Quinone and Hydroquinone Metabolites from the Ascidians of the Genus Aplidium

Ascidians of the genus Aplidium are recognized as an important source of chemical diversity and bioactive natural products. Among the compounds produced by this genus are non-nitrogenous metabolites, mainly prenylated quinones and hydroquinones. This review discusses the isolation, structural elucidation, and biological activities of quinones, hydroquinones, rossinones, longithorones, longithorols, floresolides, scabellones, conicaquinones, aplidinones, thiaplidiaquinones, and conithiaquinones. A compilation of the 13C-NMR spectral data of these compounds is also presented.

problems. Aplidium are clearly prolific producers of bioactive prenylated quinones and hydroquinones in the marine environment, and many other Aplidium species need to be investigated.

Quinones
Quinones can be derived by the oxidation of appropriate phenolic compounds, with 1,2-dihydroxybenzenes and 1,4-dihydroxybenzenes, yielding ortho-quinones and para-quinones respectively. Therefore, quinones can be formed from phenolics compounds by either the acetate or shikimate pathways, affording a catechol or quinol system. A range of quinone derivatives and related structures that contain a terpenoid fragment or shikimate-derived portion are also wide spread. For example, ubiquinones (coenzyme Q) have important biochemical functions in electron transport systems for respiration [13]. Figure 1 describes quinones that occur in Aplidium. The simple linear Prenylquinone (1) was isolated from Aplidium californicum [14]. The verapliquinones A-D (2)(3)(4)(5) were isolated from an unidentified Aplidium sp. (Ascidiacea) collected off the Breton coast. Verapliquinones A/B and C/D were characterized in a mixture of E,Z-isomers by NMR spectroscopy. The NMR data revealed that Verapliquinone B and D had a neryl group at C-2 instead of geranyl [15]. Davis and co-workers [16] reported a simple and versatile route to 1,4-benzoquinones based on the Claisen rearrangement, and applied to the synthesis of verapliquinones A and B, which had not previously been synthesized [16]. Chan and co-workers (2011) also reported verapliquinone A from A. scabellum [17].
Glabruquinone A (6) (or desmethylubiquinone Q 2 ) and Glabruquinone B (7) were isolated from Aplidium glabrum and synthesized. The main difference between 6 and 7 is that compound 6 contains a geranyl side chain instead of neryl as in 7 [18]. Glabruquinone A (6) displayed cancer preventive activity in the anchorage-independent transformation assay against mouse JB6 P + Cl 41 cells transformed with an epidermal growth factor, with an inhibition of the number of colonies C 50 (INCC 50 ) value of 7.3 µM. The INCC 50 values for 6 were 12.7, 17.5, and 50.5 µM against HCT-116, MEL-28, and HT-460 human tumor cells, respectively. Compound 6, at 10 µM, increased the UVB-induced p53 transcriptional activity of JB6 P + Cl 41 cells [18,19]. Glabruquinone A was also evaluated in vivo on mice inoculated with Ehrlich carcinoma tumors and found to inhibit tumorgrowth. Compound 6 inhibited the phenotype expression of HT-460, HCT-116, and SK-MEL-28 human tumor cells and induced apoptosis of these cell lines, as well as that of HL-60 and THP-1 tumor cells [20].

