Phytoceramides from the Marine Sponge Monanchora clathrata: Structural Analysis and Cytoprotective Effects

In our research on sphingolipids from marine invertebrates, a mixture of phytoceramides was isolated from the sponge Monanchora clathrata (Western Australia). Total ceramide, ceramide molecular species (obtained by RP-HPLC, high-performance liquid chromatography on reversed-phase column) and their sphingoid/fatty acid components were analyzed by NMR (nuclear magnetic resonance) spectroscopy and mass spectrometry. Sixteen new (1b, 3a, 3c, 3d, 3f, 3g, 5c, 5d, 5f, 5g, 6b–g) and twelve known (2b, 2e, 2f, 3b, 3e, 4a–c, 4e, 4f, 5b, 5e) compounds were shown to contain phytosphingosine-type backbones i-t17:0 (1), n-t17:0 (2), i-t18:0 (3), n-t18:0 (4), i-t19:0 (5), or ai-t19:0 (6), N-acylated with saturated (2R)-2-hydroxy C21 (a), C22 (b), C23 (c), i-C23 (d), C24 (e), C25 (f), or C26 (g) acids. The used combination of the instrumental and chemical methods permitted the more detailed investigation of the sponge ceramides than previously reported. It was found that the cytotoxic effect of crambescidin 359 (alkaloid from M. clathrata) and cisplatin decreased after pre-incubation of MDA-MB-231 and HL-60 cells with the investigated phytoceramides. In an in vitro paraquat model of Parkinson’s disease, the phytoceramides decreased the neurodegenerative effect and ROS (reactive oxygen species) formation induced by paraquat in neuroblastoma cells. In general, the preliminary treatment (for 24 or 48 h) of the cells with the phytoceramides of M. clathrata was necessary for their cytoprotective functions, otherwise the additive damaging effect of these sphingolipids and cytotoxic compounds (crambescidin 359, cisplatin or paraquat) was observed.


Introduction
Ceramide is a complex lipid, which consists of a sphingoid base attached to a fatty acid via an amide bond. Ceramides are present in membranes in small amounts only, but they serve as central mediators, regulating many fundamental cellular responses [1][2][3]. These sphingolipids trigger a number of tumor suppressive and anti-proliferative cellular programs, for example, apoptosis, autophagy, senescence, and necroptosis [4,5]. Ceramidedependent effects may hold implications for the progression of many diseases including cancer, diabetes, atherosclerosis, Alzheimer's and Parkinson's diseases [1].
Ceramides vary appreciably in the compositions of both long-chain alkyl (sphingoid and fatty acid) components, depending on their biological origins. In particular, phytoceramides, consisting of phytosphingosine-type backbones N-acylated with 2-hydroxy fatty acids, are common for marine sponges [6]. In our research on sphingolipids from marine invertebrates, phytoceramides, mainly containing iso-methyl-branched chains, were found in the sponge Monanchora clathrata, collected in Australian waters.
In the previous study of the phytoceramides of M. clathrata, these compounds were shown to be cytotoxic to MES-SA, MCF-7, and HK-2 cell lines [7]. The results from other studies suggest that phytoceramides may help to prevent neurodegeneration in vitro and in vivo [9,10]. The neuroprotective effect of phytoceramides is consistent with a possible therapeutic role of these compounds in managing cognitive impairment, associated, in the first turn, with Alzheimer's disease. We aimed to investigate mainly the cytoprotective effects of the phytoceramides from M. clathrata, in particular their effect on paraquat-induced neurotoxicity using in vitro model of Parkinson's disease. The neuroprotective activity of phytoceramides in Parkinson's disease has not been studied, although this pathology is the second most common neurodegenerative disorder behind Alzheimer's disease [11]. Alkaloid crambescidin 359 (7), previously isolated from the sample of M. clathrata analyzed here [12], was also used in our bioassays. The cytotoxic and/or cytoprotective effects of total ceramide, crambescidin 359 and their combinations were tested on MDA-MB-231 (breast adenocarcinoma) and HL-60 (leukemia) cells. The cytotoxic effect of cisplatin in combinations with the phytoceramides of M. clathrata was also tested on these cells. Cisplatin is employed for the treatment of various tumors, and there is an urgent need for therapeutic agents reducing cisplatin-induced toxicity. Based on the results of the present study, it is concluded that, depending on the time of cell pre-treatment by phytoceramides, these sphingolipids can increase or decrease the cytotoxicity of crambescidin 359, cisplatin or paraquat.

