Natural Products from the Lithistida: A Review of the Literature since 2000

Lithistid sponges are known to produce a diverse array of compounds ranging from polyketides, cyclic and linear peptides, alkaloids, pigments, lipids, and sterols. A majority of these structurally complex compounds have very potent and interesting biological activities. It has been a decade since a thorough review has been published that summarizes the literature on the natural products reported from this amazing sponge order. This review provides an update on the current taxonomic classification of the Lithistida, describes structures and biological activities of 131 new natural products, and discusses highlights from the total syntheses of 16 compounds from marine sponges of the Order Lithistida providing a compilation of the literature since the last review published in 2002.


Introduction
The Order Lithistida is a polyphyletic assemblage of sponges grouped together based on interlocking siliceous spicules called desmas that make up their skeleton [1,2]. The degree to which the desmas interlock result in lithistid sponges having a firm or rock-hard consistency [1,2]. Many lithistid families and genera have skeletal characteristics that suggest a closer phylogenetic relationship with other sponge taxa. For example, the family Corallistidae is also characterized by the presence of microscleres (amphiasters) that are similar to those found in the Order Astrophorida, Family Pachastrellidae. For many lithistid families and genera, however, skeletal similarities are not as OPEN ACCESS obvious, and until other characters (e.g., molecular data) can be evaluated and analyzed, these desma-bearing sponges will continue to be grouped in the Order Lithistida. A summary of the current taxonomic classification is shown in Figure 1.  [3].
Lithistid sponges occur world-wide in both shallow and deep water environments [4]. They are known to produce over 300 different interesting and diverse compounds comprising of polyketides, cyclic and linear peptides, alkaloids, pigments, lipids, and sterols [1,5,6]. Some of their compound diversity has been attributed to the symbiotic microorganisms that reside within the sponge [1,[7][8][9]. There have been a few excellent reviews published previously which have highlighted the compounds isolated from lithistid sponges through the year 2000 [1,5,6,10]. A recent review was published that highlighted a subset of bioactive lithistid compounds along with their mechanisms of action [10]. Since the last full review published in 2002, 131 compounds have been reported and substantial success has been achieved in the synthesis of compounds reported from the Lithistida. This review provides an update on the isolation of compounds reported from marine sponges of the Order Lithistida and highlights some of the total synthetic efforts reported since the last review published in 2002 [5].

Cyclic Peptides
The literature has been dominated by isolations of cyclic peptides from different species within the genus Theonella [11]. Since 2000, 51 new cyclic peptides have been reported from five different genera of lithistid sponges [11].
Theopapuamide A (1) is a cytotoxic undecapeptide isolated from Theonella swinhoei collected off Milne Bay, Papua New Guinea (Figure 2) [12]. It is the first natural peptide containing β-methoxyasparagine and 4-amino-5-methyl-2,3,5-trihydroxyhexanoic acid residues. It was tested in the CEM-TART (T-cells that express both HIV-1 tat and rev) and HCT-116 colorectal carcinoma cell lines with IC 50 values of 0.5 and 0.9 µM, respectively. In 2009, theopapuamide A (1) was reported along with six new cyclic peptides, theopapuamides B-D (2)(3)(4) and celebesides A-C (5-7), from an extract of Siliquariaspongia mirabilis collected off Sulawesi Island, Indonesia ( Figure 2) [13]. Compounds 2, 3, 5, and 7 were tested against HCT-116 cells giving IC 50 values of 2.5, 1. 3, 9.9, and >31 µM, respectively. The ability to inhibit HIV-1 entry was also evaluated for 2, 5 and 7 with IC 50 values of 0.5, 2.1, and >62 µM. Interestingly for celebesides A and C (5,7), in both biological assays, loss of activity correlated with the loss of the phosphate group. 1-3 were evaluated for their ability to inhibit the growth of both wild type and amphotericin B-resistant strains of Candida albicans. 1 inhibited the growth of both strains with zones of inhibition of 8 mm at 1 µg/disk while 2 and 3 displayed zones of 10 mm against both strains at 5 µg/disk. A specimen of Siliquariaspongia mirabilis collected off Nama Island, southeast of Chuuk Lagoon, in the Federated States of Micronesia yielded the mirabamides A-D (8)(9)(10)(11), which are potent inhibitors of HIV-1 entry (Figure 3) [14]. 8-11 were tested in an HIV-1 neutralization assay which tests a compound's ability to neutralize the biological effects of the HIV-1 virus on the TZM-bl cell line and used two different viral strains: HXB2 (T-cell tropic) and SF162 (macrophage-tropic). Against the HXB2 virus, IC 50 values for 8, 10, and 11 were 140, 140, and 189 nM and against the SF162 virus, 8, 10, and 11 were slightly less active with IC 50 values of 0.40, 1.01, and 1.31 µM. 8-11 were also tested in an HIV-1 fusion assay that tests the ability of a compound to inhibit envelope-mediated cell fusion against the LAV (T-cell tropic) viral strain. In the fusion assay, IC 50 values for 8, 10, and 11 were 0.041, 1.3, and 3.9 µM. 9 did not show inhibition in any of the antiviral assays under the conditions tested. 8, 10, and 11 were tested against the neutralization assay host cell line, TZM-bl showing IC 50 values of 1.8, 2.2, and 3.9 µM, respectively. 9 was tested against the HCT-116 cell line with an IC 50 value of 2.22 µM. Mirabamides E-H (12-15) along with 10 were recently isolated from a sponge, Stelleta clavosa, from the Order Astrophorida collected in the Torres Strait, Queensland, Australia ( Figure 3) [15]. 