Progress in Research on Bioactive Secondary Metabolites from Deep-Sea Derived Microorganisms

Deep sea has an extreme environment which leads to biodiversity of microorganisms and their unique physical and biochemical mechanisms. Deep-sea derived microorganisms are more likely to produce novel bioactive substances with special mechanism of action for drug discovery. This article reviews secondary metabolites with biological activities such as anti-tumor, anti-bacterial, anti-viral, and anti-inflammatory isolated from deep-sea fungi and bacteria during 2018–2020. Effective methods for screening and obtaining natural active compounds from deep-sea microorganisms are also summarized, including optimizing the culture conditions, using genome mining technology, biosynthesis and so on. The comprehensive application of these methods makes broader prospects for the development and application of deep sea microbial bioactive substances.


Introduction
Deep sea is one of the latest extreme environments developed on earth. The deep sea is an environment with extreme features including: (1) For every 10 m of increase in depth, the pressure increases by one atmosphere, so the water pressure is higher than 1000 atmospheres in the deep sea trench; (2) The temperature decreases with depth, which is usually around 2 • C on the deep sea bottom; (3) The seawater oxygen concentration mainly depends on the absorption of oxygen at the sea-air interface, the photosynthesis rate of autotrophs in the true light layer, and rate of consumption of marine life respiration; (4) The light intensity is close to zero below the depth of 250 m [1]. In conclusion, deep sea has the characteristics of extreme ecological environment, including high pressure, low temperature, lack of oxygen and darkness. The cold seeps, hydrothermal and seamounts of the world deep-sea locations may worth favoring for bioprospection. Deep-sea microbes have unique biological metabolic pathways to deal with extreme ecological environments, especially stress. Many deep-sea microbes are hypertrophic or pressure-sensitive. Existing research methods limit the cultivation of these steps [2].
Over the past fifty years, more than 30,000 marine natural products have been discovered, of which about 2% are derived from deep-sea microorganisms [3]. Based on our review of the literature, the number of marine natural products from deep-sea have increased since then, but they are still a small percentage of the total amount found. Also, we found that most recent researches on     ).  Chaetomium globosum HDN151398 was isolated from deep-sea sediments in the South China Sea, and its metabolites chaetomugilin A (5) and chaetomugilin C (6) (Figure 2) showed broad-spectrum cytotoxic activities. Compound 5 exhibited significant cytotoxic activity against human promyelocytic leukemia cell line HL-60 and human colorectal cancer cells HCT-116, with IC50 values of 6.4 and 6.1 µM, respectively. While compound 6 exhibited IC50 values of 6.6 and 5.7 µM against HL-60 and HCT-116, respectively [13].
Penigrisacid D (20) is a sesquiterpene isolated from the extract of Penicillium griseofulvum, which showed a weak inhibitory activity against esophageal cancer cell line ECA-109, with an IC50 value of 28.7 µM [16]. Aphidicolin A8 (21) is a diterpene isolated from the extract of Botryotinia fuckeliana derived from Western Pacific seawater samples, which induced human bladder cancer cell T24 and human promyelocytic leukemia cell HL-60 apoptosis by DNA damage, with the IC50 values of 2.5 µM and 6.1 µM, respectively [17]. Photeroids A (22) and B (23) were isolated from the extract of the fungus Phomopsis tersa. They are both heteroterpenes containing a 6/6/6/6 tetracyclic system which forms Ortho-Quinone methides (o-QMs) intermediates through a rare Diels-Alder reaction. Compounds 22 and 23 ( Figure 3) showed moderate cytotoxicity against four human cancer cell lines: SF-268, MCF-7, HepG-2 and A549 [18]. N-glutarylchaetoviridins (24) is an azaphilone alkaloid containing glutamine residues isolated from Chaetomium globosum HDN151398. Compound 24 exhibited significant cytotoxic activity against human gastric cancer cell line MGC-803 and human ovarian cancer cell line HO-8910, with IC50 values of 6.6 and 9.7 µM, respectively [13]. Apiosporamide (25) is also an alkaloid, isolated from the extract of Arthrinium sp. UJNMF0008 derived from sediments of the South China Sea. It exihibited cytotoxity against two human osteosarcoma cell lines (U2OS and MG63) with IC50 values of 19.3 and 11.7 µM [19].
Penicillenol A2 (47) ( Figure 12) was isolated from the extract of Penicillium biourgeianum isolated from the sediments of the South China Sea with an inhibitory effect on MSSA. The diameter of inhibitory zone (ZD) is 6.75 ± 0.25 mm. Besides, the synergy of compound 47 with penicillin G sodium (Pen), cefotaxime sodium (Ctx) and oxacillin sodium (Oxa) was studied by plate count and Kirby-Bordisk diffusion method. It was found that, in comparison with the control group, the reduction of bacteria in the experimental group using Pen (10 U mL −1 ), Ctx (15 U mL −1 ) and Oxa (1 U mL −1 ) was less than 1 log CFU/mL . Compared with using compound 30 alone, the reduction of viable bacteria in the experimental group using both the above drugs and compound 47 was greater than or equal to 2 log CFU/mL. Therefore, the combination of compound 30 and β-lactam antibiotics had a synergistic effect, which can increase the sensitivity of MRSA to β-lactam antibiotics [27].
