Solvolysis Artifacts: Leucettazoles as Cryptic Macrocyclic Alkaloid Dimers from a Southern Australian Marine Sponge, Leucetta sp.

Chemical analysis of a southern Australian sponge, Leucetta sp., led to the discovery of a pair of solvolysis adducts, leucettazoles A1 (1a) and B1 (2a), as artifacts of an unprecedented family of 15-membered macrocyclic alkaloid dimers featuring a pair of imino bridged 2-aminoimidazoles, together with a putative monomeric precursor, leucettazine A (3). The dimeric alkaloids 1a and 2a, and monomer 3, were identified by detailed spectroscopic analysis, supported by chemical transformations, analytical mass spectrometry, and biosynthetic considerations. Global natural product social networking (GNPS) molecular analysis of crude sponge extracts and solvent partitions, supported by single ion extraction (SIE) and diagnostic MS/MS fragmentations, revealed the associated natural products, leucettazoles A (1) and B (2). This study highlights that the study of natural product artifacts can be useful, and can on occasion serve as a pathway to discover cryptic new classes of natural products.

During our studies into sponges, and other Australian marine biodiversity, we regularly encounter natural products that are prone to structural transformation during extraction, isolation, and/or handling. Such compounds, typically denoted as artifacts, enjoy a questionable presence in the field of natural products. For example, where some natural product chemists tend to ignore artifacts, opting instead to describe every isolated compound as a natural product, others are willing to acknowledge the presence of artifacts, but nevertheless relegate them to inconsequential footnotes. In contrast to both these somewhat dismissive stances, we share a deeper appreciation of artifacts, in common with those of our colleagues who recognize, document, report on, and exploit knowledge of their chemical and biological properties. Appreciating the potential value of artifacts served us, and others, very well. For example, in our hands, they informed our understanding of many rare (ii) the solvolytically reactive P-glycoprotein inhibitory diketomorpholine shornephine A, isolated from a marine sediment-derived Aspergillus sp. CMB-M081F [13]; (iii) a family of hydrazine Schiff base prolinimines, recovered from cultivations of a marine fish gastrointestinal tract-derived Trichoderma sp. CMB-F563 [14]; (iv) acid-mediated intramolecular cycloaddition products derived from macrocyclic polyketide cytochalasins, isolated from a marine sediment-derived Phomopsis sp. CMB-M0042F [15]; (v) interconverting spiroketal-polyketide antifungal reveromycins from both marine and terrestrial Streptomyces spp. MST-MA568 and MST-RA7781 [16]; and (vi) oxygen-and light-mediated cycloaddition products from polyene macrolactams, heronamides, isolated from a marine beach sand-derived Streptomyces sp. CMB-M0406 [17,18]. Building on prior experience, this current report demonstrates how the knowledge of solvolysis artifacts isolated from a southern Australian marine sponge, Leucetta sp. (CMB-01047), informed the discovery of an unprecedented family of macrocyclic alkaloid dimers.

Results and Discussion
A portion of the aqueous ethanol (EtOH) extract of a sponge sample Leucetta sp. (CMB-01047) was decanted, concentrated in vacuo, and subjected to a sequence of solvent partitions and triturations, followed by reversed-phase HPLC, to yield leucettazoles A1 (1a) and B1 (2a), and leucettazine A (3) (Figure 1). The structure elucidation of 1a, 2a, and 3 was achieved by detailed spectroscopic analysis, supported by biosynthetic considerations and chemical transformation. The latter included trans-solvolysis of 1a in methanol to leucettazole A2 (1b). With 1a and 2a designated as ethanolysis adducts (i.e., artifacts), the putative natural product leucettazoles A (1) and B (2) were detected and identified in the crude aqueous EtOH extract and n-butanol (n-BuOH) partitions by a combination of (i) ultra-high-performance liquid chromatography/quadrupole time-of-flight (UPLC-QTOF) analysis with single ion extraction (SIE), (ii) global natural product social (GNPS) molecular networking, and (iii) MS/MS analysis. Identification of leucettazole A (1) was further confirmed by spectroscopic analysis of a small, mixed 3:1 sample of 1 and 3.

