Miniaturized Cultivation Profiling (MATRIX)-Facilitated Discovery of Noonazines A–C and Noonaphilone A from an Australian Marine-Derived Fungus, Aspergillus noonimiae CMB-M0339

Subjecting the Australian marine-derived fungus Aspergillus noonimiae CMB-M0339 to cultivation profiling using an innovative miniaturized 24-well plate format (MATRIX) enabled access to new examples of the rare class of 2,6-diketopiperazines, noonazines A–C (1–3), along with the known analogue coelomycin (4), as well as a new azaphilone, noonaphilone A (5). Structures were assigned to 1–5 on the basis of a detailed spectroscopic analysis, and in the case of 1–2, an X-ray crystallographic analysis. Plausible biosynthetic pathways are proposed for 1–4, involving oxidative Schiff base coupling/dimerization of a putative Phe precursor. Of note, 2 incorporates a rare meta-Tyr motif, typically only reported in a limited array of Streptomyces metabolites. Similarly, a plausible biosynthetic pathway is proposed for 5, highlighting a single point for stereo-divergence that allows for the biosynthesis of alternate antipodes, for example, the 7R noonaphilone A (5) versus the 7S deflectin 1a (6).


Introduction
As part of our ongoing investigations into the chemistry of Australian microbes, we applied a global natural products social (GNPS) [1] molecular network analysis to a library of Australian terrestrial and marine bacteria and fungi to identify those most likely to produce natural products with unprecedented (unreported) molecular formulas.Among the many microbes prioritized by this approach was Aspergillus noonimiae CMB-M0339, isolated from a marine sediment collected off Perth, Western Australia.The GNPS analysis revealed that an M1 agar plate cultivation of CMB-M0339 was capable of producing new chemistry, tentatively attributed to the class of indole diterpenes (IDT).To access this new chemistry, we employed a cultivation profiling strategy using a miniaturized 24-well plate approach (MATRIX) to map the metabolomes of CMB-M0339 under 33 different cultivation conditions.Using this approach, we determined that D400 solid-phase cultivation was optimal for IDT production.Following scaled-up cultivation and chemical analysis, we subsequently reported on the new IDT amino acid conjugates, noonindoles A-F (2022 [2]), and related glycosides, noonindoles G-L (2023 [3]).While this initial investigation employed a MATRIX analysis to identify conditions optimal for the production of noonindoles, it also became apparent that CMB-M0339 was capable of producing additional chemistry.For example, during cultivation in either Sabouraud dextrose (SD) or potato dextrose (PD) media, under either solid or broth conditions, the production of noonindoles was downregulated in favor of a new set of natural products featuring a different and distinctive UV-vis chromophore and a unique molecular formula (C 19 H 16 N 2 O 4 and C 19 H 16 N 2 O 5 ).In this report, we describe the chemical analysis of a scaled-up SDA cultivation of CMB-M0339, which resulted in the production, isolation, characterization and structure elucidation of noonazines A-C (1)(2)(3) as new examples of a rare class of 2,6-diketopiperazines, along with the known coelomycin (4), and a new azaphilone, noonaphilone A (5) (Figure 1).dextrose (SD) or potato dextrose (PD) media, under either solid or broth conditions production of noonindoles was downregulated in favor of a new set of natural prod featuring a different and distinctive UV-vis chromophore and a unique molecular form (C19H16N2O4 and C19H16N2O5).In this report, we describe the chemical analysis of a sca up SDA cultivation of CMB-M0339, which resulted in the production, isola characterization and structure elucidation of noonazines A-C (1-3) as new examples rare class of 2,6-diketopiperazines, along with the known coelomycin (4), and a azaphilone, noonaphilone A (5) (Figure 1).

