Secondary Metabolites of the Genus Amycolatopsis: Structures, Bioactivities and Biosynthesis

Actinomycetes are regarded as important sources for the generation of various bioactive secondary metabolites with rich chemical and bioactive diversities. Amycolatopsis falls under the rare actinomycete genus with the potential to produce antibiotics. In this review, all literatures were searched in the Web of Science, Google Scholar and PubMed up to March 2021. The keywords used in the search strategy were “Amycolatopsis”, “secondary metabolite”, “new or novel compound”, “bioactivity”, “biosynthetic pathway” and “derivatives”. The objective in this review is to summarize the chemical structures and biological activities of secondary metabolites from the genus Amycolatopsis. A total of 159 compounds derived from 8 known and 18 unidentified species are summarized in this paper. These secondary metabolites are mainly categorized into polyphenols, linear polyketides, macrolides, macrolactams, thiazolyl peptides, cyclic peptides, glycopeptides, amide and amino derivatives, glycoside derivatives, enediyne derivatives and sesquiterpenes. Meanwhile, they mainly showed unique antimicrobial, anti-cancer, antioxidant, anti-hyperglycemic, and enzyme inhibition activities. In addition, the biosynthetic pathways of several potent bioactive compounds and derivatives are included and the prospect of the chemical substances obtained from Amycolatopsis is also discussed to provide ideas for their implementation in the field of therapeutics and drug discovery.


Introduction
Antibiotics produced by microorganisms have made a significant contribution to human health. Among them, Actinomycetes are the most important sources for drug lead compounds. However, researchers have been turned to rare Actinomycetes to develop novel antibiotics with the emergence of multidrug-resistant bacteria [1]. In 1986, Lechevalier et al. defined Amycolatopsis as a new genus to accommodate nocardioform Actinomycetes having type IV cell wall composition and lacking mycolic acids [2]. Up to now, by searching in the List of Prokaryotic names with Standing in Nomenclature website (http://www. bacterio.net, accessed on 20 September 2020), this genus covered 94 verified species and 4 subspecies, and forms a unique branch in the evolutionary tree of Pseudonocardiaceae. Among 26 species covered in this review, most of them colonize in a wide variety of soil and a few species survival in terrestrial (insect, lichen, island, plant) and marine (sponge, sediment) environment. The various habitats allow Amycolatopsis to produce abundant secondary metabolites.

Polyphenols
Polyphenolic compounds are a large family of natural products and some of them show a series of excellent function in health [56], such as anti-allergenic, anti-inflammatory, anti-microbial, antioxidant, antithrombotic, cardio protective, and vasodilatory effects [57]. The investigation on secondary metabolites of Amycolatopsis sp. ML630-mF1 from the soil sample collected in Toba of Japan led to the isolation of five new compounds named kigamicins A-E (1-5). These compounds showed potent effects to resist methicillin-resistant Staphylococcus aureus (MRSA) with the IC 50 values ranging in 0.03-0.22 µM. Besides, they inhibited PANC-1 cell survival under a nutrient-starved condition. Typically, kigamicin D was found to suppress diverse mouse cancer cell line growth, and the IC 50 value was about 0.95 µM [13]. In the absence of nutrition, kigamicin D exhibited preferential cytotoxicity to cancer cells and could inhibit the PI3K/Akt pathway [58]. A total of 10 novel pentangular polyphenols defined as amexanthomycins A-J (6-15) were obtained from the fermentation products of Amycolatopsis mediterranei S699 ∆rifA (the A. mediterranei S699 mutant strain). These compounds were produced through deleting polyketide synthase genes related to rifamycin biosynthesis. In this study, the effects of the above compounds on suppressing topoisomerases IIα (Topo IIα) were examined. The results showed that compounds 6-8 exhibited moderate inhibitory activity against Topo IIα (500 µM), while compounds 9-15 showed no activities [14].