Mar. Drugs 2020, 18, x FOR PEER REVIEW 10 of 28 Penicillenol A2 (47) ( Figure 12) was isolated from the extract of Penicillium biourgeianum isolated from the sediments of the South China Sea with an inhibitory effect on MSSA. The diameter of inhibitory zone (ZD) is 6.75 ± 0.25 mm. Besides, the synergy of compound 47 with penicillin G sodium (Pen), cefotaxime sodium (Ctx) and oxacillin sodium (Oxa) was studied by plate count and Kirby-Bordisk diffusion method. It was found that, in comparison with the control group, the reduction of bacteria in the experimental group using Pen (10 U mL −1 ), Ctx (15 U mL −1 ) and Oxa (1 U mL −1 ) was less than 1 log CFU/mL . Compared with using compound 30 alone, the reduction of viable bacteria in the experimental group using both the above drugs and compound 47 was greater than or equal to 2 log CFU/mL . Therefore, the combination of compound 30 and β-lactam antibiotics had a synergistic effect, which can increase the sensitivity of MRSA to β-lactam antibiotics [27].   (Figure 14) were isolated from the extract of Aspergillus sp. SCSIO06786 from deep-sea sediments in the Indian Ocean. Compounds 52 and 53 with 50 µg/disc showed inhibition zones against S. aureus, MRSA and E. faecalis. Compound 54 with 50 µg/disc inhibited the growth of MRSA. In addition, their MIC was tested and the results showed that it was between 3.13-12.5 µg/mL [30]. Penixylarins B-C (55-56), 1, 3-dihydroxy-5-(12hydroxyheptadecyl)benzene (57), and 1, 3dihydroxy-5-(12-sulfoxyheptadecyl)benzene (58) ( Figure  14) were isolated from a mixed culture of the Antarctic deep-sea-derived fungus Penicillium crustosum Penicillenol A2 (47) (Figure 12) was isolated from the extract of Penicillium biourgeianum isolated from the sediments of the South China Sea with an inhibitory effect on MSSA. The diameter of inhibitory zone (ZD) is 6.75 ± 0.25 mm. Besides, the synergy of compound 47 with penicillin G sodium (Pen), cefotaxime sodium (Ctx) and oxacillin sodium (Oxa) was studied by plate count and Kirby-Bordisk diffusion method. It was found that, in comparison with the control group, the reduction of bacteria in the experimental group using Pen (10 U mL −1 ), Ctx (15 U mL −1 ) and Oxa (1 U mL −1 ) was less than 1 log CFU/mL . Compared with using compound 30 alone, the reduction of viable bacteria in the experimental group using both the above drugs and compound 47 was greater than or equal to 2 log CFU/mL . Therefore, the combination of compound 30 and β-lactam antibiotics had a synergistic effect, which can increase the sensitivity of MRSA to β-lactam antibiotics [27].   (Figure 14) were isolated from the extract of Aspergillus sp. SCSIO06786 from deep-sea sediments in the Indian Ocean. Compounds 52 and 53 with 50 µg/disc showed inhibition zones against S. aureus, MRSA and E. faecalis. Compound 54 with 50 µg/disc inhibited the growth of MRSA. In addition, their MIC was tested and the results showed that it was between 3.13-12.5 µg/mL [30]. Penixylarins B-C (55-56), 1, 3-dihydroxy-5-(12-hydroxyheptadecyl)benzene (57), and 1, 3dihydroxy-5-(12-sulfoxyheptadecyl)benzene (58) (Figure 14) were isolated from a mixed culture of the Antarctic deep-sea-derived fungus Penicillium crustosum PRB-2 with a fungus Xylaria sp. HDN13-249. Compounds 55-58 showed activities against Bacillus tuberculosis, B. subtilis or Vibrio parahaemolyticus, MIC values ranging from 6.25 to 100 µM. Among them, the MIC value of compound 56 against B. tuberculosis was 6.25 µM, showing the anti-tuberculosis potential [31]. The tyrosine phosphatase (Mptp) secreted by Mycobacterium tuberculosis is an important virulence factor of Mycobacterium tuberculosis and recognized to be an important target to treat tuberculosis. Tyrosine phosphatase is secreted by Mycobacterium tuberculosis, which has two functional phosphatases, PTP A and B (MptpA and MptpB) and enters the cytoplasm of macrophages, preventing the activation of the host's immune system and regulating the survival of the bacilli in the host [32]. Compounds 59-62 ( Figure 15) were polyacrylate derivatives with long hydrophobic chains, isolated from the extract of Aspergillus fischeri derived from deep sea sediments in Indian Ocean. Compounds 59-62 inhibited M. tuberculosis protein tyrosine phosphatase B (MptpB) through non-competitive inhibition, with IC50 values of 5.1, 12, 4.0 and 11 µM, respectively [33]. The tyrosine phosphatase (Mptp) secreted by Mycobacterium tuberculosis is an important virulence factor of Mycobacterium tuberculosis and recognized to be an important target to treat tuberculosis. Tyrosine phosphatase is secreted by Mycobacterium tuberculosis, which has two functional phosphatases, PTP A and B (MptpA and MptpB) and enters the cytoplasm of macrophages, preventing the activation of the host's immune system and regulating the survival of the bacilli in the host [32]. Compounds 59-62 ( Figure 15) were polyacrylate derivatives with long hydrophobic chains, isolated from the extract of Aspergillus fischeri derived from deep sea sediments in Indian Ocean. Compounds 59-62 inhibited M. tuberculosis protein tyrosine phosphatase B (MptpB) through non-competitive inhibition, with IC 50 values of 5.1, 12, 4.0 and 11 µM, respectively [33]. functional phosphatases, PTP A and B (MptpA and MptpB) and enters the cytoplasm of macrophages, preventing the activation of the host's immune system and regulating the survival of the bacilli in the host [32]. Compounds 59-62 ( Figure 15) were polyacrylate derivatives with long hydrophobic chains, isolated from the extract of Aspergillus fischeri derived from deep sea sediments in Indian Ocean. Compounds 59-62 inhibited M. tuberculosis protein tyrosine phosphatase B (MptpB) through non-competitive inhibition, with IC50 values of 5.1, 12, 4.0 and 11 µM, respectively [33].