Leucettazole A1 (1a)
High-resolution electrospray ionization mass spectrometry (HRESI(+)MS) analysis of 1a revealed a molecular ion attributed to a salt with a molecular formula (C22H23N6O6 + , ∆mmu −0.3) incorporating 15 double-bond equivalents (DBE). The 1 H NMR (dimethyl sulfoxide, DMSO-d6) data for 1a (Table 1; Figure S1, Supplementary Materials) revealed resonances attributed to a diastereotopic OEt moiety (OCH2CH3 H 3.31, dq and 3.45, dq). Although this diastereotopic character was suggestive of proximity to a chiral center, the absence of an optical rotation required that 1a be racemic. This, together with long-term storage in EtOH, alerted us to the possibility that 1a was an ethanolysis adduct (i.e., an artifact). The two-dimensional (2D) NMR data for 1a (Table 1 and
Tautomerization notwithstanding, HMBC correlations from (i) H2-6' to C-4' and C-5', and (ii) H-5' to C-2' and C-4' (Figure 2) permitted assembly of the ring C heterocycle in 1a, including linkage to ring D. Finally, to account for the molecular formula, rings B and C were linked via an imino bridge, thereby assembling the macrocyclic structure for leucettazole A1 (1a), as shown. Although 1a was isolated as the trifluoroacetate (TFA) salt, the additional NH resonance was undetected, most likely due to tautomeric equilibrium (arbitrarily represented as the 3-NH tautomer). Of note, ROESY correlations between H-5' and H-12', between H-12' and H-8, and between 3-NH and H-12 suggest that 1a adopts a basket-like (rather than planar) configuration.
As noted above, the presence of a racemic tertiary ethoxy moiety in 1a was taken as evidence that prolonged storage in EtOH induced ethanolysis of a putative natural product, leucettazole A (1). To test this hypothesis, separate samples of crude EtOH extract, and the pure ethoxy adduct 1a were heated in sealed tubes at 40 °C and 60 °C, in either methanol (MeOH) or 50% MeCN/H2O. Consistent with our hypothesis, UHPLC-QTOF with single ion extraction (SIE) analysis of either the crude EtOH extract or pure 1a treated with MeOH yielded a new peak attributed to a methanolysis adduct, leucettazole A2 (1b) (C21H21N6O6 + , mmu +0.1) (Figures S16-S19, Supplementary Materials). As expected, the MeOH adduct was not observed when either the crude EtOH extract or pure 1a was treated with MeCN/H2O ( Figures S20 and S21  Tautomerization notwithstanding, HMBC correlations from (i) H 2 -6 to C-4 and C-5 , and (ii) H-5 to C-2 and C-4 ( Figure 2) permitted assembly of the ring C heterocycle in 1a, including linkage to ring D. Finally, to account for the molecular formula, rings B and C were linked via an imino bridge, thereby assembling the macrocyclic structure for leucettazole A1 (1a), as shown. Although 1a was isolated as the trifluoroacetate (TFA) salt, the additional NH resonance was undetected, most likely due to tautomeric equilibrium (arbitrarily represented as the 3-NH tautomer). Of note, ROESY correlations between H-5 and H-12 , between H-12 and H-8, and between 3-NH and H-12 suggest that 1a adopts a basket-like (rather than planar) configuration.
As noted above, the presence of a racemic tertiary ethoxy moiety in 1a was taken as evidence that prolonged storage in EtOH induced ethanolysis of a putative natural product, leucettazole A (1). To test this hypothesis, separate samples of crude EtOH extract, and the pure ethoxy adduct 1a were heated in sealed tubes at 40 • C and 60 • C, in either methanol (MeOH) or 50% MeCN/H 2 O. Consistent with our hypothesis, UHPLC-QTOF with single ion extraction (SIE) analysis of either the crude EtOH extract or pure 1a treated with MeOH yielded a new peak attributed to a methanolysis adduct, leucettazole A2 (1b) (C 21     a phenol with those for a methoxy (δ H 3.71; δ C 59.9), with the C-10 regiochemistry confirmed by an HMBC correlation from the OCH 3 to C-10 . As with 1a, the structure assigned to leucettazole B1 (2a) was suggestive of an ethanolysis adduct ( Figure 3).

