Nonribosomal Peptides from Marine Microbes and Their Antimicrobial and Anticancer Potential

Marine environments are largely unexplored and can be a source of new molecules for the treatment of many diseases such as malaria, cancer, tuberculosis, HIV etc. The Marine environment is one of the untapped bioresource of getting pharmacologically active nonribosomal peptides (NRPs). Bioprospecting of marine microbes have achieved many remarkable milestones in pharmaceutics. Till date, more than 50% of drugs which are in clinical use belong to the nonribosomal peptide or mixed polyketide-nonribosomal peptide families of natural products isolated from marine bacteria, cyanobacteria and fungi. In recent years large numbers of nonribosomal have been discovered from marine microbes using multi-disciplinary approaches. The present review covers the NRPs discovered from marine microbes and their pharmacological potential along with role of genomics, proteomics and bioinformatics in discovery and development of nonribosomal peptides drugs.


MARINE ECOSYSTEM RESOURCES FOR NEW DRUG DISCOVERY
The marine ecosystem is most complex and largest aquatic systems on earth. It includes oceans, intertidal ecology, salt marsh, lagoons, estuaries, coral reefs, mangroves, deep sea, sea floor etc. Marine ecosystem has a enormous variety of organisms that are different in their physiology and adaptations and most of the marine life is found in coastal habitats (Hedgepeth, 1957). According to the Global Biodiversity Assessment by the United Nations Environment Program, oceans consist of 178,000 marine species in 34 phyla. It is estimated that 10 2 fungi, 10 3 bacteria and 10 7 viruses are likely to exist in one milliliter of seawater (Kubanek et al., 2003). Marine organisms comprise around 50% of the total biodiversity on earth. These organisms have shown remarkable contribution in the discovery and production of novel biomolecules (Jimeno et al., 2004;Vignesh et al., 2011). During 1981-2002 50% of US-FDA approved drugs are reported from either marine bioactive compounds or their synthetics analogs (Vinothkumar and Parameswaran, 2013). Cytosine arabinoside, Ara-C (anticancer) and adenine arabinoside, Ara-A (antiviral) were first discovered in the early 1950s and approved by Food and Drug Administration (US-FDA). These drugs were isolated from Caribbean sponge (Cryptotheca crypta), as spongouridine and spongothymidine. Blunt et al. (2015) reported more than 20,000 natural bioactive compounds have been obtained from marine environment in last 50 years (Blunt et al., 2015). Out of these 9 were approved as drugs and many of them are still in clinical trials. It is well documented that more than 50% of drugs that are in clinical use today belong to the nonribosomal peptides or mixed polyketide-NRP families (Hranueli et al., 2010;Agrawal et al., 2016; Table 1). Marine microbes contributes 70% of discovery of NRPs with antimicrobial, antiviral, cytostatic, immunosuppressant, antimalarial, antiparasitic, animal growth promoters and natural insecticides activities etc. (Vinothkumar and Parameswaran, 2013). Which makes marine microbial an important bioresource for getting NRPs with numerous pharmaceutical applications. The examples of some NPR based drugs which are now in the market are Daptomycin (antibiotics), Bleomycin (antitumor), Bacitracin (antibiotics for skin infections), Cyclosporin (antifungal and immunosuppressant drugs) (Figure 1) (Strieker et al., 2010). Norine is the first database entirely dedicated to NRPs and contains more than 1186 entries (Caboche et al., 2008(Caboche et al., , 2009. In this review we focus on antimicrobial and anticancer NRPs reported from marine microbes with their biological targets.

NONRIBOSOMAL PEPTIDE AND THEIR BIO COMBINATORIAL SYNTHESIS
An extensive literature on biosynthesis of non-ribosomal peptides is available in previous reviews (Sieber and Marahiel, 2003;Finking and Marahiel, 2004;Caboche et al., 2009;Strieker et al., 2010;Pfennig and Stubbs, 2012). Here we just summarized how NPRs are synthesized biologically, biomolecular structural architecture and enzymatic machinery of non-ribosomal peptide synthetases (NRPSs). NRPs are peptide secondary bioactive metabolites synthesized by a multi-modular enzyme complex called nonribosomal peptide synthetases (NRPSs) found only in bacteria, cyanobacteria and fungi (Matsunaga and Fusetani, 2003;Nikolouli and Mossialos, 2012). NRPs are formed from a series of enzymatic transformations employing a much more diverse set of precursors and biosynthetic reactions. NRPSs utilize both proteinogenic and nonproteinogenic amino acids (not encoded by DNA) as building blocks for the growing peptide chain (Finking and Marahiel, 2004;Felnagle et al., 2008). Moreover, these secondary bioactive metabolite peptides contain unique structural features, such as D-amino acids, N-terminally attached fatty acid chains, N-and C-methylated residues, N-formylated residues, heterocyclic elements, and glycosylated amino acids, as well as phosphorylated residues etc.; (Sieber and Marahiel, 2003). As a result, NRPs exhibit a broad spectrum of biological activities, ranging from antimicrobial to anticancer (Hur et al., 2012). The macrocyclic structure is a common feature of nonribosomally synthesized bioactive peptides, which is responsible for reduction in structural flexibility and may, therefore, constrain them into the biologically active conformation (Sieber and Marahiel, 2003;Grünewald and Marahiel, 2006).