Results
The EtOAc extract of a scaled up SDA (120 plates) cultivation of CMB-M0339 was subjected to sequential solvent trituration across n-hexane, CH 2 Cl 2 and MeOH, with the CH 2 Cl 2 solubles subsequently fractionated by preparative and semi-preparative reversedphase HPLC to yield 1-5 (Scheme S1).Based on their UV-vis (UPLC-DAD) chromophores, the metabolites 1-4 were attributed to a common structure class that differed from that of 5. Significantly, HRESI(+)-MS analysis of 4 revealed a molecular formula (M+Na + , ∆ppm −0.4,C 18 H 14 N 2 O 4 ) which, together with an analysis of the NMR (CDCl 3 ) data (Table S5, Figures S29-S33), identified it as the known 2,6-diketopiperazine coelomycin.First reported from a PD broth cultivation of the Thai seed-derived fungus, Menisporopsis theobromae BCC 3975 (2006 [4]), 4 was subsequently reported as coelomycin from an unidentified Spanish salt marsh leaf litter-derived Coelomycete fungus (2010 [5]).What follows is an account of the structure elucidation of the new metabolites 1-3 and 5.
A plausible polyketide synthetase (PKS) biosynthetic pathway leading to 5 (Scheme 2) could proceed via a typical PKS cascade assembly of three acetate and two propionate residues to arrive at the acyclic intermediate (i), with intramolecular cycloaddition to (ii), and dehydration and oxidation to (iii), followed by acylation with a parallel polyketide polyene (iv) to yield the ester (v).The latter is set up to undergo a second intramolecular cycloaddition to yield (vi) with concomitant loss of water to (vii), followed by keto-enol tautomerization to (viii) and cyclization-mediated termination (and reduction) to (ix), with a facile dehydration yielding 5.This pathway highlights the importance of a stereochemical outcome of the C-7 oxidation in the transformation from (ii) to (iii), which sets up the 7R absolute configuration for 5.A comparable pathway with the opposite C-7 oxidation stereochemical outcome, and with acylation of (iii) with a shorter saturated polyketide, could explain the 7S absolute configuration of deflectin 1a (6).
A plausible polyketide synthetase (PKS) biosynthetic pathway leading to 5 (Scheme 2) could proceed via a typical PKS cascade assembly of three acetate and two propionate residues to arrive at the acyclic intermediate (i), with intramolecular cycloaddition to (ii), and dehydration and oxidation to (iii), followed by acylation with a parallel polyketide polyene (iv) to yield the ester (v).The latter is set up to undergo a second intramolecular cycloaddition to yield (vi) with concomitant loss of water to (vii), followed by keto-enol tautomerization to (viii) and cyclization-mediated termination (and reduction) to (ix), with a facile dehydration yielding 5.This pathway highlights the importance of a stereo- The natural products 1-5 did not inhibit the growth (IC 50 > 30 µm) of the fungus Candida albicans ATCC10231, the Gram-negative bacterium Escherichia coli ATCC11775, Grampositive Bacillus subtilis ATCC6633 and Staphylococcus aureus ATCC25923 (Figure S43), or human colorectal (SW620) or lung (NCI-H460) carcinoma cells (IC 50 > 30 µm) (Figure S44), with the exception of 4, which exhibited modest inhibition of S. aureus (IC 50 3.4 µm).

Global Natural Products Social (GNPS) Molecular Networking
Aliquots (1 µL) of CMB-M0339 cultivation extract (100 µg/mL in MeOH) were analysed on an Agilent 6545 Q-TOF LC/MS equipped with an Agilent 1290 Infinity II UPLC system (Zorbax C 8 , 0.21 µm, 1.8 × 50 mm column, gradient elution at 0.417 mL/min over 2.5 min from 90% H 2 O/MeCN to MeCN with an isocratic 0.1% formic acid/MeCN modifier).UPLC-QTOF-(+) MS/MS data acquired for all samples at a collision energy of 35 eV were converted from Agilent MassHunter data files (.d) to an mzXML file format using MSConvert software (ProteoWizard 3.0.2319964-bit) and transferred to the GNPS server (gnps.ucsd.edu).Molecular networking was performed using the GNPS data analysis workflow [22] employing the spectral clustering algorithm with a cosine score of 0.5 and a minimum of 6 matched peaks.The resulting spectral network was imported into Cytoscape version 3.7.1 [23] and visualized using a ball-stick layout, where nodes represent parent mass, and cosine score was reflected by edge thickness.Also, group abundances were set as pie charts, which reflected the intensity of MS signals.MS/MS fragmentation analysis was performed on the same machine for ions detected in the full scan range at an intensity above 1000 counts at ten scans/sec, with an isolation width of ~4 m/z using fixed collision energy and a maximum of 3 selected precursors per cycle.General instrument parameters included gas temperature at 325 • C, drying gas at 10 L/min, nebulizer at 20 psig, sheath gas temperature at 400 • C, fragmentation Volta at 180 eV and skimmer at 45 eV.