Secondary Metabolites with Other Bioactivities
In addition to antitumoral and antimicrobial activity, secondary metabolites of deep-sea fungi reported in recent years also have anti-inflammatory and anti-food allergic activities. Compounds 64-69 were secondary metabolites from deep-sea fungi with anti-inflammatory activity.
Trieffusols C (64) and D (65) (Figure 17) were isolated from the extract of Trichobotrys effuse from deep-sea sediments of the South China Sea with inhibition of nitric oxide (NO) production in murine macrophages. Their IC50 values were 51.9 and 55.9 µM, which is equivalent to the positive control aminoguanidine (IC50: 24.8 µM) [34].

Secondary Metabolites with Other Bioactivities
In addition to antitumoral and antimicrobial activity, secondary metabolites of deep-sea fungi reported in recent years also have anti-inflammatory and anti-food allergic activities. Compounds 64-69 were secondary metabolites from deep-sea fungi with anti-inflammatory activity.
Trieffusols C (64) and D (65) (Figure 17) were isolated from the extract of Trichobotrys effuse from deep-sea sediments of the South China Sea with inhibition of nitric oxide (NO) production in murine macrophages. Their IC 50 values were 51.9 and 55.9 µM, which is equivalent to the positive control aminoguanidine (IC 50 : 24.8 µM) [34].