General Procedures
1 H-, 13 C-NMR, 1 H, 1 H-COSY and HSQC spectra (in C 5 D 5 N or CDCl 3 ) were recorded on Bruker Avance III HD 500 and Bruker Avance III 700 spectrometers (Bruker BioSpin, Bremen, Germany) at 125 MHz ( 13 C), 500 ( 1 H), and 700 ( 1 H) MHz. A Bruker Impact II Q-TOF mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany) equipped with ESI ionization source was employed to record MS and MS/MS spectra. The operating parameters for ESI-MS were the following: a capillary voltage of 4.0 kV, nebulization with nitrogen at 0.4 bar, and a dry gas flow of 4 L/min at a temperature of 200 • C. GC analyses were performed on an Agilent 6850 Series GC System chromatograph (Agilent Technologies, Santa Clara, CA, USA) equipped with an DB-1 (J&W Scientific, Folsom, CA, USA) capillary column (30 m × 0.32 mm), the carrier gas was helium (flow rate 1.7 mL/min), and the detector temperature was 300 • C. GC-MS analyses were carried out on a Hewlett-Packard HP6890 GC System (Hewlett-Packard Company, Palo Alto, CA, USA) with an HP-5MS (J&W Scientific, Folsom, CA, USA) capillary column (30.0 m × 0.25 mm), helium as the carrier gas, and 70 eV ionizing potential. The GC and GC-MS analyses of fatty acid esters and peracetylated sphingoid bases were performed using the injector temperature of 270 • C and the temperature program 100 • C (1 min) − 10 • C/min − 280 • C (30 min). Optical rotation was measured on a Perkin-Elmer polarimeter, model 343 (Perkin-Elmer GmbH, Überlingen, Germany). Column chromatography was performed using silica gel (50/100 µm, Sorbpolimer, Krasnodar, Russia). RP-HPLC separations were performed using a Du Pont Series 8800 Instrument (DuPont, Wilmington, DE, USA) with a RIDK-102 refractometer (Laboratorni pristroje, Praha, Czechoslovakia). An Agilent ZORBAX Eclipse XDB-C8 column (4 × 150 mm; Agilent Technologies, Santa Clara, CA, USA) was used for the HPLC.
The acid hydrolysis of ceramide molecular species (Fractions I-VII), obtained by RP-HPLC, was carried out under the same conditions as for total ceramide. Water (0.2 mL) was added to hydrolysate, and MeCN-H 2 O layer was extracted with hexane (5 × 0.5 mL). Hexane extract was dried in vacuo, acetylated with Ac 2 O in pyridine (1:1, v/v, 0.2 mL, overnight) and ethylated with N-nitroso-N-ethylurea. The resulting tetraacetates of sphingoid bases and ethyl esters of 2-acetyloxy acids were analyzed by GC-MS. The ethyl esters dominated in the mixture, and their signals and some signals of tetraacetylated sphingoid bases overlapped in GC profiles. To obtain a pure subfraction of the acetates of sphingoid bases, the mixture was separated by chromatography on silica gel (column: 3.0 cm × 1.5 cm) using hexane/ethylacetate (5:1 → 2:1, v/v). Elution with hexane/ethylacetate, 2:1 (60 mL) gave tetraacetates of sphingoid bases that were again analyzed by GC-MS.

Cell Viability Assay for MDA-MB-231 and HL-60 Cells
The effect of compounds on cell viability was evaluated using reduction of MTS into formazan product. The cells were cultured in 96-well plates (5000 cells/well) with the corresponding medium (100 µL/well, containing 10% FBS) for 12 h. The cells were treated with compounds at various concentrations (0-200 µM in DMSO) for 24 or 48 h. Then, MTS reagent (20 µL) was added into each well, and, after 4 h, MTS reduction was measured spectrophotometrically at 492 and 690 nm as background, using a Power Wave XS microplate reader (BioTek, Winooski, VT, USA). Cisplatin was used as a positive control.