Hydroquinones
Prenylhydroquinone (9), Figure 2, isolated from A. californicum, exhibited activity in vivo against P388 lymphocytic leukemia. The potential cancer protective properties of prenylhydroquinone (9) were also evaluated by employing a modified-Ames assay for mutagenicity against Salmonella typhimurium; when prenylhydroquinone was added to experiments, the mutagenic effects of the carcinogens were drastically reduced [14].
Compound 9 was able to form stable semiquinone radicals according to Cotelle and collaborators, and in presence of glutathione, 9 was involved in a redox cycle with the consumption of oxygen. This process triggered the formation of free radicals and decreased the glutathione content, which is considered to be one of the major defenses against oxidative damage. Even although not fully elucidated, the antitumor properties of 9 can be correlated with its redox properties and reactivity toward glutathione [22]. Prenylhydroquinone also inhibits superoxide anion production in rat alveolar macrophages and in the xanthine/xanthine oxidase system. The antioxidant activity of 9 may be attributed to a direct reaction of the superoxide anion rather than to an enzymatic inhibition or a membrane signal transfer [23]. Prenylhydroquinone (9) and geranylhydroquinone (10), isolated from Aplidium sp., have exibited antiproliferative activity (IC 50 : 41 and 9.5 μM, respectively) in a P388 murine leukemia cell-line. Geranylhydroquinone (10) at doses up to 30 μM was inactive against the solid tumor cell lines A375 (human melanoma), A549 (human breast), HepG2 (human hepatic), and HT-29 (human colon), as well as normal human liver cells (WRL-68) [24]. Both prenylhydroquinone (9) and geranylhydroquinone (10) exhibited anti-inflammatory activity in an in vitro anti-inflammatory assay with activated human peripheral blood neutrophils by inhibing superoxide production. Compound 10 was also tested in the DPPH radical scavenging assay, and was considered inactive [24].
Geranylhydroquinone exhibited cytotoxicity against the leukemia cell lines of Rous sarcoma and mammary cincinoma in vivo [33]. Cytotoxic activity was also observed for 10 against P-388 leukemia (IC 50 0.034 µg/mL) and KB human epidermoid carcinoma cells (IC 50 4.3 µg/mL). Geranylhydroquinone (10) has also demonstrated antibacterial activity, with a minimum inhibitory concentration (MIC) of 64 µg/mL against Staphylococus aureus and S. faecalis, and an MIC of 128 µg/mL against Serratia marcescens. The minimum bactericide concentrations (MBCs) were determined to be 2, 1, and ˃4 µg/mL, respectively [27]. Additionally, Compound 10 was more potent than two standard antioxidants in terms of its inhibitory effects on lipid peroxide formation in rat liver microsomes and on soy-bean 15-lipoxygenase [34].
Geranylhydroquinone (10) and hydroxydiprenylhydroquinone (11) displayed significant cytotoxicity against four tumor cell lines, in particular, against P-388 mouse lymphoma suspension culture (IC 50 0.81 and 4.5 µM, respectively) indicating that the hydroxylation of the prenyl chain in 11 may result in a marginal decrease in its cytotoxicity [26].
Methoxyconidiol (16) was isolated from A. aff. densum [35] and its effects were tested on sea urchin embryos during cell division; compound 16 disturbed the mitotic spindle assembly leading to a cell cycle arrest during the metaphase/anaphase transition [36]. The antibacterial activity of 16 on Escherichia coli and Micrococcus luteus was also evaluated by microtiter broth dilution method; it had no antibacterial effect [37]. Methoxyconidiol inhibited egg division, with IC 50 values of 0.80 μM for Paracentrotus lividus eggs and 4.30 μM for Sphaerechinus granularis eggs. When tested against the human carcinoma cell lines: MCF7 (breast), PA1 (ovary), PC3 (prostate), CEM-WT (acute lymphoblastic leukemia), and L-929 murine immortalized cells, as well as normal human fibroblasts, 16 was non-toxic to all these cell lines with an IC 50 > 100 µM [35,37]. Methoxyconidiol caused antimitotic action during the first division of sea urchin embryos, and the mechanism of action of methoxyconidiol may be mediated by the disruption of microtubule dynamics [29]. Hence, Simon-Levert et al. 2010 [37] concluded that methoxyconidiol was ineffective against human cancer cells but effective in sea urchin cells. This finding could be explained by a difference between these two type of cells in either membrane permeability and/or intracellular transport of 16.
The compound 2-geranyl-6-methoxy-1,4-hydroquinone-4-sulfate (17) was isolated from A. scabellum and inhibited superoxide production by PMA-stimulated human neutrophils in vitro, with an IC 50 of 21 μM. To determine the effect of different treatments on cell survival, drug-treated neutrophils were stained with the fluorescent markers for necrosis (propidium iodide) and apoptosis (Annexin V-FITC), and analyzed by flow cytometry. Treatment with 17 had no effect on neutrophil viability, but the results indicated that 17 did inhibit neutrophil superoxide production [17].