Secondary Metabolites with Other Bioactivities
In addition to antitumoral and antimicrobial activity, secondary metabolites of deep-sea fungi reported in recent years also have anti-inflammatory and anti-food allergic activities. Compounds 64-69 were secondary metabolites from deep-sea fungi with anti-inflammatory activity.
Trieffusols C (64) and D (65) (Figure 17) were isolated from the extract of Trichobotrys effuse from deep-sea sediments of the South China Sea with inhibition of nitric oxide (NO) production in murine macrophages. Their IC50 values were 51.9 and 55.9 µM, which is equivalent to the positive control aminoguanidine (IC50: 24.8 µM) [34].     The occurrence of food allergic diseases may be related to excessive immune response. Allergens are usually harmless foods such as milk, eggs, fish, peanuts and grains [39]. Acute hypersensitivity is triggered by factors released by mast cells when allergens interact with membrane-bound immune proteins (IgE) [40].
Polyketides 76-78( Figure 20) were isolated from Graphostroma sp. MCCC 3A00421 derived from hydrothermal sulfide, which showed antifood allergic activity. Reticulol (76) showed effective inhibition of immunoglobulin E-mediated rat basophilic leukemia-2H3 cells (RBL-2H3) degranulation, with an IC50 value of 13.5 µM, which was about seven times stronger than the The occurrence of food allergic diseases may be related to excessive immune response. Allergens are usually harmless foods such as milk, eggs, fish, peanuts and grains [39]. Acute hypersensitivity is triggered by factors released by mast cells when allergens interact with membrane-bound immune proteins (IgE) [40].

Secondary Metabolites from Deep-Sea Derived Bacteria
Bacteria from deep-sea sediments are a good source of marine natural products, and their secondary metabolites are usually novel in structure with significant biological activities [41][42][43][44][45]. In particular, actinomycetes are currently proven to be the most important sources of biologically active natural products with clinical or pharmaceutical applications [46]. According to the references, 40 (16 novel) natural products were discovered from deep-sea derived bacteria in the past three years, 19 among which showed biological activities ( Table 2).

Secondary Metabolites from Deep-Sea Derived Bacteria
Bacteria from deep-sea sediments are a good source of marine natural products, and their secondary metabolites are usually novel in structure with significant biological activities [41][42][43][44][45]. In particular, actinomycetes are currently proven to be the most important sources of biologically active natural products with clinical or pharmaceutical applications [46]. According to the references, 40 (16 novel) natural products were discovered from deep-sea derived bacteria in the past three years, 19 among which showed biological activities ( Table 2).

Antitumoral Secondary Metabolites
Compounds 80-87 all showed potent cytotoxic activity. Cebulactam A2 (80) (Figure 22) was a polyketide isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments, which had a weak antiproliferative effect on human cervical cancer cell Hela and human lung cancer cell H1299, the inhibition rates (20.00 µg/mL) were 35.0 and 31.0%, respectively [47].

Antitumoral Secondary Metabolites
Compounds 80-87 all showed potent cytotoxic activity. Cebulactam A2 (80) (Figure 22) was a polyketide isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments, which had a weak antiproliferative effect on human cervical cancer cell Hela and human lung cancer cell H1299, the inhibition rates (20.00 µg/mL) were 35.0 and 31.0%, respectively [47].

Antitumoral Secondary Metabolites
Compounds 80-87 all showed potent cytotoxic activity. Cebulactam A2 (80) (Figure 22) was a polyketide isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments, which had a weak antiproliferative effect on human cervical cancer cell Hela and human lung cancer cell H1299, the inhibition rates (20.00 µg/mL) were 35.0 and 31.0%, respectively [47].

Antitumoral Secondary Metabolites
Compounds 80-87 all showed potent cytotoxic activity. Cebulactam A2 (80) (Figure 22) was a polyketide isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments, which had a weak antiproliferative effect on human cervical cancer cell Hela and human lung cancer cell H1299, the inhibition rates (20.00 µg/mL) were 35.0 and 31.0%, respectively [47].

Antimicrobial Secondary Metabolites
Compounds 88-96 all showed potent antibacterial activity. Aborycin (88) was a lasso peptide isolated and identified from the deep-sea-derived microbe Streptomyces sp. SCSIO ZS0098 which was isolated from the deep-sea sediments of the South China Sea. Shao et al. [52] identified the aborycin biosynthetic gene cluster (abo) on the basis of genomic sequence analysis, and then heterologously expressed in Streptomyces coelicolor to obtain compound 86. The compound had moderate bacteriostatic activity against 13 Staphylococcus aureus strains from various sources, with MIC values between 8.0-128 µg/mL. The MIC values of compound 88 against Enterococcus faecalis and Bacillus thuringiensis.were 8.0 µg/mL and, 2.0 µg/mL, respectively. In addition, compound 88 had significant antibacterial activity against the poultry pathogen Enterococcus enterococci (MIC = 0.5 µg/mL) [44]. Atratumycin (89) was also a peptide isolated from the extract of Streptomyces atratus from deep-sea sediments of the South China Sea, which is a cyclic dipeptide that has activity against Mycobacterium tuberculosis, whose MIC values were 3.8 and 14.6 µM against M. tuberculosis H37Ra and H37Rv [53].
Compounds 90-94 are all polyketides. Anthracimycin B (90) and anthracimycin (91) (Figure 26) were isolated from the extract of Streptomyces cyaneofuscatus isolated from a gorgonian coral collected in the 1500 m Avilis submarine canyon. They were sensitive to Gram-positive pathogens MRSA, MSSA, vancomycin-sensitive Enterococcus faecium and vancomycin-sensitive Enterococcus faecalis and all showed strong antibacterial effects. The MIC value of compound 90 was less than 0.03 µg/mL, and the MIC value of compound 91 was between 0.125-8 µg/mL. Compound 90 also had anti-tuberculosis activity, with the MIC value of 1-2 µg/mL [54].