Paraquat Induced In Vitro Model of Neurotoxicity
Crambescidin 359 and ceramide were dissolved in DMSO to obtain stock solutions (10 mM concentrations). Cells SH-SY5Y and Neuro-2a (1 × 10 4 cells/well) were supplemented with the test compounds (20 µL in PBS) at final concentrations of 0.1, 1.0, or 10.0 µM, and preincubated for 1, 24 or 48 h. Then, the cells were treated with 1.5 mM paraquat. The cells incubated without paraquat were used as a positive control, and the cells incubated with this inducer were used as a negative control. Cell viability was measured after 24 h. For this, the medium with tested substances was changed by 100 µL of fresh culture medium containing 10 µL of MTT solution (5 mg/mL). After that, the microplate was incubated for an additional 4 h. Then, 100 µL of SDS-HCl solution (1 g SDS/10 mL dH 2 O/17 µL 6N HCl) was added and incubated for 18 h. Dye absorbance was measured using a plate format spectrophotometer at a wavelength of 570 nm (Thermo Scientific, Waltham, MA, USA). All the experiments were carried out three times. The toxic activity of paraquat was expressed as a percentage of control (untreated) cells.

Analysis of ROS (Reactive Oxygen Species) Level
Neuro-2a cells were transferred to a microplate for adhesion for 24 h, then incubated with ceramide or crambescidin 359 at various concentrations for 1h. To increase ROS in cells, paraquat was added at a concentration of 1mM per 3h. In order to study ROS formation, 20 µL of H2DCFDA solution (100 µM concentration) were added to each well (to a final concentration of 10.0 µM), and the microplate with cells was incubated for an additional 10 min at 37 • C. The intensity of fluorescence was measured using PHERAstar FS plate reader (BMG Labtech, Ortenberg, Germany) at λex = 485 nm and λem = 520 nm.

Statistical Analysis
All experiments were carried out in three or more independent experiments. Plot data were presented as mean ± standard deviation (SD). Student's t-test was performed using SigmaPlot 14.0 (Systat Software Inc., San Jose, CA, USA) to determine statistical significance.

Analysis of the 2-Hydroxy Fatty Acids and Sphingoid Bases Obtained from the Total Ceramide of M. clathrata
We applied methanolysis and hydrolysis for the chemical degradation of total ceramide. Methanolysis procedure was used to prepare fatty acid methyl esters. Hydrolysis was mainly used to release sphingoid bases because sphingoid bases could be harmed by vigorous conditions of methanolysis [16].

Structure Elucidation of the Phytoceramides of M. clathrata
The total ceramide of M. clathrata was separated into seven molecular species by RP-HPLC (Fractions I-VII, Table 5).

