Cytotoxic Naphthoquinone Analogues, Including Heterodimers, and their Structure Elucidation Using LR-HSQMBC NMR Experiments

Abstract

Approximately 1700 naphthoquinones have been reported from a range of natural product source materials, but only 283 have been isolated from fungi, fewer than 75 of those were dimers, and only two were heterodimers with a head to tail linkage. During a search for anticancer leads from fungi, a series of new naphthoquinones (1-4), including two heterodimers (3 and 4), were isolated from Pyrenochaetopsis sp. (strain MSX63693). In addition, the previously reported 5-hydroxy-6-(1-hydroxyethyl)-2,7-dimethoxy-1,4-naphthalenedione (5), misakimycin (6), 5-hydroxy-6-[1-(acetyloxy)ethyl]-2,7-dimethoxy-1,4-naphthalenedione (7), 6-ethyl-2,7-dimethoxyjuglone (8) and kirschsteinin (9) were isolated. While the structure elucidation of 1-9 was achieved using procedures common for natural products chemistry studies (HRESIMS, 1D and 2D NMR), the elucidation of the heterodimers was facilitated substantially by data from the LR-HSQMBC experiment. The absolute configuration of 1 was established by analysis of the measured vs calculated ECD data. The racemic mixture of 4 was established via X-ray crystallography of an analogue that incorporated a heavy atom. All compounds were evaluated for cytotoxicity against the human cancer cells lines MDA-MB-435 (melanoma), MDA-MB-231 (breast) and OVCAR3 (ovarian), where the IC₅₀ values ranged between 1 and 20 μM.

Graphical Abstract

While nature has been a proven source for drug leads in the anticancer realm,^{1, 2} it is a sobering fact that cancer remains the number two cause of mortality, accounting for over 9 million deaths annually in the world,³ including over 600,000 in the United States.⁴ Although fungi have been a source of many drug leads,^{1, 5–7} an anticancer drug of fungal origin remains to be developed. With this goal in mind, and given the chemical space afforded by compounds of fungal origin,^{8, 9} our team has been investigating the Mycosynthetix library of filamentous fungi.^{10, 11} An organic extract from the fungus Pyrenochaetopsis sp. (strain MSX63693) was selected due to the extract’s cytotoxic activity against the human cancer cells lines MDA-MB-435 (melanoma), MDA-MB-231 (breast) and OVCAR3 (ovarian). Nine naphthoquinone analogues were isolated and characterized, including four new compounds (1-4), along with the known compounds 5-hydroxy-6-(1-hydroxyethyl)-2,7-dimethoxy-1,4-naphthalenedione (5),^{12, 13} misakimycin (6),¹⁴ 5-hydroxy-6-[1-(acetyloxy)ethyl]-2,7-dimethoxy-1,4-naphthalenedione (7),¹³ 6-ethyl-2,7-dimethoxyjuglone (8),^{12, 15} and kirschsteinin (9).¹³

As their name indicates, naphthoquinones, a commonly encountered class of secondary metabolites, are characterized by a naphthalene skeleton with two carbonyl groups.^{16, 17} They have been reported from substrates across the natural products sciences, including plants (e.g., juglone, lapachol and plumbagin),¹⁸ marine organisms (e.g., toluhydroquinone),¹⁹ actinobacteria (e.g., naphterpin, marinone and neomarinone),¹⁶ and fungi (e.g., mompain, herbarin and fusarubin).²⁰ The broad range of bioactivities reported for these compounds indicates that they are antiallergic, antibacterial, antifungal, anti-inflammatory, antiplatelet, antiprotozoal, antithrombotic, antiviral, and cytotoxic.^{16, 21} The latter activity of these has been attributed to the modulation of radical anion species by generation of reactive oxygen species (ROS), including hydrogen peroxide and superoxide anions, which damage DNA and certain essential proteins.^{22, 23} While the Dictionary of Natural Products reports 1688 metabolites with a naphthoquinone core, only 283 have been reported from fungi, and only 74 of those are dimers, with only two (kirschsteinin¹³ and deacetylkirschsteinin²⁴) being heterodimers with a head-to-tail linkage (Figure S1).²⁵ We were particularly intrigued with the heterodimers, and in addition to uncovering new chemical diversity, a goal was to test emerging NMR pulse sequences^26–29 as a means to enhance the structure elucidation of such molecules.

