A gene cluster encoding a nonribosomal peptide synthetase-like enzyme catalyzes γ-aromatic butenolides

Abstract Sea cucumber-derived fungi have attracted much attention due to their capacity to produce an incredible variety of secondary metabolites. Genome-wide information on Aspergillus micronesiensis H39 obtained using third-generation sequencing technology (PacBio-SMRT) showed that the strain contains nonribosomal peptide synthetase (NRPS)-like gene clusters, which aroused our interest in mining its secondary metabolites. 11 known compounds (1–11), including two γ-aromatic butenolides (γ-AB) and five cytochalasans, were isolated from A. micronesiensis H39. The structures of the compounds were determined by NMR and ESIMS, and comparison with those reported in the literature. From the perspective of biogenetic origins, the γ-butyrolactone core of compounds 1 and 2 was assembled by NRPS-like enzyme. All of the obtained compounds showed no inhibitory activity against drug-resistant bacteria and fungi, as well as compounds 1 and 2 had no anti-angiogenic activity against zebrafish.


Introduction
Marine microorganisms are important sources of active natural products with new structures and a treasure house of wealth for screening active molecules against drugresistant bacteria [1], cancer cells [2] and plasmodial organisms [3].For example, ten experienced alkaloids from the sea cucumber derived Aspergillus fumigatus M580 have significant antimicrobial activity against Enterococcus faecalis.Ten known alkaloids from the sea cucumber-derived Aspergillus fumigatus M580 exhibited significant antimicrobial activity against the Gram-positive E. faecalis [4].Chaetoglobosin B showed moderate activity against methicillin-resistant Staphylococcus aureus (MRSA) and S. aureus (SA), with MICs of 94.6 and 47.3 mM, respectively [5].
Studies have revealed that genome sequencing technology has facilitated access to antitumor drug lead compounds and other compounds effective against drug-resistant bacteria [6].The third-generation sequencing technology (PacBio-SMRT) was used to sequence the genome of sea cucumber-derived fungi Aspergillus micronesiensis H39.We analyzed the genomic data of A. micronesiensis H39 using anti-SMASH, and the results showed that the fungus contains rich nonribosomal peptide synthetase-like (NRPS-like) gene clusters that can encode NRPS-like enzymes.NRPS-like enzymes have within an adenylation (A), a thiolation (T), and a thiesterase (TE) domain.TE domain can catalyze the dimerization of two activated a-keto carboxylic acids molecules to form hydroxybenzoquinones or c-AB [7].c-AB is a kind of secondary metabolite originating from fungi with diverse biological activity.According to the literature, Aspergillus, Acanthus, Eutypella, Penicillium, Auxarthron, and Clitoria can produce c-AB, and Aspergillus sp. is the main source of c-AB [8].
As part of seeking biologically active compounds from A. micronesiensis H39, rice medium inoculated with the fungus was thoroughly extracted with ethyl acetate.11 compounds were isolated from the extract by a series of column chromatography methods.Here, we report the large-scale fermentation, isolation, structural identification, cytotoxicity, and the possible biosynthetic pathway of compounds 1 and 2. Our research shows that the sea cucumber-derived fungus has the potential to produce butyrolactones and cytochalasans.
We conducted MIC assays to test the antibacterial activity of compounds 1-11 against drug-resistant bacteria and fungi, the results showed that compounds 1-11 had no antibacterial activity against any of these microorganisms (Supplementary Material Result S1).Compounds 1 and 2 were tested for their ability to inhibit angiogenesis using the zebrafish model, and compounds 1 and 2 showed no signs of inhibiting the growth of zebrafish intersegmental vessels of embryos, indicating a lack of anti-angiogenic activity (Supplementary Material Result S2).According to the report, compound 2 exhibited anti-TMV activity with an IC 50 value of 64.2 mM [10].At a concentration of 5.55 mM, compound 10 exhibited substantial suppression of AChE activity with a noticeable IC 50 value [18].Compounds 3, 5, and 6 exhibited modest cytotoxicity.Specifically, aspochalasins I (3) and J (5) were active against human CNS cancer, with IC 50 values of 19.9 and 13.4 mM, respectively [13].The IC 50 value of 20.0 mM showed that aspochalasin M (6) effectively targeted and inhibited the growth of human leukemic HL-60 cells [14].