Leucettazole A (1)
Although leucettazoles A (1) and B (2) were detected in the crude extract (see above), neither could be isolated in a pure form. Notwithstanding, spectroscopic analysis of a small quantity of a 3:1 mixed sample of 1 and 3 did provide valuable supporting evidence. More specifically, excellent concordance between the 1D NMR (DMSO-d6) data for 1 with 1a (Table 4; Figures S12-S15, Supplementary Materials), and for key 13 C NMR resonances with 1a, 2a, and 4 ( Figure 2) supported assignment of the structure for leucettazole A (1) as shown.

Leucettazole Biology
Monomeric 2-aminoimidazoles recovered from Leucetta spp. are attributed a range of biological activities including mammalian cytotoxicity [22] and leukotriene B4 (LTB4) receptor antagonism with possible anti-inflammatory potential [23], and inspired synthetic analogs (i.e., leucettines) that exhibit selective kinase inhibitory activity with potential application against Alzheimer's disease and Down

Leucettazole A (1)
Although leucettazoles A (1) and B (2) were detected in the crude extract (see above), neither could be isolated in a pure form. Notwithstanding, spectroscopic analysis of a small quantity of a 3:1 mixed sample of 1 and 3 did provide valuable supporting evidence. More specifically, excellent concordance between the 1D NMR (DMSO-d6) data for 1 with 1a (Table 4; Figures S12-S15, Supplementary Materials), and for key 13 C NMR resonances with 1a, 2a, and 4 ( Figure 2) supported assignment of the structure for leucettazole A (1) as shown.

Leucettazole Biology
Monomeric 2-aminoimidazoles recovered from Leucetta spp. are attributed a range of biological activities including mammalian cytotoxicity [22] and leukotriene B4 (LTB4) receptor antagonism with possible anti-inflammatory potential [23], and inspired synthetic analogs (i.e., leucettines) that exhibit selective kinase inhibitory activity with potential application against Alzheimer's disease and Down

Leucettazole A (1)
Although leucettazoles A (1) and B (2) were detected in the crude extract (see above), neither could be isolated in a pure form. Notwithstanding, spectroscopic analysis of a small quantity of a 3:1 mixed sample of 1 and 3 did provide valuable supporting evidence. More specifically, excellent concordance between the 1D NMR (DMSO-d 6 ) data for 1 with 1a (Table 4; Figures S12-S15, Supplementary Materials), and for key 13 C NMR resonances with 1a, 2a, and 4 (Figure 2) supported assignment of the structure for leucettazole A (1) as shown.

Leucettazole Biosynthesis
A plausible biosynthesis of the leucettazoles could proceed as outlined in Figure 7. In this hypothesis, leucettazine A (3) dimerizes with an oxidized analog via (A) a Schiff base and (B) phenolic coupling, accompanied by (C) C-4 H 2 O addition, to yield the leucettazole macrocycle, with (D) regiospecific 10 -OH methylation occurring pre-or post-dimerization. Supportive of this hypothesis, a Palauan marine sponge Leucetta microraphis was reported to yield a comparable oxidized analog in the form of leucettamine B (6) (Figure 7) [23].

Leucettazole Biosynthesis
A plausible biosynthesis of the leucettazoles could proceed as outlined in Figure 7. In this hypothesis, leucettazine A (3) dimerizes with an oxidized analog via (A) a Schiff base and (B) phenolic coupling, accompanied by (C) C-4 H2O addition, to yield the leucettazole macrocycle, with (D) regiospecific 10'-OH methylation occurring pre-or post-dimerization. Supportive of this hypothesis, a Palauan marine sponge Leucetta microraphis was reported to yield a comparable oxidized analog in the form of leucettamine B (6) (Figure 7) [23].