Antimicrobial Secondary Metabolites
Compounds 88-96 all showed potent antibacterial activity. Aborycin (88) was a lasso peptide isolated and identified from the deep-sea-derived microbe Streptomyces sp. SCSIO ZS0098 which was isolated from the deep-sea sediments of the South China Sea. Shao et al. [52] identified the aborycin biosynthetic gene cluster (abo) on the basis of genomic sequence analysis, and then heterologously expressed in Streptomyces coelicolor to obtain compound 86. The compound had moderate bacteriostatic activity against 13 Staphylococcus aureus strains from various sources, with MIC values between 8.0-128 µg/mL. The MIC values of compound 88 against Enterococcus faecalis and Bacillus thuringiensis were 8.0 µg/mL and, 2.0 µg/mL, respectively. In addition, compound 88 had significant antibacterial activity against the poultry pathogen Enterococcus enterococci (MIC = 0.5 µg/mL) [44]. Atratumycin (89) was also a peptide isolated from the extract of Streptomyces atratus from deep-sea sediments of the South China Sea, which is a cyclic dipeptide that has activity against Mycobacterium tuberculosis, whose MIC values were 3.8 and 14.6 µM against M. tuberculosis H37Ra and H37Rv [53].
Compounds 90-94 are all polyketides. Anthracimycin B (90) and anthracimycin (91) (Figure 26) were isolated from the extract of Streptomyces cyaneofuscatus isolated from a gorgonian coral collected in the 1500 m Avilis submarine canyon. They were sensitive to Gram-positive pathogens MRSA, MSSA, vancomycin-sensitive Enterococcus faecium and vancomycin-sensitive Enterococcus faecalis and all showed strong antibacterial effects. The MIC value of compound 90 was less than 0.03 µg/mL, and the MIC value of compound 91 was between 0.125-8 µg/mL. Compound 90 also had anti-tuberculosis activity, with the MIC value of 1-2 µg/mL [54].

Antimicrobial Secondary Metabolites
Compounds 88-96 all showed potent antibacterial activity. Aborycin (88) was a lasso peptide isolated and identified from the deep-sea-derived microbe Streptomyces sp. SCSIO ZS0098 which was isolated from the deep-sea sediments of the South China Sea. Shao et al. [52] identified the aborycin biosynthetic gene cluster (abo) on the basis of genomic sequence analysis, and then heterologously expressed in Streptomyces coelicolor to obtain compound 86. The compound had moderate bacteriostatic activity against 13 Staphylococcus aureus strains from various sources, with MIC values between 8.0-128 µg/mL. The MIC values of compound 88 against Enterococcus faecalis and Bacillus thuringiensis.were 8.0 µg/mL and, 2.0 µg/mL, respectively. In addition, compound 88 had significant antibacterial activity against the poultry pathogen Enterococcus enterococci (MIC = 0.5 µg/mL) [44]. Atratumycin (89) was also a peptide isolated from the extract of Streptomyces atratus from deep-sea sediments of the South China Sea, which is a cyclic dipeptide that has activity against Mycobacterium tuberculosis, whose MIC values were 3.8 and 14.6 µM against M. tuberculosis H37Ra and H37Rv [53].
Compounds 90-94 are all polyketides. Anthracimycin B (90) and anthracimycin (91) (Figure 26) were isolated from the extract of Streptomyces cyaneofuscatus isolated from a gorgonian coral collected in the 1500 m Avilis submarine canyon. They were sensitive to Gram-positive pathogens MRSA, MSSA, vancomycin-sensitive Enterococcus faecium and vancomycin-sensitive Enterococcus faecalis and all showed strong antibacterial effects. The MIC value of compound 90 was less than 0.03 µg/mL, and the MIC value of compound 91 was between 0.125-8 µg/mL. Compound 90 also had anti-tuberculosis activity, with the MIC value of 1-2 µg/mL [54].

