Structure elucidation and absolute configuration of metabolites from the soil-derived fungus Dictyosporium digitatum using spectroscopic and computational methods

Abstract

Following the discovery of a new class of compounds that inhibit the mucosa-associated lymphoid tissue lymphoma translocation 1 (MALT1) protease in a prior study, further chemical investigation of the Dictyosporium digitatum fungus resulted in the identification of 16 additional metabolites, including 12 undescribed compounds (1-12). The constitution and relative configuration of these new molecules were established by comprehensive NMR and HRMS analyses. Their absolute configurations were determined by employing Mosher’s ester analysis and TDDFT ECD calculations. Two sesquiterpenes, dictyosporins A (1) and B (2), possess an undescribed eudesmen-type of structural scaffold. The ability of the isolated compounds to inhibit MALT1 proteolytic activity was evaluated, but none of them exhibited significant inhibition.

A comprehensive suite of contemporary spectroscopic and computational methodologies for small molecule structure determination were employed to define the structural motifs of twelve new fungal metabolites. These studies illustrate the broad utility and power of NMR experiments combined with computational methods to verify the constitution and configuration of new molecular scaffolds.

Graphical Abstract

The structure and absolute configuration of twelve metabolites from the fungus Dictyosporium digitatum were established by NMR and HRMS, in conjunction with Mosher’s ester analysis and quantum mechanical ECD calculations.

1. Introduction

Fungi are widely distributed from terrestrial environments to freshwater and marine habitats (Rédou et al., 2015), and it is estimated that about 5.1 million unique fungal species exist, but only approximately 100,000 of these have been taxonomically described (Blackwell, 2011). These eukaryotic microbes produce specialized metabolites involved in a variety of ecological functions such as quorum sensing, chemical defense, allelopathy, and maintenance of symbiotic interactions (Spiteller, 2015). A recent chemoinformatic study revealed that fungal metabolites often contain unique structural scaffolds and functional group arrays that are not represented in most commercial chemical libraries which contain both natural and semi-synthetic products (Gonzalez-Medina et al., 2017). Historically, most of the focus was on terrestrial saprophytic fungi as sources of new metabolites, but more recently attention has broadened to include fungi from marine habitats and endophytic strains living in plants. Therefore, fungi remain a vast and promising resource for bioactive natural products screening and discovery efforts.

Fungi in the genus Dictyosporium belong to the family Dictyosporiaceae (Tanaka et al., 2015) and they are commonly found on submerged, decaying wood (Goh et al., 1999). This genus of fungi has only been sparsely evaluated in prior chemical investigations (Prasher and Verma, 2015). Our previous chemical study on a soil isolate of Dictyosporum sp. (now identified as Dictyosporum digitatum) resulted in identification of the first non-quinone natural products which can inhibit the mucosa-associated lymphoid tissue lymphoma translocation 1 (MALT1) protease (Tran et al., 2019). Inspired by the previous findings, we searched for other minor components of the fungal extract and isolated 12 undescribed and four known compounds. Herein their isolation, structure elucidation, and biological evaluation are reported.

2. Results and Discussion

The Dictyosporum digitatum extract was chromatographed on a Diol MPLC column and subsequently separated by repeated C₁₈ HPLC to yield 12 new metabolites (1-12), along with four known compounds (13-16) (Fig. 1). Dictyosporin A (1) was obtained as a white powder. The (+)-HRESIMS spectrum displayed a sodium adduct ion [M + Na]⁺ at m/z 273.1464 corresponding to a molecular formula of C₁₅H₂₂O₃ with a hydrogen deficiency index of five. NMR data (Table 1) confirmed that compound 1 contained 15 carbons including one sp² quaternary carbon (δ_C 146.8), two sp² methines (δ_C 129.7 and 128.9), one sp² methylene (δ_C 111.2), two oxygenated sp³ quaternary carbons (δ_C 86.4 and 81.0), one sp³ quaternary carbon (δ_C 31.4), one oxygenated sp³ methine (δ_C 69.9), one sp³ methine (δ_C 45.4), one oxygenated sp³ methylene (δ_C 66.6), three sp³ methylenes (δ_C 40.0, 39.0, and 27.4), and two methyls (δ_C 28.1 and 19.4). COSY data identified four isolated proton-proton spin systems of –C¹–C²=C³–, –C⁵–C⁶–, –C⁸(OH)–C⁹–, and –C¹⁵(OH) (Fig. 2A). HMBC data revealed the presence of an iso-propenyl group C¹³=C¹¹(C¹²)–. HMBC correlations of H-1/C-10, C-5, and C-9; H-5/C-10 and C-9; H-9/C-10, C-1, and C-5; and H-14/C-1, C-5, C-9, and C-10 established the connections of C-1 to both C-5 and C-9 via C-10, and the location of the isolated methyl C-14 (δ_C 28.1) at C-10. The quaternary carbon C-4 was linked to C-3, C-5, and C-15 by HMBC correlations of H-15/C-3, C-4, and C-5; and H-3/C-4 and C-5. The connections of C-6, C-8, and the iso-propenyl group via C-7 were determined by HMBC correlations from H-6, H-8, 8-OH, H-12, and H-13 to C-7. C-4 and C-7 were connected via an oxygen atom to provide a fused tetrahydrofuran-type of ring which fulfilled the remaining unsaturation equivalent and molecular formula requirements. Therefore, the planar structure of dictyosporin A (1) was established as shown in Fig. 2A.

Fig. 1.

Structures of compounds 1-16.

Table 1.

¹H NMR (600 MHz) and ¹³C NMR (150 MHz) data for compounds 1-4 in DMSO-d₆.