RESULTS AND DISCUSSION

An organic extract of the fungus Pyrenochaetopsis sp. (strain MSX63693, Figure S2) was prioritized based on cytotoxicity (i.e., less than 30% survival when tested at 20 μg/mL) against the human cancer cells lines MDA-MB-435 (melanoma), MDA-MB-231 (breast) and OVCAR3 (ovarian); the sample was shown to contain no known cytotoxic compounds when compared to a database with >570 fungal metabolites.^{30, 31} Bioactivity-directed purification led to the isolation of four new compounds as described below. Additionally, the known compounds 5-hydroxy-6-(1-hydroxyethyl)-2,7-dimethoxy-1,4-naphthalenedione (5),^{12, 13} misakimycin (6),¹⁴ 5-hydroxy-6-[1-(acetyloxy)ethyl]-2,7-dimethoxy-1,4-naphthalenedione (7),¹³ 6-ethyl-2,7-dimethoxyjuglone (8)^{12, 15} and kirschsteinin (9)¹³ were isolated and identified based on comparisons to literature NMR and HRESIMS data; their 1H NMR spectra are included in the Supporting Information (Figures S3-S7).

Compound 1 (0.7 mg) was isolated as an optically active yellow powder. The molecular formula was determined by HRESIMS data as C₁₄H₁₈O₅, consistent with an index of hydrogen deficiency of 6. 1H, 13C and HSQC NMR data indicated the presence of one aromatic methine (δ_H/δ_C 6.74/102.1), two oxygenated methines (δ_H/δ_C 4.86/69.3 and 3.94/78.4), two methoxy groups (δ_H/δ_C 3.92/56.0 and 3.45/57.1), two methylenes (δ_H/δ_C 3.11, 2.76/39.1 and 2.64/15.8), one methyl (δ_H/δ_C 1.07/13.2), one hydroxy proton (δ_H 2.69), and six fully substituted carbons, including a carbonyl (δ_C 199.4), and five aromatic signals (δ_C 164.1, 161.7, 142.5, 119.3, and 110.1) (Table 1). These data, especially δ_C-1 199.4, and the signals for the penta-substituted aromatic ring (δ_H-5/δ_C-5 6.74/102.1, δ_C-6 164.1, δ_C-7 119.5, δ_C-8 161.7, δ_C-9 110.1, and δ_C-10 142.5), were suggestive of a naphthalenone core.^32–34 Specifically, the molecular formula and NMR data for 1 resembled those for the naphthalenone, 3S*,4S*−7-ethyl-4,8-dihydroxy-3,6-dimethoxy-3,4-dehydronaphthalen-1(2H)-one.³² The 2D structure of compound 1 was corroborated by the two isolated spin systems (H-2/H-3/H-4 and H-1’/H-2’) observed by coupling constants in the 1H NMR spectrum, which were connected via HMBC correlations (Figure 1). This indicated that compound 1 was a stereoisomer of the known naphthalenone, based on slight differences in the chemical shifts and J values for the signals at positions 2, 3, and 4 (Table S1). The relative configuration of 1 was established based on the 3J_H-3/H-4 coupling constant of 3.2 Hz, suggesting a cis-pseudoaxial-pseudoequatorial arrangement (Figure 2);²⁴ in contrast, the trans-pseudoequatorial-pseudoequatorial arrangement would have a 3J_H-3/H-4 value less than 2.5 Hz.³⁵ The absolute configuration of 1 was established by analyzing the ECD spectrum, which displayed negative and positive Cotton effects at 220 and 280 nm, respectively. Compared with the analogue 4,8-dihydroxy-3-methoxy-3,4-dihydro-1(2H)-naphthalenone, which had the 3S,4R configuration,²⁴ the ECD data were opposite, suggesting the absolute configuration of 1 as 3R,4S. This was confirmed via ECD calculations, using a time-dependent DFT (TDDFT) method with the B3LYP/6–31G+(d) level of theory, demonstrating concordance between the experimental and calculated spectra (Figure 2). These data were used to characterize 1 as 3R,4S-7-ethyl-4,8-dihydroxy-3,6-dimethoxy-3,4-dehydronaphthalen-1(2H)-one, which was ascribed the trivial name botryosphaerone E (1) due to similarities with botryosphaerone A-D.³³

Table 1.