Cytoprotective and/or Cytotoxic Effects of the Phytoceramides and Crambescidin 359 from M. clathrata
The cytotoxic activity of the total ceramide of M. clathrata against MDA-MB-231 and HL-60 cells was extremely weak (IC50 values ≥ 200 µM, incubation for 48 h; Figure S11). Crambescidin 359 (7)    Combinations of cisplatin and total ceramide were slightly more cytotoxic for MDA-MB-231 cells (Figure 5a) than only cisplatin. However, these combinations insignificantly increased the survival of HL-60 cells (Figure 5b). After 24 h pre-incubation of MDA-MB-231 and HL-60 cells with total ceramide (especially, at a concentration of 50 µM), the cytotoxic effect of cisplatin was noticeably reduced (Figure 5c,d).  Combinations of cisplatin and total ceramide were slightly more cytotoxic for MDA-MB-231 cells (Figure 5a) than only cisplatin. However, these combinations insignificantly increased the survival of HL-60 cells (Figure 5b). After 24 h pre-incubation of MDA-MB-231 and HL-60 cells with total ceramide (especially, at a concentration of 50 µM), the cytotoxic effect of cisplatin was noticeably reduced (Figure 5c,d).
Combinations of cisplatin and total ceramide were slightly more cytotoxic for MDA-MB-231 cells (Figure 5a) than only cisplatin. However, these combinations insignificantly increased the survival of HL-60 cells (Figure 5b). After 24 h pre-incubation of MDA-MB-231 and HL-60 cells with total ceramide (especially, at a concentration of 50 µM), the cytotoxic effect of cisplatin was noticeably reduced (Figure 5c,d). Total ceramide at the concentrations of 1 and 10 µM significantly reduced ROS formation (by 29.8 ± 6.7% and 32.5 ± 3.2%, respectively) in neuroblastoma cells exposed to paraquat ( Figure 6). In this experiment, phytoceramides in non-cytotoxic concentrations were incubated with the cells for 1 h before paraquat was introduced into the culture. After analogous short-time pre-treatment (1 h) of the neuroblastoma cells with total ceramide, the neurotoxicity of paraquat unexpectedly increased (Figure 7). In contrast, after a longer pre-treatment (24 or 48 h) of the cells with phytoceramides, the neurotoxic activity of paraquat decreased (Figure 8). For example, total ceramide at the concentrations of 0.1 and 1.0 µM increased the viability of Neuro 2a cells by 28.7 ± 6.3 and 23.1 ± 1.2%, respectively, compared to the control cells exposed to only paraquat (Figure 8a). Furthermore, the pre-treatment (for 48 h) of SH-SY5Y cells with total ceramide at the concentration of 10 µM prevented cell death induced by paraquat (Figure 8b). Total ceramide at the concentrations of 1 and 10 µM significantly reduced ROS formation (by 29.8 ± 6.7% and 32.5 ± 3.2%, respectively) in neuroblastoma cells exposed to paraquat ( Figure 6). In this experiment, phytoceramides in non-cytotoxic concentrations were incubated with the cells for 1 h before paraquat was introduced into the culture. After analogous short-time pre-treatment (1 h) of the neuroblastoma cells with total ceramide, the neurotoxicity of paraquat unexpectedly increased (Figure 7). In contrast, after a longer pre-treatment (24 or 48 h) of the cells with phytoceramides, the neurotoxic activity of paraquat decreased (Figure 8). For example, total ceramide at the concentrations of 0.1 and 1.0 µM increased the viability of Neuro 2a cells by 28.7 ± 6.3 and 23.1 ± 1.2%, respectively, compared to the control cells exposed to only paraquat (Figure 8a). Furthermore, the pre-treatment (for 48 h) of SH-SY5Y cells with total ceramide at the concentration of 10 µM prevented cell death induced by paraquat (Figure 8b).
Crambescidin 359 reduced ROS formation (maximum 17.9 ± 4.6%, 10 µM concentration) in Neuro 2a cells exposed to paraquat ( Figure 5) and exhibited neuroprotective effects after 1, 24, and 48 h pre-incubation with these cells ( Figure S13). However, we did not consider crambescidin 359 a promising antiparkinsonic agent because this alkaloid showed inhibitory activities in electrophysiology experiments on human α7 nicotinic acetylcholine receptors [23]. The stimulation (not inhibition) of these receptors, promoting cognitive functions, may be a perspective strategy for the treatment of Parkinson's disease [24].
After analogous short-time pre-treatment (1 h) of the neuroblastoma cells with total ceramide, the neurotoxicity of paraquat unexpectedly increased (Figure 7). In contrast, after a longer pre-treatment (24 or 48 h) of the cells with phytoceramides, the neurotoxic activity of paraquat decreased (Figure 8). For example, total ceramide at the concentrations of 0.1 and 1.0 µM increased the viability of Neuro 2a cells by 28.7 ± 6.3 and 23.1 ± 1.2%, respectively, compared to the control cells exposed to only paraquat (Figure 8a). Furthermore, the pre-treatment (for 48 h) of SH-SY5Y cells with total ceramide at the concentration of 10 µM prevented cell death induced by paraquat (Figure 8b). Figure 6. Effects of total ceramide and crambescidin 359 on ROS formation in Neuro-2a cells. The data are presented as m ± se (n = 3). * p < 0.05 compared to cells exposed to paraquat alone. . The data are presented as m ± se (n = 3). * p < 0.05 compared to cells exposed to paraquat alone.
Crambescidin 359 reduced ROS formation (maximum 17.9 ± 4.6%, 10 µM concentration) in Neuro 2a cells exposed to paraquat ( Figure 5) and exhibited neuroprotective effects after 1, 24, and 48 h pre-incubation with these cells ( Figure S13). However, we did not consider crambescidin 359 a promising antiparkinsonic agent because this alkaloid . The data are presented as m ± se (n = 3). * p < 0.05 compared to cells exposed to paraquat alone.