The discovery of NRPs began when Tatum and colleagues (Mach et al., 1963) provided first evidence that tyrocidine, a cyclic decapeptide produced by Bacillus brevis, was biosynthesized by a mechanism independent of the ribosome (Mankelow and Neilan, 2000). They found that protein synthesis in B. brevis was inhibited by using ribosome targeting antibiotics like chloramphenicol and chlortetracycline, however, the biosynthesis of tyrocidine was not obstructed by the same. Additional biochemical analyses demonstrated that gramicidin S, a cyclic decapeptide produced by B.brevis, did not include tRNA molecules or aminoacyl-tRNA-synthetases (Nikolouli and Mossialos, 2012; Figure 2). Further work by Lipmann established that the production of cyclic decapeptide, gramicidin is an ATP-dependent reaction, catalyzed by these enzymes incorporating amino acids in a two-step process by their modules and their respective domains. The first step involves release of pyrophosphate (PPi) and the second step releases adenosine monophosphate (AMP), with the end result being an amino acid covalently linked to the enzyme (Wu et al., 2003). These finding suggested that tyrocidine and gramicidin S peptide synthesis did not involve ribosomal machinery for their synthesis, which leads to discovery of the NRPs and NRPSs. These data also gave the first indication of an amino acid as a "carrier" being involved in NRPS enzymology (Felnagle et al., 2008;Condurso and Bruner, 2012; Figure 3).
The biosynthetic study of NRP compounds is challenging if we consider their complexity and biological activities. Each nonribosomal peptide synthetase is composed of an array of distinct modular sections, each of which is responsible for the incorporation of one defined monomer into the final peptide product. Biosynthesis of a nonribosomal peptide by NRPSs involves a series of repeating reactions that are catalyzed by the coordinated actions of modules and their core catalytic domains. Each enzyme module contains three catalytic domains: adenylation domain (A), peptidyl-carrier (PCP) domain and condensation domain (C). A final peptide product released from the enzyme through cyclization or hydrolysis that takes place by thioesterase domain (TE) which is located in the final NRPSs module (Figures 4A,B; Mankelow and Neilan, 2000;Finking and Marahiel, 2004). For recent example, Thiocoraline, an anticancer nonribosomal peptide (NRP) synthesis by marine bacteria Cromonospora marina contains peptidic backbone of two S-methylated Lcysteine residues. S-Methylation occurs very rarely in nature, and is observed extremely rarely in nonribosomal peptide scaffold. The four modules TioJ, TioO, TioR, and TioS of thiocoraline NRPSs are responsible for the thiocoraline-backbone biosynthesis. TioR and TioS would most probably constitute the NRPSs involved in the biosynthesis of the thiocoraline, according to the colinearity of the respective modules ( Figure 5; Lombó et al., 2006;Al-Mestarihi et al., 2014). The potentials of marine microbes to produce NRP's with antimicrobial and anticancer activity are reported in this review. The data referring to these activities are depicted in Tables 2-4 and the structures are given in Supplementary Materials (Figures S1-S17).  (Newman and Cragg, 2004;Fenical, 2006;Jimenez et al., 2009;Petit and Biard, 2013 effective prevention and treatment of an ever-increasing range of infections caused by them (Organization, 2014). Natural products are the principal source for primary health care. Natural products are observed as a diverse group of molecules which have evolved to interact with a wide variety of protein targets for specific purposes. Also the same protein structure with little or no variation serves different purposes in different organisms. As a result, it is believed that the search for novel antimicrobial entity from natural sources will yield better results than from combinatorial chemistry and other synthetic procedures (Ngwoke et al., 2011). Here we described NRPs from marine microbial sources with antimicrobial potential.

