Novel Marine Secondary Metabolites Worthy of Development as Anticancer Agents: A Review

Secondary metabolites from marine sources have a wide range of biological activity. Marine natural products are promising candidates for lead pharmacological compounds to treat diseases that plague humans, including cancer. Cancer is a life-threatening disorder that has been difficult to overcome. It is a long-term illness that affects both young and old people. In recent years, significant attempts have been made to identify new anticancer drugs, as the existing drugs have been useless due to resistance of the malignant cells. Natural products derived from marine sources have been tested for their anticancer activity using a variety of cancer cell lines derived from humans and other sources, some of which have already been approved for clinical use, while some others are still being tested. These compounds can assault cancer cells via a variety of mechanisms, but certain cancer cells are resistant to them. As a result, the goal of this review was to look into the anticancer potential of marine natural products or their derivatives that were isolated from January 2019 to March 2020, in cancer cell lines, with a focus on the class and type of isolated compounds, source and location of isolation, cancer cell line type, and potency (IC50 values) of the isolated compounds that could be a guide for drug development.


Introduction
For decades, marine natural products have been sources of bioactive compounds with remarkable biological activities. These include anticancer, antimicrobial, antidiabetic, antiinflammatory, antiparasitic, antihypertensive, and antioxidant activities. They can be traced of the microtubules. Additionally, podophyllotoxin is an antineoplastic that has a similar mechanism of action to Vinca alkaloids by binding tubulin and preventing microtubule assembly [24]. The taxanes prevent microtubule disassembly with paclitaxel acting at the G2-M border whereas docetaxel acts at S-phase. All these drugs by their activities on the microtubules prevent the cancer cells from completing mitosis and they equally have an effect on angiogenesis which is an important process that promotes tumor growth and metastasis [25].
For clarity purposes, the drugs described above are used for general treatment, and are not molecules developed from the marine sources being reviewed.
Anticancer compounds isolated from terrestrial and marine sources have a variety of mode of action for inhibiting cancer cell proliferation and/or by inducing apoptotic cell death. Dolastatin derived from Dolabella auricularia (a shell-less mollusk), for example, prevented cancer cells from entering the metaphase stage and triggered apoptosis in lymphoma cells [28]. By inhibiting Hsp90 in the HL60 cancer cell line, diterpene 5-episinuleptolide acetate, a non-cembranoid derived from Sinularia sp., triggered downstream apoptosis [29]. Flavonoids, tannins, and curcumins are polyphenolic chemicals with anticancer action [30]. Polyphenols also have antioxidant activity and are known for their apoptosis-inducing potential which they initiate by regulating the mobilization of copper ions which are bound to chromatin, causing DNA fragmentation [31]. Curcumin, a polyphenol, suppressed tumor necrosis factor (TNF) when incorporated into cancer cells in various cells lines through interaction with various stimuli [32]. Flavonoid compounds have been shown to cause cancer cell apoptosis via intrinsic and extrinsic signaling pathways, lowering mitochondrial membrane potential and inhibiting the expression of NF-κB needed for cancer cell survival, angiogenesis, and proliferation [33].
The desired goal of anticancer drugs or chemotherapeutic agents is to selectively target the cancerous cells while sparing normal cells. Unfortunately, this ideal situation is far from reality because both normal and cancerous cells share the same metabolic processes, thus, normal cells are damaged. As a result, anticancer agents are not free from toxicity. The normal cells prone to damage by chemotherapy are blood-producing cells, hair follicles, and cells in the oral cavity, alimentary canal, and reproductive system [35]. The toxic effects of these drugs are numerous and include fatigue, peripheral neuropathy, nephrotoxicity, sexual dysfunction, diarrhea, and bone marrow depression, among others [35].
Development of a drug with potential anticancer activity involves preclinical screening of the product to determine its cytotoxicity on cancer cell lines. Cancer cell lines are necessary tools for cancer research, and provide a handy, cost-effective model for assessing cellular behavior and biological response. However, selection of a product as a cytotoxic substance involves its ability in inhibiting the proliferation of or killing 50% of the population of cancer cells (the IC 50 ). In other words, a candidate drug that could inhibit the proliferation of or kill 50% of the population of cancer cells at a lower concentration is considered highly potent, and this relates to its intrinsic ability. Furthermore, any drug that can achieve its maximum efficacy at a much lower dose is a candidate worthy of optimization and development. Therefore, potency is a factor that is considered during preclinical screening of potential anticancer agents. From a pharmacological point of view, it is a measure of drug action expressed in relation to the amount required to produce an effect of a given intensity [36]. Drugs with very high potency produce responses at Molecules 2021, 26, 5769 5 of 28 low concentrations while the less potent drugs evoke the same amount of response at very high doses. Sometimes, a less potent drug could predispose to more adverse effects than a highly potent drug because of the extended higher dose required to achieve the desired response.