¹H and ¹³C NMR Data for Compounds 1 and 2 (400 and 100 MHz, Respectively in CDCl₃).

position	δC, type	1	δC, type	2
		δH (J in Hz)		δH (J in Hz)
1	199.4, C		180.7, C
2	39.1, CH2	3.11, dd (17.5, 6.1) 2.76, dd (17.5, 3.4)	160.5, C
3	78.4, CH	3.94, m	109.7, CH	6.03, s
4	69.3, CH	4.86, dd (8.0, 3.2)	190.1, C
5	102.1, C	6.74, s	161.7, C
6	164.1, C		123.8, C
7	119.3, C		163.5, C
8	161.7, C		103.2, CH	7.24, s
9	110.1, C		131.5, C
10	142.5, C		109.0, C
1ˈ	15.8, CH2	2.64, q (7.5)	70.7, CH	5.07, q (6.8)
2ˈ	13.2, CH3	1.07, t (7.5)	19.0, CH3	1.59, d (6.8)
3-OCH3	57.1, CH3	3.45, s
2-OCH3			56.8, CH3	3.90, s
6-OCH3	56.0, CH3	3.92, s
7-OCH3			56.6, CH3	3.99, s
1ˈ-OCH3			57.0, CH3	3.27, s
4-OH		2.69, d (8.1)
5-OH				12.8, s

Figure 1.

Key spin systems (from 1H NMR data) and HMBC correlations for 1 and 2.

Figure 2.

Key coupling constants for 1 (left) and comparison of the calculated vs experimental ECD spectra in CH₃OH for compound 1 (right).

Compound 2 (3.5 mg) was isolated as yellow powder, and the molecular formula was determined as C₁₅H₁₆O₆, yielding an index of hydrogen deficiency of 8 based on the HRESIMS data, including both for the protonated molecule and the Na⁺ adduct. Analysis of 1H, 13C and HSQC NMR experiments indicated the presence of a chelated phenolic proton (δ_H 12.78), two aromatic methines (δ_H/δ_C 7.24/103.2 and 6.03/109.7), one oxygenated methine (δ_H/δ_C 5.07/70.7), three methoxy groups (δ_H/δ_C 3.99/56.6, 3.90/56.8, and 3.27/57.0), one methyl (δ_H/δ_C 1.59/19.0), two ketones (δ_C 190.1 and 180.7) and six fully substituted aromatic carbons (δ_C 163.5, 161.7, 160.5, 131.5, 123.8, and 109.0) (Table 1). These data resembled those for compound 7,^{12, 15} which was also isolated. The key difference was that 2 had a methoxy group connected to the ethyl moiety in position 6. Specifically, the spin system between H₃-2ˈ and H-1ˈ (δ_H 1.59 and 5.07, 3J= 6.8) and HMBC correlations of H₃-2ˈ with C-1ˈ and C-6, of H-1ˈ with C-2ˈ, OCH₃-1ˈ, C-6, C-7, and C-5, and of OCH₃-1ˈ with C-1ˈ confirmed the presence of the methoxy group and established the connection of the 1ˈ-methoxyethyl moiety with the 6 position of the naphthoquinone (Figure 1). To probe the absolute configuration, the specific rotation data were calculated for 2, 5, and 7, and in all cases the [α]_D values were either zero or very close to zero (data not shown). Additionally, Mosher’s esters analysis was performed on 5,³⁶ and the generation of a 1:1 ratio of two diastereoisomers was observed when reacted with either the S or R-3,3,3-trifluoro-2-methoxy-2-phenylpropanoic acid further supported a racemic mixture (Figure S16). These results indicated that the new compound 2 and the known compounds 5 and 7 were present as racemic mixtures. Compound 2 was ascribed the trivial name (1’-methoxy)-6-ethyl-2,7-dimethoxyjuglone (2).¹²

