Meroterpenes from Marine Invertebrates: Structures, Occurrence, and Ecological Implications

Meroterpenes are widely distributed among marine organisms; they are particularly abundant within brown algae, but other important sources include microorganisms and invertebrates. In the present review the structures and bioactivities of meroterpenes from marine invertebrates, mainly sponges and tunicates, are summarized. More than 300 molecules, often complex and with unique skeletons originating from intra- and inter-molecular cyclizations, and/or rearrangements, are illustrated. The reported syntheses are mentioned. The issue of a potential microbial link to their biosynthesis is also shortly outlined.


Introduction
Quinones are ubiquitous in nature, occurring as secondary metabolites in many organisms; often they are molecules essential to life, being intimately related to the oxidative processes in cells [1]. Polyprenylated 1,4-benzoquinones and hydroquinones, such as ubiquinones, plastoquinones, and tocopherols, are widespread in plants and animals, in which they play important roles in electron transport, photosynthesis, and as antioxidants [1,2]. Terpene quinone/hydroquinone natural products differing from the above-mentioned groups frequently occur as secondary metabolites in many organisms; they form a class of complex metabolites, generally called meroterpenes, of mixed OPEN ACCESS biosynthetic origin which are partially derived from terpenoids. In addition to their wide occurrence, meroterpenes display a huge range of structural diversity, with structures varying from simple compounds comprising a prenyl unit linked to the hydroquinone unit to unique architectural scaffolds, often linked with varied functionalities, arising from intra-and intermolecular ring closures and/or rearrangements of the terpene chains. Moreover, they display important biological activities, undoubtedly related to their most prominent chemical feature, that is their ability to undergo redox cycling to generate reactive oxygen species (ROS) which can damage cells [3,4].
In the marine environment, meroterpenes have been isolated mainly from brown algae and microorganisms, but another important source are marine invertebrates, mainly sponges and tunicates. The present review provides an update on the meroterpenes so far isolated from marine invertebrates; it describes the structures and biological activities of 300 natural products, thus highlighting the structural diversity generated in this class of natural products and their potential in drug discovery. The issue of a potential microbial link to their biosynthesis is also shortly outlined.

Meroterpenes from Ascidians
While the majority of metabolites isolated from ascidians are nitrogen-containing compounds (alkaloids or peptide related compounds), ascidians belonging to the genus Aplidium are known as a rich source of meroterpenes [5][6][7]. The first biologically active tunicate metabolite was indeed geranylhydroquinone (2), isolated from an Aplidium sp. and, successively, found in many others Aplidium species; it was shown to offer protection against leukemia and tumor development in test animals [8]. Several linear diprenylquinones/hydroquinones (compounds 1-17, Figure 1) have been then reported from diverse Aplidium species (A. multiplicatum [9], A. savigny [10], A. conicum [11], A. glabrum [12], A. scabellum [13], Aplidium spp. [14,15]); the majority of them possess a linear side chain of the geranyl type with rare examples of neryl derivative such as verapliquinone D (13), verapliquinone B (14), and glabruquinone B (15). A. californicum has been the source of the simple monoprenyl derivatives 1 and 9 which were identified as anticancer and antimutagenic agents [16]. Glabruquinone A (3-demethylubiquinone Q2, 11) is closely related to the ubiquinones, although lacking the methyl group in the quinoid moiety; it is not a cytotoxin but demonstrated good cancer preventive activity on JB6 Cl 41 cell transformation activated by epidermal growth factor (EGF). Structure-activity relationships studies on its synthetic analogs demonstrated that this activity depend on the length of the side chain and on the position of the methoxyl groups in the quinone part of the molecule [17]. In vivo, anticancer properties of 11 and its synthetic analogs, as well as the molecular mechanism of its action against tumor cells, have been