Phylogenetic Comparison of CMB-M0339 with Related Strains
A phylogenetic tree obtained by PhyML Maximum Likelihood analysis was constructed using the top similar 18S rRNA sequences displayed after BLAST on the Refseq RNA NCBI database using CMB-M0339 18S rRNA as queries (Figures S2-S5).The JC69 model was used to infer phylogeny sequences [25].Sequence alignments were produced with the MUSCLE program [26].A phylogenetic tree was constructed using the UGENE program using the aforementioned models and visualized using Ugene's tree view [27].

X-ray Crystallography
Crystallographic data (CuKα radiation 1.54184A • , 2θmax = 125 • ) were collected on an Oxford Diffraction Gemini S Ultra CCD diffractometer with the crystal cooled to 190 K with an Oxford Cryosystems Desktop Cooler.Data reduction and empirical absorption corrections were carried out with the CrysAlisPro program (Oxford Diffraction version 171.38.46).The structure was solved by direct methods with SHELXT and refined with SHELXL [28].The thermal ellipsoid diagrams were generated with Mercury [29].All crystallographic calculations were carried out within the WinGX graphical user interface [30].The crystal data for 4 and 5 in CIF format were deposited in the CCDC database (CCDC 2352835 and 2352836, respectively) (Tables S7-S10).

Antifungal Assay
The fungus Candida albicans ATCC10231 was streaked onto a Luria-Bertani (LB) agar plate and was incubated at 37 • C for 48 h, after which a colony was transferred to fresh LB broth (15 mL) and the cell density adjusted to 10 4 -10 5 CFU/mL.Test compounds were dissolved in DMSO and diluted with H 2 O to prepare 600 µM stock solutions (20% DMSO), which were serially diluted with 20% DMSO to provide concentrations from 600 µM to 0.2 µM in 20% DMSO.An aliquot (10 µL) of each dilution was transferred to a 96-well microtiter plate, and freshly prepared fungal broth (190 µL) was added to prepare final concentrations of 0.01-30 µM in 1% DMSO.The plates were incubated at 27 • C for 48 h, and the optical density of each well was measured spectrophotometrically at 600 nm using a POLARstar Omega plate (BMG LABTECH, Offenburg, Germany).Amphotericin B was used as the positive control (40 µg/mL in 10% DMSO).The IC 50 value was calculated as the concentration of the compound or antibiotic required for 50% inhibition of the bacterial cells using Prism 9.0 (GraphPad Software Inc., La Jolla, CA, USA) (Figure S43).

Antibacterial Assay
The test bacteria (Gram-negative Escherichia coli ATCC11775, and Gram-positive Staphylococcus aureus ATCC25923 and Bacillus subtilis ATCC6633) were streaked onto an LB agar plate and incubated at 37 • C for 24 h, after which a colony was transferred to fresh LB broth (15 mL), and the cell density was adjusted to 10 4 -10 5 CFU/mL.Test compounds were dissolved in DMSO and diluted with H 2 O to produce 600 µM stock solutions (20% DMSO),

, mult. (J in Hz) δ C δ H , mult. (J in Hz) δ C δ H , mult. (J in Hz)
A,B Resonances with the same superscript in a column overlap, and assignments can be interchanged due to low sample size resonances not detected.