Current Status and Future Prospects of Marine Natural Products (MNPs) as Antimicrobials

The marine environment is a rich source of chemically diverse, biologically active natural products, and serves as an invaluable resource in the ongoing search for novel antimicrobial compounds. Recent advances in extraction and isolation techniques, and in state-of-the-art technologies involved in organic synthesis and chemical structure elucidation, have accelerated the numbers of antimicrobial molecules originating from the ocean moving into clinical trials. The chemical diversity associated with these marine-derived molecules is immense, varying from simple linear peptides and fatty acids to complex alkaloids, terpenes and polyketides, etc. Such an array of structurally distinct molecules performs functionally diverse biological activities against many pathogenic bacteria and fungi, making marine-derived natural products valuable commodities, particularly in the current age of antimicrobial resistance. In this review, we have highlighted several marine-derived natural products (and their synthetic derivatives), which have gained recognition as effective antimicrobial agents over the past five years (2012–2017). These natural products have been categorized based on their chemical structures and the structure-activity mediated relationships of some of these bioactive molecules have been discussed. Finally, we have provided an insight into how genome mining efforts are likely to expedite the discovery of novel antimicrobial compounds.


Introduction
Infectious diseases caused by bacteria, fungi and viruses pose a major threat to public health despite the tremendous progress in human medicine. A dearth in the availability of effective drugs and the on-going threats posed by antimicrobial resistant organisms further worsen the situation particularly in developing countries. Antimicrobial resistance accounts for at least 50,000 deaths each year in Europe and the United States and it is anticipated that drug resistant infections will be responsible for the deaths of 10 million people worldwide by 2050 [1,2]. Continuously evolving antibiotic-resistance of microbial pathogens has raised demands for the development of new and effective antimicrobial compounds [3]. For generations, humans have turned to nature as a source of invaluable medicinal products, where terrestrial and marine organisms traditionally provide the most effective remedies [4,5]. It was only after the discovery of penicillin in 1928 that microbial sources were explored as sources of new therapeutic molecules. Developments in microbial culture techniques and diving expeditions in the 1970s have largely directed the drug discovery program towards the Chrysophaentin A (1), a macrocyclic natural product (comprising two polyhalogenated, polyoxygenated ω,ω′-diarylbutene units connected by two ether bonds), was isolated from the chrysophyte alga Chrysophaeum taylori. This compound inhibits the growth of clinically relevant Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus (MRSA; MIC50 1.5 ± 0.7 μg/mL), multiple drug resistant S. aureus (MIC50 1.3 ± 0.4 μg/mL), and Vancomycin-Resistant Enterococcus faecium (VREF; MIC50 2.9 ± 0.8 μg/mL). Chrysophaentin A inhibits the GTPase activity of the bacterial cytoskeletal protein FtsZ (IC50 value 6.7 μg/mL), and GTP-induced formation of FtsZ protofilaments [30]. Interestingly, this compound was found to be relatively more active among its congeners, Chrysophaentins B-G (2)(3)(4)(5)(6)(7). Analysis of bioactivity of these molecules provided insights into the pharmacophoric features of Chrysophaentins relevant to antimicrobial activity. Phenolic groups in compound 1 were determined to be crucial for activity as a hexaacetate derivative of 1 was found to be inactive at a concentration 25 μg/disk. An approximate 12-fold decrease in the MIC50 value of chrysophaentin D (4) compared to compound 1 was observed on replacement of chlorine with bromine on rings A and C. The significance of the macrocyclic structure was established following higher MIC50 values of Chrysophaentin E (5) being observed towards S. aureus and MRSA compared Chrysophaentin A (1), a macrocyclic natural product (comprising two polyhalogenated, polyoxygenated ω,ω -diarylbutene units connected by two ether bonds), was isolated from the chrysophyte alga Chrysophaeum taylori. This compound inhibits the growth of clinically relevant Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus (MRSA; MIC 50 1.5 ± 0.7 µg/mL), multiple drug resistant S. aureus (MIC 50 1.3 ± 0.4 µg/mL), and Vancomycin-Resistant Enterococcus faecium (VREF; MIC 50 2.9 ± 0.8 µg/mL). Chrysophaentin A inhibits the GTPase activity of the bacterial cytoskeletal protein FtsZ (IC 50 value 6.7 µg/mL), and GTP-induced formation of FtsZ protofilaments [30]. Interestingly, this compound was found to be relatively more active among its congeners, Chrysophaentins B-G (2)(3)(4)(5)(6)(7). Analysis of bioactivity of these molecules provided insights into the pharmacophoric features of Chrysophaentins relevant to antimicrobial activity. Phenolic groups in compound 1 were determined to be crucial for activity as a hexaacetate derivative of 1 was found to be inactive at a concentration 25 µg/disk. An approximate 12-fold decrease in the MIC 50  chlorine with bromine on rings A and C. The significance of the macrocyclic structure was established following higher MIC 50 values of Chrysophaentin E (5) being observed towards S. aureus and MRSA compared to MIC 50 values observed for the chlorinated cyclic bisdiarylbutene ethers 1 and 6. However, compound 5 was also found to be inactive toward E. faecium and VREF at concentrations as high as 25 µg/mL. In the case of the symmetrically linked dimers 6 and 7, replacing a chlorine atom on ring C with bromine confers compound 6 with at least three times better activity than compound 7. Among the tetrachlorinated macrocyles 1 and 6, compound 1 was found to be 3-5 times more potent than compound 6, indicating that the position of the ether linkage relative to the 2-butene unit affects activity. In fact, ortho-linked chrysophaentin A has been found to be more potent than the para-linked chrysophaentin F [31].
A small cyclopropane-containing fatty acid, lyngbyoic acid (8), was found to be a major metabolite of the marine cyanobacterium, Lyngbya cf. majuscula, isolated in Florida [32]. This molecule exerts antimicrobial action by disrupting quorum sensing in Pseudomonas aeruginosa. At a concentration of 100 µM, it inhibits the N-acyl homoserine lactone (HSL) receptor proteins in the organism (LasR in particular), reducing the expression of important virulence factors in the wild-type strain. The molecules inhibit the response of LasR-based QS reporter plasmids to 3-oxo-C 12-HSL. The AHL-binding site of LasR was not essential to this effect, but competition experiments indicated that compound 8 is likely to have a dual mechanism of action acting both through the AHL-binding site and independently of it. Comparison of compound 8 with related compounds (dodecanoic acid, 9; malyngolide, 10; and lyngbic acid, 11; methyl ester of dodecanoic acid, 12 and butyric acid, 13) revealed a structure-activity relationship. While compounds 9, 10 and 11 had a similar potency in pSB1075 compared to 8, either esterification (12) or shortening of the alkyl chain (13) reduced activity [32].
Two sulfated sterols, geodisterol-3-O-sulfite (14) and 29-demethylgeodisterol-3-O-sulfite (15), were isolated from the marine sponge Topsentia sp. These sulfated sterols demonstrated reverse efflux pump-mediated fluconazole resistance. They enhanced the activity of fluconazole in a Saccharomyces cerevisiae strain overexpressing the Candida albicans efflux pump MDR1, and in a fluconazole-resistant C. albicans clinical isolate known to overexpress MDR1. No activity for non-sulfated sterol in fluconazole-resistance reversal assay had been observed highlighting the relevance of sulfate group for MDR1 inhibition and synergy with fluconazole. Investigation of the geodisterols had provided insight into the clinical utility of combining efflux pump inhibitors with current antifungals to combat the resistance associated with opportunistic fungal infections caused by C. albicans [33].
In the following sections, we present a systematic overview of the marine natural products which have gained the attention of chemists and biologists over the last five years (2012-2017) as potential antimicrobial agents. The molecules are categorized according to their chemical class based on their associated structural units. Pharmacophoric features responsible for antimicrobial activity are also discussed. Table 1 lists MNPs and describes their antimicrobial potential in terms of MICs and zone of inhibition.