Compounds 3 (16.5 mg) and 4 (10.6 mg) were isolated as optically inactive powders. The molecular formulas were calculated by HRESIMS based on the protonated molecules as C₂₇H₂₄O₁₀ (3) and C₂₅H₂₀O₁₀ (4). Analysis of the 13C NMR spectra for 3 and 4 both showed paired signals, which were consistent with the presence of two naphthoquinone subunits; these data resembled those for the structure of the dimeric naphthoquinone, kirschsteinin (9), which was also isolated. Comparison of the 13C NMR spectra for 3 and 4 with 9 showed similarities for the signals assignable for subunit A, indicating that 3 and 4 each consisted of a hydroxy-dimethoxy-naphthoquinone moiety; this was confirmed by HSQC and HMBC experiments (Figure 3 and Table 2). Thus, the key difference between 3 and 4 with 9 was the substitution pattern for subunit B. Specifically, 3 and 4 had an extra methyl group, noted as a methoxy moiety at the 7’ position (δ_H/δ_C 3.94/56.5 and 3.88/56.5, respectively), which was confirmed via HMBC correlations (Figure 3). Additionally, compounds 3 and 4 lacked the acetyl group at position C-6ˈ. For 3, this was replaced by aliphatic C-11ˈand C-12ˈ [δ_H/δ_C 2.69, q (J = 7.4)/16.6 and 1.07, t (J = 7.4)/13.0], where HMBC correlations between H-8ˈto C-7ˈ; H₃-12ˈ to C-11ˈ and C-6ˈ; and H₂-11ˈ to C-12ˈ, C-6ˈ, C-5ˈ, and C-7ˈ confirmed the attachment of the ethyl moiety to the naphthoquinone ring (Figure 3). Alternatively, the lack of any substituent at position C-6ˈ in 4 was confirmed by the presence of the AB system between the protons at δ_H-6’ 6.63, d (J = 2.5) and δ_H-8’ 7.17, d (J = 2.5), consistent with a tetra-substituted aromatic ring (Figure 3). To conclude the elucidation of the heterodimers, the bonding of the A and B units for 3 and 4 was clear via the ethylidene fragment at positions C-11 and C-12 [δ_H/δ_C 4.95, q (J = 7.4)/28.1 and 1.68, d (J = 7.4)/17.3, respectively], which had strong HMBC correlations with C-5, C-6, C-7, C-2ˈ, C-3ˈ, and C-4ˈ, supporting the head-to-tail linkage (Figure 3).

Figure 3.

Spin systems from the 1H NMR spectra and key HMBC correlations for 3 and 4.

Table 2.

¹H and ¹³C NMR Data for Compounds 3 and 4 (400 and 100 MHz, Respectively in CDCl₃).

	3				4
Position	δC, type	δH (J in Hz)	HMBC	LR-HSQMBC	δC, type	δH (J in Hz)	HMBC	LR-HSQMBC
1	179.7, C				179.7, C
2	160.6, C				160.6, C
3	109.5, CH	6.01, s	1, 4, 2, 10	5, 6, 9	109.5, CH	6.01, s	1, 4, 2, 10	5, 6, 9
4	190.1, C				190.1, C
5	161.0, C				160.9, C
6	127.4, C				127.3, C
7	162.9, C				162.8, C
8	103.6, CH	7.23, s	1, 6, 7, 10	4, 12	103.6, CH	7.23, s	1, 6, 7, 9, 10	4, 12
9	130.4, C				130.4, C
10	109.1, C				109.1, C
11	28.1, CH	4.95, q (7.4)	2’, 3’, 4’, 5, 6, 7, 12	2’	28.1, CH	4.95, q (7.4)	2’, 3’, 4’, 5, 6, 7, 12	2’
12	17.3, CH3	1.68, d (7.4)	3’, 6, 11	3’	17.3, CH3	1.68, d (7.4)	3’, 6, 11	3’
2-OCH3	56.5, CH3	3.88, s	2	3	56.5, CH3	3.88, s	2	3
7-OCH3	56.7, CH3	3.94, s	7	8	56.7, CH3	3.94, s	7	8
5-OH		12.82, s	5, 6, 10	4, 7		12.83, s	5, 6, 10	4, 7
1ˈ	181.7, C				181.6, C
2ˈ	153.9, C				154.0, C
3ˈ	124.4, C				124.6, C
4ˈ	190.4, C				189.8, C
5ˈ	161.0, C				164.0, C
6ˈ	128.8, C				107.9, CH	6.63, d (2.5)	5’, 7’, 8’, 10’,	4’, 9’
7ˈ	161.7, C				164.8, C
8ˈ	103.6, CH	7.19, s	1’, 6’, 7’, 10’	4’, 5’	107.8, CH	7.17, d (2.5)	1’, 6’, 7’, 10’	4’, 5’
9ˈ	127.9, C				130.6, C
10ˈ	109.4, C				108.9, C
11ˈ	16.6, CH2	2.69, q (7.4)	5’, 6’, 7’, 12’	9’	-
12ˈ	13.0, CH3	1.07, t (7.4)	6’, 11’	9’	-
7ˈ-OCH3	56.2, CH3	3.93, s	7’	8’	56.1, CH3	3.86, s	7’	6’, 8’
2ˈ-OH		7.63, s	-	3’, 9’, 4’		7.63, s	-	3’, 9’, 4’
5ˈ-OH		12.70, s	5’, 6’, 10’	7’		12.60, s	5’, 10’	7’