Pos.	1		2		3		4
	δC	δH	δC	δH	δC	δH	δC	δH
1	40.0	Hα: 1.73, d (17.4)	53.9	Hα: 2.29, d (14.4)	31.0	Hα: 1.30 (m)	38.7	2.69, dt (4.4, 13.2)
		Hβ: 1.67, dd (5.4, 17.4)		Hβ: 1.99a		Hβ: 2.24, dt (5.4, 13.8)
2	128.9	5.75, dd (5.1, 10.2)	209.5		37.4	Hα: 2.57, dt (7.1, 13.8)	26.7	Hα 1.74 (m)
						Hβ: 2.04, ddd (1.8, 5.4, 13.8)		Hβ: 1.48, dq (4.7, 13.2)
3	129.7	5.68, d (10.2)	49.8	Hα: 2.44, d (17.4)	212.3		25.0	Hα: 2.31, m
				Hp: 2.55, d (17.4)				Hβ: 2.59, m
4	81.0		83.7		42.4	2.92, pent (6.6)	138.0
5	45.4	1.92, d (4.8)	50.9	2.03, d (4.2)	48.1	1.55, m	131.5	7.63, s
6	27.4	Hα: 2.45, d (12.0)	28.5	Hα: 2.59, d (12.0)	33.1	Hα: 1.83, m	144.5
		Hβ: 1.74, dd (4.8, 12.0)		Hβ: 1.99a		Hβ: 0.81, t (13.8)
7	86.4		86.4		72.2		140.6
8	69.9	3.61, dd (4.2, 4.6)	69.6	3.60, dd (4.3, 5.6)	31.4	Hα: 1.82, m	198.2
						Hβ: 1.50, dt (4.8, 13.8)
9	39.0	Hα: 1.12, d (14.8)	40.0	Hα: 1.26, d (14.4)	36.8	Hα: 1.29, m	47.0	Hα: 2.24a
		Hβ: 2.07, dd (5.3, 14.4)		Hβ: 1.53, dd (5.6, 14.4)		Hβ: 1.33, m		Hβ: 2.64, dd (5.4, 16.2)
10	31.4		35.3		33.0		31.6	2.24a
11	146.8		146.9		146.4		26.0	3.16, heptet (6.6)
12	19.4	1.71, s	19.4	1.72, s	18.5	1.66, s	20.9	1.13, d (6.6)
13	111.2	Hα: 4.84, s	111.1	Hα: 4.87, s	112.1	Hα: 4.92, s	21.4	1.17, d (6.6)
		Hβ: 4.76, s		Hβ: 4.78, s		Hβ: 4.85, s
14	28.1	1.15, s	28.8	1.17, s	26.9	1.19, s	12.5	0.81, d (6.6)
15	66.6	Hα: 3.23, dd (5.4, 10.2)	30.0	1.18, s	11.4	0.83, d (6.6)	167.7
		Hβ: 3.13, dd (4.8, 10.2)
7-OH						4.47, s
8-OH		4.52, d (3.8)		4.75, d (3.6)
15-OH		4.47, t (5.9)						12.76, br s

^aOverlapping signals

Fig. 2.

A: Key HMBC and COSY correlations of 1. B: Key NOESY correlations of 1. C: Mosher’s ester analysis of MTPA-1 (irregular Δδ^S-R signs in bold). D: Key HMBC and COSY correlations for dictyosporin C (3). E: Key NOESY correlations of 3. F: Octant rules applied for 3. G: Key HMBC and COSY correlations of dictyosporin D (4). H: Key NOESY correlations of 4. I: Experimental ECD spectrum of 4 and calculated ECD spectra of (1S, 10S)-4 and (1R, 10R)-4.

A series of 1D NOESY experiments were employed to define the relative configuration of 1. A 1,3-diaxial NOESY correlation between H-1α and H-5 allowed a half-chair conformation of the cyclohexene to be established. A chair conformation and cis-ring fusion of the A and B rings were assigned from 1,3-diaxial NOESY correlations of H-6α, 8-OH and H₃-14; and NOESY correlations from H-5 to H₃-14 and H-6α; and from H-1β to H-9β (Fig. 2B). The Mosher’s ester method was employed to establish the absolute configuration of C-8, and by inference, the rest of the stereogenic centers in 1 (Hoye et al., 2007; Cimmino et al., 2017). Treatment of dictyosporin A (1) with (R)-(−)-α-methoxy-α-(trifluoromethyl)phenylacetyl chloride [(R)-MTPA-Cl] or (S)-(+)-α-methoxy-α-(trifluoromethyl)phenylacetyl chloride [(S)-MTPA-Cl] afforded the bis (S)-MTPA ester (1a) and bis (R)-MTPA ester (1b), respectively. The proton chemical shift differences between 1a and 1b indicated that most protons with Δδ^S-R > 0 were located on the left side of the MTPA plane and the ones with Δδ^S-R < 0 were on the right side (Fig. 2C). It was noted that Δδ^SR signs of H-3, H-5, H-6α and H-15β were inconsistent with those of their neighboring protons. The inconsistent Δδ^SR values of H-5 and H-6α were likely due to their lying on the MTPA plane, which causes unreliable results (Ohtani et al., 1991). The protons H-3 and H-15β were too far from the MTPA group at C-8 and were affected by the MTPA group at C-15 which led to their irregular Δδ^SR values (Ohtani et al., 1991). Based on proton chemical shift analysis between 1a and 1b, the (S)-configuration was established at C-8, thus allowing the full absolute configuration of (4S, 5R, 7R, 8S, 10S) to be assigned for 1. This assignment was further supported by similar patterns between the experimental ECD spectrum of 1 and the calculated ECD of the (4S, 5R, 7R, 8S, 10S)-stereoisomer using time-dependent density functional theory (TDDFT) methods (Supplementary data Fig. S94A). Therefore, compound 1 was determined as (4S, 5R, 7R, 8S, 10S)-dictyosporin A.