Alkaloids
Alkaloids constitute the largest number of antimicrobial compounds reported from marine species ( Figure 2). Polyoxygenated dihydropyrano[2,3-c]pyrrole-4,5-dione derivatives, pyranonigrin A (16) and F (17) were isolated and identified from Penicillium brocae MA-231, which is an endophytic fungus obtained from the fresh tissue of the marine mangrove plant Avicennia marina. These compounds possess a wide array of antimicrobial activities against a human-, aqua-, and plant-pathogens [34]. Indole diketopiperazine compounds (18)(19)(20)(21) identified from the culture extract of Eurotium cristatum EN-220 (endophytic fungus species obtained from the marine alga Sargassum thunbergii) possess antimicrobial activities. While rubrumazine B (18) exhibited moderate activity (MIC 64 µg/mL) against the plant-pathogenic fungus Magnaporthe grisea, echinulin (19), dehydroechinulin (20) and variecolorin H (21) showed mild activity (MIC 256 µg/mL) against the human pathogen S. aureus. This particular trend in antimicrobial activity led to an assumption that indole diketopiperazine alkaloids formed by condensation of a tryptophan residue with a second amino acid such as L-alanine might have low relevance to the observed antimicrobial activity [35]. In a separate study by Du et al., a set of indole alkaloids including echinulin (19), cristatumin A (22), cristatumin D (23) and tardioxopiperazine A (24) were found to be active against Escherichia coli and S. aureus bacteria but were unable to inhibit the growth of the plant-pathogenic fungi, Alternaria brassicae, Valsa mali, Physalospora obtuse, Alternaria solania, and Sclerotinia miyabeana. The antibacterial activity of cristatumin A (22) appeared to be related to the serine residue in the 2,5-diketopiperazine moiety compared to that of neoechinulin A (25), which contained an alanine residue. The structural difference between isoechinulin A (26) and tardioxopiperazine A (24) was found at the C-8/C-9 position, where the single bond between C-8/C-9 in compound 24 was essential for its antibacterial activity [36].
Xinghaiamine A (38), an alkaloid isolated from the marine-derived actinomycete Streptomyces xinghaiensis NRRL B24674, has exhibited a broad-spectrum of antibacterial activities against both Gram-negative (Acinetobacter baumannii, P. aeruginosa and E. coli) and Gram-positive pathogens (S. aureus and B. subtilis). The inhibition of these pathogenic bacteria and clinical MRSA isolates demonstrates the potential of compound 38 to be an effective antibiotic against multi-drug resistant pathogens, particularly S. aureus and A. baumanii. However, the molecule displayed no obvious antifungal activity against C. albicans when tested at concentrations up to 176.64 µM. The sulfoxide moiety present in Xinghaiamine A (38) is unusual among the metabolites produced by marine actinomycetes and confers compounds with a broad spectrum of biological activities, including potent antimicrobial activities [41]. Bioassay-directed fractionation performed on the ethyl acetate fraction of an organic extract from the Red Sea sponge Hyrtios species yielded the alkaloid compounds, hyrtioerectines D-F (39-41), which possessed antimicrobial activities against C. albicans, S. aureus and P. aeruginosa but not against E. coli. The relatively higher antimicrobial activity exerted by compounds 39 and 41 with respect to compound 40 could be attributed to the presence of diphenolic moieties in their structure. Amidation of the carboxylic moiety in compound 41 exerted a slight effect on activity when compared to compound 39 [42]. The diterpene alkaloids, ageloxime B (42) and (−)-ageloxime D (43), were isolated from the marine sponge Agelas mauritiana, isolated from Yongxing island in the South China Sea. Both compounds (42 and 43) were able to inhibit the growth of C. neoformans. Compound 42 also exhibited antibacterial activity against S. aureus and MRSA [43]. A manzamine alkaloid, i.e., zamamidine D (44), was isolated from an Okinawan Amphimedon sp. marine sponge. Compound 44 is the first manzamine alkaloid possessing a 2,2 -methylenebistryptamine unit as its aromatic moiety instead of a β-carboline unit, which affected growth of both bacteria and fungi [44]. Chemical investigation of the marine-sponge derived fungus Penicillium adametzioides AS-53 yielded the bisthiodiketopiperazine derivatives, adametizines A (45) and B (46), differing in the presence of a chlorine group at C-7 in compound 45 and hydroxyl in compound 46. The presence of a chlorine atom at C-7 conferred compound 45 with better antibacterial activity than compound 46 [45].  Xinghaiamine A (38), an alkaloid isolated from the marine-derived actinomycete Streptomyces xinghaiensis NRRL B24674, has exhibited a broad-spectrum of antibacterial activities against both Gram-negative (Acinetobacter baumannii, P. aeruginosa and E. coli) and Gram-positive pathogens (S. aureus and B. subtilis). The inhibition of these pathogenic bacteria and clinical MRSA isolates demonstrates the potential of compound 38 to be an effective antibiotic against multi-drug resistant pathogens, particularly S. aureus and A. baumanii. However, the molecule displayed no obvious antifungal activity against C. albicans when tested at concentrations up to 176.64 μM. The sulfoxide moiety present in Xinghaiamine A (38) is unusual among the metabolites produced by marine actinomycetes and confers compounds with a broad spectrum of biological activities, including potent antimicrobial activities [41]. Bioassay-directed fractionation performed on the ethyl acetate fraction of an organic extract from the Red Sea sponge Hyrtios species yielded the alkaloid compounds, hyrtioerectines D-F (39-41), which possessed antimicrobial activities against C. albicans, S. aureus and P. aeruginosa but not against E. coli. The relatively higher antimicrobial activity exerted by compounds 39 and 41 with respect to compound 40 could be attributed to the presence of diphenolic moieties in their structure. Amidation of the carboxylic moiety in compound 41 exerted a slight effect on activity when compared to compound 39 [42]. The diterpene alkaloids, ageloxime B (42) and (−)-ageloxime D (43), were isolated from the marine sponge Agelas mauritiana, isolated from Yongxing island in the South China Sea. Both compounds (42 and 43) were able to inhibit the growth of C. neoformans. Compound 42 also exhibited antibacterial activity against S. aureus and MRSA [43]. A manzamine alkaloid, i.e., zamamidine D (44), was isolated from an Okinawan Amphimedon sp. marine sponge. Compound 44 is the first manzamine alkaloid possessing a 2,2′methylenebistryptamine unit as its aromatic moiety instead of a β-carboline unit, which affected growth of both bacteria and fungi [44]. Chemical investigation of the marine-sponge derived fungus Penicillium adametzioides AS-53 yielded the bisthiodiketopiperazine derivatives, adametizines A (45) and B (46), differing in the presence of a chlorine group at C-7 in compound 45 and hydroxyl in Diterpene alkaloids from the sponge Agelas nakamurai collected from the Xisha Islands in the South China Sea have demonstrated significant antimicrobial potential in vitro against S. aureus, E. coli, Proteusbacillus vulgaris and C. albicans. Iso-agelasidine B (47) and (−)-agelasidine C (48) exhibited pronounced antifungal activities (MIC 2.34 µg/mL) against C. albicans and weakly inhibited bacterial growth. In addition, diterpene alkaloids containing a 9-N-methyladeninium unit: iso-agelasine C (49), agelasine B (50), agelasine J (51) and nemoechine G (52) also possessed strong antifungal activities (MIC 0.6 µg/mL) against C. albicans and weak to moderate antibacterial activities [46]. Brevianamide F (53) was isolated from the marine-derived fungus, Penicillium vinaceum. This compound demonstrated antimicrobial activity against S. aureus and antifungal activity towards C. albicans [47]. A new secondary metabolite N-(2-hydroxyphenyl)-2-phenazinamine (54) was isolated from the saline culture broth of Nocardia dassonvillei, a marine actinomycete recovered from a sediment in the Arctic Ocean. This new compound has shown significant antifungal activity against C. albicans, with a MIC 64 µg/mL [48].
sponge Hippospongia sp. was found to possess antibacterial activity against MRSA, wild-type S. aureus and VREF and displayed antifungal activity against amphotericin-resistant C. albicans [56].