Cyanobacteria
Lobocyclamide B (31) ( Figure S3) a cyclododecapeptide containing five beta-hydroxy-alpha-amino acid residues, was discovered from Lyngbya confervoides which was active against fluconazole-resistant C. albicans. The absolute stereochemistry was determined by chiral chromatography of Marfey's reaction (MacMillan and Molinski, 2002). Brunsvicamides A-C (32-34) ( Figure S3), three new cyclic hexapeptides have been isolated from cyanobacterium Tychonema sp. Brunsvicamide C contains FIGURE 2 | Tyrocidine biosynthesis in bacteria B. brevis nonribosomal peptide synthetases of tyrocidine synthesis mainly consist, three NRPSs TycA, TycB, and TycC, which contain 10 modules (TycA comprises one module, TycB three, and TycC six modules) each of those responsible for the incorporation of a cognate amino acid into the growing chain with the help of their domains. The Te domain at the last module of TycC catalyzes peptide cyclization and thereby release of the final product (Mootz et al., 2000).
FIGURE 3 | The Gramicidin S biosynthetic machinery the enzymatic assembly consists of two NRPSs (GrsA and GrsB) and their modules, respectively. Each module is responsible for the incorporation of one monomeric amino acid. The thioesterase domain (TE domain) catalyzes the dimerization of two assembled pentapeptides and subsequent cyclization, resulting in gramicidin S (Hoyer et al., 2007). an N-methylated N'-formylkynurenine moiety. Brunsvicamide B selectively inhibits the Mycobacterium tuberculosis protein tyrosine phosphatase B (MptpB, IC 50 7.3 µM) (Müller et al., 2006).

Fungi
An extraction of a saltwater culture of an unidentifiable sponge-derived fungus leads to discovering two novel cyclic depsipeptides, Guangomides A (35) and B (36) ( Figure S4). Both compounds had weak antibacterial activity against Staphylococcus epidermis (MIC = 100 µg/mL, each) and Enterococcus durans (MIC = 100 µg/mL, each) (Amagata et al., 2006). A marine-derived Aspergillus fumigatus yielded to 11-O-methylpseurotin A (37) ( Figure S4) (PKS/NRPS), which selectively inhibited a Hof1 deletion strain of the yeast Saccharomyces cerevisiae (Boot et al., 2007). Marine-derived  (C-domain). The first modules always lacks a C domain and is used to initiate nonribosomal peptide synthesis, while those harboring a C-domain qualify for elongation and modules with thioesterase domains (TE) usually in the last domain, for termination of peptide product from enzyme through cyclization or hydrolysis (Prieto et al., 2012). (B) Mechanism of nonribosomal peptide (NRP) synthesis Adenylation domain (A) activates amino acid as aminoacyl-AMP and transfer to PCP domain which condenses coming amino acids by forming peptide bonds. Structural modifications mostly operate by epimerization domains which converts L-amino acid to D-amino acid and vice a versa. Peptide chain thus transfers to TE domain by transesterification reaction by PCP. Finally, TE domain catalyzed product release (NRPs) by either hydrolysis or macrocyclization (Condurso and Bruner, 2012).
fungus Emericella sp., and marine actinomycete Salinispora arenicola were co-cultured to induce production of Emericellamides A (38) and B (39) ( Figure S4) by fungi. Emericellamides A and B displayed modest antibacterial activities against MRSA with MIC values of 3.8 and 6.0 µM, respectively (Oh et al., 2007).

NRPS WITH ANTICANCER POTENTIAL
Cancer is the second leading cause of death worldwide. Present therapies cause serious side effects. Therefore there is need to employ alternative concepts including natural products therapy, complementary or alternative medicine, surgery, radiation therapy used alone or in combination to the prevention of cancer (Reddy et al., 2003). Here we focus on the marine natural products specially NRPs that have been evaluated for cancer prevention.
Urukthapelstatin A has also shown growth inhibition of human lung cancer A549 cells in dose-dependent manner with an IC value of 12 nM (Matsuo et al., 2007). The culture of Salinispora arenicola isolated from sea sediment (Great Astrolabe Reef, Fijiy) yielded three new cyclohexadepsipeptides, Arenamides A-C (75-77) ( Figure S6). The absolute structures and configuration of these compounds were established by the spectroscopic technique. Arenamides A (75) and B (76) blocked tumor necrosis factor (TNF)-induced activation with IC 50 values of 3.7 and 1.7 µM respectively. In addition, they also inhibited nitric oxide and prostaglandin E2 production and were moderately cytotoxic to HCT-116 cells (Asolkar et al., 2008). Bacillus silvestris that was isolated from a Pacific Ocean (southern Chile) crab yields two new cyclodepsipeptides, Bacillistatins 1-2 (78-79) ( Figure S6) with strong anti-cancer (GI 50 of 10 −4 -10 −5 µg/mL) activity (Pettit et al., 2009). The epimeric cyclic peptides Turnagainolides A (80) and B (81) ( Figure S6), isolated from marine Bacillus sp. (sediment, Turnagain Is., British Columbia, Canada), had indirect inhibitory effect on phosphatidylinositol-3-kinase (PI3K) pathway . A Streptomyces sp. obtained from marine sediment produced two highly modified linear tetrapeptides, Padanamides A (82), and B (83) ( Figure S6). They inhibit cysteine and methionine biosynthesis and are cytotoxic to Jurkat cells (IC 50 of 20 µg/mL) respectively . Chemical genomics was performed to discover the mode of action of compounds, which suggested that padanamide A inhibits cysteine and methionine biosynthesis.