Dictyosporin B (2) was isolated as a white powder and the molecular formula was deduced from HRESIMS data to be C₁₅H₂₂O₃, which indicated it was isomeric with 1. The ¹H NMR data of 2 (Table 1) differed notably from that of 1 by the absence of two olefinic protons at C-2 and C-3, and a hydroxymethyl group at C-15. In addition, two nonequivalent methylene protons at C-3 and an isolated methyl group at C-15 were observed. The ¹³C NMR data of 2 revealed one carbonyl carbon which showed HMBC correlations with H-1 and H-3, suggesting this carbonyl was located at C-2. All of the other structural components and connectivities established for 1 were also observed in 2. Detailed ROESY analysis indicated that 2 also had chair-chair conformations and cis-fusion of the A and B rings, similar to those of 1. Mosher’s ester analysis, supported by ECD calculations, (Supplementary data Fig. S94B, S95) gave the same conclusion of (4S, 5R, 7R, 8S, 10R) absolute stereochemistry for dictyosporin B (2). Dictyosporins A (1) and B (2) are the first sesquiterpenes in the eudesmanoid family, consisting of more than 1,000 natural product compounds (Wu et al., 2006), which possess a 4,7-epoxy-11(13)-eudesmen skeleton.

Compound 3 was isolated as a white powder and the molecular formula was established as C₁₅H₂₄O₂ based on (+)-HRESIMS measurements. Careful NMR analyses (Table 1) indicated 3 shared the same 11(13)-eudesmene skeleton with 1 and 2, but lacked the tetrahydrofuranoid ring. A ketone (δ_C 212.3) was assigned at C-3 based on HMBC correlations of H-2/C-3, H-4/C-3, and H-15/C-3 (Fig. 2D). The deshielded resonance of C-7 (δ_C 72.2) together with HMBC correlations from a singlet hydroxy proton (δ_H 4.47) to C-6, C-7, C-8, and C-11 allowed the hydroxy group to be located at C-7. The 1,3-diaxial NOESY correlations between H-2α, H-4, and H-14 and correlations between H-1β, H-6β, and H-8β revealed chair conformations for the hexanone and hexane rings in 3. A cis-fusion of the two rings was supported by NOESY correlations between H-5, H-14, H-9α, and H-13_b (Fig. 2E). With the relative configuration of 3 established, the ECD spectrum was recorded and it displayed a negative Cotton effect at 288 nm (Supplementary data Fig. S94C). This indicated that ring B resided in a negative CD region (upper right rear octant) on the basis of the ketone octant rule for carbonyl n-π* ECD bands (Fig. 2F) (Murphy, 1975). Thus, the absolute configuration of 3 was revealed as 4R, 5S, 7R, 10R, and this assignment was further supported by quantum chemical ECD calculations which closely matched the experimentally measured ECD spectrum (Fig. S94C, Supplementary data). Therefore, compound 3 was defined as 4R, 5S, 7R, 10R-dictyosporin C.

Compound 4 gave a protonated adduct ion [M + H]⁺ in the (+)-HRESIMS spectrum at m/z 249.1487 consistent with the molecular formula C₁₅H₂₀O₃. Examination of the COSY spectrum of 4 revealed two spin systems including a chain from H-3 to H-9 through contiguous protons H-2, H-1, and H10, and an iso-propyl group from H-12 to H-13 (Fig. 2G). HMBC correlations from a singlet olefinic proton H-5 to C-4, C-6, and C-7 allowed a diene system from C-4 to C-7 to be established. HMBC correlations of H-2/C-4; H-3/C-4 and C-5; H-1/C-5, C-6, and C-7; H-9/C-7 and C-8; and H-10/C-8 facilitated direct connections from C-3 to C-4, and C-1 to C-6; and a connection from C-9 to C7 via a ketone at C-8. The position of the iso-propyl group and a carboxylic acid at C-15 were revealed by HMBC correlations from both H-12 and H-13 to C-7; and from both H-3 and H-5 to C-15. A large ³J_1,2β coupling of 13.2 Hz indicated an axial orientation for H-1. The presence of NOESY correlations for H-1/H-2α, H-1/H-3α, and H-2β/H-14 together with the absence of NOESY correlations between H-1/H-14 supported all axial relationships between H-1, H-2β, H-3α, and H-14 (Fig. 2H). The calculated ECD of (1S, 10S)-4 was in good agreement with the experimentally measured ECD spectrum (Fig. 2I), which allowed assignment of 4 as (1S, 10S)-dictyosporin D.

Compound 5 was obtained as a white powder. The (+)-HRESIMS spectrum displayed a sodium adduct ion [M + Na]⁺ at m/z 263.0536 corresponding to a molecular formula of C₁₁H₁₂O₆. ¹H and ¹³C NMR signals characteristic of a tetra-substituted benzene ring were apparent (Table 2). The 7.8 Hz coupling between H-6 and H-7 together with HMBC correlations of H-6/C-4 and C-7a, and H-7/C-3a and C-5 helped establish the substitution pattern of the benzene ring. HMBC correlations of H-7/C-1; 4-OH/C-3a, C-4, and C-5; and 5-OH/C-4, C-5, and C-6 supported placement of an ester carbonyl and two hydroxy groups at C-7a, C4, and C-5, respectively. The position of a 1-methoxy-2-hydroxypropyl group was assigned based on COSY correlations from H-8 to H-9 and 8-OH, and HMBC correlations from H-9 and 3-OCH₃ to C-3. The molecular formula requirement of one additional double-bond equivalent and one more oxygen atom enabled establishment of a γ-lactone ring fused to the benzene ring in 5. The deshielded chemical shift of C-3 (δ_C 110.0) was consistent with a ketal group. The ECD spectrum of 5 showed three positive Cotton Effects (CEs) at 269, 238, and 204 nm and one negative CE at 225 nm which corresponded closely to calculated ECD spectra of the two stereoisomers (3R, 8S)-5 and (3R, 8R)-5 (Fig. S96, Supplementary data). Further DFT calculations for the ¹H and ¹³C chemical shifts of (3R, 8S)-5 and (3R, 8R)-5 indicated that the experimental NMR data of 5 was in agreement with that of the (3R, 8R)-isomer with a high level of confidence (Table S6, Supplementary data) (Smith and Goodman, 2010). Therefore, the structure of 5 was suggested as (3R, 8R)-dictyophthalide A.