Lipids
In the past 5 years alone, 15 lipids possessing antibacterial activities were isolated from marine species (Figure 4). Two cerebrosides, penicillosides A (81) and B (82), were isolated from Red Sea marine-derived fungi and the fungus Penicillium species isolated from the tunicate Didemnum species. Penicilloside A (compound 81) displayed antifungal activity towards C. albicans displaying an inhibition zone of 23 mm, while Penicilloside A (82) possessed antibacterial activities against S. aureus and E. coli displaying inhibition zones of 19 mm and 20 mm respectively, at 100 μg/disk concentration [57]. Glycolipopeptides-ieodoglucomides A-C (83)(84)(85), and a monoacyldiglycosyl glycerolipidiedoglycolipid (86) isolated from the fermentation broth of the marine sediment-derived bacterium Bacillus licheniformis displayed promising antimicrobial activities against Gram-positive and Gramnegative bacteria. Moreover these two pencillosides were also active against the plant pathogenic fungi Aspergillus niger, Rhizoctonia solani, Colletotrichum acutatum, B. cenerea and C. albicans [58,59]. Gageotetrins A-C (87-89), fall under a unique class of linear lipopeptides (di-and tetrapeptides) isolated from a marine-derived Bacillus were found to be comparatively more active against fungi

Lipids
In the past 5 years alone, 15 lipids possessing antibacterial activities were isolated from marine species (Figure 4). Two cerebrosides, penicillosides A (81) and B (82), were isolated from Red Sea marine-derived fungi and the fungus Penicillium species isolated from the tunicate Didemnum species. Penicilloside A (compound 81) displayed antifungal activity towards C. albicans displaying an inhibition zone of 23 mm, while Penicilloside A (82) possessed antibacterial activities against S. aureus and E. coli displaying inhibition zones of 19 mm and 20 mm respectively, at 100 µg/disk concentration [57]. Glycolipopeptides-ieodoglucomides A-C (83)(84)(85), and a monoacyldiglycosyl glycerolipid-iedoglycolipid (86) isolated from the fermentation broth of the marine sediment-derived bacterium Bacillus licheniformis displayed promising antimicrobial activities against Gram-positive and Gram-negative bacteria. Moreover these two pencillosides were also active against the plant pathogenic fungi Aspergillus niger, Rhizoctonia solani, Colletotrichum acutatum, B. cenerea and C. albicans [58,59].
Gageotetrins A-C (87)(88)(89), fall under a unique class of linear lipopeptides (di-and tetrapeptides) isolated from a marine-derived Bacillus were found to be comparatively more active against fungi than to bacteria with MIC values of 0.02-0.04 µM. Moreover, compounds 88 and 89 at a concentration of 0.02 µM displayed potent motility inhibition and lytic activity against the late blight pathogen Phytophthora capsici, which causes enormous economic damage in cucumber, pepper, tomato and beans [60]. Lauramide diethanolamine (90) was isolated from the marine bacterial strain Streptomyces sp. (strain 06CH80). Although this compound (90) showed moderate antimicrobial activities against clinical pathogens, its chemical structure is particularly unusual containing a unique carbon skeleton which is different from any other existing antimicrobial agents. This unusual structure provides researchers with the exciting opportunity of exploring various chemical modifications of the compound with the aim of developing of potentially more efficient antimicrobial agents [61]. The unsaturated fatty acids, linieodolides A (91) and B (92), were isolated from the culture broth of a marine Bacillus and the mechanism of their antimicrobial activity was proposed through the inhibition of bacterial fatty acid synthesis [62]. Two highly acetylated steroids, dysiroid A (93) and dysiroid B (94), were isolated from the marine sponge, Dysidea sp. Compounds 93 and 94 showed potent activity against bacterial strains with MICs ranging from 4 to 8 µg/mL [63]. Halistanol sulfate A (95) was obtained from the sponge Petromica ciocalyptoides and determined to be an effective antibacterial compound against Streptococcus mutans. The compound inhibited S. mutans biofilm formation by down regulating the expression of the gtfB, gtfC and gbpB virulence genes. Compound 95 inhibited biofilm formation in two S. mutans strains at low MIC, but did not inhibit initial colonization by S. sanguinis. Such activity is highly desirable in preventative treatments that inhibits pathogenic bacteria via the disruption of biofilm formation without affecting the healthy normal microflora [64].  [60]. Lauramide diethanolamine (90) was isolated from the marine bacterial strain Streptomyces sp. (strain 06CH80). Although this compound (90) showed moderate antimicrobial activities against clinical pathogens, its chemical structure is particularly unusual containing a unique carbon skeleton which is different from any other existing antimicrobial agents. This unusual structure provides researchers with the exciting opportunity of exploring various chemical modifications of the compound with the aim of developing of potentially more efficient antimicrobial agents [61]. The unsaturated fatty acids, linieodolides A (91) and B (92), were isolated from the culture broth of a marine Bacillus and the mechanism of their antimicrobial activity was proposed through the inhibition of bacterial fatty acid synthesis [62]. Two highly acetylated steroids, dysiroid A (93) and dysiroid B (94), were isolated from the marine sponge, Dysidea sp. Compounds 93 and 94 showed potent activity against bacterial strains with MICs ranging from 4 to 8 μg/mL [63]. Halistanol sulfate A (95) was obtained from the sponge Petromica ciocalyptoides and determined to be an effective antibacterial compound against Streptococcus mutans. The compound inhibited S. mutans biofilm formation by down regulating the expression of the gtfB, gtfC and gbpB virulence genes. Compound 95 inhibited biofilm formation in two S. mutans strains at low MIC, but did not inhibit initial colonization by S. sanguinis. Such activity is highly desirable in preventative treatments that inhibits pathogenic bacteria via the disruption of biofilm formation without affecting the healthy normal microflora [64].

Peptides
Marine derived antimicrobial compounds of peptide origin are shown in Figure 5. Modified diketopiperazines, rodriguesines A and B (96 and 96a) isolated as an inseparable mixture from the ascidian Didemnum sp. were found to possess broad antimicrobial activities inhibiting both oral streptococci and pathogenic bacteria. These diketopiperazines are reported to modulate the LuxR-mediated quorum-sensing systems of Gram-negative and Gram-positive bacteria and are considered to influence cell-cell bacterial signaling, offering alternative ways to control biofilms by interfering with microbial communication [64]. Diketopiperazines, cyclo-(L-valyl-D-proline) (97) cyclo-(L-phenylalanyl-D-proline (98), were isolated from a Rheinheimera japonica strain KMM 9513 collected from shores off of the Sea of Japan. These diketopiperazines inhibited the growth of B. subtilis, Enterococcus faecium and S. aureus but were inactive against E. coli, S. epidermidis, Xanthomonas sp. pv. badrii and C. albicans [65]. Compounds cyclo-(S-Proline-R-Leucine) (99) and (97) were isolated from Bacillus megaterium LC derived from the marine sponge Haliclona oculata. These compounds showed antimicrobial activity at MIC values ranging from 0.005 to 5 µg/mL against Gram-negative bacteria V. vulnificus and V. parahaemolyticus, together with the Gram-positive bacteria B. cereus and M. luteus [66]. Cyclohexadepsipeptides of the isaridin class including isaridin G (100), desmethylisaridin G (101), desmethylisaridin C1 (102) and isaridin E (103), were identified in the marine bryozoan-derived fungus Beauveria felina EN-135. These compounds possessed inhibitory activities against E. coli with MICs in the range of 8-64 µg/mL [67]. Desotamide (104) and desotamide B (105) are cyclic hexapeptides isolated from the marine microbe Streptomyces scopuliridis SCSIO ZJ46. When these compounds were explored for their antibacterial potential, notable antibacterial activities against strains of S. pnuemoniae, S. aureus, and methicillin-resistant Staphylococcus epidermidis (MRSE) were observed [68].