Biosynthetic pathways of compounds 1 and 2
One of the functions of genome sequencing is to propose the biogenic synthesis pathway of 1 and 2. Characterization of the A. micronesiensis H39 genome revealed a total size of 32.37 Mbp, N50 of 150,476 and 107 scaffolds (Supplementary Material Table S1).Anti-SMASH 6.1.1 analysis showed that the genome of A. micronesiensis H39 contains 45 putative gene clusters with catalytic secondary metabolite functions, such as NRPS, NRPS-like, terpene, type T1 PKS, and type T3 PKS (Supplementary Material Table S3).
As reported, heterologous expression of NRPS-like gene ATEG_02815 (btyA) from Aspergillus terreus in Saccharomyces cerevisiae expression system resulted in the formation of butyrolactone IIa [19,20], and compounds 1 and 2 share the same chemical skeleton as butyrolactone IIa, both containing a c-butylactone core.In isotope labeling experiments, the radioactivity of phenylalanine-2- 14 C was detected in the C-3 and C-5 regions of the c-butylactone core [21], phenylalanine has the same skeleton as the biosynthetic precursors of compounds 1 and 2, 4-hydroxyphenylpyruvate (HPPA).These evidences suggest that BtyA catalyzes c-AB with the same skeleton as compounds 1 and 2 [22].Using the BtyA amino acid sequences of A. terreus as probes for a local blast in the whole genome sequence of A. micronesiensis H39, we determined that scaffold 000002 F contains a homologous BtyA [23].We then used 2ndFind and NCBI to functionally annotate scaffold 000002 F. This cluster contained three genes located on scaffold 000002 F at 16,704 to 22,762, separately encoding an aromatic prenyltransferase (AmabpB), a methyltransferase (AmbtyB), and a nonribosomal peptide synthetase-like enzyme (AmbtyA) (Figure 2), and thus it is anticipated that these enzymes will play a role in the creation of both compounds 1 and 2 (Figure 3).Sequence alignment revealed that these three enzymes have 33.96%,53.61%, and 57.18% homology with AbpB (UniProtKB No. Q0CX54), BtyB (UniProtKB No. Q0CU18), and BtyA (UniProtKB No. Q0CU19), respectively (Supplementary Material Table S2).
Figure 4 provides a visual representation of the suggested process for the biosynthesis of 1-2.The homologous protein of BtyA, AmbtyA, may be responsible for the biosynthesis of butyrolactone II.AmbtyA contains an A, a T, and a TE domain.The A structural domain of Ambty activates HPPA via adenylation, the T structural domain attaches the activated HPPA to Ambty through a 4 0 -phosphopantetheinyl linker, and the TE structural domain promotes the condensation of HPPA units on Ambty [24].Initially, the HPPA that was activated became associated with the 4 0phosphopantetheine arm of the T domain and subsequently moved to the TE domain [22].When the T domain captures another activated p-hydroxyl PPA through its 4 0phosphopantetheinyl linker, the TE structural domain can facilitate the formation of an enolate between the two HPPA units.The union of aldehyde and alcohol leads to the formation of a carbon-carbon bond, which subsequently undergoes ester cyclization, resulting in the creation of a five-membered (unsaturated) lactone ring that features two carbonyl groups.The unstable five-membered (unsaturated) lactone ring partially undergoes keto-enol tautomerism to obtain a c-butyrolactone core [8].The c-butyrolactone core attached to the TE domain undergoes hydrolysis, and finally, the exposed carboxy group undergoes esterification to produce butyrolactone II.Butyrolactone II generates butyrolactone I under the action of prenyltransferase (Figure 4).Compound 2 was reported to generate compound 1 via a redox reaction [25].

General experimental methods
The Bruker Ascend 400 Avance III HD (Bruker, Karlsruhe, Germany) spectrometer was used to record the NMR spectra of 1D and 2D in DMSO-d 6 .Strains were maintained in an ultracold refrigerator 907 (Thermo fisher, Waltham, United States).The medium was placed in an autoclaved culture cooker, MLS-3020 (Autoclave Engineers, Pennsylvania, United States).The silicone for chromatography separation was from Qingdao Marine Chemical Co., Ltd, Qingdao, China.Fermented bacteria were extracted in an ultrasonic cleaner SB-3200DTNMR (Ningbo Xinzhi Biotechnology Co., Ltd, Ningbo, China).HRESI-MS and ESI-MS data were obtained on the Agilent 6520 Q-TOF (Agilent Technologies, Santa Clara, United States).The Shimadzu LC-20AT device (Shimadzu Corporation, Kyoto, Japan), featuring a UA sensor and a YMC HPLC column Prep C-18 column (10 � 250 mm, S-5 mm), was utilized for conducting semi-preparative HPLC.The compound solution was isolated by a vacuum concentrator scan speed 40 (Gene Company, Hong Kong, China).

