Cytotoxic compounds from marine actinomycetes: sources, structures and bioactivity

Marine actinomycetes produce a substantial number of natural products with cytotoxic activity. Actinomycete strains have been isolated from sources including fishes, coral, sponges, seaweeds, mangroves and sediments. These cytotoxic compounds can be broadly categorized into four classes: polyketides; non-ribosomal peptides and hybrids; isoprenoids and hybrids; and others, among which the majority are polyketides (146 of 254). Twenty-two of the 254 compounds show potent cytotoxicity, with IC 50 values at the ng/mL or nM level. This review highlights the sources, structures and antitumor activity of 254 natural products isolated from marine actinomycetes and first reported between 1989 and 2020.


CONCLUSION
From 1989 until the end of 2020, 254 new cytotoxic compounds have been obtained from marine actinomycetes. This review summarized the structures, strain sources, and cytotoxicity of these secondary metabolites ( Table 1). Most of the compounds (206) were reported from 2010 to 2020 (Figure 8). The numbers of newly reported compounds have increased since 1989, peaked in the mid-2010s (2013-2017) and decreased in the following years. However, we expect the numbers to increase after the COVID-19 pandemic ends. Of these 254 compounds, most are moderately active, but approximately 20 compounds show potent cytotoxicity with IC 50 values at the ng/mL/nM level (see the Prospects section). The articles reporting these compounds have been published in 30 different journals, and the "Journal of Natural Products" (72) published more articles than any other single journal, followed by "Marine Drugs" (36), "Organic Letters" (27), the Journal of Antibiotics" (21), and the "Journal of Organic Chemistry" (18; Figure 9). Interestingly, beyond these prominent natural-product journals, "Phytochemistry" published seven articles, although it is a peer-reviewed scientific journal covering pure and applied plant chemistry, plant biochemistry and molecular biology. This review classified the compounds into four classes: polyketides; non-ribosomal peptides, and hybrids of polyketides and peptides; isoprenoids, terpenoids, sterols, and hybrids of isoprenoids and peptides (or polyketides); and heterocyclic, (hetero) aromatic and other compounds. These cytotoxic compounds have diverse chemical structures, and most are polyketides (146) making up 58% of the 254 new antitumor compounds (Figure 10). Among these 146 polyketides, most are categorized as either macrolides (lactones), lactams and α/γ-pyrones (57), or benzoquinones, naphthoquinones, anthraquinones and other aromatic compounds (57), which together accounted for 45% of the total 254 compounds. Marine actinomycetes produce different biologically active secondary metabolites. In 2012, Subramani and Aalbersberg published an article in "Microbiological Research" indicating that marine actinomycetes are an ongoing source of novel bioactive metabolites [144]. In 2009, a review article reported antitumor compounds from marine actinomycetes [145]. In 2020 and 2021, we reported the sources of marine actinomycetes, chemical structures and biological activities of 127 halogenated compounds and 313 antimicrobial compounds from multiple marine actinomycetes [146,147]. Marine actinomycetes are a promising source of lead compounds for drug discovery.
Despite the discovery of many cytotoxic compounds from marine actinomycetes, several drawbacks of natural product anticancer drug discovery exist. Some cytotoxic compounds have been obtained through assayguided separation, but in many cases, no assay-guided separation was performed, and cytotoxic compounds were identified simply through purification followed by cytotoxic evaluation. Most of the cytotoxic compounds have not been tested for their selectivity toward different cancer cell lines and normal human cell lines, mainly because of insufficient financial support to researchers. Bioassay-guided separation is sometimes very tedious, and dereplication does not always work well, as researchers expect. Because naturally occurring compounds in their original forms may not always be patentable in the USA, although simple derivatives can be patent protected, natural product chemists' enthusiasm for anticancer drug discovery from natural sources has been diminished.
Selection of strains, culturing strategies and analytical techniques for natural-product-library establishment and natural-product dereplication will be of great help in anticancer drug discovery from marine actinomycetes. A future direction may involve advancing genome mining and gene manipulation, as discussed below.

PROSPECTS
Some of the reviewed compounds have demonstrated potent cytotoxic activity, with IC 50 values at ng/mL or nM levels, for example, compounds 3 [12], 6-8 [14], 55 [38], 73-76 [49], 81 [53], 101 [63], 121 and 122 [75], 124 [77], 167-169 [102], 212 and 213 [126], 216 [127] and 224 [131]. However, the selectivity of some potent cytotoxic compounds has not been investigated. Selectivity study is important, because identifying cytotoxic drugs with a high selectivity toward cancer cells is critical to increase the low survival rates of patients with cancer. One approach to avoiding adverse effects of cytotoxic agents is targeted drug delivery. For instance, a cytotoxic drug can be hung on an antibody scaffold to form an antibody-drug conjugate. Subsequently, the complex targeted agent can overcome the unspecific toxic effects of conventional drug delivery, thereby decreasing the amount of drug required for therapeutic efficacy. Most of the secondary metabolites reviewed herein have been evaluated for their antimicrobial and cytotoxic activities. Other biological properties could be identified through testing of actinomycete secondary metabolites in other biological settings. For example, the fungal metabolites sinuxylamides A and B have shown no antibacterial activity or cytotoxicity at 40 μM, but when tested for their antithrombotic activity, have demonstrated strong inhibition of the binding of fibrinogen to purified integrin IIIb/IIa in a dose-dependent manner, with IC 50 values of 0.89 and 0.61 μM, respectively [148].
Novel molecules with unprecedented structural and/ or functional attributes usually have unique bioactivities. Some of the reviewed compounds in this article have unique structures, for example, compounds 1, 2, 9, 10, 23, 101, 123, 140, 141, 144-146 and most of the compounds classified as heterocyclic, (hetero)aromatic and other compounds (196-254). Compounds 101, 212, 213 and 224 are not only structurally interesting (particularly 101) but also exhibit potent cytotoxicity. The cytotoxicity of 101 arises from the induction of double-strand breaks in DNA [149]. Compound 101 has a molecular formula of C 38 H 26 N 4 O 14 . Its molecular weight is 762 Daltons, and the numbers of hydrogen-bond donors and acceptors in the molecule clearly violate Lipinski's role of five; these findings, together with the compound's structural complexity, suggest low druggability of 101. However, structural modification and/or formulation have made many undruggable compounds druggable. For example, halichondrin B (molecular formula: C 60 H 86 O 19 ; molecular weight: 1111 Daltons) is a complex polyether macrolide originally isolated from the marine sponge Halichondria okadai [150], which was believed to be undruggable by many researchers. However, Eisai Co. has structurally simplified halichondrin B, and eribulin (brand name Halaven) was approved by the U.S. Food and Drug Administration on November 15, 2010, with an indication to treat metastatic breast cancer [151].
Most of these 254 compounds are analogs of previously reported molecules. In general, structurally unique compounds represent a decreasing percentage of the total number of compounds isolated from natural sources in the past few decades. However, exploring unexplored and unusual source organisms, or those from unique environments, could provide opportunities for finding novel natural products.
Currently, the genomes of actinomycete strains are routinely sequenced, and a host of bioinformatics tools are increasingly available for identifying potential biosynthetic gene clusters of actinomycete natural products. Developing universal expression systems for small-molecule biosynthesis with high yield, constructing genetic tools to access the biosynthetic potential of cultured marine actinomycetes and awakening "silent" biosynthetic pathways will be important approaches for discovery of small molecules from marine actinomycetes. Investigations aimed at understanding how the biosynthetic pathways operate at the genetic and biochemical levels in marine actinomycetes will open new doors to designing molecules with improved anticancer properties.