Crambescidin 359 reduced ROS formation (maximum 17.9 ± 4.6%, 10 µM concentration) in Neuro 2a cells exposed to paraquat ( Figure 5) and exhibited neuroprotective effects after 1, 24, and 48 h pre-incubation with these cells ( Figure S13). However, we did not consider crambescidin 359 a promising antiparkinsonic agent because this alkaloid showed inhibitory activities in electrophysiology experiments on human α7 nicotinic ac- . The data are presented as m ± se (n = 3). * p < 0.05 compared to cells exposed to paraquat alone.
Phytoceramides 2b, 2e, 2f, 4a-c, 4e, 4f with unbranched backbones may be found in many organisms, including terrestrial mushrooms, plants, and animals [25]. However, phytoceramides with methyl-branched backbones 1, 3, 5, and 6 are far less common. The compounds of this group were mainly obtained from marine organisms including sponges ( i-t17:0 and i-t19:0). All the unknown phytoceramides found in the present study are also characterized by these isoor anteiso-methyl-branched backbones. Among the new variants of N-acylation of the above-mentioned marine sphingoid bases, compounds 3d, 5d, and 6d from M. clathrata have methyl branching at both the sphingoid and fatty acid (i-C 23 ) chains.
Compounds 3b (monanchoramide B) and 5b (monanchoramide C) have previously been found in the Philippine sample of M. clathrata, along with monanchoramides A (i-t20:0/(2R)-2-OH-22:0) and D (i-t21:0/(2R)-2-OH-22:0) [7], which were not detected in the Australian sample of M. clathrata studied here. However, although monanchoramides do not contain anteiso-methyl-branched chains, the signals of unidentified minor anteiso-forms were found in their 13 C-NMR spectra ( [7]: Supplementary data). Apparently, the differences in the ceramide compositions of the Philippine and Australian samples of M. clathrata may be connected with geographical and seasonal influences on their fatty acid compositions. Fatty acids are known to serve as precursors in ceramide biosynthesis, and the sensitivity of the fatty acid profiles of sponges to environmentally induced (seasonal and geographical) variations was noted earlier [34].
The sponges of the family Crambeidae and related species were shown to contain several polycyclic guanidine alkaloids including crambescidins ( [35] and references cited therein). Co-occurrence of crambescidin-type alkaloids and phytoceramides was found in three crambeids, including M. clathrata studied here. In particular, monanchoramides A-D and alkaloids of the crambescidin group were isolated from the Philippine sample of M. clathrata [7,36]. In addition, ceramide i-t17:0/2-OH-24:0 and crambescidins were obtained from the related sponge Crambe crambe [27]. In our study, the pretreatment of human tumor-derived cells with the phytoceramides from M. clathrata decreased cell death induced by crambescidin 359 (7). We suggest that, similarly, these phytoceramides may help to protect sponge cells against injury, caused by their own cytotoxic crambescidin(s). Most of the crambescidin alkaloids exhibited a wide range of biological activities (including potent cytotoxicity), but little information regarding the true interaction of their polycyclic core with biological targets is known [35]. Therefore, a possible relationship between cytoprotective phytoceramides and cytotoxic crambescidins needs further investigation.
Our work showed a decrease of the cytotoxic effect of cisplatin on MDA-MB-231 and HL-60 cells after their pre-incubation with the phytoceramides of M. clathrata (Figure 5c,d).
Whether or not phytoceramides may be helpful for reducing cisplatin-induced toxicity in combination therapy, depends on the further experiments with these compounds. On the other hand, phytoceramides can be not only cytoprotective, but also cytotoxic agents for some tumor cells. The "dual" properties of the phytoceramides should be taken into account in an evaluation of their antitumor potential.
Using an in vitro paraquat model of Parkinson's disease, we found that the phytoceramides from M. clathrata influenced the neurodegenerative effect induced by paraquat. After a short time of pre-incubation (1 h), these phytoceramides decreased ROS formation and potentiated the neurodegenerative effect of paraquat. This looks contradictory because paraquat is known to exert deleterious effects through oxidative stress ( [37] and references cited therein). Thus, the reason for the observed additive damaging effect of phytoceramides and paraquat was unclear. However, we admitted that the absorption of phytoceramides by cells during 1h pre-incubation was insufficient to cause the cytoprotective properties of these compounds, and longer periods of time were required for this, as in the cases with crambescidin 359 and cisplatin (Figures 4 and 5). Indeed, the neuroprotective effect of phytoceramides was then revealed in the result of their 24 and 48 h pre-incubation with neuroblastoma cells (Figure 8). This suggests that phytoceramides, reducing neurodegeneration caused by paraquat, may be potential prophylactic agents for decreasing the risk of Parkinson's disease.
In general, the long-term preliminary treatment of the cells with the phytoceramides of M. clathrata was necessary for their cytoprotective functions; otherwise, an additive damaging effect of these sphingolipids and cytotoxic compound (crambescidin 359, cisplatin or paraquat) was observed. Therefore, exogenous phytoceramides can exert dual effects on cell survival, but their "delayed" effect may be cytoprotective.