10 and 12-15 were tested in the HIV-1 neutralization assay using the viral strain YU2-V3 with IC 50 values of 123, 121, 62, 68, and 42 nM, respectively. Mirabamides A-D (8)(9)(10)(11) are very similar in structure to papuamide A (16) but differ by the presence of the novel 4-chlorohomoproline and β-methoxytyrosine 4′-Ο-α-L-rhamnopyranoside residues ( Figure 3) [16]. 9, which had no biological activity, contains one additional alteration in which the 2,3-diaminobutanoic acid residue is replaced with 2-amino-2-butenoic acid. Mirabamides E-H (12-15) differ from 8-11 by the replacement of the threonine residue with 2-amino-2-butenoic acid. Papuamide A (16) was also tested in the fusion assay and against the HCT-116 cell line with IC 50 values of 73 nM and 3.5 µM. The anti-HIV activity of 16 has recently been determined to be through a membrane targeting mechanism in which the hydrophobic tail of the molecule inserts into the viral membrane and the tyrosine residue interacts with cholesterol [17].  Barangamides A-D (17)(18)(19)(20) were isolated from a specimen of Theonella swinhoei collected at Baranglompo Island, Indonesia along with theonellapeptolide IIe (21) and a series of previously reported theonellapeptolides from the I and II series ( Figure 4) [18,19]. The amino acid sequence of 17 is the same as the cyclic portion of the theonellapeptolide II series but 17 lacks the amino acid side chain. In the barangamide series, macrocyclization occurs through peptide bond formation of the amine of the threonine rather than through lactonization of the hydroxyl group of threonine as found in the theonellapeptolide series. Because theonellapeptolides Ia-Ie were previously known to be moderately cytotoxic against the L1210 mouse lymphocytic leukemia cell line, 17 was tested but no cytotoxicity was observed at concentrations up to 9.4 μM [18][19][20]. Cyclic undecapeptides share structural similarities with the cyclosporins which are used as immunosuppresants after organ transplants [21]. The immunomodulatory activity of the known theonellapeptolides Ia, Id, and IId as well as 17 were analyzed in the mixed lymphocyte reaction (MLR) assay [20]. Barangamide A (17) showed no activity even at the highest concentration of 94 μM while theonellapeptolide IId showed the strongest immunosuppressive activity.
Nagahamide A (22) was isolated from Theonella swinhoei collected near Nagahama, Kamikoshiki-jima Island, Japan ( Figure 5) [22]. It was purified using bioassay-guided fractionation following anti-fungal activity. Once purified, 22 showed weak antibacterial activity against Escherichia coli and Staphylococcus aureus with 7 mm zones of inhibition when tested at 50 μg/disk but no antifungal activity was observed against Saccharomyces cerevisiae or Mortierella ramanniana at the same dose. Of the seven residues in 22, two were unusual: 8,10-dimethyl-9-hydroxy-7methoxytrideca-2,4-dienoic acid (DHMDA) and 4-amino-3-hydroxybutanoic acid (or gamma-aminobeta-hydroxybutyric acid, GABOB), which is also found in the cytotoxic, anti-fungal microsclerodermins [23,24].  Microsclerodermins A-E were discussed in earlier reviews yet further studies on Microscleroderma sp. from a deep-water specimen collected off Short Dropoff, Koror, Palau afforded microsclerodermins F-I (23-26) (Figures 5 and 6) [25]. 23-26 showed very similar cytotoxicity against the HCT-116 cell line with IC 50 values of 1.1, 1.2, 2.0, and 2.6 µM, respectively. They were also tested for the ability to inhibit the growth of C. albicans using a paper disk diffusion assay with a minimum concentration in which inhibition was observed of 1.5, 3, 12, and 25 mg/disk, respectively. 23-26 differ from previously published microsclerodermins by alterations in the ω-aromatic 3-amino-2,4,5trihydroxyacid residue and 24 and 26 also have a modification in the tryptophan moiety. The dehydromicrosclerodermins C-D (27)(28) were isolated as the major constituents from an Okinawan collection of Theonella cupola ( Figure 6) [26]. From the same family as Microscleroderma, specimens of Scleritoderma nodosum were collected from the northwest side of Olango Island, Cebu, Philippines and from Milne Bay, Papua New Guinea. These specimens yielded the cytotoxic cyclic peptide, scleritodermin A (29) (Figure 7) [27]. 29 was tested against the HCT-116, the HCT-116/VM46 multidrug-resistant colon cancer, the A2780 human ovarian carcinoma, and the SKBR3 breast carcinoma cell lines with IC 50 values of 1.9, 5.6, 0.940, and 0.670 µM, respectively. Cell cycle analysis in A2780 cells treated with scleritodermin A (29) for 24 h at a concentration of 1.3 µM yielded a G2/M block. As a G2/M block is characteristic of compounds that target tubulin, 29 was studied further and found to inhibit GTP-induced tubulin polymerization by 50% at a concentration of 10 µM. 29 caused a 5.5-fold increase in the induction of apoptosis over the control after a 24 h drug exposure at a concentration close to its cytotoxic IC 50 . Since scleritodermin A (29) had significant in vitro cytotoxicity in human tumor cell lines as well as an O-methyl-N-sulfoserine, a novel conjugated thiazole moiety, and an R-ketoamide group, its total synthesis was undertaken [28,29]. The initial structure was assigned based upon NOESY data as the 2Z, 4E configuration for the conjugated thiazole moiety. The structure was revised to the 2E, 4E configuration following synthesis based upon the observation of the methine protons, CH-3 and CH-5, 1.0 ppm further upfield than in the natural product [28,29]. In addition, a difference was observed for the chemical shifts of the two methyl groups in the keto-Ile moiety revising the 14R assignment to a 14S-configuration.