Other Bioactive Secondary Metabolites
Acantimycic acid (97) was an alkaloid with good neuroprotection. It was isolated from the extract of Alcanivorax sp. SHA4 from deep-sea sediments of the Western Pacific and could inhibit the cell damage caused by glutamic acid to PC12 cells. The protective effect was more obvious at low concentration [57]. Indol-3-carbaldehyde (98) was isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments. It showed weak anti-allergic effect with the IC50 value of 55.75 µg/mL [47]. The structures of compounds 97 and 98 are shown in Figure 29.

Other Bioactive Secondary Metabolites
Acantimycic acid (97) was an alkaloid with good neuroprotection. It was isolated from the extract of Alcanivorax sp. SHA4 from deep-sea sediments of the Western Pacific and could inhibit the cell damage caused by glutamic acid to PC12 cells. The protective effect was more obvious at low concentration [57]. Indol-3-carbaldehyde (98) was isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments. It showed weak anti-allergic effect with the IC50 value of 55.75 µg/mL [47]. The structures of compounds 97 and 98 are shown in Figure 29.

Other Bioactive Secondary Metabolites
Acantimycic acid (97) was an alkaloid with good neuroprotection. It was isolated from the extract of Alcanivorax sp. SHA4 from deep-sea sediments of the Western Pacific and could inhibit the cell damage caused by glutamic acid to PC12 cells. The protective effect was more obvious at low concentration [57]. Indol-3-carbaldehyde (98) was isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments. It showed weak anti-allergic effect with the IC 50 value of 55.75 µg/mL [47]. The structures of compounds 97 and 98 are shown in Figure 29.

Other Bioactive Secondary Metabolites
Acantimycic acid (97) was an alkaloid with good neuroprotection. It was isolated from the extract of Alcanivorax sp. SHA4 from deep-sea sediments of the Western Pacific and could inhibit the cell damage caused by glutamic acid to PC12 cells. The protective effect was more obvious at low concentration [57]. Indol-3-carbaldehyde (98) was isolated from the extract of Saccharopolyspora cebuensis derived from Atlantic deep-sea sediments. It showed weak anti-allergic effect with the IC50 value of 55.75 µg/mL [47]. The structures of compounds 97 and 98 are shown in Figure 29.

Isolation and Cultivation of Deep-Sea Microorganisms
Marine microorganisms have the following characteristics: (1) can grow and/or form spores in the marine environment; (2) form a symbiotic relationship with other marine organisms; or (3) adapt and evolve at the genetic level or have metabolic activity in the marine environment [58]. It is estimated that the diversity of marine fungi exceeds 10,000 species [59,60], but so far only about 1250 species have been described [61,62]. However, deep-sea microbial research starts late for its difficulties in collection and cultivation, so people face more challenges in the exploration of its secondary metabolites.

Sample Pretreatment
In natural samples without pretreatment, the isolation frequency of bacteria is higher than that of fungi [63]. Different pretreatment methods should be adopted for different target strains to improve the isolation efficiency.
Because the actinomycete spores have a certain heat resistance, dry and wet heat treatment can effectively reduce other bacterial contamination [64][65][66][67][68]. Dry heat treatment can inactivate bacteria, and at the same time induce the germination of actinomycetes spores to a certain extent; the principle of wet heat treatment is to denature and inactivate non-target strain proteins in the sample by heating in a water bath. In addition to heat treatment, the commonly used pretreatment methods include chemical reagent treatment [69], differential centrifugation [70] and so on.
Microwave treatment can not only significantly increase the number of isolated alkaliphilic and halophilic marine actinomycetes, but also significantly increase the isolation of rare marine actinomycetes. Ding et al. [71] used 120 W, 2450 MHz microwave and ice-water mixture to process one part of the suspension in the treatment of sea mud samples. After gradient dilution, they were applied to three separate media. In the seven samples after microwave treatment, the number of rare alkaliphilic marine actinomycetes in four samples and the halophilic marine actinomycetes in three samples increased significantly. Therefore, microwave processing also has certain application value.