Methods
In this review, we examined papers published (in the English language only) from January 2019 to March 2020 on marine natural products or derivatives, structurally elucidated and tested for anticancer activity. The articles used were found by searching PubMed/MEDLINE and Google Scholar using Boolean operators (AND, OR, NOT) and a combination of related terms such as marine natural products or marine-derived compounds with anticancer or cytotoxic activity. Additionally, the search criteria were expanded using cross-referencing and "related articles" functions. A total of 476 articles' abstracts were examined and those articles without isolated compounds and those with isolated compounds but not from marine sources were excluded, bringing down the total number to 195 articles. Further screening to exclude the compounds without or with less potent (IC 50 > 20 µg/M) anticancer activity was carried out, thus, 76 articles were included in the review. From the 76 articles, we considered the IC 50 values (≤20 µM) of 731 isolated compounds/derivatives comprising a total of 331 novel and 400 known compounds. The cytotoxic activities of these compounds were screened against a total of 120 cancer cell lines by using these assay methods: 3-(4, 5-dimethylthiazol-2-y1)-2, 5-diphenyltetrazolium bromide (MTT), Cell Counting Kit-8 (CCK-8) CellTiter-Glo (CTG), and sulforhodamide-b (SRB) assays.
Some compounds were tested against more than one cancer cell line. However, the isolated compounds with an IC 50 ≤ 20 µM are in the supplementary section. The structures of the novel compounds that inhibited the proliferation of some cancer cell lines with IC 50 values ≤ 1.0 µM are shown in the text. The IC 50 values were chosen to enable the exclusion of compounds with less potent antiproliferative activity.

Marine-Derived Alkaloids Cytotoxic to Cancer Cell Lines
A novel compound, subereamolline D (Cpd. 1; Figure 1), a bromotyrosine-derived alkaloid isolated from a sea sponge Fascaplysinopis reticulate from the Xisha Islands, south China, inhibited the growth of the Jurkat cell line (IC 50 of 0.88 µM) [37].
The following three new azaphilone alkaloids that contain glutamine residues, N-glutarylchaetoviridins A, B, and C (Cpds. 2-4), and Cpds. 5 and 6, which are related, were isolated from Chaetomium globosum HDN151398 fungal extract. N-glutarylchaetoviridins A-C (Cpds. 2-4) represent the first compounds with glutamate residues. The incubation of the fungal strain with amino acids produced five azaphilone-loaded diverse amino acid residues (Cpds 7-11). Thus, this method can improve the structurally diverse nature of this strain by culturing with amino acids. Unfortunately, none of these compounds exhibited a potent cytotoxic activity (IC 50 5.7-19.4 µM) against a panel of eleven human cancer cell lines that were screened (Table S1) when compared with a standard, adriamycin, with the range 0.1-0.6 µM IC 50 [38].
Following the LC-MS/MS molecular networking-based metabolomics and cytotoxic activity-guided study, two new discorhabdin-type alkaloids, tridiscorhabdin (Cpd. 45) and didiscorhabdin (Cpd. 46), were isolated from Latrunculia biformis, a sponge from the Weddell Sea (Antarctica). As novel compounds, a unique C-N bridge (C-1/N-13) between discorhabdin monomers was present in them, with Cpd. 45 being the first trimeric discorhabdin molecule isolated from the natural environment. Additionally, this compound was cytotoxic to human colon cancer cell line HCT 116 at 0.31 µM IC 50 . It equally inhibited the non-cancerous human keratinocyte cell line HaCaT with a similar IC 50 value (IC 50 = 0.92 µM), thus, suggesting its general toxicity and low selectivity for cancer cells; the structure is in Figure 2 [43].
Following the LC-MS/MS molecular networking-based metabolomics and cytotoxic activity-guided study, two new discorhabdin-type alkaloids, tridiscorhabdin (Cpd. 45) and didiscorhabdin (Cpd. 46), were isolated from Latrunculia biformis, a sponge from the Weddell Sea (Antarctica). As novel compounds, a unique C-N bridge (C-1/N-13) between discorhabdin monomers was present in them, with Cpd. 45 being the first trimeric discorhabdin molecule isolated from the natural environment. Additionally, this compound was cytotoxic to human colon cancer cell line HCT 116 at 0.31 µ M IC50. It equally inhibited the non-cancerous human keratinocyte cell line HaCaT with a similar IC50 value (IC50 = 0.92 µM), thus, suggesting its general toxicity and low selectivity for cancer cells; the structure is in Figure 2 [43].  Their cytotoxic activities can be compared to that of sorafenib (IC 50 = 8.2), the positive control, although compound 10 was more potent. The disulfide bridge at C-2/C-2' seems to be crucial for the anticancer activity [44].