General Experimental Procedures
Specific optical rotation ([]D) measurements were acquired on a JASCO P-1010 polarimeter in a 100 × 2 mm cell at room temperature (22 °C). Ultraviolet-visible (UV-Vis) spectra were obtained on

General Experimental Procedures
Specific optical rotation ([α] D ) measurements were acquired on a JASCO P-1010 polarimeter in a 100 × 2 mm cell at room temperature (22 • C). Ultraviolet-visible (UV-Vis) spectra were obtained on a Cary 50 UV-visible spectrophotometer. NMR spectra were obtained on a Bruker Avance DRX600 spectrometer equipped with either a 5-mm PASEL 1 H/D-13 C Z-Gradient probe or 5-mm CPTCI 1 H/ 19 F-13 C/ 15 N/DZ-Gradient cryoprobe, controlled by TopSpin 2.1 software. Chemical shifts (ppm) were referenced internally against residual solvent signals (DMSO-d 6 : δ H 2.50, δ C 39.5). High-resolution electrospray ionization mass spectra (HRESIMS) were obtained on a Bruker micrOTOF mass spectrometer by direct infusion in MeCN at 3 µL/min using sodium formate clusters as an internal calibrant. HPLC-MS data were acquired using an Agilent 1100 series separation module equipped with a diode-array multiple wavelength detector coupled to an Agilent 1100 series LC/mass selective detector (MSD), operating in positive and negative modes using electrospray ionization (ESI) mode. Ultra-high-performance liquid chromatography (UHPLC) was performed on an Agilent 1290 infinity UHPLC system composed of a 1290 infinity quaternary pump, thermostat, autosampler, and photodiode-array detector. Analytical, semi-preparative, and preparative HPLC analyses were performed on Agilent 1100 series LC modules, with corresponding detectors and fraction collectors. All solvents used for HPLC separation and purification were chromatographic grade.
Leucettazole A (1): light-yellow powder, as a mixture with 3; UV (MeCN) λ max 318 nm, 284 nm; 1D and 2D NMR (DMSO-d 6 ), see Table 4 and Figures S12-S15 [25]. Molecular networks were generated using the online Global Natural Products Social molecular networking web-platform (GNPS) (gnps.ucsd.edu). MS-Cluster with a precursor ion mass tolerance of 2.0 Da and an MS/MS fragment ion tolerance of 0.5 Da were selected to create consensus spectra [26]. A minimum cluster size of two, cosine score 0.65, and minimum number of fragments of six were selected for molecular networking. The spectral networks were imported into Cytoscape 3.5.1 [27] and visualized using force-directed layout where nodes represented parent masses and edge thickness corresponded to cosine score.

Antibacterial Assays
The bacterium to be tested was streaked onto a tryptic soy agar plate and was incubated at 37 • C for 24 h. One colony was then transferred to fresh tryptic soy broth (15 mL) and the cell density was adjusted to 10 4 -10 5 colony-forming units (CFU)/mL. The compounds to be tested were dissolved in DMSO and diluted with H 2 O to give a 600 µM stock solution (20% DMSO), which was serially diluted with 20% DMSO to give 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 microbial broth (190 µL) was added to each well to give final concentrations of 30-0.01 µM in 1% DMSO. The plates were incubated at 37 • C for 24 h, and the optical density of each well was measured spectrophotometrically at 600 nm using POLARstar Omega plate (BMG LABTECH, Offenburg, Germany). Each test compound was screened against the Gram-negative bacteria Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 10145 and the Gram-positive bacteria Staphylococcus aureus ATCC 25923 and Bacillus subtilis ATCC 6633. Rifampicin was used as a positive control (40 µg/mL in 10% DMSO). For Pseudomonas aeruginosa, a mixture (1:1) of rifampicin and ampicillin (40 µg/mL in 10% DMSO) was used as a positive control. The IC 50 value was calculated as the concentration of the compound or antibiotic required for 50% inhibition of the bacterial cells using Prism 7.0 (GraphPad Software Inc., La Jolla, CA). The antibacterial results are presented in Figure S31 (Supplementary Materials).

Antifungal Assay
The fungus Candida albicans ATCC 10231 was streaked onto a Sabouraud agar plate and was incubated at 37 • C for 48 h. One colony was then transferred to fresh Sabouraud 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 give a 600 µM stock solution (20% DMSO), which was serially diluted with 20% DMSO to give 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 each well to give final concentrations of 30-0.01 µM in 1% DMSO. The plates were incubated at 37 • C for 24 h and the optical density of each well was measured spectrophotometrically at 600 nm using POLARstar Omega plate (BMG LABTECH, Offenburg, Germany). Amphotericin B was used as a positive control (30 µg/mL in 10% DMSO). Where relevant, IC 50 values were calculated as the concentration of the compound or antifungal drug required for 50% inhibition of the fungal cells using Prism 7.0 (GraphPad Software Inc., La Jolla, CA). The antifungal results are presented in Figure S32 (Supplementary Materials).