Peptides
Marine derived antimicrobial compounds of peptide origin are shown in Figure 5. Modified diketopiperazines, rodriguesines A and B (96 and 96a) isolated as an inseparable mixture from the ascidian Didemnum sp. were found to possess broad antimicrobial activities inhibiting both oral streptococci and pathogenic bacteria. These diketopiperazines are reported to modulate the LuxRmediated quorum-sensing systems of Gram-negative and Gram-positive bacteria and are considered to influence cell-cell bacterial signaling, offering alternative ways to control biofilms by interfering with microbial communication [64]. Diketopiperazines, cyclo-(L-valyl-D-proline) (97) cyclo-(Lphenylalanyl-D-proline (98), were isolated from a Rheinheimera japonica strain KMM 9513 collected from shores off of the Sea of Japan. These diketopiperazines inhibited the growth of B. subtilis, Enterococcus faecium and S. aureus but were inactive against E. coli, S. epidermidis, Xanthomonas sp. pv. badrii and C. albicans [65]. Compounds cyclo-(S-Proline-R-Leucine) (99) and (97) were isolated from Bacillus megaterium LC derived from the marine sponge Haliclona oculata. These compounds showed antimicrobial activity at MIC values ranging from 0.005 to 5 μg/mL against Gram-negative bacteria V. vulnificus and V. parahaemolyticus, together with the Gram-positive bacteria B. cereus and M. luteus [66]. Cyclohexadepsipeptides of the isaridin class including isaridin G (100), desmethylisaridin G (101), desmethylisaridin C1 (102) and isaridin E (103), were identified in the marine bryozoan-derived fungus Beauveria felina EN-135. These compounds possessed inhibitory activities against E. coli with MICs in the range of 8-64 μg/mL [67]. Desotamide (104) and desotamide B (105) are cyclic hexapeptides isolated from the marine microbe Streptomyces scopuliridis SCSIO ZJ46. When these compounds were explored for their antibacterial potential, notable antibacterial activities against strains of S. pnuemoniae, S. aureus, and methicillin-resistant Staphylococcus epidermidis (MRSE) were observed [68].

Halogenated Compounds
The chemical structures of halogenated compounds with antimicrobial properties isolated between 2012 and 2017 are presented in Figure 6. Compounds demonstrating a broad spectrum of antibacterial activities were polybrominated diphenyl ethers, 2-(2 ,4 -dibromophenoxy)-3,5dibromophenol (106) and 2-(2 ,4 -dibromophenoxy)-3,4,5-tribromophenol (107) isolated from the marine sponge Dysidea granulosa; and 2-(2 ,4 -dibromophenoxy)-4,6-dibromophenol (108) from Dysidea spp. These brominated ethers exhibited in vitro antibacterial activity against MRSA, methicillin-sensitive Staphylococcus aureus (MSSA), E. coli O157:H7, and Salmonella. Structurally compound 106 differed from compound 108 at a bromo-substituted position, and from compound 107 by containing an additional bromo group at the C-4 position. From the structure-activity relationships it was suggested that bromination and para-substitution decreases the antimicrobial activities of bromophenols, except against the human pathogen, Listeria monocytogenes [69]. Bromotyrosine-derived alkaloids were isolated by bioassay-guided fractionation of extracts from the Australian marine sponge Pseudoceratina purpurea. Aplysamine 8 (109) was not found to have any notable activity against E. coli or S. aureus while hexadellin (110), aplysamine 2 (111) and 16-debromoaplysamine 4 (112) displayed activity against S. aureus at concentrations ranging from 125-250 µg/mL [70]. The brominated diterpene, sphaerodactylomelol (113) which belongs to the rare dactylomelane family, bromosphaerol (114) and 12R-hydroxybromosphaerol (115) were isolated from cosmopolitan red algae, Sphaerococcus coronopifolius, collected in the Atlantic. These compounds inhibited the growth of S. aureus but were inactive against E. coli and P. aeruginosa [71]. A brominated compound, 6-bromoindolyl-3-acetic acid (116) displayed varied activities against both Gram-positive and Gram-negative bacteria: V. harveyi, Photobacterium damselae subsp. damselae. A. hydrophila, S. aureus and B. subtilis [72]. The bromopyrrole alkaloids, nagelamides X-Z (117-119), isolated from a marine sponge Agelas sp. exhibited antimicrobial activity against a range of bacteria and fungi. Compounds 117 and 118 are dimeric bromopyrrole alkaloids with a novel tricyclic skeleton, which consists of spiro-bonded tetrahydrobenzaminoimidazole and aminoimidazolidine moieties, while compound 118 is the first dimeric bromopyrrole alkaloid involving the C-8 position in dimerization [73]. The antibacterial bromotyrosine-derived metabolites, ianthelliformisamines A-C (120-122), were isolated from the marine sponge Suberea ianthelliformis. Compound 120 displayed selective inhibitory activity against P. aeruginosa while compound 121 showed little inhibition. The spermine moiety associated with compounds 120 and 122 appeared to be important for activity against P. aeruginosa, since replacement of spermine by spermidine in compound 121 reduced the activity significantly. Furthermore, the addition of an extra cinnamyl derivative in compound 122 to the terminal amine of the spermine chain decreased the antibacterial selectivity between P. aeruginosa and S. aureus; however the observed selectivity may be due to differential cell permeability between the Gram-negative and the Gram-positive bacteria [74].