During an LC-MS screening study by the Crews group to predict which phenotypes of Theonella swinhoei contain swinholide A or motuporin, they isolated another series of compounds called the isomotuporins A-D (30-33) (Figure 7) [30][31][32][33][34][35]. The loss of the methoxy group in the 3-amino-9-methoxy-10-phenyl-2,6,8-trimethyldeca-4,6-dienoic acid (ADDA) residue in isomotuporin D (33) is the first report of that variation from a natural source. Most importantly, this study provides further assistance for researchers probing for the swinholide or motuporin biosynthetic pathways and also demonstrates that some populations produce either 2S-motuporin A or 2R-motuporin A (30). Homophymines A-E (34-38) and A1-E1 (39-43) are a series of cyclodepsipeptides isolated from Homophymia sp. collected from shallow waters off the east coast of New Caledonia ( Figure 8) [36,37]. They are similar in structure to the previously published antiviral marine cyclodepsipeptides, callipeltin A, neamphamide A, papuamides, theopapuamides (1)(2)(3)(4), and mirabamides (8)(9)(10)(11)(12)(13)(14)(15) [12][13][14][15][16]38,39]. 39-43 differ from 34-38 due to an amide moiety in place of the carboxylic acid on the 4-amino-2,3-dihydroxy-1,7-heptandioic acid residue. The anti-viral properties of 34 were tested in an assay with peripheral blood mononuclear cells (PBMC) infected with the III B strain of HIV-1. 34 had cytoprotective properties by inhibiting the production of an infection with an IC 50 value of 75 nM. Homophymine A (34) was cytotoxic against uninfected PBMC cells with an IC 50 of 1.19 µM but it was almost sixteen times more effective against infected cells. 34-43 were evaluated against a panel of cell lines including human cancer and the Vero green monkey kidney cell lines. 34-43 exhibited potent cytotoxicity with IC 50 values ranging from 2 to 100 nM. They were the most potent in the PC3 human prostate adenocarcinoma and the SK-OV3 human ovarian adenocarcinoma cell lines. Further studies were performed on 34-43 to determine if they were toxic or antiproliferative. They were found to undergo apoptosis through a caspase independent pathway but were ultimately determined to exert their toxicity through an acute direct and non specific mechanism.
Paltolides A-C (44-46) were isolated from a deep-water specimen of Theonella swinhoei collected off Uchelbeluu Reef in Palau ( Figure 9) [40]. These are anabaenopeptin-type compounds structurally similar to patented compounds isolated from an Australian sponge Melophlus sp. [41]. Compounds within the anabaenopeptin class contain an N-methylated amino acid adjacent to and before the C-terminal residue which is cyclized to the ε-amine of the lysine residue. 44 contains a standard leucine residue at this site and is the first report of an anabaenopeptin-type peptide lacking an N-methyl group at this site. These compounds are part of a rare subgroup of the anabaenopeptins since they contain a C-terminal tryptophan residue linked to the ε-amine of the N-terminal lysine residue. The other compounds within this subgroup are known to inhibit carboxypeptidase U. 44 and 45 did not have biological activity in the HIV-1 entry assay or against HCT-116 but their ability to inhibit carboxypeptidase U has not been evaluated.
in a dose-dependent manner in both early (0-6 h) and late (24-96 h) phases. 47 and 48 displayed a 60% and 46% reduction of edema at 0.3 mg/kg. Based on current NSAIDs on the market such as naproxen (ED 50 40 mg/kg), 47 is nearly 100 times more potent. Perthamides E (49) and F (50) were isolated from a Theonella swinhoei collected on a reef at a depth of 22 m from the western coast of Malaita Island, Solomon Islands ( Figure 10) [46]. 47 and 50 have a 3-amino-2-hydroxy-6methyloctanoic acid (AHMOA) residue in place of the 3-amino-2-hydroxy-6-methylheptanoic acid (AHMHA) residue. Based on the anti-inflammatory activity in the mouse models, 47-50 were evaluated for their antipsoriatic effects on TNF-α and IL-8 release using primary human keratinocytes (PHK) cells. Although 48 and 50 were too cytotoxic at concentrations up to 10 µM, 47 showed a dose-dependent response to inhibit the release of both TNFa and IL-8 and 49 significantly inhibited the release of IL-8 [46].