Cytotoxicity Assays
Adherent human colorectal carcinoma (SW620) and human embryonic kidney (HEK293) cells were cultured in Roswell Park Memorial Institute (RPMI) medium 1640. All cells were cultured as adherent monolayers in flasks supplemented with 10% foetal bovine serum, L-glutamine (2 mM), penicillin (100 unit/mL), and streptomycin (100 µg/mL), in a humidified 37 • C incubator supplied with 5% CO 2 . Briefly, cells were harvested with trypsin and dispensed into 96-well microtiter assay plates at 3000 cells/well after which they were incubated for 18 h at 37 • C with 5% CO 2 (to allow cells to attach as adherent monolayers). Test compounds were dissolved in 20% DMSO in phosphate-buffered saline (PBS) (v/v), and aliquots (10 µL) were applied to cells over a series of final concentrations ranging from 10 nM to 30 µM. After 48 h of incubation at 37 • C with 5% CO 2 , an aliquot (20 µL) of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) in PBS (5 mg/mL) was added to each well (final concentration 0.5 mg/mL), and microtiter plates were incubated for a further 4 h at 37 • C with 5% CO 2 . After final incubation, the medium was aspirated, and precipitated formazan crystals dissolved in DMSO (100 µL/well). The absorbance of each well was measured at 580 nm with a PowerWave XS Microplate Reader from Bio-Tek Instruments Inc. (Vinooski, VT, USA). IC 50 values were calculated using Prism 7.0 (GraphPad Software Inc., La Jolla, CA, USA), as the concentration of analyte required for 50% inhibition of cancer cell growth (compared to negative controls). Negative controls comprised 1% aqueous DMSO, while positive controls used vinblastine as the test sample. All experiments were performed in duplicate. The cytotoxicity results are presented in Figure S33 (Supplementary Materials).

Conclusions
Knowledge of leucettazoles A (1) and B (2) is an intriguing late addition to our understanding of the natural product scaffolds embodied within southern Australian marine sponges, and the proposed biosynthesis provides useful insights into a prospective biomimetic synthesis. Significantly, our discovery of the leucettazoles showcases three important aspects of modern natural product science. The first is the importance of valuing and exploring the potential of artifacts, such as leucettazoles A1 (1a) and B1 (2a), as a window into the world of otherwise reactive (and cryptic) natural products. The second is the remarkable capacity of modern UHPLC-QTOF methods, supported by techniques such as GNPS and SIE visualization, and MS/MS analysis, to give access to new and exciting chemical diversity, overlooked by earlier generations of researchers. The third is the potential relationship between natural products, artifacts, and biological activity. The fact that the leucettazole ethanolysis adducts 1a and 2a did not exhibit cytotoxic properties against a panel of bacterial, fungal, and mammalian cells is perhaps not surprising, and should not discount the possibility that natural products 1 and 2 are bioactive. Much as is the case with Michael acceptors and other bioactive molecules that form covalent bonds with biological targets, it is possible that 1 and 2 are bioactive due to their innate solvolytic reactivity,. If correct, the solvolysis products 1a and 2a could be viewed as inactivated adducts, much as Michael adducts deactivate Michael acceptors. While the ecological purpose and pharmacological potential of the leucettazoles remains a mystery, we contend that this scaffold is deserving of further investigation. That said, given their rarity in nature, and inherent chemical reactivity, further exploration will likely require the input of synthetic and medicinal chemists.
Author Contributions: R.J.C. conceptualized the research, including the focus on artifacts, and assembled the marine sponge collection; P.P. carried out the isolation and spectroscopic characterization of Leucetta chemicals; K.S. performed GNPS analyses; Z.G.K. and M.Q. performed bioassays; P.P., A.A.S., and R.J.C. assigned molecular structures to the Leucetta natural products and artifacts; P.P., A.A.S., Z.G.K., and R.J.C. constructed the Supplementary Materials; R.J.C. reviewed all data and drafted the manuscript, with support from P.P. and A.A.S.

Funding:
This research was supported in part by the University of Queensland and the Institute for Molecular Bioscience.