Streptomyces sp. isolated from volcanic island produced new cyclic peptides Ohmyungsamycin A (84) and B (85) ( Figure S7). The presence of unusual amino acid units, including Nmethyl-4-methoxytrytophan, β-hydroxyphenylalanine, and N, N-dimethylvaline in compound (84-85) have been determined by interpretation of the NMR, UV, and IR spectroscopic and MS data. Both exhibited inhibitory activities against diverse cancer cells with IC 50 values ranging from 359 to 816 nM and 12.4 to 16.8 µM respectively. However, compound (84) was more active in this regard interestingly; these compounds exhibit relatively selective anti-proliferative activity against cancer cells compared to normal cells. This may be due to the consequence of genetic background or of the biologically various characteristics between cancer and normal cells. However, the exact molecular mechanism behind the selectivity should be further investigated (Um et al., 2013). Proximicins A-C (86-88) ( Figure S7) are novel aminofuran antibiotics with anticancer activity, isolated from marine strains of verrucosispora sp. Compounds (86-88) showed inhibitory activity against gastric adenocarcinoma (AGS, IG 50 = 0.6, 1.5, 0.25 µg/mL respectively), hepatocellular carcinoma (HepG2, IG 50 = 0.8, 9.5, 0.7 µg/mL respectively) and breast carcinoma cells (MCF 7, IG 50 = 7.2, 5.0, 9.0 µg/mL respectively). A cell-cycle analysis in AGS cells revealed that Proximicin C produced cell arrest in the G0/G1 phase after incubation for 24 h. After 40 h, there was an increase in the number of cells in the sub-G1 phase, that is, apoptotic cells (+2.9%). It was also found that proximicin C induce upregulation of p53 and of the cyclin kinase inhibitor p21 in AGS cells (Fiedler et al., 2008).

Cyanobacteria
An assemblage of Lyngbya majuscula and Phormidium gracile collected in Papua New Guinea produced a cyclic depsipeptide Hoiamide A (89) ( Figure S8). The highly unusual structure of hoiamide A synthesized by mixed peptide-polyketide biosynthetic pathway showed moderate cytotoxicity to cancer cells and partial agonist of site 2 on the voltage-gated sodium channel as it produced a rapid and concentration-dependent elevation of neuronal [Na + ] in neocortical neurons (IC 50 = 92.8 nM) (Pereira et al., 2009). An assemblage of the marine cyanobacteria L. majuscula and Schizothrix species collected from Fiji was the source of cyclic depsipeptides Yanucamides A (90) and B (91) ( Figure S8), which contain a 2, 2-dimethyl-3-hydroxyoct-7-ynoic acid moiety. Both compounds exhibited strong brine shrimp toxicity (LD 50 , 5 ppm) (Sitachitta et al., 2000). The cyclic depsipeptides named Lyngbyabellins A (92) (Figure S8), contain a 7,7dichloro-2,2-dimethyl-3-hydroxyoctanoic acid moiety have been isolated from the cytotoxic fraction of L. majuscula collected from Guam and the Dry Tortugas National Park, Florida. Compound (92) have moderate cytotoxicity against human nasopharyngeal carcinoma cell line (KB cells) and human colon adenocarcinoma cell line (LoVo cells), with IC 50 values of 0.03 and 0.50 µg/mL, respectively and also showed cellular microfilament network in A-10 cells at 0.01-5.0 µg/mL concentrations (Luesch et al., 2000). Another collection from Tortugas National Park, Florida was the source of cytotoxic and    antifungal cyclic depsipeptide Lyngbyabellin B (93) ( Figure S8). Lyngbyabellin B was toxic to brine shrimp (LD 50 = 3.0 ppm) (Milligan et al., 2000). A marine cyanobacterium Microcystis aeruginosa contained the cyclic hexapeptide Microcyclamide (94) (Figure S8), which showed moderate cytotoxicity against P388 murine leukemia cells at 24-30 µg/mL (Ishida et al., 2000). The cyanobacterium L. majuscule collected from Guam was the source of Apratoxin A (95) ( Figure S8). This cyclodepsipeptide of mixed peptide-polyketide