Linear Polyketides
Through genomic analysis, the strain Amycolatopsis orientalis ATCC 43,491 was found to be the producer of vancomycin, which possessed genetic loci to produce over 10 secondary metabolites apart from vancomycin. It was estimated that a gene cluster containing the type I polyketide synthase mediated the biosynthesis for a new glycosidic polyketide ECO-0501 (42) [24]. Compound 42 exhibited stronger antibacterial activity than van-comycin against S. aureus ATCC TM 6538P in pH 5.0 and 6.0 with the MIC values of 0.125 and 0.25 µg/mL. This compound had potent effect on resisting Gram-positive bacteria MRSA and vancomycin-resistant Enterococci (VRE) strains. The mechanistic studies proved that ECO-0501 may impact on either cell wall or membrane biosynthesis [60]. In addition, compound 42 chemical modified analogs, including esterified 43-45, N-acetylated 46, and hydrogenated 47 were reported. Compound 46 showed antibacterial activity against S. aureus ATCC TM 6538P with MIC values of 0.25, 0.5 and 2 µg/mL in pH 5, 6 and 7, respectively. The novel antibiotic vancoresmycin (48) was obtained from the culture broth of Amycolatopsis sp. ST 101170. It showed a potent effect on resisting the Gram-positive strains of E. faecium, S. aureus, S. pneumonia, S. epidermidis, S. pyogenes, together with a variety of drug-resistant microorganisms. The IC 50 values were found to be less than 0.05 µM. By a non-pore forming and concentration-dependent depolarization mechanism, compound 48 selectively targeted the cytoplasmic membrane of gram-positive bacteria [61]. No inhibitory effect against gram-negative bacteria or anti-fungal activity was observed [25]. All 7 linear polyketides described above are presented in Figure 2.

Macrolactams
Macrolactams have been used in clinical trials since 1940 [62], in which penicillin and cephalosporins are the representative antibiotics. For better exploiting the rifamycin diversity, the Amycolatopsis mediterranei S699 strain was cultured on the YMG agar media.  [30]. A novel compound, ansamycin (79), was produced by Amycolatopsis alba DSM 44262. However, this compound exhibited no antimicrobial activity for S. aureus, B. subtilis, P. aeruginosa and C. albicans [31]. All 21 macrolactams described above are presented in Figure 3.

Cyclic Peptides
Cyclic peptides always possess antibacterial, antitumor, hypotoxic, immunosuppressive activities and have a merit of favorable binding affinity and selectivity for certain receptors [63]. The limited conformational freedom conferred by cyclization enables cyclic peptides to span large surfaces while retaining the conformational restriction that yields high selectivity and affinity. Such advantages render them the ideal selection for developing therapeutics [64]. While screening antibiotics against MRSA and VRE, the novel cyclic peptide, PRG-A (98) containing the distinct piperazic acid, was obtained from the fermentation broth of Amycolatopsis sp. ML1-hF4 isolated from a soil sample collected at Shinagawa, Tokyo, Japan [37].   [41]. A new bioactive antibiotic, eremomycin B (117), along with one known antibiotic, eremomycin (118), were isolated from the culture broth of Amycolatopsis orientalis subsp. Eremomycini [42]. All 13 glycopeptides described above are presented in Figure 5.

Enediyne Derivatives
Amycolamycins A (156) and B (157), two new enediyne derivatives, were isolated from Amycolatopsis sp. HCa4, which was collected from locust. Compound 156 could inhibit M231 cell lines by inducing apoptosis through activation of caspase-3 with the IC 50 value of 7.9 µM [55]. The two enediyne derivatives described above are presented in Figure 7.

Biofunction of Amycolatopsis Species
The Amycolatopsis species have potential for biological degradation, bioconversion and biosorption, which might solve the problem of environmental pollution in the future [68].

Biological Degradation
ZJ0273 was a widely used broad-spectrum herbicide and left in soil in large numbers. Cai et al. found that ZJ0273 could be utilized by Amycolatopsis sp. M3-1 as the sole carbon and energy source with higher degrading activity. At 30 • C and pH 7.0, the efficiency of ZJ0273 degradation by Amycolatopsis sp. M3-1 was 59.3% and 68.5% in 25 and 60 days, respectively [69]. Naproxen was a drug utilized by humans; however, it was detected in surface waters and sanitary effluents in 71 countries, causing toxic effects on biota and further destroying the ecological environment [70]. At the concentration of 50 mg/L, naproxen could be used as the sole carbon and energy source of Amycolatopsis sp. Poz 14 and completely degraded in 18 days, while it will affect the growth of Amycolatopsis when its concentration was more than 50 mg/L [71]. It takes a long time for plastics to degrade in nature and the resulting environmental pollution problems are becoming more and more serious. Some Amycolatopsis strains possessed a polylactic acid (PLA) degradation capability including Amycolatopsis sp. HT-32, Amycolatopsis sp. 3118, Amycolatopsis sp. KT-s-9, A. mediterranei ATCC 27649, Amycolatopsis sp. 41, Amycolatopsis sp. K104-1, A. orientalis ssp. orientalis, Amycolatopsis thailandensis CMU-PLA07T and Amycolatopsis sp. SCM_MK2-4 [72]. Tan et al. proved that A. mediteranei was capable to hydrolyze the aliphatic plastics poly(ε-caprolactone) and poly(1,4-butylene succinate) via a extracellular lipase [73]. In addition, Amycolatopsis sp. ATCC 39,116 could depolymerize high molecular weight lignin species and catabolize a significant portion of the low molecular weight aromatics and may become a mature route for biological lignin valorization in the future [74].