To confirm the position for the methoxy groups and the connectivity between the two naphthoquinone units for compounds 3 and 4, the LR-HSQMBC NMR experiment was utilized to observe long range correlations (4–6J_CH). This relatively new technique was reported by Williamson and colleagues in 201426 and has been applied for the elucidation of natural products with proton-deficient scaffolds.^27–29 The data from this experiment (Table 2 and Figure 4) showed cross-peaks that were not observed in the conventional HMBC spectra, facilitating the unequivocal assignments of some positions. Specifically, the correlations between 2-OCH₃ to C-3, and 2’-OH to C-4’, confirmed that subunit A had a methoxy moiety at position C-2, while in subunit B there was a phenol at the equivalent position (Figure 4). In contrast, in the HMBC spectrum, the correlations for these methoxy moieties were overlapped. In addition, the assignments for the 5-OH, 5’-OH, 7-OCH₃, and 7’-OCH₃ positions were further confirmed in both 3 and 4 via key 4J_CH correlations (Figure 4). Williams and others^27–29 have noted the value of this experiment for proton deficient molecules, and in this case, we found it beneficial to verify assignments of closely related signals using data orthogonal to HMBC correlations. This experiment was straightforward to set up and required approximately the same amount of time as a conventional HMBC experiment on the same amount of sample. However, the processing of these data was somewhat challenging, since defaults for commonly used software (i.e., Mnova-Mestrelab) have not been created yet. As such, key steps and tips for data processing are outlined in the Supporting Information (Figure S27 and Table S2).

Figure 4.

Key LR-HSQMBC correlations for 3 and 4.

There are only three natural asymmetrical dimers reported,^{13, 24, 37} one of those being 9, with the same type of linkage as 3 and 4, and the configuration of C-11 has never been determined. Based on the [α]_D values being close to zero, it was possible that these dimers were present as racemic mixtures. Therefore, to probe the presence of the racemates, compound 4 was reacted with 4-bromobenzoyl chloride to generate an analogue suitable for X-ray crystallography^{38, 39} that incorporated a heavy atom. The structure of the reaction product was confirmed via HRESIMS and 1H NMR data. Crystals formed readily after approximately 4 days from a 3:7 CHCl₃-CH₃OH solution, and X-ray crystallographic analysis confirmed the presence of both the 11S and 11R enantiomers for compound 4 (Figure S30 and Table S3). Given the similar specific rotation data for all of these dimers [i.e., ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ +2, (c 0.1, CHCl₃, 3), ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ +3, (c 0.1, CHCl₃, 4) and ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ −2, (c 0.1, CHCl₃, 9)], the racemic mixture was presumed for both 3 and 9 as well. Thus, compound 3 was identified as 6-ethyl-​3-​[1-​(5,​8-​dihydro-​1-​hydroxy-​3,​6-​dimethoxy-​5,​8-​dioxo-​2-​naphthalenyl)​ethyl]​-​2,​5-dihydroxy-​7-methoxy-1,​4-​naphthalenedione, and ascribed the trivial name kirschsteinin B (3), and compound 4 as 3-​[1-​(5,​8-​dihydro-​1-​hydroxy-​3,​6-​dimethoxy-​5,​8-​dioxo-​2-​naphthalenyl)​ethyl]​-​2,​5-dihydroxy-​7-methoxy-1,​4-​naphthalenedione, and ascribed the trivial name kirschsteinin C (4).

Owing to the fact that a series of racemates were isolated, the chemistry of the fungal culture was analyzed in situ using the droplet probe,⁴⁰ so as to examine the authenticity of these as natural products.^{41, 42} This technique has been used to analyze the chemistry of a variety of natural products substrates in situ, including fungal cultures,^43–46 cyanobacteria,⁴⁷ and plants.⁴⁸ Compounds 2-5, 7, and 9 were confirmed by analyzing the droplet probe data (Figure S31). While this does not rule out the possibility of racemization during the processing of the fungal extracts, it does validate that these compounds were present in the fungus, including the heterodimers 3, 4, and 9. Upon peer review, a further question was raised regarding the possibility that compounds 2, 5, and 7 could be generated from 8. To test this, compound 8 was exposed to CH₃OH, CH₃OH-H₂O (7:3) and CH₃OH with acetic acid (10%) over 24 h at 35 °C. UPLC-HRESIMS analysis of the reaction products did not indicate the production of 2, 5, or 7 (Figure S32). In total, these experiments provide evidence that compounds 2-5, 7, and 9 were biosynthesized by the fungus.