biogenesis exhibited in vitro cytotoxicity against human tumor cell lines at IC 50 of 0.36-0.52 nM. Apratoxin A induces G1 phase cell arrest and apoptosis, which is at least particularly initiated through antagonism of FGF signaling via STAT3 (Luesch et al., 2001b). Another collection of L. majuscule from Guam gave two cyclic depsipeptides, Pitipeptolides A (96) and B (97) ( Figure S8) with anti-mycobacterial and weak cytotoxicity against LoVo cells with IC 50 values of 2.25 and 1.95 µg/mL, respectively. Pitipeptolides A and B also stimulated elastase activity. It is suggested that this activity is due to the presence of hydrophobic portions in the molecule (Luesch et al., 2001a). Marine cyanobacterium Lyngbya sp. collected from Palauan was the source of six new β-amino acid-containing cyclic depsipeptides, the Ulongamides A-F (98-103) ( Figure S8). All peptides were found to be weakly cytotoxic against KB and LoVo cells with IC 50 values of ca. 1 µM and ca. 5 µM respectively except compound Ulongamides F (Luesch et al., 2002). Examination of a L. confervoides collection from Saipan, Commonwealth of the Northern Mariana Islands, led to the isolation of a novel cytotoxic cyclic depsipeptide Obyanamide (104) ( Figure S8). Obyanamide was cytotoxic against KB cells with an IC 50 of 0.58 µg/mL. According to the results, the β-amino acid residue was found to play a critical role in the biological activities. Additionally, the ester bond along with the Ala (Thz) moiety was also essential for biological activities (Williams et al., 2002a). Malevamide D (105) ( Figure S8), a highly cytotoxic peptide ester have been isolated from marine cyanobacterium Symploca hydnoides (Horgen et al., 2002). A culture Symploca sp. yielded Tasiamide (106) (Figure S8), an acyclic peptide. Tasiamide demonstrated cytotoxic activity against KB and LoVo cells with IC 50 values of 0.48 and 3.47 µg/mL, respectively (Williams et al., 2002b). A new cytotoxic peptide Tasiamide B (107) ( Figure S9) which contain the unusual amino acid-derived residue 4-amino-3-hydroxy-5-phenylpentanoic acid (Ahppa) have been isolated from cyanobacterium Symploca sp. This peptide displayed an IC 50 value of 0.8 µM against KB cells (Williams et al., 2003b).
A Papua New Guinea collection of the marine cyanobacterium L. majuscule was the source of six cyclic depsipeptides, Guineamides A-F (108-113) ( Figure S9). The presence of betaamino or beta-hydroxy carboxylic acid residues in all peptides was determined using a combination of chemical manipulations as well as Marfey's method. Guineamides B and C showed moderate cytotoxicty to a mouse neuroblastoma cell line with IC 50 values of 15 and 16 µM, respectively (Tan et al., 2003b). A new bioactive cyclic depsipeptide, Homodolastatin 16 (114) ( Figure S9) have been isolated from L. majuscula, collected from Wasini Island off the southern Kenyan coast. Homodolastatin 16 showed moderate activity against oesophageal (IC 50 = 4.3 µg/mL) and cervical cancer cell lines (IC 50 = 1 µg/mL) (Davies-Coleman et al., 2003). An examination of an organic extract of a cyanobacterium L. majuscula, collected from Guam, led to the isolation of the cyclic peptide Lyngbyastatin 3 (115) ( Figure S9). The presence of two unusual amino acid units, 3-amino-2-methylhexanoic acid (Amha) and 4-amino-2, 2-dimethyl-3-oxopentanoic acid units (Ibu) was determined by standard methods. Lyngbyastatin 3 displayed in vitro activity against KB and LoVo cell lines with IC 50 values of 32 and 400 nM respectively (Williams et al., 2003a).
The marine cyanobacterium Okeania sp. collected from the coast near Jahana, Okinawa, was the source of Kurahyne B (198) ( Figure S14). It showed growth inhibition against HeLa and HL60 cells, with IC 50 values of 8.1 and 9.0 µM, respectively (Okamoto et al., 2015).

