Bioactive Sesterterpenes and Triterpenes from Marine Sponges: Occurrence and Pharmacological Significance

Marine ecosystems (>70% of the planet’s surface) comprise a continuous resource of immeasurable biological activities and immense chemical entities. This diversity has provided a unique source of chemical compounds with potential bioactivities that could lead to potential new drug candidates. Many marine-living organisms are soft bodied and/or sessile. Consequently, they have developed toxic secondary metabolites or obtained them from microorganisms to defend themselves against predators [1]. For the last 30–40 years, marine invertebrates have been an attractive research topic for scientists all over the world. A relatively small number of marine plants, animals and microbes have yielded more than 15,000 natural products including numerous compounds with potential pharmaceutical potential. Some of these have already been launched on the pharmaceutical market such as Prialt® (ziconotide; potent analgesic) and Yondelis® (trabectedin or ET-743; antitumor) while others have entered clinical trials, e.g., alpidin and kahalalide F. Amongst the vast array of marine natural products, the terpenoids are one of the more commonly reported and discovered to date. Sesterterpenoids (C25) and triterpenoids (C30) are of frequent occurrence, particularly in marine sponges, and they show prominent bioactivities. In this review, we survey sesterterpenoids and triterpenoids obtained from marine sponges and highlight their bioactivities.


Introduction
Terpenes include primary and secondary metabolites, all biosynthesized from the five carbon isoprene building units [2]. Structural modification of these isoprene units leads a massively diverse range of derivatives with a wide array of chemical structures and biological properties. While higher plants' terpenoids were already studied and ethnopharmacologically rationalized centuries ago, those from marine counterparts were not explored until the first half of the 20 th century.
Steroidal terpenoids were the first marine isoprenes to be discovered by Bergmann during the 1930s-1940s, particularly sterols that were obtained from various marine macroorganisms [3]. Secondary metabolites, including terpenes, play an important ecological role in marine organisms. Being sessile and soft bodied, marine organisms face a harsh competition for space, reproduction, maintenance of an unfouled surface and deterrence of predation [4]. Therefore, marine organisms have developed bioactive secondary metabolites as a potential defensive means against competitors and/or predators [1]. These compounds are rapidly diluted after being released into the water and hence have to be of outstanding potency to retain their efficacy. These bioactivity(ies) proved appealing for chemical ecologists as well as for pharmacologists in their search for new drugs to treat or cure serious ailments such as inflammatory, infectious and cancerous diseases.
Marine terpenoids dominate much of the literature expression with a huge number of derivatives having been obtained from marine resources. It seems pointless to compile a review that includes all major classes of marine terpenoids. Therefore, in this review we concentrate on two major classes of marine isoprenes from sponges, namely the sesterterpenoids (C 25 ) and triterpenoids (C 30 ) with particular attention placed on their biological activities.
Manoalide was further investigated and found to be a potent inhibitor of phospholipase A 2 (PLA 2 ) [31][32][33][34][35][36][37][38]. Subsequently, many structurally related metabolites with PLA 2 inhibitory activity were also reported [8,[39][40][41][42][43][44][45]. PLA 2 is an enzyme that specifically catalyzes the hydrolysis of phospholipids at the S N -2 position to produce a lysophospholipid and arachidonic acid, which in turn provides the substrate for proinflammatory mediators such as leukotrienes, prostaglandins and thromboxanes, collectively known as the eicosanoids [41]. Since manoalide revealed an irreversible inhibition of phospholipase A 2 (PLA 2 ) [33], the structure-activity relationships (SAR) of this compound attracted scientific interests to study and to understand both PLA 2 function and mechanism of action in the whole cell. Therefore, several studies were successfully performed to determine the contributions of the various functional groups incorporated in 1 and its analogs, such as the γ-hydroxybutenolide, α-hydroxydihydropyran and trimethylcyclohexenyl ring systems, to the efficacy as PLA 2 inhibitors [36,41,45]. These studies indicated that (1) the existence of the hemiacetal in the α-hydroxydihydropyran ring is crucial for irreversible binding, (2) the γ-hydroxybutenolide ring is involved in the initial interaction with PLA 2 and (3) the hydrophobic nature of the trimethylcyclohexenyl ring system allows non-bonded interactions with the enzyme that enhances the potency of these analogs. These studies suggested that the closed ring form of manoalide is the predominant molecular moiety that accounts for the selective and potent inhibition of PLA 2 [36].