Linear Peptides
Miraziridine A (57) is a linear five amino acid peptide isolated from a specimen of Theonella aff. mirabilis collected off the Amami and Tokara Islands, Japan ( Figure 13) [49]. It contains the rare aziridine-2,3-dicarboxylic acid residue that has only been reported one other time from a Streptomyces sp. and also a vinylogous arginine residue that has never before been reported from a natural source. 57 inhibited the enzymatic activity of cathepsin B with an IC 50 value of 2.1 µM. The total synthesis of 57 has been completed (discussed later in this review [50,51]).

57
Koshikamide A 2 (58) was isolated from a Theonella sp. collected off Shimo-koshiki-jima Island, Kagoshima, Japan ( Figure 14) [52]. The structure of 58 was determined to be a close structural homolog of the previously described Koshikamide A 1 with an additional arginine residue added to the C-terminus. 58 was cytotoxic to P388 murine leukemia cell line with an IC 50 value of 4.6 µM whereas Koshikamide A 1 was more cytotoxic with an IC 50 value of 1.7 µM.  [42]. Their structures were determined by extensive NMR and mass spectrometry. 59 was observed to exist as two stable conformers due to cis/trans isomerization. They were tested in a single round HIV-1 infectivity assay against a CCR5-using viral envelope but no biological activity was observed. Polytheonamides A (62) and B (63) were isolated from a specimen of Theonella swinhoei collected off Hachijo-jima Island, Japan and initially reported in 1994 by the Fusetani group at Tokyo University ( Figure 16) [53,54]. During their studies to determine the configuration of 62 and 63, they realized the initially proposed structure was incorrect and determined the new structure by spectral and chemical methods, relying heavily on 2D NMR experiments [53]. 62 and 63 are 48 amino acid residue polypeptides with numerous unprecedented structural features such as the unusual N-terminal blocking group, the first report from a natural source of the 5,5-dimethyl-2-oxoheptanoyl group, the presence of eight rare t-Leu residues from a marine source, and the first report of the residue β,β-dimethylmethionine sulfoxide. 62 and 63 are isomeric at the sulfoxide moiety on the 44th residue. 62 and 63 are the largest non-ribosomal peptides with an alternating D/L stereochemistry throughout the chain seen only once before in gramicidin A, which is a linear 15-residue peptide produced by Bacillus brevis [55].

Polyketides and Macrolides
Hurghadolide A (64) and Swinholide I (65) were isolated from a specimen of Theonella swinhoei collected in Hurghada at the Egyptian Red Sea coast ( Figure 17) [56]. Hurghadolide A (64) is one acetate unit shorter than 65 and therefore has an unprecedented asymmetric 42-membered dilactone moiety which represents a new macrolide carbon skeleton. Swinholide I (65) is similar in structure to swinholide A and is the first swinholide derivative with hydroxylation on the side chain [57].  In 2005, the Gerwick group reported the isolation of swinholide A (initially isolated in 1985 from the sponge Theonella swinhoei) and two glycosylated derivatives, ankaraholide A (67) and B (68), for the first time from field collections of two species of cyanobacteria ( Figure 18) [7]. Specimens of Symploca cf. sp. that produced swinholide A were collected from the Fiji Islands while specimens of Geitlerinema sp. collected from Nosy Mitso-ankaraha Island, Madagascar produced 67 and 68. This discovery as well as the fact that the swinholides are produced by three taxonomically unrelated sponges suggests that symbiotic microorganisms may be the true producers of these metabolites. Since the swinholides cause cytotoxicity by disrupting the actin cytoskeleton, the effects of the added sugar moiety were evaluated. 67 caused cytotoxicity in NCI-H460 human large cell lung cancer, Neuro-2a mouse brain neuroblast, and MDA-MB-435 human melanoma cell lines with IC 50 values of 119, 262, and 8.9 nM, respectively. In A-10 aortic smooth muscle cells, 68 caused the complete loss of filamentous actin at 30 and 60 nM which is consistent with the disruption of the actin cytoskeleton. A new calyculinamide-related congener called swinhoeiamide A (69) was isolated from Theonella swinhoei collected off the coast of the Karkar Island, Papua New Guinea (Figure 19) [59]. The structure of the new compound was assigned on the basis of 1D and 2D NMR spectroscopy and HRMS data. 69 differs from calyculinamide A by replacement of the complex side chain attached to C-29 in the oxazole ring system with a methyl substituent and 69 is also the first calyculin congener in which the double bond at C-4 and C-5 is hydrogenated [60]. 