Pityriacitrin is an alkaloid of marine origin, with a typical beta-carboline scaffold. It has diverse biological functions and was isolated from Chinese Burkholderia sp. NBF227. A series of novel beta-carboline analogues, 9a-p (Cpds. 69-84) derived from pityriacitrin, were tested for anticancer activity against the following human cancer cell lines: SGC-7901, A875, HepG2, and MARC145. Some of these beta-carboline derivatives exhibited moderate to high cytotoxic activities. Compound 9o (Cpd. 83) with a sulfonyl group had the highest inhibitory activity against all the cell lines with the IC 50 values of 6.82, 8.43, 7.69, and 7.19 µM, respectively. The substitution of the amide scaffold attached on the compounds was carried out with the following substituents: a phenylethylamine for compounds 9a-f (Cpds. 69-74), or a benzylamine for compounds 9g-l (Cpds. 75-80), or an arylamine for compounds 9m-n (Cpds. 81, 82), and that did not yield analogues more active than compound 9o (Cpd. 83), that had a sulfonyl substituent, and compound 9p (Cpd. 84) with an amino acid unit. Therefore, this suggested that sulfonyl group attachment on the beta-carboline scaffold is responsible for cytotoxic activity. Interestingly, some of these beta-carboline derivatives, especially 9e (Cpd.  (Table S1) [46].
Three new cembranolides (Cpds. 160-162) and six known cembranolide diterpenoids (Cpds. 163-168) were isolated from an Okinawan soft coral Lobophytum sp. The new compounds were screened for anticancer activity against HeLa, A459, B16-F10, and RAW 264.7 cells and anti-inflammatory effect in LPS-stimulated inflammatory RAW 264.7 macrophage cells. The result revealed that only cembranolide diterpene derivative 1 (Cpd. 160) was antiproliferative against the cancer cells below 20 µM IC 50 (Table S2) and anti-inflammatory (IC 50 = 7.75 µM) by the inhibition of nitric oxide production in the LPS-induced RAW 264.7 macrophage cells, more than the other compounds. Further studies on structural elucidation revealed all the compounds contain an alpha-methylene-gamma-lactone ring adjacent to a cembrane and Cpds. 160, 165, 156, and 157 have an epoxide ring that may be responsible for the mild bioactivities of Cpd. 160 [56].
A good source of erythrolide-chlorinated briarane diterpenoids could be traced to the Caribbean soft coral Erythropodium caribaeorum. The ecological role of these compounds as feeding deterrents is a fact. They have a wide variation in their composition based on the location of the sample collection. This soft coral can be found in many locations of the Caribbean Sea in Colombia, including Santa Marta, Islas del Rosario, and Providencia, which make up several coral reef zones in the south and southwest Caribbean Sea. The authors evaluated the differences in erythrolide composition with the metabolic profiles of the samples collected from each of these locations by HPLC-MS analyses. The principal component analysis showed a difference in the diterpene composition according to the origin of the sample collected. In addition, diterpenes from the collected samples from each location were isolated to describe the three chemotypes. Thus, the chemotype from Santa Marta was highly varied, with new erythrolides W (Cpd. 265) and X (Cpd. 266) together with eight known erythrolides (Cpds. 267-274). The sample from Islas del Rosario had a low-diversity chemotype constituted by large amounts of erythrolide A (Cpd.  (Table S2) [60]. Notably, erythrolides R and V (Providencia Island chemotype bearing a free OH in C-5) did not exhibit any cytotoxicity, suggesting the role of the acetyl group at C-5 in the recorded activity.
A new sesquiterpene, (+)-19-methylaminoavarone (Cpd. 290), and six known compounds (Cpds. 291-296) were isolated from the Dysidea sp., a marine sponge from the Xisha Islands. There was a revision of the carbon spectrum data of Cpds. 291, and the absolute configurations of Cpds. 290 and 291 were confirmed by electronic circular dichroism (ECD) analysis. Cpds. 290-292 and 294-296 had moderate to good cytotoxic activity against several human cancer cell lines (Table S2) [62].
Penicindopene A (Cpd. 297), a new indole diterpene, together with seven known compounds (Cpds. 298-304), became isolated from the deep-sea fungus Penicillium sp. YPCMAC1. From the structural analysis studies, penicindopene A was the first example of an indole diterpene that possesses a 3-hydroxyl-2-indolone moiety; it was moderately cytotoxic to A549 and HeLa cell lines with IC 50 values of 15.2 and 20.5 µM, respectively [63].

Marine-Derived Amino Acids, Peptides, and Polyketides Cytotoxic to Cancer Cell Lines
The following four new polyketide ansamycins: divergolides (macrolides) T, U, V, and W (Cpds. 323-326) and two known analogues (Cpds. 327 and 328) remained isolated from the fermentation broth of the mangrove-derived actinomycete Streptomyces sp. KFD18. By using spectroscopy and single-crystal X-ray diffraction analyses, their structures and the absolute configurations of their stereogenic carbon were determined. Cpds. 323-326 exhibited cytotoxic activity against the human gastric cancer cell line SGC-7901, the human leukemic cell line K562, the cervical cell line HeLa, and the lung carcinoma cell line A549. Cpd. 325 was the most active; Cpds. 327 and 328 were inactive against all the tested cancer cell lines (Table S3). Cpds. 323 and 325 had potent and specific cytotoxic activity against the SGC-7901 cells (IC 50 2.8 and 4.7 µM, respectively). They were more cytotoxic to the cell line than the two positive controls, imatinib (IC 50 = 86.8 µM) and adriamycin (IC 50 = 6.9 µM), used in the study. They induced apoptosis in SGC-7901 cells after double-staining with acridine orange-ethidium bromide (AOEB) and 4', 6-diamidino-2-phenylindole (DAPI). This is the first time the apoptosis-inducing potential of divergolides is reported [66].