Table 2.

¹H NMR (600 MHz) and ¹³C NMR (150 MHz) data for compounds 5-9 in DMSO-d₆.

Pos.	5		6		7		8		9
	δC	δH	δC	δH	δC	δH	δC	δH	δC	δH
1	168.0
2			153.5		158.2		157.0		157.9
3	110.0		102.0	6.27, s	108.1	6.39, s	101.8	6.98, s	103.1	7.05, s
3a	130.1		119.9		143.1		120.2		129.6
4	140.6		192.1		36.4	2.91, m	127.7		119.1
5	151.2		77.1	3.89, dd (4.6, 7.8)	26.7	Hα: 2.10, m	115.3	7.25, s	131.8	7.55, d (8.1)
						Hβ: 1.82, m
6	118.0	7.01, d (7.8)	70.7	3.95, m	36.2	2.43, m	154.6		110.4	6.73, d (8.1)
7	116.6	7.16, d (7.8)	30.4	Hα: 3.15, dd (4.8, 16.8)	184.2		103.6	7.21, s	150.5
				Hβ: 2.80, dd (7.8, 16.8)
7a	119.5		162.4		145.9		155.7		143.5
8	67.3	4.26, q (6.6)	13.1	2.27, s	13.7	2.34, s	13.8	2.44, s	13.8	2.47, s
9	17.1	1.12, d (6.6)			63.2	Hα: 3.64, m	192.4	10.07, s	189.7	9.83, s
						Hβ: 3.52, m
3-OCH3	50.8	2.97, s
4-OH		9.39, s
5-OH		10.51, s		5.50, d (4.6)
6-OH				5.44, d (4.2)				a
7-OH										a
8-OH		5.63, br s
9-OH						4.87, t (5.4)

^aNot observed

Compound 6 provided a sodium adduct ion [M + Na]⁺ at m/z 205.0472 in the HRESIMS spectrum that indicated it had a molecular formula of C₉H₁₀O₄. A proton spin system –C⁵(OH)–C⁶(OH)–C⁷– was deduced from COSY data, while the C-4 carbonyl carbon was assigned by HMBC correlations from H-5 and OH-5 to C-4. HMBC correlations of H-3/C-2, C-3a, and C-7a; and H-8/C-2 and C-3 facilitated assignment of a 2-methylfuran moiety in 6. HMBC correlations of H-6/C-7a; and H-7/C-3a and H-7a allowed the dihydroxy ketone system to be attached to the methylfuran at C-7a. No HMBC correlations from H-5 to C-3a or from H-3 to C-4 were observed in a HMBC experiment with ⁿJ_HC = 8.3 Hz; however, when the HMBC was optimized for ⁿJ_HC = 2.0 Hz a week correlation from H-8 to C-4 was observed, which supported a connection between C-3a and C-4. The relative configuration of 6 was established by the ³J_HH coupling constant values and selective 1D NOESY data. The large ³J_5,6 and ³J_6,7β values of 7.8 Hz and the small ³J_6,7α value of 4.8 Hz indicated an axial-axial-axial-equatorial relationship for H-5, H-6, H-7β, and H-7α. The 1,3-diaxial orientation of H-5 and H-7β was confirmed by NOESY correlations observed between these two protons. The ECD spectrum calculation for the (5R,6S)-isomer of 6 showed a similar pattern with the experimental ECD spectrum of 6 (Fig. S97, Supplementary data) suggesting the absolute configuration of 6 as (5R,6S)-dictyofuran A.

Compound 7 was obtained as a white powder and gave a sodium adduct ion [M + Na]⁺ in the (+)-HRESIMS spectrum at m/z 203.0679 consistent with a molecular formula of C₁₀H₁₂O₃. Detailed NMR analyses indicated that 7 possessed the same 2-methylfuran moiety as 6. COSY correlations revealed a spin system from H-6 to 9-OH through contiguous H-5, H-4, and H-9 protons. Three-bond HMBC correlations of H-9/C-3a; H-5/C-3a and C-7; and H-6/C-7a supported connections from C-4 to C-3a and from C-6 to C-7a via C-7 forming a planar structure as shown in Fig. 1. The experimental ECD spectrum of 7 matched the calculated ECD spectrum of the (4R)-isomer (Fig. S98, Supplementary data) allowing the assignment of (4R)-dictyofuran B (7).

Dictyofuran C (8) was obtained as white powder. The (−)-HRESIMS spectrum provided a [M - H]⁻ ion at m/z 175.0400 that established a molecular formula of C₁₀H₈O₃ with seven degrees of unsaturation for 8. HMBC analysis indicated that 8 also had a 2-methylfuran moiety and that this moiety was fused to a tetra-substituted benzene ring at C-3a and C-7a, based on HMBC correlations of H-5/C-3 and C-3a; and H-7/C-3a and C-7a (Table S9, Supplementary data). An aldehyde group was located at C-4 due to HMBC correlations from the aldehyde proton to C-3a, C-4, and C-5. A remaining hydroxy group was substituted at C-6 (δ_C 154.8) based on its deshielded chemical shift, which completed the structure assignment of dictyofuran C (8).

Dictyofuran D (9) was obtained as white powder and its molecular formula of C₁₀H₈O₃ was identical to that of 8. The ¹H NMR spectrum of 9 was also similar to that of 8, except for replacement of two singlet aromatic protons in 8 by two doublet aromatic protons with a J coupling value of 7.8 Hz in 9 (Table 2). Detailed 2D NMR analysis indicated that structure of 9 only differed from that of 8 by the location of a hydroxy group at C-7. This is the first time that dictyofuran D (9) has been isolated from a natural source, however it was previously reported as a synthetic intermediate in the synthesis of phosphodiesterase type 4 inhibitors, N-(3,5-dichloro-1-oxo-4-pyridyl)-6-difluoromethoxybenzo[4,5]furo[3,2-c]pyridine-9-carboxamide and its derivatives (Gharat et al., 2008).