69 exhibited insecticidal activity toward the neonate larvae of the insect Spodoptera littoralis in a chronic feeding bioassay with an effective dose, 50% (ED 50 ) value of 2.11 ppm and a median lethal dose (LD 50 ) value of 2.98 ppm. 69 was found to inhibit the growth of C. albicans and Aspergillus fumigatus with MIC values of 1.2 and 1.0 µg/mL, respectively. 69 also exhibited dose-dependent cytotoxicity against various undisclosed cell lines and tissues with IC 50 values ranging between 20 and 90 ng/mL. Another calyculin A derivative, hemicalyculin A (70), was isolated from Discodermia calyx collected off Sikine-jima Island, Japan ( Figure 19) [61]. The structure of 70 is comprised of just the southern hemisphere of calyculin A and allowed the Fusetani group to pursue structure-activity relationships providing further insight into the binding of calyculin A with protein phosphatases 1 and 2A (PP1 and PP2A) [61][62][63][64]. 70 was tested alongside calyculin A to determine its ability to inhibit PP1 and PP2A with IC 50 Figure 20) [65]. The structure of 71 was confirmed by a single-crystal X-ray diffraction study. Bitungolides are structurally similar to pironetin reported previously from a Streptomyces sp. [66]. Pironetin shows moderate in vivo anti-tumor activity and arrests cells at the M-phase of the cell cycle [67]. 71-76 were tested against a number of phosphatases and showed no activity against protein tyrosine phosphatase-S2 (PTP-S2), PP1, or PP2A. 71-76 showed weak activity against dual-specificity protein phosphatase vaccinia H1-related (VHR). The total synthesis of bitungolide F has been completed and will be discussed in a later section [68]. Recently, a total synthesis of the proposed structure for leiodolide B (78) was published, but its spectroscopic data did not match those of the authentic sample [70]. This will be discussed in a later section.   [72,73]. This data indicated that alterations in the δ-lactone ring have a minor contribution toward the activity but changes at the diene end of the molecule had no significant decrease in activity. Complete loss of activity was observed when alterations were made in the middle section of the molecule. This data was in agreement with earlier findings that the C-7 through C-17 backbone, which adopts a hairpin conformation, is essential to the cytotoxicity of discodermolide [74,75]. is structurally homologous to the potent antiproliferative compound leucascandrolide A and the cytochrome bc 1 complex was found to be the primary cellular target of both compounds [77]. Based on its potent biological activity and similarity to leucascandrolide A, 85 was targeted by numerous synthetic groups to complete its total synthesis and absolute configuration. The relative configuration of 85 was suggested as 11R, 13R based on coupling constant analysis, 2D-NOESY, and a series of double-pulsed field gradient spin echo (DPFGSE) NOE experiments but once it was synthesized, its configuration was reassigned to 11S, 13S [78,79]. The total synthesis of 85 will be discussed in a later section.
Mirabalin (86), initially reported as mirabilin, was isolated from Siliquariaspongia mirabilis collected southeast of Chuuk lagoon in the Federated States of Micronesia ( Figure 24) [80,81]. 86 is characterized by the presence of a 35-membered macrolide lactam ring bearing a pentadiene conjugated system, a tetra-substituted tetrahydropyran ring, and a linear polyketide moiety attached to the macrolide ring via an amide linkage. 86 inhibited the growth of the HCT-116 cell line with an IC 50 value of 0.27 µM and was not cytotoxic to several other cell lines tested. 86 is the first macrolide of the chondropsin family with a conjugated pentadiene and a tetrasubstitued tetrahydropyran ring as well as being the first such macrolide to be isolated from a member of the Theonellidae family. Mirabalin was difficult to work with as it was unstable in ambient light and degraded within 3-4 h. Dictyostatin-1 (87) was initially isolated in 1994 from a marine sponge in the genus Spongia sp. but in 2003, it was isolated from a Corallistidae (now revised to be Neopeltidae) collected using the Johnson-Sea-Link submersible at a depth of 442 m off the north coast of Jamaica ( Figure 25) [82,83]. In initial studies, dictyostatin-1 arrested cells in the G2/M phase of the cell cycle [82]. 87 has been shown to be a potent promoter of tubulin assembly similar to paclitaxel and discodermolide (133) [84]. It has recently been synthesized and is also being used in work by the Paterson and Curran/Day Groups to make discodermolide-dictyostatin analogs [85][86][87][88][89][90][91].
Theopederin F-J (88-92) were isolated from T. swinhoei collected off the Kerama Islands, Ryukyu Archipelago, Japan ( Figure 25) [92]. 88 inhibited the growth of wild type S. cerevisiae with an 11 mm inhibitory zone at 10 pg/disk and against the erg6 mutant with a 12 mm inhibitory zone at 1 µg/disk. 88 was also found to be cytotoxic to the P388 cell line with an IC 50 [94,95].