examined; it was shown to inhibit the growth of the solid Ehrlich carcinoma in mice and to induce apoptosis in various human tumor cell lines [18]. Both 2-geranyl-6-methoxy-1,4-hydroquinone-4-sulfate (17) and the triprenylated (farnesyl) hydroquinone rossinone A (8) were found to be active in an anti-inflammatory assay, in vitro, with activated human peripheral blood neutrophils; they inhibit the superoxide production when either N-formyl-methionyl-leucyl-phenilalanine (fMLP) or phorbol myristate acetate (PMA) were used to activate the respiratory burst [13,15]. The monoprenylhydroquinone 1 and geranylhydroquinone 2 were also active in the same assay [15], indicating that prenyl quinones could indeed hold promise for the development of new anti-inflammatory agents. Several interesting classes of cyclic, polycyclic, macrocyclic, and/or dimeric prenyl quinones/hydroquinones are reported to occur in marine ascidians. The biosynthetic origin of these compounds has been largely speculated about; they clearly arise from corresponding derivatives with linear prenyl chains that are usually isolated concomitantly from the natural source. Ortho-prenylated quinones can undergo different chemical transformations, ranging from cascade cyclization reactions involving carbocation species [19] to pericyclic reactions, such as electrocyclizations or cycloadditions, via ortho-quinones methide intermediates [20]. However, with the data available, it is not possible to argue whether the above mentioned transformations occur in the organism, prior to its extraction, either enzymatically (in nature dehydrogenase enzymes appear to catalyze such processes [21]) or not, or that they take place during isolation and/or chromatographic purification.
Cyclodiprenyl hydroquinones/quinones ( Figure 2) have been isolated from A. aff. densum (methoxyconidiol, 18) [22] and from A. conicum (conitriol, 19, and conidione, 20) [11]. Methoxyconidiol (18) displayed an antimitotic action on the first division of sea urchin embryos, disrupting M-phase progression and completely blocking cytokinesis without having any effect on DNA replication [23]. The occurrence of chromene (2H-benzopyran) derivatives as natural products has been reported in A. californicum (21) [14], A. costellatum (22) [24], A. multiplicatum (23) [9], A. scabellum (24) [13], and A. solidum (25 and 26) [25]. The chromane derivative 27 has been isolated from Synoicum costellatum [26], a species closely allied to the genus Aplidium, and successively found in A. conicum together with its C-1′ epimer, conicol (28) [11]. Didehydroconicol (29) has been isolated from A. aff. densum [22]. Tuberatolides (30 and 31) and sargachromenols (32 and 33), isolated from Botryllus tuberatus along with their putative linear precursor yezoquinolide (16), antagonized the chenodeoxycholic acid (CDCA)-activated human farnesoid X receptor (hFXR), a ligand-dependent transcription factor in the nuclear receptor superfamily which has been recently identified as a promising drug target in the treatment of atherosclerosis [27]. Longithorol E (34), from A. longithorax, is a unique macrocyclic chromenol containing a new 14 membered carbocycle [28] (Figure 3). Rossinone B (35) is a triprenylated (farnesyl) quinone first isolated from an Antarctic Aplidium sp. [15] and successively recovered in the viscera extract of A. fugiense, also collected in Antarctica, along with the related compounds 36-38 [29] (Figure 4). Rossinone B has a rather novel structural architecture featured by a linearly fused 6-6-5-ring core, which so far has been found in only three plant-derived natural products, pycnanthuquinones A-C [19,30]. The tricyclic framework of 35 supposedly derive from the corresponding linear hydroquinone derivative rossinone A (8) which has been reported to co-occurr in Aplidium sp. [15]. Interestingly, neither acyclic hydroquinones nor putative quinone-containing precursors of compounds 35-38 were detected in A. fugiense extract [29]. Rossinone B (35) exhibited anti-inflammatory, antiviral, and antimicrobial activities [15]. Attracted by its novel chemical structure, promising biological properties and potentially intriguing biosynthetic pathway, a biomimetic total synthesis of (±)-rossinone B has been achieved through a highly efficient strategy featuring a series of rationally designed reactions, including an intramolecular vinyl quinone Diels-Alder