Compound 10 was isolated as a white powder and (+)-HRESIMS analysis indicated it had a molecular formula of C₁₂H₁₆O₄. The ¹H and ¹³C NMR data for 10 were consistent with the molecular formula (Table S11, Supplementary data). The COSY experiment revealed the presence of a propenyl group with an E-configuration at Δ⁸ due to the J coupling value of 15 Hz for the olefinic protons. The HMBC spectrum showed correlations from two oxygenated methylene protons H-7 (δ_H 4.79 and 4.82) and a methyl singlet (δ_H 1.96) to an ester carbonyl C-1’ (δ_C 170.3) which established an acetate substituent at C-7. A cyclohexenone moiety was deduced based on COSY correlations of H-3/H-4/H-5 and key HMBC correlations of H-3/C-1 and C-2; and H-5/C-6. Long-range HMBC correlations from H-7 to C-2 and C-6; and from H-9 to C-2 supported the connection of C-7 to C-1 and the propenyl group to C-2. Substitution of a hydroxy group at C-4 (δ_C 64.3) fulfilled the molecular formula requirements of 10. Due to the limited supply of 10, Mosher’s ester analysis to determine the absolute configuration at C-4 was not performed. However, based on the comparison between experimental and calculated ECD spectra (Fig. S99, Supplementary data), the structure of (4R)-dictyosporone A (10) was proposed.

Compound 11 was obtained as a colorless oil. A sodium adduct ion [M + Na]⁺ at m/z 205.0834 in the (+)-HRESIMS spectrum indicated 11 had a molecular formula of C₁₀H₁₄O₃. Two proton spin systems –C²=C³– and –C⁶–C⁷(OH)–C⁸–C⁹ were assigned based on COSY correlations. HMBC correlations of H-2/C-1, C-4 and H-3/C-1, C-5 together with the requirements of one remaining degree of unsaturation and one additional oxygen atom resulted in the establishment of a disubstituted α-pyrone ring. HMBC correlations of H-7/C-4 and H-3/C-6 allowed the aliphatic chain to be attached to C-4. A relatively deshielded singlet methyl C-10 was assigned as the second substituent on the α-pyrone ring due to a HMBC correlation of H-10/C-5. The 7S absolute configuration for 11 was assigned as a result of Mosher’s ester analysis (Fig. S100, Supplementary data) and thus the compound was established as (7S)-xylariolide E (11).

Compound 12 was obtained as colorless oil and it had the same molecular formula of C₁₀H₁₄O₃ as xylariolide E (11). NMR data of 12 was largely analogous to those of 11 (Table 3). The major differences between the ¹H of 12 and 11 were that the former had a doublet methyl proton signal H-9 and a relative downfield resonance of the C-8 methine (δ_H/δ_C 3.55/64.8) leading to the assignment of a hydroxy group (δ_H 4.48) at C-8. HMBC correlations from 8-OH to C-7 and C-8 further supported its location. The Mosher’s ester analysis (Fig. S101A, Supplementary data) led to the establishment of (8S)-xylariolide F (12).

Table 3.

¹H NMR (600 MHz) and ¹³C NMR (150 MHz) data for compounds 11-13 in DMSO-d₆.

Position	11		12		13
	δC	δH	δCa	δH	δC	δH
1	161.7		161.6		161.5
2	111.8	6.11, d (9.6)	112.1	6.13, d (9.6)	112.7	6.20, d (9.6)
3	149.0	7.40, d (9.6)	147.8	7.41, d (9.6)	144.7	7.56, d (9.6)
4	113.1		115.1		119.2
5	159.2		158.3		157.6
6	36.4	Hα: 2.39, dd (4.2, 14.4)	25.0	Hα: 2.30, m	66.2	4.47, t (6.6)
		Hβ: 2.28, dd (8.4, 14.4)		Hβ: 2.36, m
7	71.3	3.42, m	38.8	1.45, q (7.2)	39.1	Hα: 1.41, m
						Hβ: 1.58, m
8	29.6	Hα: 1.42, m	64.8	3.55, m	18.3	Hα: 1.21, m
		Hβ: 1.33, m				Hβ: 1.32, m
9	10.1	0.89, dd (7.2, 7.8)	23.6	1.07, d (6.0)	13.9	0.87, t (7.2)
10	17.3	2.19, s	16.8	2.19, s	16.6	2.21, s
6-OH						5.16, br s
7-OH		4.56, d (5.4)
8-OH				4.48, d (4.8)

^aDetermined from HSQC and HMBC spectra

The structure of compound 13 was elucidated as xylariolide D based on NMR and MS analyses. This compound was previously isolated from the endophytic fungal strain Xylaria sp. NCY2; however, its absolute configuration was not determined (Hu et al., 2010). Compound 13 in this study was determined as (6S)-xylariolide D based on the Mosher’s ester analysis (Fig. S101B, Supplementary data). Other known compounds, 2,6-dihydroxy-4-methylacetophenone (14) (Yu et al., 1998), 2,5-dimethyl-7-hydroxychromone (15) (Hu et al., 2014), and emidin (16) (Uno et al., 2001) were assigned by spectroscopic data comparisons (NMR and MS) with the literature values.

