicillin and kanamycin. The cells were maintained at 37 °C in a humidified atmosphere containing 5 % CO₂.

Cell Viability Assay

HSC-5 or WI-38 cells were seeded into 96-well plates (90.0 μL, 3.0×10³ cells). After 24 h, cells were treated with 10.0 μL of compounds in 10 % DMSO-medium and incubated for 72 h at 37 °C and in 5 % CO₂ in air. Cell viability was evaluated using the MTT assay. 10.0 μL of 5.00 mg/mL MTT dissolved in PBS was added to each well. After incubation for 3 h at 37 °C, the medium was aspirated and 100 μL of DMSO was added to each well, and color intensity was measured using SpectraMax i3 (Molecular Devices) microplate reader at 540 nm.

Flow Cytometry Analysis of Cell Death by Annexin V/PI Double Staining

HSC-5 cells were seeded into small flasks (6.5×10⁵ cells). After 24 h, cells were treated with compounds 3 and 23 (final concertation is 30 μM) and incubated for 24 h. After the incubation time, the medium was collected and then adherent cells were harvested by treatment with 0.2 % trypsin-EDTA and centrifuged for 5 min at 3000 rpm at 4 °C. The cell pellet was resuspended in 1 mL PBS and stained with 10 μL of Annexin V-FITC and Propidium iodide for 15 min. Imaging was done using flow cytometer BD Accuri^TM C6.

DOPC Model Membrane Fluidity

To a solution of DOPC in chloroform (9.6 mM) were added a solution of Nile red (30 μL, 100 μM) in chloroform and a solution of S-MEL−A (n=9) 3 or R-MTL−A (n=9) 23 in chloroform (100 μM, 0, 1.2, 4, 12 mM) in 5 mL glass vial. The chloroform solvent was removed by gently passing dry nitrogen gas. The traces of the solvent were then moved by keeping the sample in a vacuum desiccator for 2 h. 400 μL Milli Q water was added to the dried lipid film until the final concentration of 3 or 23 is 0.3, 1, 3 mM, stood for 2 h, then vortexed for 1 min to produce multilamellar vesicles (MLV). The maximum wavelength of MLV suspensions was measured using JASCO Spectrofluorometer FP-8500 (excitation wavelength: 553 nm, fluorescence wavelength: 637 nm).

Cell Membrane Fluidity Assay

Cell membrane fluidity was measured using a membrane fluidity kit (Abcam, UK) according to the manufacturer's instructions. HSC-5 and WI-38 cells were seeded into clean bottom 96-well black microtiter plates (1.0×10⁴ cells/well in 10 % FBS DMEM and MEM) and incubated at 37 °C in 5 % CO₂ in air for 24 h. After 24 h, S-MEL−A (n=9) 3 and R-MTL−A (n=9) 23 were added to each well for 24 h and 36 h, then regents were added to each well for 1 h. After incubation, the cells were analyzed using excitation at 360 nm and recording the fluorescence emission at both 400 and 470 nm. Relative membrane fluidity was calculated by the ratio of intensity of excimer to monomer fluorescence (470/400 nm).

Recovery Effects on Damaged Skin Assay

The potential effects of MELs on the skin were evaluated using a three-dimensional cultured human skin model (LabCyte EPI-MODEL24, J-TEC, Japan), where the readout was cell viability.⁷ An aqueous solution of 1 wt % sodium dodecyl sulfate (SDS) was applied to the cell surface and incubated for 5 min at 37 °C to induce damage similar to dry skin conditions. After washing 10 times with PBS (0.5 mL), olive oil fraction (50 μL) containing 6 mM of each compound was directly applied, and then the cells were incubated for 24 h at 37 °C. Then, the surface of the cells was washed with PBS. The viability of the cells was determined using the MTT assay. 0.5 mL of 0.5 mg/mL MTT dissolved in culture medium (J-TEC, Japan) was added to each well. After incubation for 3 h at 37 °C, cells were removed from the well. The absorbance of the mixture was measured using a microplate reader at 570 nm. The cells cultured without the SDS treatment were used as a control.

Acknowledgments

This research was supported in part by JSPS KAKENHI Grant Number JP23H01966 in Scientific Research (B), JST SPRING Grant Number JPMJSP2123, JST CREST Grant Number JPMJCR20R3, AMED-CREST Grant Number JP22gm1610010, and the KOSE Cosmetology Research Foundation.

Conflict of Interests

The authors declare no conflict of interest.

Open Research

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.