reaction to construct the linear 6-6-5 tricyclic core of 35 [31]. A. longithorax has been the source of longithorones and longithorols, a group of farnesylated quinone/hyroquinones featuring unprecedented macrocylic skeletons derived formally by the rarely encountered cyclization of farnesyl quinones/hydroquinones to give [9]-and [10]metacyclophane, as well as [12]paracyclophane structures [28,[32][33][34][35]. Longithorones B-D (39)(40)(41), J (42), and K (43) together with longithorols C (44) and D (45), are monomeric C 21 compounds ( Figure 5). Longithorol C (44) could undergo an intramolecular cyclization, followed by dehydration, to yield the chromenol longithorol E (34), which is possibly an artifact of the isolation process. A short synthetic approach to the macrocyclic framework of longithorone C has been described via ring-closing metathesis using the Grubbs second generation catalyst [36]. Longithorone J (42) is the first example of a γ-hydroxy-cyclohexenone in this class of compounds. Floresolides A-C (46)(47)(48) are three further monomeric cyclofarnesylated hydroquinones isolated from an Indonesian Aplidium sp. They are unique members of the longithorone/longithorol class of meroterpenes, having an endocyclic ε-lactone bridging the aromatic ring and a [10]metacyclophane moiety. Floresolides showed moderate cytotoxicity against KB cells [37]. Longithorones A (49) and E-I (50-54) are dimeric compounds ( Figure 6). Compounds 49-54, as well as the monomeric longithorones 39-43, exhibit atropisomerism arising from hindered rotation of quinone ring through their macrocyclic rings. 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 39-43 provides some support for this proposal. The stereochemistry of the central carbocyclic rings in 35 and 39-43 is consistent with such a fusion [34]. An enantioselective biomimetic synthesis of longithorone A has been accomplished, which demonstrates the feasibility of the reactions proposed for the biosynthesis-albeit using non-enzymatic conditions. The challenge of a synthesis of longithorone A was heightened by the presence of two forms of chirality: the stereogenic centers present in the tricyclic core portion of the molecule and atropisomerism. The synthesis presents a unique example of chirality transfer in complex molecule synthesis involving the use of stereogenic centers to control atropisomerism, removal of the stereogenic centers, and transfer of the atropisomerism back to stereogenic centers in the natural product [38]. Longithorols A (55) and B (56) represent the first examples of hydroquinones in the [12]paracyclophane structure class. They were isolated as their pentaacetates because of their rapid decomposition occurring under purification conditions [35]. To date, the only biological activity reported for longithorones/longithorols class of marine metabolites pertains to longithorone A (49), which was shown to display cytotoxicity against P388 murine leukaemia cells.
Further examples of pseudodimeric meroterpenoids are scabellones A-D, (57-60) isolated from A. scabellum [13], possessing a benzo[c]chromene-7,10-dione scaffold particularly rare among natural products ( Figure 7). Scabellone B (58) was found to inhibit the superoxide production by PMA-stimulated human neutrophils in vitro; it was also evaluated against the neglected disease parasites targets Trypanosoma brucei rhodesiense, T. cruzi, Leishmania donovani, and Plasmodium falciparum and exhibited selectivity toward Plasmodium falciparum (K1 chloroquine-resistant strain) with IC 50 4.8 μM and only poor cytotoxicity (L6 rat myoblast cell line, IC 50 65 μM) [13].  A sample of A. conicum collected along Sardinia coasts gave rise to the isolation of a large group of unique prenylated thiazinoquinones, namely conicaquinones A and B (61 and 62) [39], aplidinones A-C (63-65) [40], and thiaplidiaquinones A and B (66 and 67) [41] (Figure 8). All these meroterpenes feature an unusual 1,1-dioxo-1,4-thiazine ring fused with the quinone portion. Aplidinones A-C (63-65) and conicaquinones A and B (61 and 62) possess a benzoquinones and a naphtoquinone moiety, respectively. Thiaplidiaquinones A and B (66 and 67) possess an unprecedented tetracyclic core, visibly composed of two geranylated benzoquinones that have fused together, whose biosynthetic origin has been speculated to stem from hypotaurine addition to tricyclic pyranoquinones derived