Bioconversion
The strains of Amycolatopsis sp. HR167 and Amycolatopsis sp. ATCC39116 were able to convert ferulic acid (cell wall component of higher plants) into vanillin (important flavor compound) with concentrations of 11.5 and 13.9 g/L, respectively. The vanillin production of vdh (encoded vanillin dehydrogenase) mutant of Amycolatopsis sp. ATCC39116 was increased 2.3 times due to the enzyme catalyzed the catabolism of vanillin [75]. Wuxistatin, a novel HMG-CoA reductase inhibitor, was transformed from lovastatin by hydroxylase (cytochrome P450) and isomerases of Amycolatopsis sp. CGMCC 1149, showing a four-fold activity, more than lovastatin [76,77].

Bioactivities of Secondary Metabolites from Amycolatopsis
The bioactivities of secondary metabolites from Amycolatopsis strains have been also presented in Table 2, including antimicrobial, cytotoxic, antioxidant, topo IIα inhibition, anti-hyperglycemic, enzyme inhibition and DNA damage.  A-102395 (121) Inhibiting bacterial translocase I (0.01 µM) [45] Epoxyquinomicins C (145) and D (146) Inhibiting type II collagen-induced arthritis [51] Among the total of 159 secondary metabolites, 41 compounds exhibited potent antimicrobial activities, a majority of which showed inhibition on Gram-positive bacteria growth. Most of them were also found to be active against various multi-drug resistant strains. Kigamicins