Rossinones
Rossinones (Figure 3), particularly rossinone B (20) and its derivatives, are linearly fused in a 6,6,5-ring core of rossinone B, which is an extremely rare skeleton, known for only three plant-derived natural products. Rossinone B was isolated for the first time from an unidentified Antarctic species of Aplidium and then from A. fuegiense [24,38]. Therefore, the isolation of 20 extended the evolutionary range of the requisite biosynthetic terpene cyclase(s) from Plant kingdom to Animalia [24]. To date, five rossinones (18)(19)(20)(21)(22) have been discovered. Rossinones A (18) and B (20) exhibited anti-inflammatory activity in an in vitro anti-inflammatory assay with activated human peripheral blood neutrophils by inhibiting superoxide production. However, in the DPPH radical scavenging assay, 18 and 20 were found to be inactive (at doses up to 30 μM), indicating that these rossinones are considerably less effective as superoxide scavengers than as suppressors of superoxide production by neutrophils. Rossinones A and B have exhibited selective antiviral activity against the DNA virus HSV-1, versus the RNA virus PV-1, with both compounds exhibiting antiviral activity at 2 μg/disk. Both compounds also exhibited antimicrobial activities against Bacillus subtilis and the fungus Trichophyton mentagrophytes [24].

Longithorones, Longithorols and Floresolides
Longithorones ( Figure 4) and longithorols ( Figure 5) are unique farnesylated quinones/hydroquinones isolated from A. longithorax (Monniot). Their complex structures are characterized by the presence of a metacycophane and/or paracyclophane system built in a farnesyl quinone or hydroquinone formally by the rarely encountered cyclization of farnesyl quinones/hydroquinones [12]. One special characteristic of those compounds is the atropisomerism that is caused by the restricted rotations of their macrocyclic rings. Eleven compounds have been isolated (longithorones A-K), including monomeric prenylated quinones (24)(25)(26) and dimeric prenylated quinones (23 and 27-31), and the cyclofarnesylated quinones (longithorones J and K; 32-33) [40,41]. The biosynthesis of dimeric longithorones, which have been supposed to originate by both intra-and intermolecular Diels-Alder reactions, has been speculated about. Fusion of the two farnesyl-quinone units can be envisioned as arising via a Diels-Alder cycloaddition of suitably unsaturated precursors, whereas rings B and C could arise by a transannular Diels-Alder reaction. The co-isolation of the monomers provides some support for this proposal [12].
Longithorone A (23) displayed cytotoxicity against P388 murine leukemia cells with an IC 50 of ~10 mg· mL −1 [42]. In addition, the longithorone J (32) was tested for cytotoxicity against the cell lines SHSY5Y (human neuroblastoma), HEK293T (SV40 T antigen transformed human embryonal kidney cells) and A549 (human non-small cell lung carcinoma). Compound 32 did not exhibit any cytotoxicity in the A549 cell assay when tested at 2 and 20 mg/mL. However, 32 displayed minimal activity at 20 mg/mL against the SHSY5Y and HEK293T cells, with cell deaths of 28% and 16%, respectively [43].
Zacarian and collaborators synthesized the highly rigid macrocyclic carbon skeleton of longithorone C (25) by exploiting quadrupolar interactions as synthetic strategy [44]. Despite the fact that cyclophanes have been extensively synthesized and evaluated because of their unique physical and chemical properties, examples of the isolation and total synthesis of cyclophane-containing natural products are rare and challenging. This has attracted the interest of synthetic chemists in search of new and efficient strategies for syntheses with a reduced number of steps [45,46].   Longithorols A (34) and B (35) are prenylated paracyclophane and metacyclophane hydroquinones, and longithorols C (36) and D (37) are para-substituted cyclofarnesylated hydroquinones. The hydroquinones (longithorols A-D, 34-37) were also isolated from A. longithorax, moreover 34 and 35 were isolated as their pentaacetate forms because of their instability [47,48].
Floresolides are monomeric cyclofarnesylated hydroquinones with an endocyclic -lactone. They are members of longithorone/longithorol class of meroterpenes. Floresolides A-C (38-40) ( Figure 6) have been isolated from Aplidium sp. collected in Indonesian. All of these floresolides exhibited moderate cytotoxicity against KB tumor cells [49]. The synthesis of floresolide B (40) hydroquinone lactone core was performed by employing ring-closing metathesis approach by Briggs and Dudley [50]. The total synthesis of racemic floresolide B was reported by Nicolaou and Xu, who used an olefin metathesis-based strategy for the formation of the macrocyclic lactone portion [51,52].

Scabellones
Chan et al., 2011 [17], have described the isolation of the pseudodimeric meroterpenoid scabellones A-D (41-44) (Figure 7). Scabellone B (42) was able to inhibit the superoxide production by PMA-stimulated human neutrophils in vitro, with an IC 50 of 125 μM. In contrast, 42 had no effect on neutrophil viability. Scabellone B (42) was also evaluated against the neglected disease parasite targets Trypanosoma brucei rhodesiense, T. cruzi, Leishmania donovani, and Plasmodium falciparum. This compound exhibited selectivity towards only P. falciparum (a K1 chloroquine-resistant strain), with an IC 50 of 4.8 μM, and demonstrated poor cytotoxicity (in a L6 rat myoblast cell line, IC 50 : 65 μM). The core benzo[c] chromene-7,10-dione scaffold of scabellones A-D is rare among natural products and has previously been associated with antiproliferative or apoptosis-inducing biological properties.