from oxa-6π-electrocyclization of ortho-quinone methide tautomers of bis-benzoquinones [42]. Based on this premise, two biomimetic synthesis of the thiaplidiaquinones scaffold have been reported [42,43]. In a first concise total synthesis of thiaplidiaquinone A, the key ring forming steps are both based on biosynthetic considerations and involve the construction of the central benzo[c]chromene quinone unit by an extremely facile oxa-6π-electrocyclic ring closure reaction of an ortho-quinone intermediate, derived by tautomerization of a bis-benzoquinone, readily accessed from two simple phenolic precursors. This is followed by the installation of the 1,4-thiazine-dioxide ring by reaction of the benzo[c]chromene quinone with hypotaurine [43]. An alternative biomimetic synthesis of both thiaplidiaquinones A and B has been reported where bis-benzoquinones construction, instead of via a Suzuki-Miyaura reaction [43], was achieved simply by base-mediated dimerization of geranylbenzoquinone. Subsequent reaction with hypotaurine yielded the dioxothiazine regioisomers of thiaplidiaquinones A and B [42]. Both conicaquinones A and B showed a marked and selective cytotoxic effects on rat glioma cells [40], and thiaplidiaquinones were strongly cytotoxic against Jurkat cell line, derived from a human T lymphoma, inducing cell death by apoptosis [41]. The pro-apoptotic mechanism of thiaplidiaquinones involves the induction of a strong production of intracellular reactive oxygen species (ROS) in the cells, likely due to the inhibition of the plasma membrane NADH-oxidoreductase (PMOR) system, an important target for anticancer drugs, through interference with the coenzyme-Q binding site [41]. In order to validate the structural assignment made for aplidinones by theoretical means a synthetic approach has been undertaken which yielded some synthetic analogs of aplidinone A in which the geranyl chain is replaced by other alkyl chains [44]; these compounds as well as the natural metabolite 63 were subjected to cytotoxicity assays and preliminary structure-activity relationships (SAR) studies. Both aplidinone A and its synthetic analogs were shown to possess interesting cytotoxic effects; SAR studies revealed that cytotoxic activity depends on the nature and the length of side chain linked to the benzoquinone ring and, mainly, on its position respect to the dioxothiazine ring. The study also evidenced one of the synthetic analogs as a potent cytotoxic and pro-apoptotic agent against several tumor cell lines, which also inhibits the TNFα-induced NF-κB activation in a human leukemia T cell line [44].

Meroterpenes from Sponges
Marine sponges have yielded a huge variety of meroterpenes having a terpenoid skeleton varying from sesqui-, di-, sester-or triterpene units.
The large class of sesquiterpene quinones isolated from various species of marine sponges, has attracted the attention of researchers because of their biological properties, including antimicrobial [45], antileukemic [46], anti-malarial [47], immunomodulatory [48,49], and anti-HIV [50] activities. Above all, the cytotoxic and antiproliferative properties of many sesquiterpene quinones/hydroquinones isolated from sponges have supported several studies for the development of new antitumor agents [51,52].
A large family of antineoplastic compounds, named metachromins (68-87), has been isolated from sponges of the genus Spongia, Thorecta and Hippospongia ( Figure 9) [53][54][55][56][57][58]. They exhibited potent antitumor activity against L1210 murine leukemia cells in vitro. All three compounds were found to be cytotoxic against all cell lines, with 86 being the most active. Surprisingly, 87, possessing a napthoquinone functionality, which is known to impart significant cytotoxic properties to various molecules [58], was significantly less active than both 85 and 86. Interestingly, metachromins N-R (78-82) [57], which contain both quinone and phenol functions, are inactive, most likely due to the bulky nature of the substituents present on the quinone portion.
In fact, most of sesquiterpene benzo(hydro)quinone isolated from sponges possess a decalin structure, consisting of a drimane or 4,9-friedodrimane skeleton, connected to the quinone/hydroquinone moiety generally via one carbon-carbon bonding. Particularly prolific sources of this sort of meroterpenes are marine sponges belonging to the Dysidea genus.