3. Concluding Remarks

As part of our ongoing efforts to discover natural product inhibitors of the MALT1 paracaspase enzyme, all of the isolated compounds (1-16) were tested for their ability to inhibit the proteolytic activity of MALT1. These assay results indicated that none of the fungal metabolites were active at a high-test concentration of 100 μM. The sixteen compounds obtained in the current study include a series of variously functionalized sesquiterpenes and benzofurans, a chromenone, a hexanone, an acetophenone, a pyranone, and an anthraquinone. These metabolites possess diverse structural scaffolds and functional group arrays, yet none of the tested compounds were active against the MALT1 enzyme. Our prior work with D. digitatum resulted in the isolation and testing of eleven additional diverse metabolites, but only two compounds with 6H-oxepino[2,3-b]chromen-6-one skeletons were able to inhibit MALT1 activity (Tran et al., 2019). These results suggest that the MALT1 enzyme is not impacted by many classes of natural products and is only subject to inhibition by a relatively narrow subset of compounds. With the identification of various novel sesquiterpenoids and new polyketides, this study broadens our understanding of the metabolic potential of the under explored fungal genus Dictyosporium. These findings reveal that D. digitatum is a prolific source of low molecular weight metabolites with previously undescribed molecular architecture.

4. Experimental section

4.1. General experimental procedures

NMR spectra were acquired on a Bruker Avance III spectrometer equipped with a 3 mm cryogenically cooled probe operating at 600 MHz for ¹H and 150 MHz for ¹³C. ¹H and ¹³C spectra were referenced to the residual deuterated solvent peaks at δ_H 7.24 and δ_C 77.2 (CDCl₃) and δ_H 2.50 and δ_C 39.5 (DMSO-d₆). HMBC experiments were optimized for ⁿJ_CH = 8.3 Hz or 2.0 Hz. All 2D NMR experiments were acquired with nonuniform sampling (NUS) set to 50%, except for HSQC which had NUS set to 25%. HRESIMS data were acquired on an Agilent 6520 Accurate Mass Q-TOF instrument. HPLC purifications were performed using a Varian ProStar 218 solvent delivery module HPLC equipped with a Varian ProStar 325 UV–VIS detector, operating under Star 6.41 chromatography workstation software. All solvents used for extraction and chromatography were HPLC grade and the H₂O used was ultrapure water. Optical rotations ([α]_D) were measured on a PerkinElmer 241 polarimeter. UV spectra were obtained on an Agilent 8453 UV–VIS spectrophotometer. ECD spectra were measured on a AVIV 420SF Circular Dichroism spectrometer.

4.2. Fungal material and extraction

The fungal isolate was obtained from a soil sample collected in Herod, Illinois, USA (GPS coordinates: 37.53°, −88.97°). The isolate grew slowly on the surfaces of both malt extract and Czapek plates. Mycelium was collected from the isolate and subjected to homogenization in TE buffer (10 mM EDTA HCl, 0.1 mM EDTA, pH 8.0) with zirconium oxide beads in a Bullet Blender Storm (MidSci #BBY24M). The isolate was identified as Dictyosporium digitatum based on analysis of gene sequence data for tef-1 (translation elongation factor 1 alpha gene) (Genbank accession number MN496466), as well as the ribosomal internal transcribed spacer region (1TS1F-5.8S-ITS4) (GenBank accession number MH882417). BLASTn analysis revealed sequence similarity values of >98% for tef-1 and >99% for 1TS1F-5.8S-ITS4 between the fungal isolate and several samples in the NCBI database corresponding to Dictyosporium digitatum. Voucher specimens of the fungus have been retained at the University of Illinois Fungarium, Illinois Natural History Survey (specimen ILLS00121147) and the University of Oklahoma Citizen Science Soil Collection, Fungal Repository (specimen 61B2, ANM413). Additionally, the fungus was deposited in the CBS Collection under specimen number ILLS00121147. For chemical studies, the fungus was grown on Cheerios breakfast cereal supplemented with a 0.3% sucrose solution and 0.005% chloramphenicol in three large mycobags (Unicorn Bags, Plano, TX, USA). The fungus was grown for 6 weeks, whereupon it had achieved complete colonization of the solid substrate. The fungal biomass was extracted by soaking overnight in 8 L of ethyl acetate. The organic extract was subjected to partitioning three times against water (1:1, vol/vol). The ethyl acetate layer was retained, and the solvent removed by evaporation in vacuo, yielding 5.5 g of vibrant orange-red organic-soluble material.

4.3. Isolation and characterization data

The extract (5 g) was fractionated on a CombiFlash Diol-MPLC column (Teledyne ISCO, Licoln, NE, USA) eluted with a flow rate of 85 mL/min using a series of organic solvents: (1) 90% hexanes/10% CH₂Cl₂ in 10 min; (2) 95% CH₂Cl₂/5% EtOAc in 10 min; (3) 100% EtOAc in 10 min; (4) 83% EtOAc/17% MeOH in 10 min and (5) 100% MeOH in 20 min, to give 14 fractions A-N. Fraction D was purified on Phenominex Luna C₁₈ HPLC column, 5 μm, 21.4 × 250 mm, (Phenomenex, Torrance, CA, USA) at a flow rate of 9 mL/min with a contiguous series of CH₃CN/H₂O: a linear gradient from 15% CH₃CN/85% H₂O to 50% CH₃CN/50% H₂O in 20 min; isocratic elution at 50% CH₃CN/50% H₂O for 10 min and then a linear gradient from 50% CH₃CN/50% H₂O to 100% CH₃CN in 30 min to yield compounds 1 (1.2 mg, t_R = 21.3 min), 2 (1.0 mg, t_R = 25.5 min) and 3 (0.4 mg, t_R = 34.0 min). Fraction F was purified on a similar C₁₈ HPLC column (5 μm, 250 × 10 mm) at a flow rate of 4 mL/min with a linear gradient from 20% CH₃CN/80% H₂O to 80% CH₃CN/20% H₂O in 25 min and then from 80% CH₃CN/20% H₂O to 100% CH₃CN in 5 min to afford 13 (3.2 mg, t_R = 10.4 min), 14 (0.9 mg, 12.5 min), 8 (2.0 mg, t_R = 15.3 min) and 16 (1.8 mg, 25.1 min). Fraction G was chromatographed on a C₁₈ HPLC column (5 μm, 250 × 10 mm) with a similar elution program used for fraction F to yield 11 (1.5 mg, t_R = 8.4 min), 13 (1.5 mg, t_R = 10.4 min), 9 (0.4 mg, t_R = 13.3 min) and 8 (0.8 mg, t_R = 15.4 min).