Medium Selection and Improvement
When the strains are separated, the medium as a nutrient source plays an important role in the growth and metabolism of the strains. Different culture media provide different carbon and nitrogen sources for different microorganisms to grow, so it is necessary to select the appropriate culture media for microorganism screening.  [72] used the above-mentioned six media (all added chloramphenicol and streptomycin sulfate to inhibit growth of bacteria) to separate samples from the deep-sea sediment samples of Yapu Trench. In their study, YM media was the best from the perspective of the isolation ability of six different media, which obtained nine kinds of fungi; followed by PDA which allowed the retrieval of eight different fungal species. The worst were CMA (three kinds) and CDA (two kinds).
For bacteria, according to the main components of the medium, it can be divided into marine agar medium (MA), synthetic medium for selective isolation of actinomycetes (Actinomycete Isolation Agar, AIA), starch medium, natural ingredient medium, high salt medium and other media (Table 3). Chen et al. [73] used the 23 media in the table to isolate bacteria in the 4000 m deep-sea sediments of the South China Sea, and most of natural products from the strains obtained from the deep sea sediment environment, are antibiotic, cytotoxins, with high efficiency enzyme activity, and tolerant for unfavorable environment, degradation of refractory pollutants and other characteristics suitable for the unique marine extreme environment.

Screening Methods of Deep-Sea Natural Products
One of the keys to develop and utilize biological resources is how to obtain bioactive natural products from cultivable deep-sea microorganisms. Traditional natural product activity screening method is also suitable for the activity screening of deep-sea microbial metabolites, which mainly tracks the active substances in the cultivation broth. In addition, commonly used methods include model screening for specific target modeling and evaluation, and gene screening based on microbial natural product synthesis gene clusters.

In Vivo Screening Methods
In vivo screening models mainly refer to animal models and Serum pharmacology models. Animal models can mimic clinical features such as physiology and pathology similar to those of patients. Serum pharmacology models can help prove the true positive compounds, whether they are original drugs or metabolites [74]. Therefore, in vivo experiments have an irreplaceable role in activity screening. However, due to its time-consuming, low throughput, and large sample consumption, it is less used in preliminary screening.

Cell and Receptor/Enzyme Model Screening Methods
Cell and receptor/enzyme model screening is used for target screening, and is usually established as a specific and effective model on pharmacology at the cellular or molecular level.
Compared with simple chemical methods, evaluation of biological activity of natural products based on cell models can not only simulate the human physiological environment, but it can also explore and evaluate the biological activity and mechanism of natural products at multiple targets; compared with animal experiments in vivo, it not only shorten the experimental time, but greatly reduces the experimental cost [75]. The most commonly used cell models are human cancer cell lines, such as: A54, MCF7, HepG2, Caov-3, PANC-1 and so on. In addition, there are other models at the cellular level to test other bioactivities. Xu et al. used hemolysis assay on sheep red blood cell to test the anti-complement activity of 42 strains of marine actinomycetes isolated from Dalian Xinghai Bay mud samples, and further isolated three small molecular compounds with weak anti-complement activity from extract of strain DUT11 [76].
Receptors or enzymes related to various physiological and pathological processes in the body are considered to be one of the main targets of drug action [74]. Liu et al. tested IC 50 values of the polypropionate derivatives against MptpB to show their antituberculosis activities [33]. ACE2 has been shown to be the main receptor for SARS-CoV S protein to infect cells [77]. Deng et al. showed that baicalin had an inhibitory activity against ACE with the IC 50 value of 2.24 mM [78].