Luffariellolide (5) is a sesterterpenoid analog of secomanoalide (2), which was first reported from a Palauan sponge Luffariella sp. [8]. Structurally, luffariellolide differed in having C-24 as methyl group instead of an aldehyde functionality as in 2 and it was obtained as the (Z) isomer as well.  In contrast to the irreversible inhibitory action of manoalide (1) towards PLA 2 , luffariellolide (5) is a slightly less potent, but a partially reversible inhibitor. This meant that 5 became a more preferable anti-inflammatory agent for potential pharmacological investigation [8].
Luffariolides A-J represent a related group of sesterterpenoidal analogs, which have been obtained from different collections of the Okinawan marine sponge Luffariella sp. [13,14,16].
Luffarin metabolites comprise another group of compounds represented by 28 derivatives. 26 of them, luffarins A-Z, have been reported from the Australian marine sponge Luffariella geometrica [12], while the other two were obtained from the Adriatic Sea sponge Fasciospongia cavernosa [28]. Based on the chemical structures, luffarins have been classified into 14 bicyclic sesterterpenes, luffarins A-N; one bicyclic bisnorsesterterpene, luffarin O; one monocyclic sesterterpene, luffarin P; and six acyclic sesterterpenes, luffarin Q-V, in addition to four diterpenoidal derivatives, luffarin W-Z [12].
Biosynthetically, a relationship could be recognized between the various luffarins as illustrated in Figure 1. Luffarins appear to belong to the same enantiomeric series as reported for manoalide-type marine natural products. It is also curious to note that no acyclic luffarins incorporated the hydroxylated butenolide functionality. Perhaps the most interesting luffarins from a biosynthetic point of view are luffarins B (21) and O (21a), which were the first examples of a hitherto unknown cyclization pattern in compounds of this class [12].   Another example of bicyclic sesterterpenes are thorectandrols A-E (31-35) that were isolated from a Palauan collection of the marine sponge Thorectandra sp. [47,48] together with the parent compounds of this group palauolide (29) and palauolol (30). Palauolide (29) was obtained first as an antimicrobial sesterterpene from a three sponge association collected in Palau [54]. While palauolol (30) was identified as an anti-inflammatory sesterterpene from the Palauan sponge Fascaplysinopsis sp. and chemically it was recognized as being a secondary alcohol that upon dehydration yields 29 [55]. All thorectandrols (31)(32)(33)(34)(35) in addition to palauolide (29) and palauolol (30) were tested for antiproliferative activity against six to twelve human tumor cell lines depending on sample availability [48]. Palauolol (30) was active against all tested cell lines except A549 (non small lung cancer), with IC 50 values in the range 1.2-1.7 µM, while palauolide (29) showed a diminished activity. On the other hand, thorectandrols A-E revealed only weak to no cytotoxicity against the tested cell lines (IC 50 's 70-100 µM). While firm deductions on the structural requirements for activity were not possible, it appeared that the presence of both the hemiacetal lactone functionality and the 16-hydroxyl group in palauolol (30) enhanced cytotoxicity compared to palauolide (29) and other thorectandrols [48].

40
Petrosaspongiolides A (41) and B (42) were the first cheilantane sesterterpene lactones to be isolated from a New Caledonian sponge incorrectly assigned to the genus Dactylospongia [57] and then reassigned as a new genus and a new species: Petrosaspongia nigra (Bergquist 1995 sp. nov., class Demospongiae; order Dictyoceratida; family Spongidae) [58].