Four new analogs of onnamide A named 21,22-dihydroxyonnamides A 1 -A 4 (95-98) were isolated from a polar fraction of Theonella swinhoei collected off Okinawa, Japan ( Figure 26) [96]. Due to isomerism around the diol, they could only be separated after conversion to isopropylidene derivatives. A 2:2:2:1 isomeric mixture inhibited the P388 cell line with an IC 50 value of 46 nM, but each one alone did not have the same cytotoxicity. Recently, the Piel group at the University of Bonn, Germany isolated the biosynthetic gene cluster for onnamide A from the metagenome of Theonella swinhoei [9]. Their research showed that onnamide A was clearly produced by a bacterial symbiont and not the host sponge [9].

Sterols, Lipids, and Fatty Acids
Lysoplasmanylinositols 1 (99) and 2 (100) were isolated from a collection of T. swinhoei with a white interior off Hachijo-jima Island, Japan ( Figure 27) [97]. 100 inhibited the growth of E. coli at 50 µg/disk with a 12 mm inhibitory zone but activity was only observed for 99 through bioautography. Related lysophosphatidylinositols have been previously isolated from the ascidian Halocynthia roretzi as antifungal constituents but this is the first report of lysoplasmanylinositols isolated from a marine organism [98]. Calyceramides A-C (101-103) were isolated from Discodermia calyx collected off Sikine-jima Island, Japan ( Figure 28) [99]. They are sulfated ceramides that inhibit neuraminidase with IC 50 values of 0.63, 0.32, and 1.3 µM, respectively. They are very closely related to the ceramide 1-sulfates, which are inhibitors of DNA topoisomerase I isolated from the bryozoan Watersipora cucullata [100]. Azoricasterol (105) is the first metabolite isolated from Macandrewia azorica ( Figure 29) [103]. The specimen was collected from a depth of 600 m off the flanks of the Gettysburg and Ormonde Sea Mount in the North Atlantic by benthic dredging. 105 has an unusual side chain that has two additional methyl groups and a quaternary center forming a rare elongated side chain for sterols.  Aurantosides G-I (113)(114)(115) were isolated from T. swinhoei collected in Milne Bay, Papua New Guinea ( Figure 31) [105]. They are similar to aurantoside A but have one less chlorinated methylene unit in the polyene side chain [106]. Also, while aurantoside A is a trisaccharide, 113 is a monosaccharide and 114 is a disaccharide. 113-115 did not inhibit the HCT-116 colorectal carcinoma cell line when tested at 152, 124, and 103 µM, respectively. They were also inactive in an anti-HIV assay when tested at concentrations of 20, 15, and 12.8 µM, respectively. Aurantoic acid (116) is a chlorinated polyene moiety that is evident in the structures of the 113-115 ( Figure 31) [30]. 116 was isolated from T. swinhoei collected at a depth of 20-50 m in Bunaken Marine Park, North Sulawesi, Indonesia along with dehydroconicasterol (120, discussed later). 116 showed minimal inhibition when tested at a concentration of 70 µM against C6 glioma, HeLa epithelial carcinoma, and H9c2 cardiac myoblast cell lines. Sponges of the genus Theonella are known to produce biosynthetically unique 4-methylene sterols and these may be ideal taxonomic markers [107]. Recently, a series of conicasterol and theonellasterol derivatives (117)(118)(119)(120)(121)(122)(123)(124)(125)(126)(127)(128)(129)(130) were isolated from Theonella swinhoei collected from Yongxing Island in the South China Sea, at Panglao Island, Bohol, Philippines, on the barrier reef of Vangunu and Malaita Island, Solomon Islands, and at Bunaken Marine Park, North Sulawesi, Indonesia ( Figure 32) [30,[107][108][109]. 9α-hydroxy-15-oxoconicasterol (117) and 8β-hydroxy-B-norconicasta-6α-aldehyde (118) have novel hydroxyl substitution at either the C-9 or C-8 position and 118 has a B-nor-framework [107]. 7α-hydroxytheonellasterol (119) has an additional hydroxyl group at C-7 and an ethyl group at C-24 when compared to the known compounds theonellasterol and swinhosterol C [110,111]. Dehydroconicasterol (120) was isolated alongside aurantoic acid (116) and showed minimal inhibition when tested at a concentration of 70 µM in C6, HeLa, and H9c2 cell lines [30]. The biological activity of 117 and 118 has not evaluated but 119 showed cytotoxicity against an undisclosed panel of eight cell lines with a mean IC 50 value of 29.5 µM [109]. Theonellasterols B-H (121-127) and conicasterols B-D (128)(129)(130) were evaluated for their ability to affect the nuclear receptors, pregnane-X-receptor (PXR) and farnesoid-X-receptor (FXR) [108]. Using a HepG2 human hepatoma reporter cell line transfected with FXR or PXR, 126 partially activated FXR at 10 µM and 121, 122, 124, and 126-130 were effective antagonists of FXR transactivation by chenodeoxycholic acid. All of the compounds (121)(122)(123)(124)(125)(126)(127)(128)(129)(130) were agonists of PXR at 10 µM. 126 is one of the first natural products that modulates FXR and is also a ligand for PXR. A molecular docking study was conducted for 121-128 and 130 to determine structure-activity relationships [108].  Figure 33) [112]. Its unique 11, 12, 13, 14-bis-secosteroid structure was determined by extensive spectroscopic analysis and DFT 13 C calculations. It is structurally similar to the other known 4-methylene sterols 117-130 but its bis-secosteroid structure has not been observed from natural sources. It was found to be a potent inducer of PXR transactivation at a concentration of 10 µM with no effect on FXR. Its PXR activity was confirmed by an increase in the gene expression of downstream PXR targets as well as in molecular docking studies.