Virtual and Gene Screening Methods
Virtual screening based on compound structural diversity, that is, using computer programs to screen bioactive compounds from existing virtual libraries. And compounds with higher chemical structure spatial diversity are more suitable for virtual library establishment [79].
Gene screening breaks through the traditional active screening model. Because the secondary metabolite synthetic gene clusters with similar structures have a certain degree of similarity, the strains that produce the target compound can be obtained from nature by screening specific gene clusters.
Polyketide compounds are catalyzed by a type of polyketide synthetase (PKS) which is widely present in nature. Polyketide synthetase can generally be divided into three types according to its protein structure and catalytic mechanism, namely type I, type II and type III [80]. Type I PKS includes type I modular PKS (bacteria) and type I repeat PKS (fungi). A typical type I module PKS is a multifunctional complex enzyme composed of modules. Each module contains a unique and non-repetitive structural domain, which mainly contains acyltransferase (acyltransferase, AT), β-keto synthase (ketosynthase, KS) responsible for catalyzing the formation of carbon-carbon bonds and extending the main chain, and acyltransferase cylcar-rier protein (acyltransferase cylcar-rier protein, ACP) responsible for receiving and transporting acyl units provided by the AT domain), these three domains constitute the smallest catalytic module and are also the three essential domains of PKS.
Non-ribosomal peptide synthetase (NRPS) is also widely present in bacteria, fungi and plants, and uses different amino acids as substrates to catalyze the production of condensed peptides. NRPS is mainly composed of different independent modules. Each module contains an adenylation structural functional domain (andeylation, A) that selects and activates special amino acids, and loads aminoacyl residues into the sulfhydryl structural functional domain (thiolation, T), and the condensation domain (C) of peptide compounds that polymerize activated amino acids to produce amides.
In fungi, the genes encoding PKS and NRPS can be aggregated to produce type I repeat PKS units (KS, AT, DH, CMeT, KR and ACP domains) and NRPS units (A, T and C domains) PKS-NRPS hybrid enzyme. PKS-NRPS has the function of catalyzing the combination of PKS products and NRPS products, thereby producing more abundant natural products-PKS-NRPS hybrid compounds [81]. Such PKS-NRPS hybrid compounds are often a class of natural products with complex and diverse structures and a wide range of biological activities. They not only play an important role in the survival and prosperity of the host in the natural environment, but also is an important source for the discovery of active lead compounds with potential applications.
Jiang et al. [82] applied the type I polyketide synthase (PKS-I) gene screening system and DNA sequence similarity comparison to select positive Actinoplanes sp. from 32 strains of marine actinomycetes. Actinoplanes sp. FIM060065 was one of them. And from its fermentation broth researchers obtained a macrolide compound homogenous to tiacumicin B by High Performance Liquid Chromatography (HPLC) preparation, which showed strong antibacterial activity towards Gram-positive bacteria, such as Clostridium difficile, Streptococcus pneumoniae, Bifidobacterium, etc. Vanessa Rédou et al. [83] isolated 124 filamentous fungi and 59 yeasts from sediments in the Canterbury basin of New Zealand. The PKS-NRPS analysis results showed that there was no PKS-NRPS hybrid gene in the yeast genome; compared with yeast, filamentous fungal isolates from deep seabed sediments have greater bioactive compound synthesis potential, but they have fewer bioactive compound genes than those isolated from shallower depths.

Secondary Metabolite Discovery Based on Synthetic Biology
Synthetic biology is a multidisciplinary disruptive study leading a new generation of biotechnology revolution, in which gene editing takes an important part. Gene editing has unique advantages in establishing an artificially regulated biosynthetic system, further mining new natural product resources of actinomycetes, solving the bottleneck of existing natural products and developing derivatives. Recently, CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 system, known as "magic scissor", was found to improve the efficiency of gene editing.
Elizabeth J. Culp et al. [84] applied CRISPR/Cas9 system to 11 actinomycete strains, knocked out the common streptomycin and streptomycin genes, produced a variety of hidden rare antibiotics, and constructed a platform that can be widely used to stimulate the potential of microbial secondary metabolism. Indra Roux et al. [85] established the first CRISPRa system for filamentous fungi and discovered the mic cluster product, dehydromicroperfuranone. Meanwhile, factors affecting the efficiency of the system was also studied.

Conclusions and Perspective
Although the research on secondary metabolites of deep-sea microbes started later than that in other environments [9], it has drawn much more attention, and natural products with novel structures and good biological activities have been discovered in the past three years. From our literature review, fungi seem to be the focus of most isolations from the deep-sea for bioprospection of metabolites with biological activities; also, producers of higher diversity and amount of compounds. Among bacteria, actinomycetes seem to be studied more deeply in natural product research, and they have shown the potential to become biological resources with novel structures and good biological activities. When it comes to structural classes, polyketides showed a broad spectrum of bioactivities, such as antitumor, antibacterial, anti-inflammatory and antifood allergic.
The rapid development of deep sea exploration and bioinformatics has provided solid technical support for the chemical diversity and bioactivity diversity of secondary metabolites of deep sea microorganisms, but there are still many challenges, such as activation of specific biosynthetic gene clusters and heterologous expression, directed transformation of synthetic gene clusters, design of virtual screening libraries for natural products, and how to solve the problem of yield of active compounds.
Deep-sea microbial natural product resources are still a virgin land that needs to be developed urgently. Reasonable and green applied research will contribute more power to drug discovery.

Conflicts of Interest:
The authors declare no conflict of interest.