Biosynthetic Pathways of Secondary Metabolites from Amycolatopsis
Research on biosynthetic pathways is essential for the further study on secondary metabolites. For example, finding the regulatory gene could increase or decrease the production of metabolites and also uncover how the concerted efforts of various enzymes to form the compound [83]. In this review, we list the hypothetical biosynthetic pathways for several potent bioactive compounds. Few studies have been conducted and need to arouse the attention of researchers.
The mutant strain A. mediterranei S699∆rifA, which was deleted for the biosynthesis gene of rifamycins, displayed the ability for producing ten new pentangular polyphenols, amexanthomycins A-J (6-15) [14]. As described in the literature, the production of as-sociated genes included polyketide synthase (PKS), glycosyltransferase, methyltransferase, monooxygenase, dehydrogenase, oxidoreductase, cytochrome P450 and epimerase ( Figure 8A). The biosynthetic pathway of amexanthomycins were proposed by Li et al. [14] and exhibited in the Figure 8B. An acetyl-CoA starter unit and 11 malonyl-CoA extender units could produce prediction intermediate, the pentacyclic xanthone core, by min-PKS synthase, cyclase, and oxidoreductase. Then, the predicted oxidase catalyzed the oxidative rearrangement reaction of intermediate. Finally, this aglycone was glycosylated by the glycosyl transferases, completing the biosynthesis of compounds 6-15 ( Figure 8B) [14]. The genome of A. orientalis ATCC 43,491 included a type I PKS which encoded by ORF 18-23 and synthesized the polyketide chain [24]. The monooxygenase and acyl-CoA ligase were encoded by ORF 7 and 25 which catalyzed arginine to 4-guanidino butyryl-CoA. D-glucose was catalyzed by oxidoreductase (ORF 13) and turned into Dglucuronic acid. Glycine and succinyl-CoA were transformed into 5-aminolevulinate by acyltransferase (ORF 16), and then turned into 5-aminolevulinate-CoA by acyl-CoA ligase (ORF 17). 5-aminolevulinate-CoA was transformed into aminohydroxycyclopentenone through cyclization reaction by the coenzyme A ester. Three ORFs (14, 15 and 24) provided glycosyltransferase, amide synthetase and acyltransferase to add 4-guanidino butyryl-CoA, D-glucuronic acid and aminohydroxycyclopentenone onto the polyketide chain which formed compound 47 (Figure 9) [24]. The genome of Amycolatopsis sp. HCa4 was analyzed by antiSMASH and 2ndfind, the cluster 19 was highly similar to the biosynthetic gene cluster of rifamycin [29]. A 3amino-5-hydroxybenzoic acid starter unit and two malonyl CoA and eight methyl malonyl CoA extender units could produce intermediate 1 on a type I polyketide synthase. The release of the polyketide chain and the formation of intramolecular amide were catalyzed by the amide synthase encoded by Rmp F and then generated proansamycin X (intermediate 2). Proansamycin X was then catalyzed by a serious of enzymes encoded by Rmp T, U, 11, 5, etc., and turned into the key intermediate 3, dimethyl-desacetyl-rifamycin S. All the above synthetic processes were the same as the synthesis of rifamycin, but Xiao   A-102395 (121) was a capuramycin-type nucleoside antibiotics possessed high specific chemical features, which were isolated from Amycolatopsis sp. SANK 60206. By synthase encoded by Cpr38, chorismate was catalyzed to form 4-amino-4-deoxychorismate (ADC), which subsequently catalyzed elimination of pyruvate by aminotransferase (Cpr12) to form para-aminobenzoic acid (PABA). Catalyzed by actinomycin synthetase (Cpr37), PABA became activated acyl-adenylate and combined with the free-standing carrier protein (Cpr36) to yield the thioester-linked PABA. Under the synergic catalyzation of ketosyn-thase (Cpr34) and chain length factor (Cpr35), the thioester-linked PABA as a recipient was decarboxylatively condensed with malonyl-S-acyl carrier protein (ACP) to form βketothioester. That β-ketothioester was reduced by 3-oxoacyl-ACP reductase Cpr33 and then hydroxylated by luciferase-like monooxygenase Cpr32 to form 3-(4-aminophenyl)-2,3-dihydroxypropanoic acid. The next step was polyamide biosynthesis, in which 3-(4-aminophenyl)-2,3-dihydroxypropanoic acid was catalyzed by a serious of enzymes including a hydrophilic amino acid (Cpr54), two carrier proteins (Cpr48 and 55), a condensation domain protein (Cpr47), and three transglutaminase-like proteins (Cpr49, 50 and 57) to form an A-102395 core. The coupling of the arylamine-containing polyamide to the A-102395 core was catalyzed by carboxyl methyltransferase (Cpr27) and MitI transacylase (Cpr51) [84]. However, the mechanism of Cpr51 has not been proven and needs further research ( Figure 11). The strain Amycolatopsis sp. HCa4 possessed acm gene of amycolamycins A and B, which spanned a ∼76 kb contiguous DNA region. The Acm A 2 , A 3 , A 4 and A 5 provided NDP-glucose dehydrogenase, glucuronic acid decarboxylase, C-methyltransferase and aminotransferase. These enzymes catalyzed the NDP-glucose to NDP activated aminosugar. The 6-methylsalicylic acid synthase, CoA ligase and C-methyltransferase were encoded by Acm B, B 2 and B 1 , which catalyzed three successive steps starting from acetyl-CoA and malonyl-CoA to 3,6-dimethylsalicylyl CoA. The genes of Acm P 1 , P 2 , P 3 , P 9 , P 6 and P 4 encoded the glycosyltransferase, NRPS A-PCP didomain protein, hydroxylase, monooxygenase, O-methyltransferase and halogenase. These enzymes catalyzed six successive steps converting p-hydroxyphenylpyruvate to 2-chloro-3-hydroxy-4,5-dimethoxymandelate moiety. Acetyl-CoA and malonyl-CoA were catalyzed to form enediyne core by a series of enzymes, which were encoded by E, E 2 -E 11 , D 2 , L, M and N. The next step needed B 3 (acetyltransferase) to connect NDP activated aminosugar to 3,6-dimethylsalicylyl CoA. The NDP activated aminosugar and 2-chloro-3-hydroxy-4,5-dimethoxymandelate moiety were then connected to the enediyne core, which needs the Acm A 6 (acetyltransferase) and Acm P 10 (type II condensation enzyme), respectively. The connection product was transformed into compounds 156 and 157 by bergman cyclization (Figure 12) [55].