Isolation, identification, and preservation of endophytic fungus
Aspergillus micronesiensis H39 was isolated from the sediment of edible sea cucumber gastrointestinal tract in Weihai, Shandong Province.PCR was carried out on the fungus's ITS region, utilizing the universal primers PJ007 (5 0 -TCGACAGAAGATGAT-ATTGAAGGAGCA-3 0 ) and PJ008 (5 0 -AAGAAGGATTACCTCTAAACAAGT-GTACCT −3 0 ) to amplify the sequence.The Omega BIO-TEK Gel Extraction Kit was utilized to purify the PCR products and determine their sequence.The sequence was analyzed by BLAST in NCBI.The nucleotide sequence of A. micronesiensis H39 has been uploaded to the NCBI GenBank database under accession number GenBank OP835917.
The fungus was stored in a 30% aqueous glycerol solution in a −80 � C freezer at the Biology Institute, Qilu University of Technology (Shandong Academy of Sciences).The microorganisms have been deposited in the China General Microbiological Culture Collection Center (CGMCC) with No. 40514.

Genome mining of biosynthetic gene clusters
Whole-genome sequencing of the fungus A. micronesiensis H39 was performed using PacBio-SMRT, a third-generation sequencing technology.To locate the biosynthetic gene cluster responsible for producing the target compound within the genome database of A. micronesiensis H39, we performed a local blast search using a nonribosomal peptide synthase-like BtyA (UniProtKB No. Q0CU19) from A. terreus as a probe.2ndFind and anti-SMASH were used to predict open reading frames (ORFs).UniProt and FGENESH were used to predict protein sequences.We employed the blastP program on the NCBI website to complete the homology alignment of the primary structure of key enzymes.Multiple sequence alignment among AmbtyA with seven similar sequences was accomplished using the ClustalW2 program and the phylogenetic tree was constructed using the MEGA 6.0 software.The genome sequencing datasets of A. micronesiensis H39 have been uploaded to GenBank under the project number PRJNA939002.The putative ORFs in the c-AB biosynthetic gene cluster from A. micronesiensis H39 are listed in Result S2.

Fermentation, extraction, and isolation
A total of sixty 500 ml Erlenmeyer flasks filled with 80 g of rice and 120 ml of deionized water were inoculated with the seed medium of A. micronesiensis H39 and incubated at 28 � C for a duration of 28 days.After that, fermentation was conducted in a dark incubator at 28 � C for 30 days.

Discussion and conclusion
Compounds 1 and 2 are classified as the typical 4,5-disubstituted c-AB, which account for about 54% of reported c-AB.Unlike other 4,5-disubstituted c-AB compounds, compounds 1 and 2 are distinct in that they contain a unique C4-C3 carbon linkage between the C6 00 -C3 00 units and an a,b-unsaturated-c-lactone that connects C-1 to C-4.Compound 1 had two distinctive hydroxyl groups linked to C-8 00 and C-9 00 in the C-3 00 isopentane side chain.Compounds 3-6 belong to the apochalasin class of the cytochalasans, compounds with a parent nucleus characterized by the presence of 2-methylpropyl.Compound 7 belongs to the cytochalasin class, and is characterized by the attachment of a benzyl group to the parent nucleus.The 11-or 12-membered carbocyclic (or oxygen-containing) rings make up compounds 3-7.
Most of the c-ABs showed abundant biological activities including antibacterial, antioxidant, antitumor, anticytotoxic, antiviral, a-glucosidase inhibitory and antiinflammatory activities.Here, a zebrafish assay showed that the 3-methoxy group can affect the anti-angiogenic activity of c-AB: c-AB with 3-hydroxy substituents has better anti-angiogenic activity than c-AB with 3-methoxy substituents [26].
The antibacterial experiment results showed that compounds 3-7 had no antibacterial activity, as reported, the absence of an unsubstituted lactam ring and an a,b-unsaturated carbonyl group in the macrolide moiety of compounds 3-7 may be the reason for the lack of antibacterial activity [27].
The abundant NRPS-like gene clusters provide a prerequisite for the biosynthesis of compounds 1 and 2. NRPS-like enzyme is a single modular NRPS enzyme that contains a TE domain instead of a peptide-forming C domain.The NRPS-like enzymes of homologous fungi have good compatibility and their domains can be successfully exchanged [7].Compounds 1 and 2 are the products of an NRPS-like enzyme, the c-butyrolactone core being biosynthesized from HPPA as a precursor under the action of the AmbtyA.
These results showed that A. micronesiensis H39 has the potential to catalyze secondary metabolites with unique structures, which is worth exploring in the future.