Fraction I was separated by C₁₈ HPLC (5 μm, 250 × 10 mm) at a flow rate of 4 mL/min with a linear gradient from 5% CH₃CN/95% H₂O to 50% CH₃CN/50% H₂O in 20 min and then from 50% CH₃CN/50% H₂O to 100% CH₃CN in 10 min to obtain compounds 6 (0.3 mg, t_R = 11.5 min), 12 (0.2 mg, t_R = 13.4 min), 11 (1.5 mg, t_R = 14.5 min) and 4 (1.0 mg, t_R = 25.2 min). Fraction J was injected into a C₁₈ HPLC column (5 μm, 250 × 10 mm) with a similar program used for fraction I to give 7 (0.4 mg, t_R = 12.4 min), 5 (0.5 mg, t_R = 14.5 min) and 10 (0.5 mg, t_R = 15.6 min). Fraction K was separated by C₁₈ HPLC (5 μm, 250 × 10 mm) with a similar program used for fraction I to give 15 (0.8 mg, t_R = 18.4 min).

Dictyosporin A (1):

white powder; [α]²²_D +68 (c 0.10, MeOH); UV (MeOH) λ_max (log ε) 220 (3.27) nm; ECD (c 2,988 × 10⁻⁶ M, MeOH) λ_max (Δε) 216 (−0.22) and 202 (+2.03) nm; ¹H and ¹³C NMR data in Table 1. (+) HRESIMS m/z 273.1464 [M + Na]⁺ (calcd for C₁₅H₂₂O₃Na, 273.1461).

Dictyosporin B (2):

white powder; [α]²²_D +17 (c 0.07, MeOH); UV (MeOH) λ_max (log ε) 236 (3.94) and 280 (3.40) nm; ECD (c 2,074 × 10⁻⁶ M, MeOH) λ_max (Δ) 320 (+0.21), 244 (+0.68) and 214 (−0.17) nm; ¹H and ¹³C NMR data in Table 1. (+) HRESIMS m/z 273.1452 [M + Na]⁺ (calcd for C₁₅H₂₂O₃Na, 273.1461).

Dictyosporin C (3):

white powder; [α]²²_D −34 (c 0.03, MeOH); UV (MeOH) λ_max (log ε) 225 (3.57) and 284 (3.40) nm; ECD (c 2,119 × 10⁻⁶ M, MeOH) λ_max (Δε) 288 (−0.62) nm; ¹H and ¹³C NMR data, Table 1; (+) HRESIMS m/z 259.1670 [M + Na]⁺ (calcd for C₁₅H₂₄O₂Na, 259.1669).

Dictyosporin D (4):

white powder; [α]²²_D +148 (c 0.09, MeOH); UV (MeOH) λ_max (log ε) 302 (4.15) nm; ECD (c 1,512 × 10⁻⁶ M, MeOH) λ_max (Δε) 350 (−0.31), 303 (+5.06), 234 (−0.47) and 209 (+0.85) nm; ¹H and ¹³C NMR data in Table 2; (+) HRESIMS m/z 249.1487 [M + H]⁺ (calcd for C₁₅H₂₁O₃, 249.1485).

Dictyophthalide A (5):

white powder; [α]²²_D +9 (c 0.04, MeOH); UV (MeOH) λ_max (log ε) 205 (4.12), 222 (3.99), 268 (3.68) and 298 (3.58) nm; ECD (c 1,852 × 10⁻⁶ M, MeOH) λ_max (Δε) 269 (+0.39), 238 (+0.86), 225 (−1.84) and 204 (+2.04) nm; ¹H and ¹³C NMR data in Table 3; (+) HRESIMS m/z 263.0536 [M + Na]⁺ (calcd for C₁₁H₁₂O₆Na, 263.0526).

Dictyofuran A (6):

white powder; [α]²²_D +59 (c 0.02, MeOH); UV (MeOH) λ_max (log ε) 203 (4.04) and 270 (3.41) nm; ECD (c 1,099 × 10⁻⁶ M, MeOH) λ_max (Δε) 303 (+0.33), 282 (−0.29), 254 (+0.44) and 212 (+1.65) nm; ¹H and ¹³C NMR data in Table S7 (Supplymentary data); (+) HRESIMS m/z 205.0472 [M + Na]⁺ (calcd for C₉H₁₀O₄Na, 205.0471).

Dictyofuran B (7):

white powder; [α]²²_D +67 (c 0.03, MeOH); UV (MeOH) λ_max (log ε) 234 (3.74) and 283 (4.31) nm; ECD (c 926 × 10⁻⁶ M, MeOH) λ_max (Δε) 307 (−1.58), 278 (+3.01) and 235 (+3.47) nm; ¹H and ¹³C NMR data in Table S8 (Supplymentary data); (+) HRESIMS m/z 203.0677 [M + Na]⁺ (calcd for C₁₀H₁₂O₃Na, 203.0679).

Dictyofuran C (8):

white powder; UV (MeOH) λ_max (log ε) 229 (3.54) and 284 (3.26) nm; ¹H and ¹³C NMR data in Table S9 (Supplymentary data); (−) HRESIMS m/z 175.0397 [M - H]⁻ (calcd for C₁₀H₇O₃, 175.0401).

Dictyofuran D (9):

white powder; UV (MeOH) λ_max (log ε) 217 (4.46), 300 (4.29) and 347 (4.47) nm; ¹H and ¹³C NMR data in Table S10 (Supplymentary data); (+) HRESIMS m/z 177.0547 [M + H]⁺ (calcd for C₁₀H₉O₃, 177.0446).