Polyketides
Antimicrobial marine natural products of polyketide origin are presented in Figure 7. Chemical investigations of cultures of the marine Streptomyces sp. 182SMLY led to the discovery of the antimicrobial polycyclic anthraquinone, N-acetyl-N-demethylmayamycin (123), which inhibited the growth of MRSA with a MIC 20 µM but displayed no inhibition against E. coli [75]. An unusual antibiotic polyketide with a new carbon skeleton, lindgomycin (124), and ascosetin (125) extracted from mycelia and culture broth of different Lindgomycetaceae strainswere two folds less active against the clinically relevant bacteria Staphylococcus epidermidis, S. aureus, MRSA, and Propionibacterium acnes than standard drug chloramphenicol. The antifungal activity of compound 124 and 125 was four times lower than nystatin against the human pathogenic yeast C. albicans. Moreover Xanthomonas campestris and Septoria tritici, were also inhibited by these metabolites but no inhibitory effects against Gram-negative bacteria was observed [76]. Among the polyketide endoperoxides, manadodioxans D (126) and E (127) isolated from the marine sponge Plakortis bergquistae in Indonesia, only manadodioxan E (127) displayed antimicrobial activities against bacteria, namely E. coli and B. cereus. The presence of a carbonyl group at C-13 position in compound 126 possibly has sequestered the antimicrobial activity. However, both compounds were found to be inactive against C. albicans, and S. cerevisiae [77]. Using antibacterial bioassay-guided fractionation, two O-heterocyclic compounds belonging to pyranyl benzoate analogues of polyketide origin 2-(7-(2-Ethylbutyl)-2,3,4,4a,6,7-hexahydro-2-oxopyrano- [ (129), were isolated from the ethyl acetate extract of B. subtilis MTCC 10407. Two additional homologs, i.e., (3-(methoxycarbonyl)-4-(5-(2-ethylbutyl)-5,6-dihydro-3-methyl-2H-pyran-2-yl)-butylbenzoate) (130) and [2-(8-butyl-3-ethyl-3,4,4a,5,6,8a-hexahydro-2H-chromen-6-yl)-ethylbenzoate] (131), were also isolated from the ethyl acetate extract from the host seaweed Sargassum myriocystum. Although compounds 130 and 131 displayed weaker antibacterial activities than compounds 128 and 129 these four compounds possessed similar structures suggesting the ecological and metabolic symbiosis between seaweeds and bacteria. It was evident from the study that the presence of dihydro-methyl-2H-pyran-2-yl propanoate system was essential to impart the antibacterial activity. Tetrahydropyran-2-one moiety of the tetrahydropyrano-[3,2b]-pyran-2(3H)-one system of compound 128 might be cleaved by the metabolic pool of seaweeds to afford biologically active methyl 3-(dihydro-3-methyl-2H-pyranyl)-propanoate moiety of compound 130 (which was shown to have no significant antibacterial activity in intact form) [78].  Fradimycins A (132) and B (133) isolated from marine Streptomyces fradiae strain PTZ0025 displayed antimicrobial activity against S. aureus [79]. Macrolactins, polyene cyclic macrolactones possess a wide range of biological activities including antimicrobial, antiviral and anticancer. The majority of macrolactins are produced by Bacillus sp., whilst glycosylated macrolactins A1 (134) and B1 (135) are frequently isolated from marine Streptomyces species. The position of the hydroxyl group (OH) or the introduction of a keto group (C=O) in the macrolactone ring affects the antimicrobial activity of macrolactins, while the introduction of ester groups at the C-7 in macrolactins leads to better antibacterial activity. Compounds 134 and 135 were less active than their corresponding ethercontaining macrolactins, probably due to the attachment of a sugar moiety at C-7 position of both compounds. However, glycosylated macrolactins are reported to have better polar solubility in comparison to the non-glycosylated macrolactins [61]. From the culture broth of a marine Bacillus sp., compounds including macrolactones (Macrolactin X-Z, 136-138) and macrolactinic acid (139) were isolated. These metabolites displayed potential antimicrobial activity against microbial strains, B. subtilis (KCTC 1021), E. coli (KCTC 1923), and S. cerevisiae (KCTC 7913). Macrolactins exhibit their antibacterial activity by inhibiting peptide deformylase in a dose-dependent manner. The position of hydroxy groups in the macrolactone ring is also important for antimicrobial activity of macrolactins. The hydroxy group at C-15 in macrolactone ring increases the antimicrobial activity of macrolactins, whereas introduction of carbonyl group at C-15 decreases antimicrobial activity. The number of ring members and the presence of a hydroxy group at C-7 and C-9 position have no effect on the antibacterial activity [62]. Fradimycins A (132) and B (133) isolated from marine Streptomyces fradiae strain PTZ0025 displayed antimicrobial activity against S. aureus [79]. Macrolactins, polyene cyclic macrolactones possess a wide range of biological activities including antimicrobial, antiviral and anticancer. The majority of macrolactins are produced by Bacillus sp., whilst glycosylated macrolactins A1 (134) and B1 (135) are frequently isolated from marine Streptomyces species. The position of the hydroxyl group (OH) or the introduction of a keto group (C=O) in the macrolactone ring affects the antimicrobial activity of macrolactins, while the introduction of ester groups at the C-7 in macrolactins leads to better antibacterial activity. Compounds 134 and 135 were less active than their corresponding ether-containing macrolactins, probably due to the attachment of a sugar moiety at C-7 position of both compounds. However, glycosylated macrolactins are reported to have better polar solubility in comparison to the non-glycosylated macrolactins [61]. From the culture broth of a marine Bacillus sp., compounds including macrolactones (Macrolactin X-Z, 136-138) and macrolactinic acid (139) were isolated. These metabolites displayed potential antimicrobial activity against microbial strains, B. subtilis (KCTC 1021), E. coli (KCTC 1923), and S. cerevisiae (KCTC 7913). Macrolactins exhibit their antibacterial activity by inhibiting peptide deformylase in a dose-dependent manner. The position of hydroxy groups in the macrolactone ring is also important for antimicrobial activity of macrolactins. The hydroxy group at C-15 in macrolactone ring increases the antimicrobial activity of macrolactins, whereas introduction of carbonyl group at C-15 decreases antimicrobial activity. The number of ring members and the presence of a hydroxy group at C-7 and C-9 position have no effect on the antibacterial activity [62].

Isocoumarins
Antimicrobial dihydroisocoumarin derivatives penicisimpins A-C (146-148) were reported from a rhizosphere-derived fungus, Penicillium simplicissimum MA-332 obtained from a marine mangrove plant Bruguiera sexangula var. rhynchopetala (Figure 8). These isocoumarins possess a broad-spectrum of antibacterial and antifungal activities. Among these three isocoumarins, penicisimpins A (146) exhibited the greatest activity against E. coli, P. aeruginosa, V. parahaemolyticus, V. harveyi, and C. gloeosprioides (each with MIC value of 4 µg/mL) while the other two congeners (147 and 148) were only moderately active against these pathogens. It was determined that the methyl group attached at C-7 was responsible for the enhanced activity observed in compound 146 relative to 147, while the double bond at C-11 was responsible for the decreased activity (compound 146 vs. 148) [83]. The marine-derived fungal species Penicillium vinaceum has been reported to produce citreoisocoumarin (149) that displayed activity against S. aureus (19 mm inhibition zone) [47].

Miscellaneous Compounds
Several compounds which have not been classified under any of the above subheadings are discussed in this section and their chemical structures are illustrated in Figure 9. Terretrione A (151), terretrione C (152) and terretrione D (153) containing a 1,4-diazepane skeleton were isolated from an organic extract of the fungus Penicillium sp. CYE-87 derived from a tunicate-Didemnum sp. collected from the Suez Canal in Egypt. These compounds displayed antimicrobial activity against C. albicans but were inactive against S. aureus and E. coli [47,85]. Marine-derived fungi, Penicillium vinaceum species when investigated for antimicrobial compounds led to discovery of α-cyclopiazonic acid (154), which was active only against E. coli with an inhibition zone of 20 mm [47]. 5methoxydihydrosterigmatocystin (155), a compound isolated from the marine-derived fungus, Aspergillus versicolor MF359, isolated from a marine sponge of Hymeniacidon perleve was active against S. aureus and B. subtilis. The compound did not display activity against MRSA and P. aeruginosa as the MIC was found to be >100 μg/mL against both the organisms [86]. Caerulomycin A (156), an antifungal compound isolated from marine actinomycetes Actinoalloateichus cyanogriseus showed potent activity against Candida isolates, C. albicans and C. albicans CO9, and two fluconazole resistant strains namely, C. glabrata HO5Fl and C. krusei GO3. The MICs of Caerulomycin A was in the range of 0.39-1.26 μg/mL. Furthermore, the MIC values obtained for Caerulomycin A against fluconazole resistant C. glabrata were comparable with the MIC values obtained for Amphotericin B [87]. Pyridones, trichodin A (157) and pyridoxatin (158), were extracted from both the mycelia and the culture broth of the marine fungus, Trichoderma sp. strain MF106 obtained from the Greenland Seas. Compounds 157 and 158 possess moderate antibiotic activities against the Gram-positive B. subtilis, S. epidermidis, MRSA and yeast, C. albicans but were inactive against Trichophyton rubrum [88]. A