From another New Caledonian collection of the same sponge, 15 additional petrosaspongiolide congeners (C-R) were isolated [59,60]. From the chloroform extract of another Dictyoceratida sponge of the genus Spongia, 21-hydroxy derivatives of petrosaspongiolides K (44a) and P (48a) were isolated in addition to four other pyridinium alkaloids named spongidines A-D (51-54) [61]. Spongidines were found to be structurally related to petrosaspongiolide L (45) particularly in the presence of pyridine ring.  Petrosaspongiolides A-L were subjected to in vitro cytotoxicity assay against the human bronchopulmonary NSCLC-N6 carcinoma cell line. They revealed IC 50 values ranging between 1.0-32.2 µM [59]. Petrosaspongiolides C (43) and K (44) exhibited the highest potency with IC 50 values of 1.0 and 3.5 µM, respectively. However, petrosaspongiolides A (41) and B (42) were the least cytotoxic congeners in vitro with IC 50 values of 28 and 32.2 µM, respectively, 41 inhibited tumoral proliferation in vivo at 20 mg/Kg without significant toxicity when tested on immunosuppressed rats carrying a bronchopulmonary tumor (NSCLC-N6) [59].
Petrosasponiolides M-R (46-50) revealed the presence of a γ-hydroxybutenolide moiety and a hemiacetal function. Due to these structural similarities to manoalide (1), petrosaspongiolides M-R have received special attention from the scientific community to study their inhibitory activity against PLA 2 from different resources to point out their specificity. Two main groups of PLA 2 enzymes have been reported [62], the secretory PLA 2 (sPLA 2 groups I, II, III, V, IX, and X with relatively small molecular weights) and the cytosolic PLA 2 (cPLA 2 groups IV, VI, VII, and VIII with higher molecular weights). Inhibition of specific PLA 2 constitutes a potentially useful approach for treating a wide variety of inflammatory disorders such as spetic shock, adult respiratory distress syndrome, arthritis, and acute pancreatitis [61].
The mechanism of action of petrosaspongiolides M-R (46-50) as anti-inflammatory marine metabolites has been the topic for many research articles [63][64][65][66][67][68]. The covalent binding of 46 to bee venom PLA 2 has been investigated by mass spectrometry and molecular modeling. The mass increment observed was consistent with the formation of a Schiff base by reaction of a PLA 2 amino group with the hemiacetal function at the C-25 atom of the petrosaspongiolide M γ-hydroxybutenolide ring [63]. The molecular mechanism of inactivating the bee venom and the human type IIA secretory PLA 2 s by petrosaspongiolides R (50) [67], and M (46) [68], respectively, has been investigated. In both cases, either covalent (imine formation) and/or non-covalent (van der Waals) interactions contributed to the inhibitory activity against PLA 2 enzymes [67,68]. Due to potent anti-inflammatory properties of petrosaspongiolides, their chemical synthesis has been interestingly investigated. Recently, the first enantioselective synthesis of petrosaspongiolide R (50) has been successfully performed [69].

Triterpenes (C 30 )
Steroidal triterpenes were the first marine isoprenes to be discovered in the 1930s. Scientific interest has been driven towards these metabolites due to the isolation of biosynthetically unprecedented derivatives possessing a broad spectrum of bioactivity(ies). Marine triterpenoids have been reported from various marine macroorganisms. In this section, we survey two examples of triterpenoidal metabolites namely isomalabaricane triterpenes and steroidal saponins obtained from marine sponges with particular attention being drawn to their pharmacological significance.
Isomalabaricane triterpenoids having polyene conjugated functionality can be classified into three groups: (1) stelletins principally possessing the γ-pyrone functionality, which could be ring-opened in some of its congeners yielding the side chain with terminal free carboxylic acid and methyl moieties, (2) stelliferins oxygenated at C-22, and 3) globostellatic acids whose main feature is a carboxyl group at C-4. In addition to triterpenoids, the isomalabaricane core has been also recognized in some sesquiand/or sesterterpenes. The isomalabaricane terpenoids were sometimes trivially named according to their sponge origin.
Upon light exposure, the isomalabaricane-type terpenes readily isomerize at the C-13 position. Therefore, during isolation and characterization processes, they rapidly equilibrate into a 1:1 mixture of the 13E and 13Z isomers [78][79][80]88,89,98,99]. Nevertheless, these compounds continue to gain a great deal of attention because of their significant cytotoxic activity [79,89], whereas the nature of the natural isomer, either 13E or 13Z or both, is still unresolved. Recently it was reported that the 1 H NMR spectrum of a crude extract obtained from the fresh sponge Rhabdastrella aff. distinca (Hainan, the South China Sea) revealed that it mostly contained isomalabaricanes with the 13E-configuration (H-15 of most derivatives appeared around 7.0 ppm). Thus, the 13Z isomers were suggested in this case to be formed through isomerization during the isolation and analytical procedures [86].