Chemical Synthesis of DHM2EQ
DHM2EQ was a derivative of epoxyquinomicin C and possessed greater strong inhibitory activity on type II collagen-induced arthritis than epoxyquinomicin C. The derivative was synthesized from 2,5-dimethoxyaniline in 5 steps via chemical synthesis. In pyridine, 2,5-dimethoxyaniline (a) and acetylsalicyloyl chloride were coupled to give salicylamide (b). In methanol, compound (b) was oxidized into quinone monoketal (c) by iodobenzenediacetate. Under deprotection of the phenolic acetyl group, epoxidation of (c) in aqueous THF with alkaline hydrogen peroxide gave epoxide (d). Compound (d) was reduced by NaBH4 yield (e) and the deprotection of compound (e) with p-TsOH gave DHM2EQ ( Figure 13) [85].

Biosynthesis of CDCHD
Chelocardin (CHD) isolated from A. sulphurea was a structurally atypical tetracycline [2]. CHD possessed excellent anti-microbial activity with little toxicity. 2-Carboxamido-2-deacetyl-chelocardin (CDCHD) was a derivative of CDH by introducing oxyD (amidotransferase) and oxyP (thiolase) genes from Streptomyces rimosus otc gene cluster into A. sulphurea. The production of CDCHD was very low when only introduced OxyD into A. sulphurea because OxyP could suppress priming of CDH by removing the competing acetyl units [88]. Then, the CDH gene cluster took over the rest of reaction [89]. These two genes worked together to change the main product from CHD to CDCHD ( Figure 15).

Conclusions
This review summarizes the various chemical structures and biological activities of 159 compounds isolated from Amycolatopsis species inhabiting soil, insects, lichen, islands, the marine and plants between 1990-2020. A total of 45 compounds possessed bioactiv-ities, of which 32 compounds have glycosides and 31 compounds have cyclic skeletons. Thus, the novel compounds with glycosides and cyclic skeletons should be considered by researchers. For example, compound 51, the homolog of 49 and 50, lacked glycoside and showed 5-to 100-fold less cytotoxicity [26]. The multitudinous secondary metabolites of the genus Amycolatopsis represent great research value and deserve further investigation. On the other hand, the genus of Amycolatosis could metabolize a variety of carbon sources and grow in a wide temperature range, which provides the possibility for them to become important biotechnological tools. It has been proven that this genus has great potential in degrading plastics, treating heavy metals, and biotransformation [68]. More researches are needed to transform these potentials into applications to solve practical problems to benefit mankind.
The study of biosynthetic pathway is a crucial process for excavating bioactive natural products. However, people are more willing to study the biosynthesis and mechanism of action of vancomycin, rifamycin and their derivatives. There are relatively fewer studies on the biosynthesis of other bioactive compounds and more attention is needed to be paid to researchers. In the course of the biosynthetic pathway study, a series of tools, for example, antiSMASH [90] or PRISM [91], have been fully exploited, which could derive a prediction of natural products, including the enzymes, regulatory genes and biosynthetic genes et al. through the genome sequencing results. We could also use these tools to reveal sufficiently more silent biosynthetic gene clusters and uncover more and more new interesting bioactive natural products. The biosynthetic potency of Amycolatopsis species is evidenced to be massive and this genus possesses many silent biosynthetic gene clusters waiting to be found [92]. Recently, Pan et al. obtained two new compounds, amycolapeptins A and B by combined-cultivating two strains of Amycoaltopsis sp. 26-4 and Tsukamurella pulmonis TP-B0596 for the first time, while they could not be discovered in a monoculture of Amycoaltopsis sp. 26-4 [93], which provided a new path for the cultivation of Amycolatopsis.
In conclusion, the research on the genus Amycolatopsis needs to be further considered in-depth. Most of all, the mechanism of action and biosynthetic regulatory genes of potent active compounds deserve to be deeply explored since they could determine the utility value of these compounds. Derivatives sometimes tend to have stronger activity so that more study might be focused on the structural modification of secondary metabolites for providing more analogues to be screened for antibiotics. In addition, compounds with excellent bioactivity that have been discovered should be solved for mass production due to their promising medicinal application. The potential ecological effects of Amycolatopsis species should be also taken seriously. The environmental pollution problem might be solved in some ways by thoroughly excavating the biofunction of the strains. In the future, we firmly believe that the genus Amycolatopsis will show its expansive utilization and serve for pharmaceutical area and environmental protection.