Conicaquinones, Aplidinones, Thiaplidiaquinones and Conithiaquinones
Two cytotoxic terpene quinones, the isomeric prenylated quinones conicaquinones A and B (45, 46) (Figure 8), which have an unusual 1,1-dioxo-1,4-thiazine ring added to their quinone moiety, have been isolated from Mediterranean ascidian A. conicum. Both compounds were evaluated in vitro on rat glioma (C6) and rat basophilic leukemia (RBL-2H3) cell lines and demonstrated selectivity against rat glioma cells [53]. A follow-up study with this species, conducted by the same research group resulted in the isolation of three new geranylated quinones, aplidinones A-C (47-49) [54] and two new prenylated benzoquinones, designated as thiaplidiaquinones A and B (50,51), with an unprecedented tetracyclic skeleton in which a chromenol unit is attached to a p-benzoquinone ring condensed to a 1,4-thiazine-dioxide ring. Thiaplidiaquinones A and B were investigated for their antitumor activity. Both compounds were able to induce apoptosis in Jurkat cell lines that were derived from a human T lymphoma through the overproduction of reactive oxygen species (ROS), which mediated the collapse of the mitochondrial potential (ΔΨ m ). The thiaplidiaquinones 50 and 51 exhibited cytotoxic activity against the human leukemia T cell line Jurkat cells, with an IC 50 of approximately 3 μM [55]. The total synthesis of thiaplidiaquinone A (50) was described by Carbone [56], while the biomimetic synthesis of thiaplidiaquinone A and B was developed by Copp around the same time [57].
The lack of 13 C-NMR resonance for the group of unsaturated carbon atoms, which were substituted by a δ-deshielded value (δ C 70.9), as in 4 [15], suggest the presence of a hydroxyl group. This finding was also observed for compounds 11 and 12.
Compound 15 and 16 presented a substituted hydroquinone nucleus, together with a 3,4-disubstituted-1-methylcyclohexene ring, and 1-hydroxyisopropyl unit, which was attached to C-6ʹ of the cyclohexene ring upon observation of the ROESY spectrum and NOE experiment, respectively. The difference between 15 and 16 was the presence of a methoxy group at 16 [31,35].
Compound 17 produced 1 H-NMR spectrum resonances attributable to two aromatic protons at δ H 6.75 (1H, d, J = 2.1 Hz, H-5) and δ H 6.54 (1H, d, J = 2.1 Hz, H-3), two olefinic protons at δ H 5.29 (1H, t, J = 6.6 Hz, H-2′) and δ H 5.09 (1H, t, J = 6.6 Hz, H-6′), three moderately deshielded methylene signals (δ H 3.27, 2.08, and 2.00), one methoxyl signal at δ H 3.83, and three allylic methyl singlets at δ H 1.69, δ H 1.65, and δ H 1.58. Direct comparison of the 1 H and 13 C chemical shifts of the aromatic ring signals of 17 with a synthetic derivative of 17 (compound which had a hydroxyl group at C-4 instead of the sulfate) allow Chan and co-worker to identified an upfield shift of C-4 and downfield shifts of C-3 and C-5 in the 13 C-NMR spectrum and downfield shifts of H-3 and H-5 in the 1 H-NMR spectrum which were consistent with the placement of the sulfate group at the C-4 [17].
The 13 C-NMR data for Rossinone A (18), Table 3, allowed for the identifications of a triprenylated (farnesyl) hydroquinone-bearing substitution in the terminal prenyl unit. A α-hydroxy ketone group (δ C 201.8) and a carbinol resonance at δ C 69.8 in the side chain of 18 were established [23].
The that could be explained by the diverse steric influence of the hydroxyl group in these two compounds; the structure was confirmed by NOE experiments [38].
For 5,6-Epoxy-rossinone B (22), revealed in 1 H-NMR spectrum the lack of the double bond in ring B when compared with 20, which is substituted by an epoxide ring in 22. The H-6 was attributed to a singlet at δ H 3.80. In the 13 C-NMR spectrum the values at δ C 61.4 and δ C 55.7, were assigned to C-6 and C-5, respectively. Additionally, due to the absence of the conjugated double bond, the C-4 quinone carbonyl in 22 was shifted down-field (δ C 193.9) with respect to 20 (δ C 185.0) [38].
The complex and unprecedented structure of longithorone A (23) was determined by crystal X-ray diffraction, while the enantioselective biomimetic synthesis of longithorone A was accomplished by Layton and coworkers [59]. The 1 H and 13 C-NMR of (23) (  (27) and F (28) are atropoisomers with respect to the para-disubstituted benzoquinone ring [40]. In turn, longithorone F (28) differs from longithorone G (29) in its C-2ʹ, C-3ʹ, and C-10ʹ stereochemistry. The NOESY data for 29 revealed correlations between the aldehyde hydrogen H-13 with H-5 and H-2ʹ suggesting that these protons were all on the same face of the molecule. These correlations were not, however, observed in the NOE experiment for 28.
The farnesyl chains in longithorone J (32) and K (33) were characterized by three shielded olefinic methyl resonances, two of these (δ C 14.9 and δ C 16.9) assigned E-geometry configuration and the other (δ H 22.8) was assigned a Z-geometry. The 1 H and 13 C-NMR spectra of longithorone J include an oxymethine proton at δ H 4.84 (δ C 71.5) in contrast to the presence of two ketone resonances at δ C 200.7 and δ C 196.8 in longithorone K (33). The absolute stereochemistry of longithorone J (32) has been determined by the advanced Mosher method, while the absolute stereochemistry for longithorone K (33) was suggested by comparison with 32 and based on biosynthesis [43].
A comparison of the 1 H and 13 C-NMR data for longithorols A-D (34-37) ( Table 5) with previous longithorones reveals that, in longithorols, signals for two substituted 1,4-hydroquinones (C-16 to C-21 and C-16ʹ to C-21ʹ) are present instead of signals for two substituted 1,4-quinones. In addition, an acetoxymethine group can be recognized by the signals at δ H 6.61 (d, J = 10.0 Hz) and δ C 67.1 (for example, in 34), which replace the methylene group corresponding to C-1 in longithorones [47]. The absolute stereochemistry of longithorol C (36) has been determined by the advanced Mosher method [48].
Floresolides skeleton can be recognized by the presence of an endocyclic ε-lactone with an α,β-epoxy-group (δ C 63.7 and δ C 59.1) in 38 and (δ C 62.5 and δ C 58.3) in 39, and a double bond at the C-2 and C-3 position (δ C 137.8 and δ C 136.1) in 40 ( Table 5). The terminal methylene group at the C-11 and C-12 positions is also typical in this structural family. The Floresolide 39 has a primary alcohol and a fully substituted benzene ring including two bromine atoms in its structure. Compound 40 has absolute stereostructure, confirmed by crystal-X-ray diffraction [49].