By using the bioactivity-guided study and the LC-MS/MS molecular networking approach, the following nine new linear lipopeptides: microcolins E, F, G, H, I, J, K, L, and M (Cpds. 329-337) and four known related compounds, microcolins A, B, C, and D (Cpds. 338-341) were isolated from Moorea producens, a marine cyanobacterium. Catalytic hydrogenation of Cpds. 338-341 produced two known compounds, 3,4-dihydromicrocolins A and B (Cpds. 342, 343), and two new derivatives, 3,4-dihydromicrocolins C and D (Cpds.  344, 345), respectively. The combination of spectroscopic and advanced Marfey's analysis was used to determine the structures of the new compounds. Surprisingly, in Cpds. 329-341 and 336, structurally unusual amino acid units, 4-methyl-2-(methylamino)-pent-3-enoic (Mpe) acid and 2-amino-4-methylhexanoic acid (N-Me-homoisoleucine), respectively, were discovered; these are rare residues in naturally occurring peptides. The analogues had cytotoxic activity against H-460 human lung cancer cells at very low IC 50 values ranging from 0.037-5.0 µM (Table S3). The structures of the compounds with IC 50 ≤ 1.0 µM are in Figure 3 [67]. Thus, the structure-activity relationship found by the authors to compare the relative cytotoxic potencies of the compounds revealed that a hydroxyl group at C-4 of Pro and a double bond in the Mdp moiety are crucial for activity. This is an observation in tandem with the findings of [68]. Additionally, the loss of the N-methyl group from the Val residue as found in microcolin M (Cpd. 337) increased cytotoxic activity from 0.910 µM in microcolin B (Cpd. 339) to 0.069 µM in microcolin J (Cpd. 334), consequently increasing the potency of the analogue. In addition, the presence of an acetyl group at C-3 of Thr, for instance, by comparing microcolin F (Cpd. 330) with microcolin G (Cpd. 331) and microcolin A (Cpd. 338) with microcolin D (Cpd. 341; IC 50 0.075 µM), increased the cytotoxicity. The propionate group at the C-3 position of Thr decreased the cytotoxicity (microcolin K (Cpd 335), which is about 32-fold less potent than microcolin A). Likewise, adding a double bond in the M-ME-Leu residue of microcolins E-G (Cpds. 329-341) led to the loss of the antiproliferative activity. Moreover, the removal of one of the pendant methyl groups in the fatty acid chain does not affect the cytotoxic activity significantly; thus, it does not seem to play a role in determining the cytotoxic activity of the series [67].
Deep-sea actinomycete Nonomuraea sp. AKA32 produced a new and two known aromatic polyketide cytotoxic compounds that inhibited the growth of a murine B16 melanoma cancer cell line. The new compound is akazamicin (Cpd. 346), and the two known compounds are actinofuranone C (Cpd. 347) and N-formylanthranilic acid (Cpd. 353). The cytotoxic IC 50 values of the compounds were 1.7, 1.2, and 25 µM, respectively [69].
Out of the four new cycloheptapeptides, fuscasins A-D (Cpds. 372-375) that were isolated from the marine sponge Phakellia fusca from the South China Sea, fuscasin A (Cpd. 372), that bears a pyrrolidine-2, 5-dione, had potent cytotoxic activity. It inhibited the growth of only HepG2 (IC 50 = 4.6 µM) from the six cancer cell lines (MCF-7, HeLa, NCI-H460, PC9, and SW480) that were screened. Interestingly, it did not inhibit the growth of cardiomyoblast H9C2, a normal cell line (IC 50 value of 100 µM), which suggests that Cpd. 372 may exhibit selective toxicity only on the cancer cell while sparing the normal cells. The planar structures of the compounds were characterized by using spectroscopic methods, and the advanced Marfey's method was used to determine the absolute configurations of amino acid residues [71].
A new polyene compound (Cpd. 424) and a new diketopiperazine (Cpd. 425), in addition to three known compounds (Cpd. 426-428), were isolated from Penicillium crustosum HDN153086, an Antarctic marine-derived fungus. The structures of Cpds. 424-428 were deduced from the MS, NMR, and TD-DFT calculations of specific ECD spectra. The compounds were tested for cytotoxic activities against the K562 cell line. However, only Cpd. 425 (fusaperazine F) is cytotoxic to cancer cells, with an IC 50 of 12.7 µM [78].