Rhabdastrellins A-F (64-69), along with stelletins L (70) and M (71), were obtained from the marine sponge Rhabdastrella aff. distinca collected from a coral reef off Hainan, in the South China Sea [86]. Four of the rhabdastrellins (64-67) exhibited a primary alcohol moiety at C-29 instead of a methyl group as for the stelletins and the other two rhabdastrellins E (68) and F (69). While all rhabdastrellins and stelletins L and M share a hydroxyl group at C-3 instead of a carbonyl group as in other stelletins [86].  [79].
Apoptotic cell death is a stress response of cells to cytotoxic agents that might be executed either through a receptor-mediated pathway that activates caspase-8 or through a receptor-independent pathway that involves the cyclin-kinase inhibitors p53/p21. Both pathways lead to a translocation of pro-apoptotic Bax protein to the mitochondria, thereby resulting in a dissipation of mitochondrial membrane potential, activation of caspase-3, and execution of the apoptotic machinery [84].
Due to the significant antiproliferative activity exhibited by stelletins and stelliferins, research efforts have been directed towards their chemical synthesis. In 1999, Raeppel et al. successfully synthesized the common trans-syn-trans perhydrobenz[e]indene moiety in the isomalabaricane-type terpenoids, which enabled the chemical synthesis of stelletins and stelliferins [91].
Globostellatic acid (82) is the prototype of the third group of isomalabaricane-type triterpenoids sharing carboxylation at C-4. It was first isolated together with three other derivatives, globostellatic acids B-D, from the marine sponge Stelletta globostellata collected off Mage Island near Kagoshima, Japan [92].
Other globostellatic acid congeners, F-M, and X methyl esters, have been reported from different collections of the Indonesian marine sponge Rhabdastrella globostellata [93,94].
For cytotoxicity toward mouse lymphoma L5178Y cells, the 3-O-deacetyl congeners, globostellatic acids H/I (83/84) were the most active with an IC 50 of 0.31 nM. However, acetylation of the C-3 hydroxyl group decreases its bioactivity abruptly, as in globostellatic acids J/K (85/86), with an IC 50 of 8.28 nM. The reverse was found for the 13Z isomer of stelliferin riboside (72a) that revealed higher activity than its 3-O-deacetyl congener with IC 50 values of 0.22 and 2.40 nM, respectively [93].
On the other hand, globostellatic acids showed only moderate or no cytotoxicity against either human cervix carcinoma HeLa or rat pheochromocytoma PC-12 cell lines [93]. Two globostellatic acid X methyl esters (87 and 88), possessing the 13E-geometry, inhibited proliferation of human umbilical vein endothelial cells (HUVECs), 80-to 250-fold greater in comparison to several other cell lines and hence inhibiting angiogenesis which if pathologically uncontrolled, accompanies several diseases such as atherosclerosis, arthritis, diabetic retinopathy, and cancer. 13E,17E-Globostellatic acid X methyl ester (87) also inhibited basic fibroblast growth factor (bFGF)-induced tubular formation and vascular endothelial growth factor (VEGF)-induced migration of HUVECs. In addition, 87 induced apoptosis of HUVECs without affecting their VEGF-induced phosphorylation of ERK1/2 kinases [94].
Jaspolide B (109) arrested HL-60 cells in the G 2 /M phase of the cell cycle and induced apoptosis in a dose-and time-dependent manner. Jaspolide B with an IC 50 value of 0.61 µM exhibited a comparable efficacy as that of paclitaxel (IC 50 = 0.78 µM). These results suggested 109 to be a promising anticancer agent for chemotherapy of leukemia by prohibiting cell cycle progression at the G 2 /M phase and triggering apoptosis [103]. In a further study with human hepatoma cells, jaspolide B (109) inhibited the growth of Bel-7402 and HepG2 cells with IC 50 values of 29.1 and 29.5 µM, respectively. Incubation with 0.5 µM of 109 caused time-dependent induction of apoptosis in Bel-7402 as confirmed by the enhancement of mitochondrial masses, cell membrane permeability, and nuclear condensation. In conclusion, the anticancer effect of jaspolide B involves multiple mechanisms including apoptosis induction, cell cycle arrest, and microtubule disassembly but these were weaker than colchicine, a well-known microtubuledisassembly agent [104]. These multiple mechanisms of jaspolide B, especially the apoptosis induction, pose interesting perspectives for further exploration of the isomalabaricane-type terpenes as potential anticancer agents.