Miscellaneous Compounds
Several compounds which have not been classified under any of the above subheadings are discussed in this section and their chemical structures are illustrated in Figure 9. Terretrione A (151), terretrione C (152) and terretrione D (153) containing a 1,4-diazepane skeleton were isolated from an organic extract of the fungus Penicillium sp. CYE-87 derived from a tunicate-Didemnum sp. collected from the Suez Canal in Egypt. These compounds displayed antimicrobial activity against C. albicans but were inactive against S. aureus and E. coli [47,85]. Marine-derived fungi, Penicillium vinaceum species when investigated for antimicrobial compounds led to discovery of α-cyclopiazonic acid (154), which was active only against E. coli with an inhibition zone of 20 mm [47]. 5-methoxydihydrosterigmatocystin (155), a compound isolated from the marine-derived fungus, Aspergillus versicolor MF359, isolated from a marine sponge of Hymeniacidon perleve was active against S. aureus and B. subtilis. The compound did not display activity against MRSA and P. aeruginosa as the MIC was found to be >100 µg/mL against both the organisms [86]. Caerulomycin A (156), an antifungal compound isolated from marine actinomycetes Actinoalloateichus cyanogriseus showed potent activity against Candida isolates, C. albicans and C. albicans CO9, and two fluconazole resistant strains namely, C. glabrata HO5Fl and C. krusei GO3. The MICs of Caerulomycin A was in the range of 0.39-1.26 µg/mL. Furthermore, the MIC values obtained for Caerulomycin A against fluconazole resistant C. glabrata were comparable with the MIC values obtained for Amphotericin B [87]. Pyridones, trichodin A (157) and pyridoxatin (158), were extracted from both the mycelia and the culture broth of the marine fungus, Trichoderma sp. strain MF106 obtained from the Greenland Seas. Compounds 157 and 158 possess moderate antibiotic activities against the Gram-positive B. subtilis, S. epidermidis, MRSA and yeast, C. albicans but were inactive against Trichophyton rubrum [88]. A pyridinium compound, 1-(10-Aminodecyl) Pyridinium (159), isolated from the marine actinomycete, Amycolatopsis alba var. nov. DVR D4 demonstrated antimicrobial activities against Gram-positive and Gram-negative bacteria with MICs in the range of 70-160 µg/mL [89]. A dibenzofuran derivative porric acid D (160) and altenusin (161) were isolated from the methanol extract of the marine derived fungus, Alternaria sp., isolated from the Bohai Sea. These compounds were reported to have antibacterial activity against S. aureus [90]. Several small molecules including p-hydroxybenzoic acid (162), trans-cinnamic acid (163) and N-hydroxybenzoisoxazolone (164) were isolated from Pseudoalteromonas flavipulchra and showed antibacterial activity against Vibrio anguillarum. Greater growth inhibition was observed in V. anguillarum when a mixture of all three compounds was used compared to when each of the compounds was used individually, suggesting that they may act in a synergistic manner [72]. The antimicrobial compounds, bacilosarcin B (165) and C (166) and amicoumacin A (167) and B (168), were isolated from the culture broth of a marine-derived bacterium B. subtilis. The C-12 amide group of amicoumacin was found to be crucial for antibacterial activity against MRSA based on structural comparisons of amicoumacin A (167) and amicoumacin B (168). It was observed that only the compound amicoumacin A (167) with C-12 amide groups exhibited antibacterial activities against B. subtilis, S. aureus and L. hongkongensis, which strongly supported the idea that the C-12 amide group of amicoumcin acts as a pharmacophore in antibacterial activities. This conclusion was further supported by comparing antibacterial activities between compounds 167/168 and 165/166. Compound 167 exhibited antibacterial activities against B. subtilis, S. aureus and L. hongkongensis, which were about six-fold higher than those of 168 [91].

Synthetic Interventions
Although naturally occurring marine natural products are bestowed with interesting structural features (both chemical and stereochemical), some pharmacophoric modifications are still required to improve their biological efficacy [93]. In this respect, several synthetic chemists are engaged in tailoring these natural products to obtain new chemical entities by modifying their natural structures [94] (Figure 10). Recently, Sakata et al. identified isatin (170), an algicidal substance produced by the marine bacterium Pseudomonas sp. C55a-2 isolated from coastal sea water of Kagoshima Bay in Japan and targeted this compound for synthetic modification. This particular compound was chosen as many strains belonging to the genera Pseudomonas, Alteromonas and Pseudoalteromonas sp. have previously been reported to use chemical defences in the form of extracellular agents like isatin. With background knowledge of the antifungal activities of isatin, several structural modifications were made to the compound including bromination of the C-5 carbon of the isatin ring, altering the length of its alkyl chain and N-protections, resulting in the development of molecules with potentially better pharmacophoric features. Structure-activity relationships revealed that a bromine substitution at the

Synthetic Interventions
Although naturally occurring marine natural products are bestowed with interesting structural features (both chemical and stereochemical), some pharmacophoric modifications are still required to improve their biological efficacy [93]. In this respect, several synthetic chemists are engaged in tailoring these natural products to obtain new chemical entities by modifying their natural structures [94] ( Figure 10). Recently, Sakata et al. identified isatin (170), an algicidal substance produced by the marine bacterium Pseudomonas sp. C55a-2 isolated from coastal sea water of Kagoshima Bay in Japan and targeted this compound for synthetic modification. This particular compound was chosen as many strains belonging to the genera Pseudomonas, Alteromonas and Pseudoalteromonas sp. have previously been reported to use chemical defences in the form of extracellular agents like isatin. With background knowledge of the antifungal activities of isatin, several structural modifications were made to the compound including bromination of the C-5 carbon of the isatin ring, altering the length of its alkyl chain and N-protections, resulting in the development of molecules with potentially better pharmacophoric features. Structure-activity relationships revealed that a bromine substitution at the C-5 carbon atom in isatin derivatives led to a decrease in antibacterial activity when compared structurally to the parent isatin molecule. The antibacterial activity remains unchanged for N-methyl and N-butyl isatin derivatives, however, the addition of a free NH group to the structure of these compounds results in a decrease in antibacterial activity. Hence, it can be deduced that the free NH moiety of isatin is necessary for its potent inhibitory activities against fouling bacteria. The antibacterial activity was found to be better in the class of compounds wherein acetonyl moiety is introduced as functionality, compared to more extended hydrophobic benzoyl, as well as electron donating ethoxy group. Compounds 171-173 from this group exhibited a greater inhibitory effect compared to the parent compound 170 and to the N-protected isatins 174-179 and the 5-bromoisatin derivatives 180-182. The presence of a 3-acetonylidene group and a free NH moiety in compounds 171-173 were determined to be crucial structural elements responsible for enhancing the antibacterial activity of these compounds [95].
Bromopyrrole alkaloids are produced by marine sponges and possess an array of diverse biological activities, including antimicrobial and antineoplastic activities [96][97][98]. Many of these molecules are readily identifiable by a 4,5-dibromopyrrole ring contained within their structure, with oroidin (183) being the best characterized of these alkaloids noted for its antibiofilm activity [99] In  (183), from sponges of the genus, Agelas, which possess significant antimicrobial activity against the bacterial strains E. faecalis, S. aureus and E. coli and C. albicans. The research group synthesized several derivatives using oroidin as a template. The most bioactive of all these derivatives was found to be 4-phenyl-2-aminoimidazole (197), which exhibited an MIC 90 value of 12.5 µM against Gram-positive bacteria and 50 µM against E. coli [103].
A new marine-derived monoterpenoid compound, penicimonoterpene (+)-1 (198), was isolated from Penicillium chrysogenum QEN-24S in 2014 and had shown antifungal activity against A. brassicae and potent antibacterial activity against marine bacteria Aeromonas hydrophila, V. harveyi, and V. parahaemolyticus. This activity pattern encouraged chemists to develop a number of derivatives of compound 198. Modifications focused on variation of the substituents at the C-8 position, the carbon-carbon double bond at the C-6/7 position, and carboxyl substituents at the C-1 position. Compounds 199-203 were synthesized according to these modifications and found to be particularly active against F. graminearum. It was determined that oxidation of the methyl to a hydroxymethyl group at the C-8 position or replacement of the methyl ester group at C-1 by an ethyl ester significantly increased the antifungal activity of these compounds. Compounds with a reduced double bond at C-6/7 also showed better inhibitory activities against gloeosporioides and F. graminearum except for those containing an aldehyde group [104].