The cytotoxic activities of compounds 1-9 were tested against the human cancer cells lines MDA-MB-435 (melanoma), MDA-MB-231 (breast) and OVCAR3 (ovarian) using protocols described previously.⁴⁹ With the heterodimers 3, 4, and 9, compound 9 was the most potent, suggesting that the acyl moiety at the 6’ position enhanced activity. The monomeric analogues were roughly equipotent, with a loss in activity observed in 1 that was likely due to the reduction to the naphthoquinol.

In summary, four new naphthoquinones were discovered (1-4), with the latter three having cytotoxic activity against a panel of human cancer cell lines. In particular, two of these compounds were heterodimers with a head to tail linkage (3 and 4), bringing the total number reported from nature to five (i.e., one from a plant³⁷ and now four from fungi^{13, 24}). In addition to prominent structure elucidation techniques, such as 1D and 2D NMR spectroscopy, mass spectrometry, and X-ray crystallography, a relatively new technique, the LR-HSQMBC NMR experiment, was utilized. Most previous reports of this technique illustrate its value for proton deficient molecules. In this study, LR-HSQMBC NMR allowed us to differentiate signals for closely related chemical shifts, as are common in heterodimers.

EXPERIMENTAL SECTION

General Experimental Procedures.

Optical rotations were obtained using a Rudolph Research Autopol III polarimeter using CHCl₃ as solvent. UV and ECD spectra were acquired on a Varian Cary 100 Bio UV-Vis spectrophotometer and an Olis DSM 17 CD spectrophotometer, respectively. IR spectra were collected using a PerkingElmer Spectrum 65 with Universal ATR attachment. NMR data were collected using either a JEOL ECS-400 NMR spectrometer equipped with a high sensitivity JEOL Royal probe operating at 400 MHz for 1H and 100 MHz for 13C; or an JEOL ECA-500 NMR spectrometer operating at 500 MHz for 1H and 125 MHz for 13C; the residual solvent signals of CDCl₃ (δ_H = 7.260 and δ_C = 77.160) were utilized for referencing. HRMS experiments utilized either a Thermo LTQ Orbitrap XL mass spectrometer or a Thermo Q Exactive Plus (Thermo Fisher Scientific); both were equipped with an electrospray ionization source. UPLC was carried out on a Waters Acquity system with data collected and analyzed using Thermo Xcalibur software. HPLC was carried out using a Varian ProStar HPLC system equipped with ProStar 210 pumps and a ProStar 335 photodiode array detector (PDA), with data collected and analyzed using Galaxie Chromatography Workstation software (version 1.9.3.2). Preparative reversed-phase HPLC purification of samples was performed on a Gemini Prep C₁₈ (5 μm; 250 × 210 mm) column using a 21 mL/min flow rate of the mobile phase consisting of a mixture of CH₃CN and H₂O (with 0.1% formic acid). Flash column chromatography was carried out with a Teledyne ISCO Combiflash Rf connected to both an evaporative light scattering detector and a PDA detector with UV set at 200–400 nm.

Fungal Strain Isolation and Identification.

Fungal strain MSX63693 was first identified using morphology by growing on malt extract agar (Difco); however, evidence of sporulation of either sexual or asexual spores was not observed. Therefore, molecular data from the internal transcribed spacer (ITS) rDNA region was used via a BLAST search and maximum likelihood analysis to determine the phylogenetic placement of strain MSX63693. The protocols for DNA extraction, PCR amplification, and Sanger sequencing were outlined previously.⁵⁰ To sequence the entire ITS region, a primer combination of ITS1F and ITS4^{51, 52} was utilized.⁵³ Two ITS sequences were obtained to ensure quality of the data and account for possible contamination. Both ITS sequences were identical and were 515 and 505 bp, respectively. Based on the NCBI BLAST search using the rRNA/ITS database with RefSeq data,⁵⁴ the closest matches (≥97–98%) were sequences of the coelomycetous Dothideomycetes fungal genus Pyrenochaetopsis.^55–58 Subsequently all type strains from this genus were downloaded from GenBank and aligned using MUSCLE in the program Seaview v. 4.5.3,⁵⁹ and a RAxML analysis⁶⁰ was performed on the CIPRES server⁶¹ using default settings with 1000 bootstrap replicates to provide statistical support to the clades on the phylogeny. The resulting analysis indicated that strain MSX63693 was placed with other species within the genus Pyrenochaetopsis,⁵⁵ and sister to P. terricola. The clades were poorly supported in our ITS phylogeny compared to the multigene phylogeny reported in Lopez et al.⁵⁶ Although, this strain showed strong bootstrap support (≥90%) as being sister to P. terricola, we err on the side of caution and identify it conservatively as Pyrenochaetopsis sp. (Pyrenochaetopsidaceae, Pleosporales, Ascomycota). Phylogenetic analysis using multiple gene regions is necessary to identify the strain accurately to the species level. The sequence data were deposited in GenBank (accession no: ITS: MT636852, MT636853).