Polyether compounds from Streptomyces species are known for antibacterial, antiviral, antiparasitic, antifungal, and antitumor activities. Some of these compounds target cancer stem cells and multi-drug-resistant cancer cells [79]. The authors isolated three polyether-type metabolites (Cpds. 429-431) from the marine-derived Streptomyces cacaoi through antimicrobial bioactivity-guided fractionation. Cpd. 431 is a new natural product. As several polyether compounds with structural similarities, such as monensin, have been associated with autophagy and cell death, they assessed the cytotoxicity of these three compounds. From the cytotoxicity screening, Cpds. 430 and 431 were cytotoxic to human cancer cell lines CaCo-2, HeLa, PC-3, and A549, but not to non-cancerous human cell lines MRC-5 and Vero. Unfortunately, the cytotoxic activity of compound 429 was not determined. Cpds. 430 and 431 were selective for CaCo-2 and PC-3 at a low IC 50 (7.4 and 11.8 µM for Cpd. 430), respectively (Table S3). In addition, Cpds. 430 and 431 induced the accumulation of autophagy flux markers, LC3-II and p62, and caused the cleavage of caspase-3, caspase-9, and poly (ADP-ribose) polymerase 1 (PARP-1). There was a dose-dependent down-regulation of the proteins that mediate the autophagosome by the compounds. This study provided an insight into the molecular mechanisms of the polyether-type polyketides and suggests their potency as inhibitors of autophagy and apoptosis inducers [79].
From Jaspis splendens, an Indonesian marine sponge, the following compounds, Jasplakinolide (Cpd. 439) (the parent compound), a new acylic jasplakinolide congener (Cpd. 440), an acyclic derivative (Cpd. 441) that requires revision (Cpd. 442), and two jasplakinolide derivatives (Cpds. 443, 444), were isolated. The chemical structures of the new and known compounds were elucidated based on HRESIMS and 1D and 2D NMR spectral analysis. The NMR data of the other compounds were compared with those of jasplakinolide (Cpd. 439). The compounds inhibited the proliferation of mouse lymphoma (L5178Y) cells with nanomolar to micromolar IC 50 values (Table S3). (+)-Jasplakinolide Z6 (Cpd. 440) is an acyclic derivative of Cpd. 441, lacking the depside bond between C-1 and C-15, and with an IC 50 of 3.2 µM, it was the least potent derivative compared to Cpds. 439, 442, and 443, whose IC 50 values were in the nanomolar range (<100 nM), but it was more potent than the positive control, kahalalide F (IC 50 = 4.3 µM) [81]. This shows that the 19-membered ring of jasplakinolide is not the main structural requirement for its activity. Therefore, lipophilicity influences the antiproliferative activities of the acyclic jasplakinolide congener. This is because of free carboxylic acid functionality in Cpd. 440 that imparts higher polarity to it more than Cpd. 442, thus rendering Cpd. 440 less lipophilic. Additionally, further proof from previous research [81] on the antiproliferative activity of three acyclic jasplakinolides Z1-Z3 against a panel of human cancer cells revealed that jasplakinolides Z2 and Z3 derivatives were the methyl and ethyl esters of jasplakinolide Z1, respectively. Results also show that the hydrolysis of the depside bond yields a free carboxylic acid that diminished its activity significantly, which reverted to being equivalent to the parent compound by converting the carboxylic acid group into ester functionality. That may be due to increased lipophilicity of the ester derivatives that may facilitate their cellular uptake and hence recover the activity of the compound [82].
The isolation of the eight new nitrogenated azaphilones (Cpds. 445−452) and two known compounds (chaetoviridin A (Cpd. 453) and chaetoviridin E (Cpd. 454)) was carried out with the culture of the fungus Chaetomium globosum MP4-S01-7 from the deep sea. The elucidation of the structures of the new compounds was carried out by HSQC-HECADE NMR data, J-based configuration analysis, and modified Mosher's method, and the verification was carried out by comparing the recorded and computed NMR chemical shifts from the quantum chemical calculations together with a DP4+ statistical procedure. The compounds were screened for an in vitro cytotoxic activity against the gastric cancer cell lines MGC803 and AGS. Most of the compounds inhibited cancer cell viability at about 10.0 µM, but Cpds. 446, 447, and 449 exhibited the most potent cytotoxic activity on the cancer cells, with IC 50 values less than 1.0 µM (Table S3). Additionally, Cpd. 446 inhibited cell cycle progression, and both Cpds. 445 and 446 caused apoptosis in gastric cancer cells in a concentration-dependent manner [83]. The structures of Cpds. 445, 446, and 449 are in Figure 3.
A novel epidithiodiketopiperazine (DC1149Ba) (Cpd. 455), isolated from Trichoderma lixii, an unidentified sponge from Mentawai, Indonesia, inhibited the proliferation of Panc-1 with an IC 50 of 0.02 µM. This cell line was cultured in a glucose-deficient medium with a selectivity index of 35,500-fold higher for cells cultured under glucose-starved conditions than those under general culture conditions. Equally, this compound inhibited the response of the endoplasmic reticulum stress signaling, an effect which could be caused by inhibiting complex II in the mitochondrial electron transport chain [84]. The structure of epidithiodiketopiperazine (DC1149Ba) (Cpd. 455) is in Figure 3.  (Table S3) [85].