Genome Mining of Marine Microorganisms-The Future of Antimicrobial Discovery
Antimicrobial compound discovery has traditionally mainly relied on bioassay-guided approaches involving the cultivation of microorganisms under a variety of growth conditions, the subsequent screening of culture extracts for bioactivity and chemical characterization of the compounds produced. Over time however, this approach has led to the frequent re-isolation of known compounds, resulting in a drastic decline in research efforts being undertaken by research groups and pharmaceutical companies [105], causing a significant deficit in the number of novel,

Genome Mining of Marine Microorganisms-The Future of Antimicrobial Discovery
Antimicrobial compound discovery has traditionally mainly relied on bioassay-guided approaches involving the cultivation of microorganisms under a variety of growth conditions, the subsequent screening of culture extracts for bioactivity and chemical characterization of the compounds produced. Over time however, this approach has led to the frequent re-isolation of known compounds, resulting in a drastic decline in research efforts being undertaken by research groups and pharmaceutical companies [105], causing a significant deficit in the number of novel, natural products available for commercial and medicinal use. In this age of antimicrobial resistance, the demand for functionally diverse, unique antimicrobials has never been greater. Fortunately, advances in sequencing and '-omics' based technologies have revived the field of natural product discovery in recent years, owing in large part to the cost effectiveness and speed associated with next-generation sequencing. With more than 99,000 sequenced bacterial genomes currently publically available in the NCBI database [106] and sequence data from thousands of metagenome projects which can be accessed at the Genomes OnLine Database [107], researchers are now focusing their efforts on genome-guided investigations as a complementary approach to traditional bioactivity-guided methods in an effort to expedite the identification of 'talented' microbes, which are likely to possess the biosynthetic machinery typically associated with antimicrobial compound production.
Biosynthetic gene clusters (BGCs) are specialised groups of genes located in close proximity to each other in bacterial genomes and encode successive steps in the biosynthesis of natural products. Sequence-based detection, analysis and functional elucidation of these clusters are paramount to unlocking the true biosynthetic potential which resides within a microorganism. However functional elucidation of the products of these BGCs is not always easy, since most are either poorly expressed or not expressed at all under common laboratory culture conditions. Such was the case observed for the model antibiotic-producing actinobacterium, Streptomyces coelicolor, which, prior to having its entire genome sequenced in 2002 [108] was thought to contain BGCs responsible for the production of six distinct metabolites, previously identified by classic molecular genetic approaches [109]. However, upon inspection of the organism's complete genome sequence, 16 additional BGCs including BGCs potentially encoding nonribosomal peptide synthetases (NRPSs) and polyketide synthases (PKSs) were identified by bioinformatics-based predictions [110,111]. These clusters were deemed likely to encode enzymes involved in the synthesis of polyketides and nonribosomal peptides which are considered to be among the more valuable classes of microbial secondary metabolites from a biopharmaceutical perspective [112,113].
At the time of writing this review, data pertaining to over one million putative gene clusters currently resides in the Atlas of Biosynthetic gene Clusters which forms part of the Integrated Microbial Genomes component of the Joint Genome Institute (JGI IMG-ABC) [114]. With such a wealth of sequence information available, the challenges researchers now face primarily centre on how to effectively mine such a huge quantity of data in order to rapidly identify biosynthetically 'talented' microorganisms and furthermore, how to prioritize which BGCs to investigate among such a vast collection of uncharacterized clusters when targeting a desired anti-microbial bioactivity.
In this respect the Secondary Metabolite Bioinformatics Portal (SMBP), functions as a useful access point for investigators; containing website links to databases and tools used in genome mining and secondary metabolism research [115]. Automated tools such as antiSMASH [116] BAGEL [117] and PRISM [118] and databases such as Bactibase [119] ClusterMine360 [120] and MIBiG [121] represent just a few of the sophisticated in silico analytical tools which have quickly established themselves as the go-to resources for BGC identification, characterization and comparison. antiSMASH (antibiotics and Secondary Metabolite Analysis Shell) in particular is one of the most extensively used open-source BGC mining tools. Integrating and cross-linking several analysis tools including BLAST+, HMMer3 and FastTree, this computational platform not only facilitates the detection of known BGCs but can also detect unknown, BGC-like regions in genomes via ClusterFinder [122] an algorithm which uses Pfam domain, pattern based predictions to detect putative BGCs. ClusterFinder works on the premise that even the biosynthetic pathways for unknown compounds are likely to use the same broad families of enzymes for the catalysis of key reactions. The antiSMASH framework has been continuously developed [116,123] since its launch in 2010 [124]. antiSMASH version 4.0 was recently released in April 2017 [125] and contains improvements in prediction software for specialized secondary metabolites, including ribosomally synthesized and post-translationally modified peptides (RiPPs) and terpene products and has been expanded from BGC mining in bacteria and fungi to BGC mining in plants [126]. Another recently developed tool is EvoMining, which integrates evolutionary concepts related to the emergence of natural product biosynthesis into genome mining. It is a newly developed phylogenomics approach toward BGC analysis and has been successfully used in the retrieval of arseno-organic BGCs, which were previously unobtainable via antiSMASH analysis [127].
With respect to the specific identification of antibacterial compounds and the pathways involved in their biosynthesis, genome mining approaches have proven useful in the identification of these types of compounds from marine derived bacteria. Figure 11 highlights a number of antimicrobial compounds discovered via genome mining approaches. The gene cluster involved in the biosynthesis of heronamide F (204) was identified following genome scanning of the deep-sea derived Streptomyces sp. SCSIO 03032. Heronamides are polyketide macrolactams which belong to a class of potent antifungal metabolites that are produced by marine-derived actinomycetes [128]. Confirmation of the involvement of this cluster in the biosynthesis of heronamide F was achieved by the functional confirmation of one of the genes in the cluster, namely herO, encoding a cytochrome P450 by gene knockout experiments [129]. Another example resides with three cyclohexapeptides, destomides B-D (205-207) which were initially isolated from the marine microbe Streptomyces scopuliridis SCSIO ZJ46. Destomide B was found to display antibacterial activity against strains of Streptococcus pneumonia, Staphylococcus aureus and methicillin-resistant Staphylococcus epidermidis (MRSE) [68]. Genome mining identified a putative 39-kb desotamide dsa gene cluster in the S. scopuliridis strain, and was determined to contain three NRPS genes. Subsequent heterologous expression of the dsa gene cluster in the heterologous host S. coelicolor M1152 confirmed its involvement of the biosynthesis of desotamides [130]. Finally, another recent example involving ribosomally synthesized and post-translationally modified peptides (RiPPs), centres on the new cinnamycin-like lantiobiotic, mathermycin (208) which has recently been isolated from the marine-derived strain, Marinactinospora thermotolerans SCSIO 00652 and possesses antimicrobial activities towards Bacillus subtilis [131]. Genome mining of the strain facilitated the identification of the BGC for mathermycin, while subsequent expression of the gene cluster in the heterologous Streptomyces lividans host system allowed its antibacterial activity to be assessed.
Another popular approach in the identification of novel marine derived antimicrobials is to couple genomics with metabolomics, whereby genome mining together and metabolic profiling are used to identify novel natural products in marine microorganisms. Our group has previously employed this approach to identify bioactive compounds from Streptomyces sp. (SM8) isolated from the sponge Haliclona simulans which displayed antibacterial and antifungal activity [132]. Similarly, Paulus and co-workers employed both metabolomic and genomic profiling approaches on the marine Streptomyces sp. MP131-18 to identify a number of new biologically active compounds, including two new members of the bisindole pyrroles spirorindimycins, spiroindimicin E (209) and F (210). In addition, two new members of the α-pyrone lagunapyrone family, namely lagunapyrone D and E were also identified using similar approaches [133]. Another example of the successful use of this coupled approach resides in marformycins A-F (211-216), which display selective anti-microbial activity against Micrococcus luteus, Propioniobacterium acnes and Propionobacterium granulosum. The marformycin gene cluster was identified following genome scanning of the deep-sea sediment-derived Streptomyces drozdowiczii SCSIO 10141 strain. Confirmation of the involvement of this cluster in the biosynthesis of this group of cyclic peptides was achieved following in vivo inactivation studies coupled with metabolite identification [134].
Other techniques have also been developed to expedite the identification of BGCs, which include mining for the presence of self-resistance mechanisms within these gene clusters, allowing investigators to deduce the antibiotic compounds which an organism is likely to produce. Resistance mechanisms are characteristic traits associated with antibiotic producers, enabling an organism to avoid suicide from self-toxicity following the biosynthesis of its own molecule [135]. Resistance mechanisms include enzymes which degrade toxic compounds, efflux pumps for the effective removal of unwanted substances from the cell and target modification [136]. Wright and colleagues first used resistance as a discriminating criterion in 2013, demonstrating that organisms resistant to glycopeptide and ansamycin antibiotics are more likely to produce similar compounds [137]. Following this resistance based hypothesis increased the discovery rate of producers of the aforementioned antibacterial compounds by several orders of magnitude. Harnessing the success of this approach, Wright and colleagues further devised a method for isolating scaffold-specific antibacterial producers by taking advantage of the innate self-protection mechanisms employed by the producing organisms, isolating strains in the presence of a selective antibiotic [138]. In 2015, Moore and colleagues developed a target-directed genome mining method to identify BGCs [139]. As previously mentioned target modification is one of several resistance strategies employed by antibiotic producing bacteria and is effective in correlating an antibiotic to its mode of action. Since it is common for antibiotic producing bacteria to mutate and duplicate genes encoding proteins for resistance, Moore's group reasoned that identifying target-duplicated genes which are co-clustered with BGCs would provide valuable information pertaining to the molecular targets of the BGC products without any prior knowledge of the molecule synthesized. Since antibiotic targets are often the product of housekeeping genes, Moore and colleagues screened 86 Salinispora bacterial genomes for duplicated copies of housekeeping genes and related them to their presence in BGCs. Using this approach they successfully identified a duplicated fatty acid synthase in the direct vicinity of an orphan hybrid PKS-NPRS gene cluster and prioritized this for investigation. Following cloning, heterologous expression and mutational analysis, the authors linked the gene cluster to the biosynthesis of thiolactomycin (217) a previously characterized fatty acid synthase inhibitor, and to the production of a group of unusual thiotetronic acid natural products [139]. In 2016, Oakley and colleagues provided experimental validation for target-directed genome mining in a fungal BGC system [140]. The group identified a proteasome subunit-encoding gene within a gene cluster in Aspergillus nidulans and hypothesized that the cluster may be responsible for the production of a proteasome inhibitor. Following a number of molecular genetic based strategies, the investigators determined that the product of the cluster was indeed a proteasome inhibitor, fellutamide B (218). Recently, Johnston and co-workers used resistance-based mining to predict natural products with new modes of action [141]. The authors used a retrobiosynthetic algorithm to mine biosynthetic scaffolds and resistance determinants to identify structures with unknown modes of action. Using this approach, the investigators determined that the telomycin family of natural products from Streptomyces canus possess a new antibacterial mode of action which targets cardiolipin, a bacterial phospholipid.
Tracanna and co-workers recently suggested another strategy as a means of prioritizing BGCs which may encode novel antibiotics, based on synergistic interactions [142]. Natural products can occur as synergistic pairs in nature, as in the case of cephamycin and clavulanic acid, two compounds which are naturally produced by Streptomyces clavuligerus. The BGCs for these compounds are intertwined in a 'supercluster' configuration [143,144]. This genetic conformation inspired the production of the antibiotic, Augmentin ® , which is comprised of a combination of the β-lactam antibiotic, amoxicillin with the β-lactamase inhibitor, clavulanic acid. Synergistic pairs of natural products are particularly attractive commodities in the ongoing attempt to tackle antimicrobial resistance, as it is more difficult for a pathogen to develop resistance to both compounds. They suggest that thorough analysis surrounding the evolutionary history of the genes associated with large, hybrid BGCs may provide a valuable insight into whether these BGCs are in fact 'superclusters', responsible for the production of a number of different compounds which may be synergistic.
It is important that in silico analytical tools keep pace with these genome-guided discovery strategies in order to expedite novel compound discovery. The Antibiotic Resistant Target Seeker (ARTS) is one such web tool which uses three criteria to detect known resistance genes as well as putative resistance house-keeping genes in actinobacterial genomes i.e., (i) duplication (ii) evidence for horizontal gene transfer (HGT) and (iii) localization within a BGC [145]. This inexpensive, computational analysis tool, facilitates high-throughput screening of bacterial genomes. Although the current focus of ARTS is on the analysis of actinobacterial genomes, the pipeline also works for other phyla and is being expanded to include reference sets for other taxa. Several other useful resources are available pertaining to the field of antimicrobial resistance, with the Comprehensive Antibiotic Resistance Database (CARD) [146] and Antibiotic Resistance Gene-ANNOTation (ARG-ANNOT) being the two most extensive databases [147].