Since the class of isomalabaricane terpenoidal metabolites has been reported in the literature from different sponge species of the genera Rhabdastrella, Stelletta, Jaspis, and Geodia as shown above, the identity of these sponges has been questioned and reevaluated. Interestingly, the taxonomic reevaluation of these sponges revealed that they all might be reassigned to Rhabdastrella globostellata (class Demospongiae; order Astrophorida; family Ancorinidae) [80]. However, it could not be ascertained for the isomalabaricane producing Stelletta sp. from Somalia [74] and Stelletta tenuis from China [77]. The latter, collected from an identical location (Hainan Island), was taxonomically recognized as R. globostellata [75].
Eryloside A (118) was the first eryloside congener isolated from the Red Sea sponge Erylus lendenfeldi (class Demospongiae; order Choristida; family Geodiidae) [107]. Twenty eight additional erylosides (A-F, F 1 -F 7 , G-V) have been reported from different species of the genus Erylus including E. goffrilleri [109,114], E. formosus [110,113], E. nobilis [111], in addition to another collection of E. lendenfeldi [112]. For eryloside A (118), antitumor activity against murine leukemia P388 cells with an IC 50 = 5.7 µM and antifungal activity against Candida albicans (MIC = 21.1 µM) have been reported [107]. Eryloside E (119), glycosylated at C-30 through an ester linkage with the rare t-butyl substitution of the side chain, was isolated from an Atlantic sponge Erylus goffrilleri [109]. It revealed immunosuppressive activity with an EC 50 of 1.8 µM and a therapeutic index (TI) of 9.5, which indicated that the immunosuppressive effect is specific and is not due to a general cytotoxic effect [109]. Eryloside F (120) was reported from two collections of the marine sponge E. formosus [110] and exhibited potent thrombin receptor antagonistic activity. Furthermore, it inhibited platelet aggregation in vitro. Against hepytocyte HepG2 cells, 120 possessed little activity [110]. Erylosides F 1 (121) and F 3 (122) were isolated along with nine other congeners from the Caribbean sponge E. formosus [113]. In contrast to its 24-epimer, eryloside F 3 (122) induced early apoptosis in Ehrlich carcinoma cells at 130 µM, while erylosides F (120) and F 1 (121) activated the Ca 2+ influx into mouse spleenocytes at the same doses [113].
Formoside A (125) was first reported by Jaspars and Crews in 1994 from the Caribbean marine sponge Erylus formosus [115]. Later, it was isolated together with formoside B (126) from another collection of the same sponge from the Bahamas [116]. Formoside A (125) and its N-acetyl galactosamine derivative, formoside B (126) possess deterrent properties against predatory fish. Therefore, they were suggested to have important ecological functions, resembling those ascribed to similar compounds present in sea stars, sea cucumbers, and terrestrial plants [116]. Nobiloside (127), a penasterol saponin, was reported from the marine sponge E. nobilis collected off Shikine-jima Island, Japan [117] and revealed the presence of a carboxylic group at C-30 in addition to uronic acid moieties. Nobiloside (127) inhibited neuraminidase from the bacterium Clostridium perfrigens with an IC 50 of 0.5 µM [117].  Both sokodosides displayed moderate antifungal activity against the fungus Mortierella ramanniana and the yeast Saccharomyces cereivisiae, but no antibacterial activity was found. Additionally, sokodosides A (128) and B (129) exhibited cytotoxic activity against P388 cells with IC 50 values of 103 and 62 µM, respectively [118].
Sarasinosides follow erylosides in the number of isolated metabolites. To date, 21 sarasinoside congeners have been reported, which all featured a carbonyl group at C-23 position. Sarasinoside A 1 (130) was the first steroidal saponin reported in the literature, even before eryloside A (118), from the Palauan marine sponge Asteropus sarasinsum, together with other eight new congeners [120][121][122]. Then, from the same sponge collected in the Solomon Islands, four additional sarasinosides D-G were reported [123]. From each of the marine sponges Melophlus isis (Guam) [124] and M. sarassinorum (Indonesia) [125], four sarasinoside congeners were isolated.