Bulbimidazoles A−C (Cpds. 463−465) inhibited the growth of P388 murine leukemia cells with their IC 50 in the micromolar range (Table S3). The compounds are new alkanoyl imidazoles that remained isolated from the culture extract of the γ-proteobacterium Microbulbifer sp. DC3-6 sourced from a stony coral of the genus Tubastraea. By using the Ohrui−Akasaka method, Cpd. 463 was determined to be a mixture of (R)-and (S)-configurations that occurred in a 9:91 ratio after the absolute configuration of the substituted anteiso-methyl was assigned [86].
Shellmycins A-D (Cpds. 466-469) were the four novel cytotoxic tetrahydroanthra-γpyrone compounds isolated from Streptomyces sp. shell-016 from a marine shell sediment sample taken from Binzhou Shell Dike Island and Wetland National Nature Reserve, China. The compounds were antitumorigenic on the five cancer cell lines A375, H1299, HepG2, HT29, and HCC1937 that were screened, with IC 50 values from 0.69-26.3 µM (Table S3). The putative biosynthetic pathways of the compounds were discussed based on their structureactivity relationship, and their structures were deduced by the interpretation of 1D and 2D NMR and HR-MS data [87]. The absolute configuration of Cpd. 466 was confirmed by single crystal X-ray diffraction, and Cpds. 468 and 469 are stereoisomers to each other. The structures of the compounds with cytotoxic IC 50 ≤ 1.0 µM are in Figure 3 [87].  (Table S4). Cpds. 471 and 475 were bound to the retinoid X receptors with Kd values of 13.8 and 12.9 µM, respectively, and could induce apoptosis by a retinoid X receptor (RXR)-αdependent regulation of RXRα transcriptional expression, thus, promoting poly-ADP-ribose polymerase (PARP) cleavage. In addition, they could be cytotoxic via cell cycle arrest at the G0/G1 phase [88].
Two novel oxygenated ergostane-type sterols (Cpds. 526, 527) and one known related compound, sinugrandisterol A (Cpd. 528), were isolated from the soft coral Sinularia sp. sourced from the Xisha Islands, South China Sea. In the antiproliferation screening, the new compounds inhibited the proliferation of MDA-MB-436, A549, Hep3B, HT-29, and H157 human cancer cell lines with moderate IC 50 values (Table S4). Furthermore, the result revealed that Cpd. 528-treated H157 cells, with Hoechst 33,258 staining, morphologically presented characteristics of apoptosis. In addition, the Western blot assay implies that Cpd. 528 could up-regulate and down-regulate the expressions of Bax and Bcl-2, respectively [90].
By following the one-strain-many-compounds approach, seven culture media of Clonostachys rosea MMS1090 were optimized to increase the yield of eburicol (Cpd. 529). After the purification and structural analysis by NMR, eburicol was tested against four human cancer cell lines, MCF-7, MDA-MB-231, NSCLC-N6-L16, and A549. However, eburicol could inhibit the growth of MCF-7 and MDA-MB-231 cells with 2.0 and 15.7 µM IC 50 values, respectively. This is the first time the accumulation of eburicol in the marine fungal strain C. rosea has been reported, and this confirms its potential to produce bioactive lipids [91].
The bioactivity-guided purification of the extract from Streptomyces bingchenggensis ULS14 from Lagos Lagoon in Nigeria gave two known compounds, ULDF4 (kigamicin) (Cpd. 603) and ULDF5 (staurosporine) (Cpd. 604). The IC 50 values of ULDF4 and ULDF5 against the proliferation of the human cancer cell line, HeLa were 0.11 and 0.24 µM/mL, respectively. This finding was the first report of the anticancer potential of actinomycetes from Lagos Lagoon, which can be used for potential drug developmental purposes [98].

Other Marine-Derived Compounds Cytotoxic to Cancer Cell Lines
The rice-solid fermentation of the marine-derived Streptomyces sp. NB-A13 produced six new (Cpd. 605-610) and nine known (Cpd. 611-619 (Table S6). All the compounds were more selective for SW-620 than the PC-3 cells, although they had a similar potency on both cancer cell lines (Table S6). A structure-activity relationship of the derivatives revealed a bisindolocarbazole, and the bridged sugar units are necessary for the cytotoxicity based on the fact that compounds 611 and 618 are the most active as well as Cpds. 609, 614, and 617 compared to the other analogues. Cpd. 611 was more potent than Cpd. 618 that served as a positive control. Additionally, Cpd. 611 more strongly inhibited the activities of protein kinase C-theta (PKC-θ) at IC 50 of 0.06 µM than the other analogues (Cpds. 605-610 with a range of IC 50 of 0.06-9.43 µM) except for Cpd. 610 with no inhibitory activity, but it was close to that of the positive control (staurosporine (Cpd.14; IC 50 = 0.01 µM)). These findings suggest that the discovered staurosporine derivatives with considerable cytotoxic and PKC-θ inhibitory activities could be important materials for the development of new anticancer agents [99]. The structures of Cpds. 606 and 609 are in Figure 4.  (Table S6) [117].