Bolinaquinone (108) is a cytotoxic sesquiterpene isolated from a Dysidea sponge in which quinone moiety is located on an unusual carbon of the 4,9-friedodrimane skeleton [86]. This compound was cytotoxic against HCT-116 human colon carcinoma (IC 50 = 1.9 μg/mL) and it has been demonstrated to act by interfering with or damaging DNA. 21-dehydroxybolinaquinone (109), isolated from the Hainan sponge Dysidea villosa, showed moderate PTP1B inhibitory activity and cytotoxicity, with IC50 values of 39.5 and 19.5 mM, respectively [87].
Wiedendiols A (117) and B (118) were isolated from the marine sponge Xestospongia wiedenmayeri, collected in the Bahamas [96]. The CETP-SPA inhibition assays carried out with these compounds revealed an IC 50 = 5 μM in both cases. Later, the inhibition of CETP was verified using a precipitation method to separate lipoproteins after incubation of HDL radiolabeled with LDL and CETP. In this assay, both 117 and 118 had an IC 50 of 1.0 and 0.6 μM, respectively. Wiedendiol B is a ten-fold stronger inhibitor of cyclooxigenase-2 than the reference compound indomethacine [97].
An aldehyde function is the distinctive feature of siphonodictyals A-D (123)(124)(125)(126), G (127), B2 (128), and B3 (129), which have been isolated from Siphonodictyon species along with the relevant alcohols siphonodictyols G (130) and H (131) and siphonodictyoic acid (132); most of them were sulfated [100][101][102]. The isolated compounds were tested for antimicrobial activity (antibacterial, antifungal); it has been suggested that the different substituents on the aromatic moieties have an impact on activity and that the ortho-hydroquinone structure may be the active center of the molecules. It is likely that the hydroquinone is oxidized in the metabolism of the assay organisms to yield the more toxic ortho-quinone.
The presence of aminoquinone compounds is not very common in natural products; however, several sesquiterpenes quinones/hydroquinones, with such a drimane or a 4,9-friedodrimane, where the aromatic fragment is substituted with simple amines and amino acids have been isolated from sponges (Figures 12 and 13). The sesquiterpene aminoquinones smenospongine (133), smenospongiarine (134), smenospongidine (135), and their corresponding 5-epimers 136-138, have been isolated from different sponge species [47,70,103,104]. Erythroid differentiation of K562 cells induced by compounds 133-138 and other related compounds has been studied. On the basis of structure-activity relationship studies, the following evidences were obtained: (a) the quinone structure is indispensable; (b) the amino group should play an important role; (c) the substituents at the amino group are not crucial; (d) the configuration at the C-5 in sesquiterpene part is not important [105,106]. Smenospongines B (139) and C (140) were isolated from D. elegans collected in Australia along with the sesquiterpene benzoxazole nakijinol B (141) and its diacetyl derivative 142 [107]. The biological activities of these compounds were established against a panel of human tumor cell lines, as well as a normal mammalian cell line. The compounds were found to have cytotoxic activities in the range 1.8-46 μM and appeared to lack selectivity for tumor versus normal cell lines. The presence of two bulky acetate moieties resulted in an approximate two-fold increase in the activity of 142 compared to the diol 141. One possible explanation for this increase in activity is that the acetate groups may contribute to greater bioavailability through enhanced membrane permeation after which metabolism, possibly hydrolysis by esterases, releases the active compounds intracellularly [108]. For 139 and 140, the additional methylene in the nitrogen-substituted side chain had the effect of reducing observed activity by a factor of 2 ( Figure 12).
Alisiaquinones A-C (217-219) and alisaquinol (220) [155], isolated from a New Caledonian deep water sponge, are related to xestoquinone, halenaquinone, adociaquinones, but they show an unusual substitution pattern on the furan ring. These new meroterpenes displayed mild activity with micromolar range on two enzymatic targets of importance for the control of malaria, the plasmodial kinase Pfnek-1 and a protein farnesyl transferase (PFT) as well as on different chloroquine-sensitive and chloroquine-resistant strains of Plasmodium falciparum [155].