Conclusions
The studies presented in this review reveal the importance of the prenylated quinone and hydroquinone metabolites, which have important biological activities and occur with frequency in ascidians of the genus Aplidium. Those non-nitrogenous metabolites, mainly prenyl quinones or hydroquinones, which can be either linear or cyclic compounds, such as rossinones, longithorones, longithorols, floresolides, scabellones, conicaquinols, aplidinones, thiaplidiaquinones, and conithiaquinones, are examples of meroterpenes. The evaluated compounds mainly presented cytotoxic, anti-inflammatory, and antimicrobial activities. Furthermore, ascidians (tunicates) are a promising source of new bioactive compounds from marine environments [61][62][63]. The Aplidium genus is able to produce meroterpenes with a large range of structural variety, of which the longithorone series is notable for containing the most complex structures with a metacyclophane and paracyclophane scaffolds. The complex and elaborate structure of some of these meroterpenoids has led several research groups to engage in performing the total synthesis of specific compounds for their structural confirmation and to clarify biosynthetic pathways. The 13 C-NMR data of a given compound are an important tool for natural products research because it is sometimes possible to propose the structure of a novel natural compound by performing comparisons with data for known compounds. For example, the structure of rossinones, were deduced by comparing 2D NMR spectrum with those from literature data reported for rossinone B. Therefore, considering the enormous diversity in their chemical structures and the biological potential of the prenyl quinones and hydroquinones found in Aplidium, it is important to conduct further studies of this genus using a multidisciplinary approach.