Conclusions
There has been a growing rate of cancer incidence worldwide due to factors such as an aging population, eating habits, environmental changes, and co-morbidities. Poor treatment outcomes and complications of cancer disease are aggravated by multi-drug The isolation of the five new decalin derivatives, altercrasins A, B, C, D, E (Cpds. 620-624), was carried out with Alternaria sp. OUPS-117D-1; a strain initially derived from Anthocidaris crassispina, a sea urchin. The absolute stereo-structure of altercrasin A (Cpd. 620) was determined by chemical transformation and the modified Mosher's method. Silica gel and reversed-phase high-performance liquid chromatography (RP-HPLC) were used to purify altercrasins B, C, D, and E (Cpds. 621-624), and their structures were elucidated by using 1D and 2D nuclear magnetic resonance (NMR) spectroscopic analyses. By comparing the NMR chemical shifts, NOESY correlations, and electronic circular dichroism (ECD) spectral analyses of Cpd. 620, the absolute configurations of Cpds 621-624 were deduced. Subsequently, the authors discovered that Cpd. 620 was the stereoisomer of Cpd. 621 and Cpd. 623 was a stereoisomer of Cpd. 624. Additionally, their cytotoxic activities were screened against the murine P388 leukemia, human HL-60 leukemia, and murine L1210 leukemia cell lines; the result revealed that Cpds. 623 and 624 exhibited a potent cytotoxic effect on the cancer cells (Table S6) [100].
The marine-derived actinobacterial strain Nocardiopsis sp. OUCMDZ-4936 produced three new p-terphenyl derivatives, nocarterphenyls A-C (Cpds. 625-627) and three known analogues (Cpds. 628-630). Cpd. 625 possessed a benzothiazole and Cpd. 627 a benzothiazine moiety. These moieties are not common in the skeleton of p-terphenyls. The compounds exhibited potent cytotoxic activity against some of the cell lines among the 26 human cancer cell lines used for the screening, with the range of 0.10-17.0 µM IC 50 values (Table S6) [101]. The structure of nocartephenyl A is in Figure 4.
The purification of the fermented broth of Streptomyces sp. RKND004, sourced from Prince Edward Island sediment, yielded two new polycyclic polyether compounds, terrosamycins A (Cpd. 631) and B (Cpd. 632). The compounds were identified by the OSMAC approach and a UPLC-HRMS-based metabolomics assay. By using NMR, HRMS, and X-ray diffraction data analyses, the structure of Cpd. 631 was deduced. From the spectral data analysis of Cpd. 632, there are attached two methoxy groups that replaced the two hydroxy groups in Cpd. 631; that implies the methoxy group is required for the anticancer activity. Cpd. 631 preferentially binds potassium over sodium. Like other polyether ionophores, Cpds. 631 and 632 inhibited the growth of two breast cancer cell lines with moderate IC 50 values in the range of 3.6-10.1 µM (Table S6). Moreover, these compounds were more selective for the cancer cells than the normal cells (human foreskin fibroblast cell line (BJ) (IC 50 = 79 and 267 µM, respectively) and healthy Cercopithecus aethiops kidney epithelial cells (Vero) (IC 50 = 36 and 93 µM, respectively)) that were screened too [102]. Thus, it is worth mentioning that the terrosamycins were more potent than salinomycin that has progressed to pilot clinical trials for the treatment of breast cancer and other cancers.
The previously isolated oxadiazine, nocuolin A (NocA) (Cpd. 633) from the cyanobacterial strain Nodularia sp. LEGE 06071, was screened for anticancer activity against a colon cancer cell line (HCT116) and an immortalized epithelial cell line (hTERT RPE-1). The cytotoxic efficacy of NocA was significant in the cancer cells undergoing exponential growth but was not for non-proliferating immortalized cells [103]. It induced apoptosis in HCT116 multi-cellular spheroids by suppressing the overexpression of the bcl genes. Amazingly, the result of the transcriptome analysis of the HCT116 cells did not relate to any compound in the CMap. However, it pointed to stress responses and cell starvation, which may be due to a decrease in adenosine triphosphate (ATP) production. Additionally, autophagy and a decrease in the mitochondrial respiration parameter within one hour of treatment were recorded. A similar study also concluded that nutritionally starved spheroid cells are sensitive to impaired mitochondrial energy production due to limited metabolic plasticity, therefore, suggesting NocA as a mitochondrial toxin to HCT-116 [104].