In the agar diffusion antimicrobial assay (10 µg/disc), sarasinoside A 1 showed strong and selective activity against the yeast S. cerevisiae but was inactive against B. subtilis and E. coli. On the other hand, sarasinoside J (132) was active against S. cerevisiae and showed moderate antibacterial activity against B. subtilis and E. coli [125].
Mycaloside A (133) and G (134) as well as the total glycoside fraction did not influence nonfertilized eggs and the developing embryo up to the 8-blastomere stage at concentrations of up to 94.6 µM. However, these compounds were effective as spermatostatics when preincubated for 15 min with sea urchin sperm with an EC 50 of 3.04 µM. The total glycoside fraction generated a less toxic effect (EC 50 = 7.03 µg/mL) [127].   [129]. The compounds are C-4 norpenasterol triterpenoidal derivatives. Later, ectyoplasides were reisolated together with feroxosides A (137) and B (138) from the same sponge collected along the coasts of Grand Bahama Island [130]. Feroxosides have been shown to be unusual C-4 norlanostane triterpenes glycosylated with a rhamnose-containing tetrasaccharide chain. Against murine fibrosarcoma WEHI164, murine leukemia P388, and murine monocyte-macrophage J774 cell lines, both ectyoplasides (135 and 136) exhibited moderate in vitro cytotoxic activity with IC 50 values ranging from 9.0 to 11.4 µM [129], whilst against the latter cell line, feroxosides (137 and 138) were mildly cytotoxic (IC 50 = 17.6 µM) [130].

Future Aspects
The enormous diversity of marine natural products combined with improved global concerns to find new therapeutic agents for the treatment of different ailments provide the stimulus to evaluate marine natural products in clinical trials. Marine drug discovery faces many obstacles including a sufficient supply, and the low concentrations of some compounds that may account for less than 10 -6 % of the wet weight [135]. However, there have been substantial advances, suggesting that sustainable sourcing could be achievable. Since the continuous and exhaustive harvesting of terrestrial drug lead resources proved to be unreliable and resulted in the frequent re-isolation of known compounds, researchers from academia and from pharmaceutical companies alike are now turning their focus to the sea in search for new lead structures from nature. Nevertheless, the large scale production of marine natural products for clinical use is a real challenge, and therefore environmentally sound and economically feasible alternatives are required.
Chemical synthesis is among the first strategies to be explored, but unfortunately the structural complexity of marine metabolites with novel mechanisms of action and high selectivity has resulted in only a few successful examples with this strategy such as the conus toxin ziconotide [136]. A second strategy, but also as labor-intensive, is to study the pharmacological significance of marine natural product pharmacophores and then attempt to define the critical pharmacophore that can result in practical drugs based on a marine prototype via chemical synthesis, degradation, modification or a combination of these.
Aquaculture of the source organisms, including sponges, tunicates, and bryozoans, with an aim at securing a sustainable supply of the active constituent(s), has progressed notably in cancer applications. However, in most cases the biomass currently generated is still far from that required, should a marine-based drug finally enter the pharmaceutical market [137]. Furthermore, the cultivation of invertebrates in their natural environment is subject to several hazards and threats, such as destruction by storms or diseases. An intriguing strategy has been to identify the true producers of bioactive compounds and to explore whether or not they are of microbial origin including bacteria, cyanobacteria, or fungi that are known to harbour within the tissues of marine invertebrates.
If bacterial or other associated microorganisms proved to produce the compounds of interest, a careful design of special culture media would be crucial for large-scale fermentation e.g., ET-743 production. Currently, only 5% or less of the symbiotic bacteria present in marine specimens can be cultivated under standard conditions [138]. Consequently, molecular approaches offer particularly promising alternatives through the transfer of biosynthetic gene clusters to a vector suitable for largescale fermentation, thereby avoiding the obstacles in culturing symbiotic bacteria.
Oceans will play a potential role in the future to control and relieve the global disease burden. In spite of the substantial development that has been achieved in disclosing novel drug leads from marine resources, more efforts are still required for more chemical entities to reach to clinical applications.