Concluding Remarks and Future Prospects
The marine environment is home to a vast number of macro and microorganisms with untapped biosynthetic activities which are used to a large extent to ensure their survival in this diverse and often hostile habitat. This unique environment facilitates the biosynthesis of an array of secondary metabolites which act as chemical defenses and display a broad range of antimicrobial bioactivities. Nevertheless, despite extensive structural and stereo chemical diversity, only seven marine-derived metabolites have to date been approved as drugs, while 12 MNPs (or derivatives thereof) are currently in different phases of clinical trials.
As previously noted, none of the newly discovered marine natural products have as yet progressed to clinical trials, although a few of them are in preclinical studies. The slower pace of MNPs towards clinical trials is due to several factors that hinder their development as clinical agents. One of the major factors is the "continuous supply". Large quantities of a compound are required to

Concluding Remarks and Future Prospects
The marine environment is home to a vast number of macro and microorganisms with untapped biosynthetic activities which are used to a large extent to ensure their survival in this diverse and often hostile habitat. This unique environment facilitates the biosynthesis of an array of secondary metabolites which act as chemical defenses and display a broad range of antimicrobial bioactivities. Nevertheless, despite extensive structural and stereo chemical diversity, only seven marine-derived metabolites have to date been approved as drugs, while 12 MNPs (or derivatives thereof) are currently in different phases of clinical trials.
As previously noted, none of the newly discovered marine natural products have as yet progressed to clinical trials, although a few of them are in preclinical studies. The slower pace of MNPs towards clinical trials is due to several factors that hinder their development as clinical agents. One of the major factors is the "continuous supply". Large quantities of a compound are required to carry out biological assays to determine the site of action, specific targets, selectivity of the compound and its cytotoxicity. Irrespective of the potential applications of a functionally promising compound, a significant challenge faced by researchers is that several hundred grams of the compound are required for preclinical development, and multi kilogram quantities required for clinical trials. This is often one of the major bottle-necks in the development of MNP for clinical applications. Synthetic chemists around the world however are continuing to develop synthetic and semi-synthetic strategies to help overcome the supply issue surrounding MNPs, in an effort to satisfy the requirements to help bring these molecules to the preclinical stage and eventual development for use in the commercial or medicinal arenas.
MNPs can be chemically modified with various biosteric structural units to develop 'drug-like' molecules. Also, developments in mariculture (farming the growth of the organism in its natural environment) and aquaculture (culturing an organism under artificial conditions) have been attempted in order to solve the problem of sustainable supply of macroorganisms. However, the unique and sometimes exclusive conditions of the sea make cultivation or maintenance of isolated samples still very challenging and often impossible.
The preclinical pipeline also demands elaborate, mechanistic and pharmacokinetic studies to develop tailored MNPs, which in itself is a hugely challenging but nevertheless exciting task. Regardless of these challenges, the preclinical pipeline continues to supply studies with several hundred novel bioactive marine compounds with the potential for use as therapeutics. From a global perspective, the marine pharmaceutical pipeline remains very active, and now appears to have sufficient momentum to deliver additional antimicrobial compounds to the marketplace in the near future. The efficiency of various marine compounds against pathogens is very encouraging and there is no doubt that their exploitation and application will continue to develop. Genome mining has ushered in a renaissance in the field of natural product discovery, providing new hope in the ongoing search for novel antimicrobial compounds. This strategy allows researchers to harness the true biosynthetic potential which resides within diverse groups of marine microorganisms and offers an invaluable insight into not only the biosynthetic, but also the evolutionary and defensive strategies these organisms employ in the marine environment. Collaborative endeavours involving marine natural products chemistry with organic chemistry, medicinal chemistry, pharmacology, biology, bioinformatics and associated disciplines will help to ensure and facilitate an increase in marine natural product reaching the market as antimicrobial therapeutics.