Fermentation, Extraction, and Isolation.

To each 250 mL Fernbach flask (8 flasks in total) containing a sample of ~10 g of rice with fungal growth, 60 mL of 1:1 CH₃OH-CHCl₃ was added. The sample was chopped with a spatula and shaken for ~16 h at ~100 rpm at rt. The eight samples were combined and filtered in vacuo, and the remaining residue was washed with CH₃OH. To the filtrate, 90 mL of CHCl₃ and 150 mL of H₂O were added. The biphasic solutions were stirred on a stir plate for 30 min and then transferred to a separatory funnel. The organic bottom layer was drawn off into a round-bottom flask, which was evaporated to dryness. The dried organic extract was then reconstituted in 100 mL of hexanes and 100 mL of 1:1 CH₃OH-CH₃CN, the biphasic solution was shaken vigorously and then transferred to a separatory funnel. The bottom layer was drawn off into a round bottom flask and evaporated to dryness under vacuum. The resulting defatted extract (1.2 g) was dissolved in a mixture of CHCl₃-CH₃OH and adsorbed onto Celite 545, prior to purification via flash chromatography using a RediSep silica column (24g) and a gradient solvent system of hexane-CHCl₃-CH₃OH at a flow rate of 35 mL/min over 39 min; 41 column volumes were collected to afford four fractions. Fraction 1 was subjected to preparative RP-HPLC using a Gemini C₁₈ column with a gradient system of 35:65 to 100:0 of CH₃CN-H₂O (acidified with 0.1 % formic acid) over 15 min at a flow rate of 21.2 mL/min to yield 6 subfractions. Subfractions 1, 3, 5 and 6 yielded compounds 5 (t_R= 7.5 min, 9.5 mg), 8 (t_R= 13.2 min, 150.4 mg), 9 (t_R= 15.0 min, 20.5 mg) and 3 (t_R= 17.0 min, 16.5 mg), respectively. Subfraction 2 was subjected to a preparative RP-HPLC using a Gemini column with a gradient system of 50:50 to 70:30 of CH₃CN-H₂O (acidified with 0.1 % formic acid) over 30 min at a flow rate of 21.2 mL/min to generate compounds 6 (t_R= 23.0 min, 6.3 mg), 1 (t_R= 25.0 min, 0.7 mg), 2 (t_R= 26.5 min, 3.5 mg), 7 (t_R= 33.6 min, 8.2 mg) and 4 (t_R= 36.0 min, 10.6 mg).

The experiments with the droplet probe were conducted in a manner analogous to recent reports.^{40, 43, 46, 62} Briefly, a droplet of 1:1 MeOH:H₂O was analyzed using a gradient of 15:85 to 0:100 of CH₃CN-H₂O (acidified with 0.1 % formic acid) over 10 min, using a BEH C₁₈ column with a flow rate of 0.3 mL/min.

Botryosphaerone E (1):

Yellow solid. ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ −63 (c 0.1, CHCl₃); UV (CHCl₃) λ_max (log ε) 215 (4.5), 291 (4.1) nm; ECD (0.1 mg/mL, CH₃OH) λ_max (Δε) 220 (−14), 280 (+10), 320 (−3) nm; 1H and 13C NMR, Table 1; HRESIMS m/z 267.1224 [M + H]+ (calcd for C₁₄H₁₉O₅, 267.1227), m/z 289.1043 [M + Na]+ (calcd for C₁₄H₁₈O₅Na, 289.1052).

(1’-Methoxy)-6-ethyl-2,7-dimethoxyjuglone (2):

Orange solid. ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{25}$ 0 (c 0.1, CHCl₃); UV (CHCl₃) λ_max (log ε) 259 (3.3), 311 (3.2), 433 (2.9) nm; IR ν_max 3316, 2924, 1684, 1624, 1590, 1487, 1412, 1360, 1303, 1252, 1208, 1180, 1108, 1080 cm–1; 1H and 13C NMR, Table 1; HRESIMS m/z 293.1016 [M + H]+ (calcd for C₁₅H₁₇O₆, 293.1025), m/z 315.0835 [M + Na]+ (calcd for C₁₅H₁₆O₆Na, 315.0845).