Fungi
A culture of marine fungi Fusarium CNL-619 was the source of a new cyclic depsipeptide N-Methylsansalvamide (199) (Figure S15), which showed weak in vitro cytotoxicity against NCI human tumor cell lines (GI 50 8.3 µM) (Cueto et al., 2000). An unidentified fungus isolated from the red alga, Ceradictyon spongiosum (Okinawa) have been shown to produce two linear dodecapeptides, Dictyonamides A (200) and B (201) ( Figure S15). Only the compound (200) showed inhibitory effect on cyclin-dependent kinase 4 with IC 50 value of 16.5 µg/mL (Komatsu et al., 2001). A culture of marine fungus, Scytalidium sp., collected from Bahamas was the source of two new cyclic heptapeptides Scytalidamides A (202) and B (203) ( Figure S15) and both compounds displayed moderate cytotoxicity to the HCT-116 cell line in vitro with IC 50 values of 2.7 and 11.0 µM, respectively (Tan et al., 2003a). A strain of Trichoderma virens was isolated from ascidian Didemnum molle and from the surface of a green alga of genus Halimeda from Papua New Guinea, which was the source of two modified dipeptides Trichodermamides A (204) and B (205) ( Figure S15). Trichodermamide B has showed significant in vitro cytotoxicity against HCT-116 cells (colon carcinoma) with an IC 50 of 0.32 µg/mL (Garo et al., 2003). A fungal strain Exserohilum rostratum associated with a marine cyanobacterial mat produced four moderately cytotoxic cyclic dipeptides Rostratins A-D (206-209) ( Figure S15). The structures and absolute configurations of peptides were determined by two-dimensional NMR techniques and Mosher method respectively. Compounds (206-209) exhibit activity against colon carcinoma (HCT-116) with IC 50 values of 8.5, 1.9, 0.76, and 16.5 µg/mL, respectively (Tan et al., 2004).
Marine microorganisms have been recognized as one of the most promising groups of organisms from which novel pharmacologically active molecules, with potential benefits against cancer, can be isolated. Recently, several compounds have been emerged as templates for the development of novel anticancer drugs. However the mechanisms implicated in the cytotoxicity of these compounds in tumor cell lines are still largely overlooked but several studies point to an implication in apoptosis. For instance, several compounds were found to inhibit cell growth in a large variety of cancer cell lines, the pathways by which cancer cells are inhibited are still poorly elucidated. In some cases, compounds were found to induce cell death by activation of the apoptotic process; nevertheless the mechanisms underlying the apoptosis still need more investigations. Some compounds were found to create an imbalance in cellular redox potential, with mitochondria representing a central role in the process. However, more studies are needed in order to clarify it. Cell cycle is another disturbed process, mainly due to disruption of the microtubules and actin filaments; however there are only a few studies connecting marine NRPs with alterations in cell cycle and more studies are needed in order to clarify the involvement of these compounds in the process. Even membrane sodium channels can establish interactions with the compounds, revealing its potentially important role in the observed effects. In summary, more investigations are needed in order to clarify the specific targets and the mechanisms that are behind cancer cell cytotoxicity, namely the involvement of the apoptotic process by the implication of functional genomics.

ROLE OF GENOMICS, PROTEOMICS AND BIOINFORMATICS IN DISCOVERY AND DEVELOPMENT OF NONRIBOSOMAL PEPTIDES DRUGS
The non-ribosomal peptides (NRPs) are an essential source of chemical diversity for drug discovery and development. At present, there are more than 1,164 different non-ribosomal peptides known in public database (NCBI) which consists of over 500 unique monomers, including both proteinogenic and non-proteinogenic L-and D-amino acids as well as carboxylic acids and amines (Caboche et al., 2010). Due to great structural diversity (linear, cyclic and branched or other complex primary structures) these complex secondary metabolites had impact on all therapeutic area, as making them suitable to be used as clinical agents. However, such potential NRPs often need to be modified to improve their clinical properties and/or bypass resistance mechanisms (Bush, 2012). For instance, FDA approved Oritavancin has been developed by using semisynthesis strategy from Vancomycin for treatment of drug resistant skin infections (Markham, 2014). Indeed, modification in the nucleotide sequence of a natural NRPS gene or combining modules of different NRPSs may potentially lead them to be more effective with unique pharmacological activity. However, this requires in-depth understanding of both the assembly line and the resulting products. Over the last few decades several bioengineering approaches have been developed to increase the yields of NRPs and generating modified peptides with altered bioactivity or improved physicochemical properties (Winn et al., 2016). Earlier, biosynthetic generation of novel NRPs analogs focused on precursor directed biosynthesis (PDB) or mutasynthesis. In PDB, a wild-type organism is provided with modified or synthetic amino acids with the prospect that the substrate specificity of the relevant NRPS shall be flexible enough to allow addition of the modified precursors into the final peptide. However, mutasynthesis is the exact opposite. The modified substrates are fed to an engineered organism which lacks the enzyme(s) required for the biosynthesis of a specific natural precursor, so that a modified substrate or precursor analog may be effectively incorporated (Weist et al., 2004). These methods are important because they generate natural product analogs rapidly.