Alkaloids
Collections of Discodermia polydiscus from the north point of the Acklins Island, Bahamas and off Grand Bahama Island yielded the discodermindole analog; 6-hydroxydiscodermindole (132) in trace amounts ( Figure 34) [113]. 132 inhibited the proliferation of the P388 cell line with an IC 50 value of 12.4 µM. Figure 34. The discodermindole analog, 6-hydroxydiscodermindole (132), was isolated from Bahamian collections of Discodermia polydiscus.

Recent Total Synthesis of Lithistid Natural Products
The total synthesis of a number of lithistid derived compounds has been achieved over the past decade. Some of these syntheses resulted in the reassignment of the initially proposed relative configurations while some allowed for the assignment of absolute configurations.
Discodermolide (133) inhibits cell proliferation by polymerizing and hyperstabilizing tubulin similar to paclitaxel but shows activity against paclitaxel resistant tumors ( Figure 35). A number of syntheses have been previously reviewed for 133 but the large-scale synthesis accomplished during this decade was very important to provide enough of the compound for clinical trials [5,73,[114][115][116]. Despite the other microtubule-stabilizing agents discovered, 133 is the most potent natural promoter of tubulin assembly discovered to date [116,117]. Though clinical trials have been halted, more recent research has shown that discodermolide and paclitaxel have a synergistic effect when used in combination [118]. Some excellent reviews have been published that discuss the different approaches used in the synthesis of discodermolide [85,116,119].

Figure 35.
A multi-gram total synthesis of discodermolide (133) was completed in order for Novartis to begin Phase I clinical trials.
Callipeltoside A (134) was isolated from a specimen of Callipelta sp. collected off New Caledonia in 1996 ( Figure 36) [128]. 134 inhibited the proliferation of the KB and P388 cell lines and, in the NSCLC-N6 non-small-cell bronchopulmonary carcinoma cell line, 134 arrested cells in the G1 phase. Based on its stereochemical and structural complexity coupled with its unique biological activity, members of the callipeltoside family became attractive targets for total synthesis. The first total synthesis of 134 was a convergent synthesis that focused on making three main pieces: the macrolide, the side chain, and the sugar with an overall yield of 0.47% in 50 steps [129]. Other syntheses have been published by the Evans group at Harvard University, the Panek group at Boston University, and the Paterson group [130][131][132]. Their different synthetic approaches increased the overall yield to a maximum of 4.8% with a longest linear sequence of 23 steps [132]. Callipeltoside C (135) was recently synthesized resulting in the revision of the absolute configuration around its sugar moiety ( Figure 36) [133,134]. It was completed with a longest linear sequence of 20 steps in 11% overall yield. Callipeltin B (136) is an eight-residue cyclic depsipeptide isolated from Callipelta sp. collected off New Caledonia ( Figure 37) [135]. 136 was found to inhibit the proliferation of the KB, P388, and NSCLC-N6 cell lines. It is structurally similar to callipeltin A whose structure was revised after its original publication although its total synthesis has not been completed [39,136,137]. The total synthesis of 136 was completed using a solid-phase support with a 15% overall yield. By comparing data with the original proposed structure, the configuration of 136 was reassigned from 18S, 21S to 18R, 21R. A series of callipeltin analogs have since been synthesized and analyzed for their ability to inhibit the HeLa cervical adenocarcinoma cell line [138].
Reidispongiolide A (137) is a polyketide containing a 26-member macrolide ring which was isolated from a specimen of Reidispongia coerulea collected off New Caledonia ( Figure 38) [139,140]. It is a member of the sphinxolide/reidispongiolide class of macrolides which are members of an emerging class of actin-binding cytotoxic macrolides [139]. The absolute configuration of reidispongiolide A had not been determined; therefore the Paterson group carried out the total synthesis defining it as shown in 137. They completed the total synthesis with a longest linear sequence of 25 steps in a 1.1% overall yield [141].  Neopeltolide (85) has received significant attention from synthetic chemists since its publication in 2007 due to its interesting structural aspects and potent biological activity ( Figure 24) [76]. The first total synthesis of 85 was completed by the Panek group and had a longest linear sequence of 19 steps with an overall yield of 1.3% [78]. It was determined from this initial synthesis that the stereochemistry, which was originally proposed as 11R, 13R, is actually 11S, 13S. A number of syntheses followed and all were in agreement with the Panek group's initial observation [77][78][79][142][143][144]. The highest yielding synthesis to date was reported by the Fuwa/Sasaki group at Tohuku University which prepared 85 with the longest linear sequence of 25 steps and an 8.3% overall yield [142]. Synthetic neopeltolide analogs have been tested against a variety of human, murine, and bovine cell lines to obtain valuable SAR information [145].