The meroterpene sulfate fascioquinol E (224), isolated from a Fasciospongia sp., (Figure 16) is a rare example of diterpene benzo(hydro)quinone with a linear terpene moiety [156]. Fascioquinol E is an inhibitor of protein tyrosine phosphatase CpsB, which unexpectedly inhibited the growth of Gram-positive pathogens. CpsB is a member of the polymerase and histidinol phosphate phosphatase (PHP) domain family. Another member of this family found in a variety of Gram-positive pathogens is DNA polymerase PolC and this competes away fascioquinol E inhibition of CpsB phosphatase activity. It was showed that fascioquinol E not only inhibits the phosphatase activity of CpsB, but also ability of PolCPHP to catalyse the hydrolysis of pNP-TMP. This suggests that PolC may be the essential target of fascioquinol E, and that the PHP domain may represent an as yet untapped target for the development of novel antibiotics [157,158].

Conclusive Remarks
The great chemical diversity generating in the group of meroterpenes isolated from marine invertebrates and their wide range of biological activities represent a useful tool for development of new therapeutics. But the biomedical potential of these compounds could be greatly enhanced by a comprehensive understanding of their biosynthetic origin combined with the recent progress in molecular biology. The occurrence of different but biosynthetically related meroterpenes in different organisms, in terrestrial sources, and/or in collections of the same organism from distinct geographical locations, strongly supports the possibility of their biosynthesis by associated microorganisms. Significantly, several meroterpenoids have been recently isolated from Aspergillus spp. derived from tissues of marine invertebrates. It is known that members of the genus Aspergillus can combine polyketide and terpene precursors to produce meroterpenoids, some of whom having important relevance to human health; this is the case of territrem B, produced by A. terreus, a potent irreversible inhibitor of acetyl cholinesterase (AChE) and a candidate for drug development for treating Alzheimer's disease [204]. Examples of meroterpenoids isolated from invertebrate-associated Aspergillus spp. are tropolactones A-D (323-326) isolated from an Aspergillus sp. derived from an unidentified sponge [205], insuetolides A-C (327-329) from A. insuetus isolated from the Mediterranean sponge Psammocinia sp. [206], terretonins E (330) and F (331), isolated from A. insuetus derived from the Mediterranean sponge Petrosia ficiformis [207], austalides M-Q (332-336) from an Aspergillus sp. derived from the sponge Tethya aurantium [208], and yanuthones (337-344) isolated from A.niger obtained from tissue homogenates of an Aplidium ascidian [209] (Figure 24). Thus, there are grounds to suppose that meroterpenoids isolated from marine invertebrates or, at least, portions of their structure are microbial products, most likely elaborated by Aspergillus fungi. If confirmed, this possibility could work to advantage the research on these compounds, both for the exploitation of their huge chemical diversity and for a potential large-scale production of the bioactive molecules. The Aspergillus genus of fungi, indeed, has been largely investigated due to its medical and commercial importance. Research on Aspergillus has contributed much knowledge about its fundamental cell biology and biochemistry and, foremost, the significance of Aspergillus was cause for the sequencing of the genomes of some of the most well-known members of this genus which are now publicly available [210]. Attempts to locate the biosynthetic genes for meroterpenoids production in the genome of some Aspergillus spp. have been performed with encouraging results; the biosynthetic pathway for some meroterpenoids (austinol, terretonin) has been proposed [211,212]. Understanding of Aspergillus secondary metabolism would greatly profit from the genome sequencing projects; sequence information greatly facilitates the identification of natural product genes, the function of which can be demonstrated by molecular biological and biochemical approaches. When a set of genes involved in the formation of the same secondary metabolite are recognized, a biosynthesis can be proposed. Down the road, such advances should be useful for enhanced production of secondary metabolites of interest and the development of second-generation compounds with improved pharmacodynamic and pharmacokinetic characteristics. Thus, advances in Aspergillus secondary metabolite research in the post-genomic era will bring an understanding of meroterpenoids biosynthesis at the genetic level which should facilitate engineering of second generation molecules and increasing production of first generation compounds.