In earlier reviews many examples of precursor directed biosynthesis of NRPs are available (Thiericke and Rohr, 1993). Other methods being adopted for the production of new nonribosomal peptides is engineering of precursor supply in vivo or introducing tailoring enzymes from other pathways with new glycosylation, halogenation and sulfation enzymes being applied outside of their native clusters to create structural diversity. Although it's similar to precursor directed biosynthesis, it focuses mainly on endogenous biosynthesis rather than exogenous feeding. The introduction of halogen unit into NRP scaffolds has been a common target. For example, when the enzyme PrnA (a favin-dependent tryptophan-7-halogenase) from Pseudomonas fuorescens Pf-5 was expressed alongside the NRPS genes for the uridyl peptide antibiotic pacidamycin, produced by Streptomyces coeruleorubidus, a new halogenated analog was generated (Roy et al., 2010). Using such a technique enduracidin analogs have been produced by altering halogenase in wild-type Streptomyces fungicidicus (Yin et al., 2010). An alternative but complicated strategy has also been developed to generate novel NRPs. It exchanges NRPS subunit, module, and domain of the core peptide itself. Initially, this method was applied by Cubist Pharmaceuticals for the development and marketing of nonribosomal peptide antibiotic daptomycin, first natural product antibiotic that gained approval for clinical use in over 30 years (Baltz et al., 2006). Unfortunately, Cubist Pharmaceuticals failed to identify any daptomycin variants with better antibacterial property than parent daptomycin. Another route that has also been explored which involves modifying the length of the peptide chain by deletion or insertion of one or more modules (Mootz et al., 2002;Butz et al., 2008). A recent study indicates that the introduction of individual or combined point mutations in the binding pocket of an NRPS adenylation domain generates new diversity of NRPs (Han et al., 2012).
A latest technique called heterologous expression offer considerable promise especially for natural hosts which are slow growing, genetically difficult to handle, unculturable, or even unknown. The transfer of biosynthetic genes from the original microbial organisms to more amenable heterologous host bacteria is more amenable to large-scale fermentation production would overcome the limitation of procurement of the drug from the ocean (which is currently limited to expensive aquaculture or field harvesting) and ensure supply (Ongley et al., 2013). The gene cluster responsible for polyketide epothilone (a potential anticancer agent) biosynthesis in the myxobacterium Sorangium cellulosum was cloned and completely sequenced by Tang et al. (2000). Concomitant expression of these genes in the actinomycete Streptomyces coelicolor produced epothilones A and B (Tang et al., 2000). After this heterologous expression system portends a plentiful supply of this medically relevant agent. Similarly A novel gene (amyZ) encoding a cold-active and salt-tolerant α-amylase (AmyZ) was cloned from marine bacterium Zunongwangia profunda (MCCC 1A01486) and the protein was expressed in Escherichia coli (Qin et al., 2014). The Ptchi19 gene of the marine Pseudoalteromonas tunicata CCUG 44952T was cloned and expressed in E.coli (García-Fraga et al., 2015). A new κ-carrageenase gene from marine bacterium Zobellia sp. ZM-2 was cloned and expressed in E.coli . Heterologous expression of the barbamide biosynthetic gene cluster from the marine cyanobacterium Moorea producens in the terrestrial also led to the production of a new barbamide congener 4-O-demethylbarbamide . The biosynthetic pathway for bacitracin was successfully transferred from Bacillus licheniformis to the related species B. subtilis (Eppelmann et al., 2001). The polyketide biosynthesis pathway for the marine-derived telomerase inhibitor griseorhodin A was productively transferred to Streptomyces lividans from an environmental Streptomyces isolate (Li and Piel, 2002). Ugai et al. (2016) got success in heterologous expression of the cryptic gene cluster found in A. solani to obtain a marinederived antifungal agent didymellamide B from the A. oryzae transformant introducing PKS-NRPS, trans-ER, and P450 genes asolSCA (Ugai et al., 2016). Likewise many other successful examples are available in literature (Fortman and Sherman, 2005;Luo et al., 2016;Winn et al., 2016).