Kirschsteinin B (3):

Orange solid. ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ +2 (c 0.1, CHCl₃); UV (CHCl₃) λ_max (log ε) 267 (3.5), 315 (3.5), 411 (3.4) nm; IR ν_max 3299, 2934, 1682, 1665, 1649, 1629, 1589, 1386, 1300, 1276, 1248, 1204, 1108, 1093 cm–1; 1H and 13C NMR, Table 2; HRESIMS m/z 509.1438 [M + H]+ (calcd for C₂₇H₂₅O₁₀, 509.1442).

Kirschsteinin C (4):

Orange solid. ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$ +3 (c 0.1.0, CHCl₃); UV (CHCl₃) λ_max (log ε) 264 (3.6), 316 (3.6), 349 (3.3), 429 (3.5) nm; IR ν_max 3309, 2934, 1670, 1628, 1590, 1383, 1302, 1271, 1249, 1204, 1110, 1093 cm–1; 1H and 13C NMR, Table 2; HRESIMS m/z 481.1128 [M + H]+ (calcd for C₂₅H₂₁O₁₀, 481.1129).

2’-(4-Bromobenzoyl)-kirschsteinin C.

Compound 4 (5.0 mg) was dissolved in dry THF (1 mL) and 1.2 mg of DMAP and 2.0 mg of 4-bromobenzoyl chloride were added. 2 mL of CH₂Cl₂ were added, and the reaction was sealed under a N₂ atmosphere and stirred at rt for 72 h.^{38, 39} The reaction product was partitioned in a separatory funnel using H₂O and CHCl₃, and the organic phase was purified by preparative TLC to obtain 2.5 mg of the brominated analog. HRESIMS m/z 663.0480 [M + H]+ (calcd for C₃₂H₂₄79BrO₁₁, 663.0502) and m/z 665.0464 [M + H]+ (calcd for C₃₂H₂₄81BrO₁₁, 665.0476) and 1H NMR data are in Figures S28 and 29, respectively.

Computational Methods.

The minimum energy structures were built with Spartan’10 software (Wavefunction, Inc.). The conformational analysis was performed using the Monte Carlo search protocol under the MMFF94 molecular mechanics force field. The resulting conformers were minimized using the DFT method at the B3LYP/6–311G+(2d,p) level of theory. Then, the time-dependent DFT (TDDFT) method at the B3LYP/6–31G+(d) level of theory was employed for ECD calculations. The calculated excitation energy (nm) and rotatory strength (R) in dipole velocity (Rvel) and dipole length (Rlen) forms were simulated into an ECD curve. All calculations were performed employing the Gaussian’09 program package (Gaussian, Inc.).

Cytotoxicity Assay.

Human melanoma cancer cells MDA-MB-435, human breast cancer cells MDA-MB-231 and human ovarian cancer cells OVCAR3 were purchased from the American Type Culture Collection. The cell line was propagated at 37 °C in 5% CO₂ in RPMI 1640 medium, supplemented with fetal bovine serum (10%), penicillin (100 units/mL), and streptomycin (100 μg/mL). Cells in log phase growth were harvested by trypsinization followed by two washes to remove all traces of enzyme. A total of 5,000 cells were seeded per well of a 96-well clear, flat-bottom plate (Microtest 96, Falcon) and incubated overnight (37 °C in 5% CO₂). Samples dissolved in DMSO were then diluted and added to the appropriate wells. The cells were incubated in the presence of test substance for 72 h at 37 °C and evaluated for viability with a commercial absorbance assay (CellTiter-Blue Cell Viability Assay, Promega Corp.) that measured viable cells. IC₅₀ values are expressed in μM relative to the solvent (DMSO) control.

Table 3.

Cytotoxic Activity of Compounds 1–9 Against a Panel of Human Cancer Cells Lines.

Compound	IC50a values (μM)
	MDA-MB-435	MDA-MB-231	OVCAR3
1	> 25	> 25	> 25
2	3.4	12	13
3	15	17	18
4	3.6	20	17
5	3.0	12	7.9
6	3.2	15	9.6
7	0.7	6.2	0.9
8	2.1	7.8	7.3
9	2.2	4.4	6.6
Positive controlb	0.4 nM	0.2 nM	5.1 nM

^aIC₅₀ values were determined as the concentration required to inhibit growth to 50% of control with a 72 h incubation at 37 °C.

^btaxol (paclitaxel)