Papuamide B (138) is a 22-membered cyclic depsipeptide connected to a complex linear tetrapeptide via an amide linkage ( Figure 39) [16]. Both papuamide A (16) and B (138) exhibit a potent inhibitory effect on the infection of human T-lymphoblastoid cells by HIV-1 RF with an EC 50 value of 4 ng/mL. The total synthesis of 138 was completed by the Ma Group at the Chinese Academy of Sciences and suggested a revision of the 2,3-diaminobutanoic acid in the side chain from 2S, 3R to 2S, 3S [146]. The papuamides are structural analogs of the mirabamides (8)(9)(10)(11)(12)(13)(14)(15) and have very similar biological activities. This synthetic scheme is amenable to complete the total synthesis of closely related compounds as well as determine SAR information. Figure 39. The absolute configuration of papuamide B (138) was revised based on the total synthesis. Superstolide A (139) is a macrolide isolated from Neosiphonia superstes collected off New Caledonia ( Figure 40) [147]. It is highly cytotoxic against several cancer cell lines including the P388, human nasopharyngeal KB, and NSCLC-N6-L16 non-small cell lung carcinoma cell lines with IC 50 values of 5, 8, and 6 nM, respectively [147]. Due to their interesting chemical structures and potent biological properties, the superstolides have attracted considerable attention as synthetic targets. The Roush group at Scripps Florida reported the total synthesis in 2008 [148]. Their synthetic route was likely biomimetic in nature and utilized an intramolecular Suzuki coupling reaction and a highly stereoselective transannular Diels-Alder cycloaddition [148]. Microsclerodermin E (140) and its series of cyclic peptides were isolated by Faulkner and co-workers at Scripps Institute of Oceanography ( Figure 40) [24]. Structurally, each of them contains several unique features, including a 23-membered cyclic hexapeptide core, featuring a very complex β-amino acid and three other unnatural amino acid residues. The Ma group at the Chinese Academy of Sciences reported the first total synthesis of a member of the microsclerodermin group with a 1% overall yield in 26 linear steps [149].
Scleritodermin A (29) was recently isolated from Scleritoderma nodosum (Figure 7) [27]. It has an interesting structure and was found to have potent biological activities as well as cause cell cycle arrest in the G2/M phase. The total synthesis was undertaken by the Nan group at the Chinese Academy of Sciences and the structure was revised from 2Z, 4E and 14R to 2E, 4E and 14S [28].
Cyclotheonamides E2 (141) and E3 (142) are cyclic pentapeptides found to be potent inhibitors of serine proteases ( Figure 41) [150]. Because of their biological properties and unusual structural features, further research was done on their mode of enzyme inhibition as well as their total synthesis. 141 and 142 are structurally similar to cyclotheonamides A and B, both of which have been synthesized previously [151][152][153]. The differences are a D-alloisoleucine in place of D-phenylalanine as well as the addition of the benzoylalanine and isovalerylalanine side chain [151]. The Wasserman group at Yale University synthesized 141 and 142 using a cyano ylide method [154]. Miraziridine A (57) is structurally similar to known compounds that are inhibitors of trypsin-like serine proteases, papain-like cysteine proteases, and pepsin-like aspartyl proteases ( Figure 13) [49]. The total synthesis of 57 was completed by Schaschke at the Max-Planck-Institut für Biochemie in order to use 57 as a blueprint to efficiently design small molecule protease inhibitors [51]. The Konno group at the Kyoto Prefectural University of Medicine produced 57 as well as a number of analogs to provide important SAR information [50]. They determined that the N-terminal aziridine is important for cathepsin B inhibition.
Leiodolide B (78) was recently isolated from a deep-water sponge Leiodermatium sp. by the Fenical group ( Figure 21) [69]. Since only a very small amount (0.8 mg) could be isolated from the natural source, its structure was determined based on comparisions with leiodolide A (77) and a proposed biosynthetic closure of the brominated tetrahydrofuran ring [69]. The Fürstner group at the Max-Planck-Institut für Kohlenforschung prepared four isomers of 78 yet none of them matched the NMR data of the natural product [70]. Further work is being performed by the Fürstner group to synthesize leiodolides A and B.
Bitungolide F (76) was isolated from an Indonesian collection of Theonella cf. swinhoei ( Figure 20) [65]. 76 is structurally similar to pironetin, a compound isolated from Streptomyces sp. which arrests cell at the M-phase of the cell cycle [65]. The structure of bitungolide A (71) had been confirmed by X-ray and the structure of 76 was based on comparison with 71. The She group from Lanzhou University, China completed the total synthesis of 76 in 17 steps with a yield of 20.1% [68].

Conclusion
From the unique activities and amazing diversity of the compounds isolated just over the past decade, sponges from the Order Lithistida will almost certainly continue to be a significant source of interesting biologically active compounds. Recent research has shown that some of the compounds isolated from sponges within this order are likely produced by the microorganisms residing within them and work is ongoing to culture these microbes [1,7,67]. A future area of research will be to understand what unique characteristics of lithistid sponges leads to such a great diversity of microbes and ultimately such a diverse array of compounds.