All studies presented above for production of novel NRPs and engineering NRPS assembly lines in the native host are laborious having low throughput and low yield. Recent advances in genome sequencing, gene synthesis, metabolomics and bioinformatics revolutionized the process of NRPS engineering. In-silico based bioprospecting of available microbial genome sequences gives us a quick look at the hidden biosynthetic capacity of natural products in the microbial species. Several active as well as silent enzymes have been identified in fungal and bacterial genomes which are involved in the biosynthesis of NRPs. The corresponding secondary metabolites of these enzymes have not been identified to date (Brakhage, 2013;Doroghazi and Metcalf, 2013). Various powerful computational algorithms and tools have been developed to analyze BGC and to determine whether they are likely to encode unique compounds (Medema and Fischbach, 2015). Comprehensive ranges of software tools are available for identification of BGC in genome sequences. These tools are generally divided into two categories: high-confidence/lownovelty and low-confidence/high-novelty. High-confidence/lownovelty includes tools such as CLUSEAN13, ClustScan14, np.searcher15, SMURF16 and antiSMASH . These tools analyze Hidden Markov Models (HMMs) with manually curated cutoffs to identify signature genes or domains that are highly specific for known classes of biosynthetic pathways. Such strategies give a quick and reliable interpretation of NRPs gene cluster of a single strain from its genome sequence. Low-confidence/high-novelty mainly focuses on the identification of new BGC types by applying three approaches; namely pattern-based mining, phylogenetic mining and comparative genomic mining. These further include Cluster Finder, EvoMining, and Algorithim, respectively. Tools for identification of BGCs with respect to metagenomes include PCR-based sequence-tag and the shotgun assembly approach. The sequence-tag approach identifies clones from selected harbor pathways in metagenomic libraries by amplifying known biosynthetic domains using PCR. This is particularly useful for identifying variants of known pathway types. This has been also used to identify gene clusters encoding close relatives of molecule such as rapamycin, teicoplanin, and thiocoraline (Owen et al., 2013). However, the tag-based approach can be used to find entirely new molecules that are produced by known BGC classes, especially when coupled with phylogenomic tools such as NaPDoS (Ziemert et al., 2012). These tools find application in identifying domains that represent new areas of the extant biosynthetic diversity. A range of systems have been developed to predict the substrate specificities of NRPS adenylation domains (Röttig et al., 2011;Prieto et al., 2012). Tools such as NP.searcher and antiSMASH individual monomer predictions are then combined to give a rough idea of the core scaffold of a nonribosomal peptide. Simultaneously, advancement in mass spectrometry gives efficient dereplication for analysis of smallmolecule products of biosynthetic pathways (Nielsen and Larsen, 2015). NRPQuest algorithm uses molecular networking approach to identify potential gene clusters for observed tandem mass spectra of NRPs (Mohimani et al., 2014). The search database for NRPquest generates all possible orders of NRPS assembly lines within each detected NRP BGC hence, predicting the amino acids encoded by each of its module using NRPSPredictor2 (Röttig et al., 2011). A chemoinformatic based library and informatic search strategy for natural products (iSNAP) has also been doveloped for true nontargeted dereplication across a spectrum of nonribosomal peptides and within natural product extracts (Ibrahim et al., 2012). It is clear that the tools and techniques discussed above have accelerated the discovery and development of novel NRPs with desirable biological activities.

CONCLUSION AND FUTURE PROSPECTS OF MARINE DERIVED NONRIBOSOMAL PEPTIDES
Marine chemicals often possess quite novel structures which in turn lead to pronounced biological activity and novel pharmacology. The study of such chemicals, therefore, is a very promising endeavor. There are three parallel branches in marine natural products chemistry: marine biomedicinals, marine chemical ecology and marine toxins. Integration of these three fields of study gives marine natural products chemistry its exclusive character and vigor. The search among marine chemicals for medically useful agents involves two steps, discovering the type of biological activity and studying the pharmacological mechanism of the activity. It is now clear that efforts to date in marine natural product chemistry have largely focused on easily collected microorganisms and their major metabolites, and while there has been a recent shift to, as detailed above, minor metabolites present in very small quantities are a challenge for analytical and biological evaluations.
As has been demonstrated in this review, the potential for nonribosomal peptides from marine as sources and/or leads to drugs that have pharmacological effects (i.e., cancer and anti-infective) is only now being realized. Combining enzyme technology and solid phase peptide synthesis, it is possible to generate a vast variety of unique peptides composed of nonproteinogenic amino acids with unique pharmacological and biotherapeutic potential. It is possible that in coming years at least one or more marine derived novel nonribosomal peptide will enter into commerce as a drug. In concluding, the huge ranges of nonribosomal peptides that have so far been identified from marine resources frequently have no comparable equivalent in terrestrial organisms. The work by (predominately) young investigators on the many aspects of nonribosomal peptides (like biosynthesis) in the commensal and/or symbiotic microbes associated with these invertebrates, or in the microbes isolated from shallow and deep sediments will increase the numbers of nonribosomal peptides from marine for further work. The marine system has hardly been scratched as yet!

AUTHOR CONTRIBUTIONS
SA collected the available bibliographic information and wrote the manuscript. AA and CB conceived the study. SD and DA reviewed the collected information critically.

FUNDING
Deakin University provided a postgraduate scholarship to SA.