The following ten new di-, tri-, and tetrasulfated triterpene glycosides, psolusoside B1 (Cpd. Cpds. 642 and 643 are highly polar and have four sulfate groups in their carbohydrate moieties, plus two sulfates in the same terminal glucose residue. Psolusoside B2 (Cpd. 636) has an atypical non-holostane aglycone with 18 (16)-lactone and a distinctive 7, 8-epoxy fragment. The cytotoxicity screening of the compounds against the following mouse cancer cell lines, Ehrlich ascites carcinoma, neuroblastoma Neuro 2-A, and erythrocytes, was carried out. Psolusoside L (Cpd. 638) (a pentaoside) that has three sulfate groups at C-6 of two glucoses and one 3-O-methylglucose residue and holostane aglycone, was the most active compound among the others (Table S6). The structure-activity relationship revealed that the sulfate group at C-2 of the terminal glucose residue attached to C-4 of the first xylose residue extensively reduced activities of the corresponding glycosides. Psolusosides of group B1 (Cpd. 634), B2 (Cpd. 635), and the known psolusoside B) did not show any activity in any of the screenings because of the attachment of non-holostane aglycones and tetrasaccharide-branched sugar chains to the sulfate group of carbon-2 of psolusoside K (Cpd. 637) [105].
By the use of 1D, 2D NMR, and HR-ESI-MS spectroscopic techniques, the structures of one new compound, named holothurin A5 (Cpd. 657), and eight known triterpene glycosides (Cpds. 658-665), that were obtained from the methanol extract of the Vietnamese sea cucumber Holothuria edulis, were deduced. Holothurin A5 (Cpd. 657) has a hydroperoxy group at C-25. That was the first report of the isolation of this group of triterpene saponin compounds from sea cucumbers [108]. In addition, the in vitro cytotoxicity of the compounds was screened against five human cancer cell lines (HepG2, KB, LNCaP, MCF7, and SK-Mel-2). The compounds, especially Cpds. 659, 663, and 664, had mild to moderate to strong cytotoxic activity on the cancer cells (Table S6) [109].
Asfour and co-authors succeeded in the large-scale production of pure terrein (Cpd. 670), a compound obtained from Aspergillus terreus strain S020, after culturing in a 21 L fermentation static broth for 40 days at room temperature on potato dextrose broth and a series of chemical treatments and chromatographic separations, in a yield of 537.26 ± 23.42 g/kg extract; thus, this represents the highest yield of terrein produced by fermentation to date. After the cytotoxic bioactivity-guided screening of the fungal extracts, the isolated compound terrein was also screened against human colorectal (HCT-116) and hepatocellular (HepG2) carcinoma cell lines. It exhibited cytotoxicity on the cells with IC 50 values of 12.13 and 22.53 µM (Table S6), respectively. This study, therefore, suggests A. terreus strain S020 as a good source for the large-scale production of terrein in industries [111].
By searching for new anticancer agents from marine samples, Fan et al. isolated these compounds: three new decalinoylspirotetramic acid derivatives, pyrenosetins A, B, and C (Cpds. 725-727), and one known decalin tetramic acid, phomasetin (Cpd. 728), from the C18-SPE fractions of the trichloromethane sub-extract of an endophyte Pyrenochaetopsis sp. FVE-001, a fungus derived from the brown alga Fucus vesiculosus. Cpds. 725 and 726 exhibited antitumor activity against the melanoma neoplasm cell line A-375 with IC 50 values of 2.8 and 6.3 µM (Table S6), respectively. That is the first report of isolated secondary metabolites from the marine-derived Pyrenochaetopsis sp. and the second from the genus Pyrenochaetopsis [116].

Conclusions
There has been a growing rate of cancer incidence worldwide due to factors such as an aging population, eating habits, environmental changes, and co-morbidities. Poor treatment outcomes and complications of cancer disease are aggravated by multi-drug resistant malignant tumors, the high cost of treatment, and adverse drug reactions/patient compliance. Therefore, there is high demand for new anticancer agents or for modifying the existing ones. This necessitated the search for potent and efficacious substances from marine sources that will improve or replace existing ones in order to contain multi-drugresistant neoplasms. This review compiled a total of 731 compounds/derivatives that belong to these classes: glycosides, alkaloids, saponins, lipids, terpenes, ribose, steroids, peptides, xanthones, ethers, lignins, coumarins, carbazoles, azaphilones, nucleosides, polyketides, and quinones (Tables S1-S6). These compounds/derivatives sourced from soft corals, bacteria, fungi, sponges, algae, sea cucumbers, seaweeds, mollusks, and sea urchins exhibited moderate to high cytotoxic activities against 121 mammalian cancer cell lines as described in 76 articles from January 2019-March 2020. The IC 50 (≤20.0 µM) values of both known and novel compounds were shown with structures of the new compounds. Interestingly, some of these compounds exhibited a broad spectrum of anticancer activity and are worth further exploration in terms of mechanism of action, structure-activity relationship, and clinical trials. However, this could serve as a guide for the selection of novel compounds/derivatives for future optimization and drug development.