Dictyosporone A (10):

white powder; [α]²²_D +2.0 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 279 (3.66); ECD (c 2,480 × 10⁻⁶ M, MeOH) λ_max (Δε) 322 (+0.04), 283 (−0.11) and 226 (−0.17) nm; ¹H and ¹³C NMR data in Table S11 (Supplymentary data); (+) HRESIMS m/z 247.0939 [M + Na]⁺ (calcd for C₁₂H₁₆O₄Na, 247.0941).

Xylariolide E (11):

colorless oil; [α]²²_D +25 (c 0.10, MeOH); UV (MeOH) λ_max (log ε) 222 (3.61) and 306 (3.81) nm; ¹H and ¹³C NMR data in Table S12 (Supplymentary data); (+) HRESIMS m/z 205.0834 [M + Na]⁺ (calcd for C₁₀H₁₄O₃Na, 205.0835).

Xylariolid F (12):

colorless oil; [α]²²_D +21 (c 0.01, MeOH); UV (MeOH) λ_max (log ε) 220 (3.29) and 304 (3.14) nm; ¹H and ¹³C NMR data in Table S13 (Supplymentary data); (+) HRESIMS m/z 205.0835 [M + Na]⁺ (calcd for C₁₀H₁₄O₃Na, 205.0835).

4.4. Preparation of the Mosher’s ester derivatives of compounds 1, 2, 11, 12 and 13

Compounds 1, 2, 11, 12, and 13 (0.1 mg each) were individually dissolved in deuterated pyridine (250 μL), transferred into separate NMR tubes and (R)-MTPA-Cl (2 μL) was added to each tube. The NMR tubes were then shaken carefully and kept in the dark at room temperature for 24 h to afford the (S)-MTPA ester derivatives. A similar procedure was performed using the (S)-MTPA-Cl to produce the (R)-MTPA ester derivatives of each compound. The ¹H NMR spectra of the R and S Mosher’s ester derivatives were individually recorded and COSY and HSQC experiments were used to confirm the ¹H NMR assignments.

4.4.1. 1-(S)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 6.00 (1H, d, J = 9.6 Hz, H-3), 5.81 (1H, m, H-2), 5.47 (1H, br s, H-8), 5.04 (1H, s, H_α−13), 5.02 (1H, s, H_β−13), 4.62 (1H, d, J = 10.8 Hz, H_α−15), 4.19 (1H, d, J = 10.8 Hz, H_β−15), 2.40 (1H, d, J = 12.0 Hz, H_α−6), 2.33 (1H, d, J = 16.2 Hz, H_β−9), 2.16 (1H, d, J = 12.0 Hz, H_β−6), 1.94 (3H, s, H-12), 1.93 (1H, m, H-5), 1.65 (2H, s, H-1), 1.33 (1H, d, J = 15.6 Hz, H_α−9), 0.80 (3H, s, H-14); (+) HRESIMS m/z 705.2263 [M + Na]⁺ (calcd for C₃₅H₃₆F₆O₇Na, 705.2258).

4.4.2. 1-(R)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 5.96 (1H, d, J = 9.0 Hz, H-3), 5.83 (1H, m, H-2), 5.54 (1H, br s, H-8), 4.84 (1H, s, H_α−13), 4.77 (1H, s, H_β−13), 4.61 (1H, d, J = 10.8 Hz, H_α−15), 4.24 (1H, d, J = 10.8 Hz, H_β−15), 2.42 (1H, m, H_β−9), 2.41 (1H, m, H_α−6), 2.09 (1H, d, J = 12.6 Hz, H_β−6), 2.02 (1H, br s, H-5), 1.91 (3H, s, H-12), 1.73 (2H, s, H-1), 1.42 (1H, d, J = 15.6 Hz, H_α−9), 1.11 (3H, s, H-14); (+) HRESIMS m/z 705.2250 [M + Na]⁺ (calcd for C₃₅H₃₆F₆O₇Na, 705.2258).

4.4.3. 2-(S)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 5.39 (1H, m, H-8), 5.02 (1H, s, H_α-13), 5.01 (1H, s, H_β-13), 2.99 (1H, d, J = 12.0 Hz, H_β-3), 2.51 (1H, d, J = 12.0 Hz, H_α-3), 2.42 (1H, d, J = 12.6 Hz, H_α-6), 2.25 (2H, s, H-1), 2.25 (1H, m, H_β-6), 1.99 (1H, m, H-5), 1.98 (1H, m, H_β-9), 1.92 (3H, s, H-12), 1.46 (1H, d, J = 16.8 Hz, H_α-9), 1.23 (3H, s, H-14), 0.85 (3H, s, H-15); (+) HRESIMS m/z 489.1863 [M + Na]⁺ (calcd for C₂₅H₂₉F₃O₅Na, 489.1860).

4.4.4. 2-(R)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 5.46 (1H, m, H-8), 4.84 (1H, s, H_α−13), 4.78 (1H, s, H_β−13), 3.01 (1H, d, J = 12.0 Hz, H_β−3), 2.53 (1H, d, J = 12.0 Hz, H_α−3), 2.44 (1H, d, J = 12.6 Hz, H_α−6), 2.32 (2H, s, H-1), 2.18 (1H, m, H_β−6), 2.06 (1H, m, H_β−9), 2.03 (1H, m, H-5), 1.90 (3H, s, H-12), 1.55 (1H, d, J = 16.8 Hz, H_α−9), 1.23 (3H, s, H-14), 1.15 (3H, s, H-15); (+) HRESIMS m/z 489.1868 [M + Na]⁺ (calcd for C₂₅H₂₉F₃O₅Na, 489.1860).

4.4.5. 11-(S)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.31 (1H, d, J = 9.0 Hz, H-3), 6.29 (1H, d, J = 9.6 Hz, H-2), 5.27 (1H, m, H-7), 2.69 (1H, dd, J = 8.4, 14.4 Hz, H_α−6), 2.55 (1H, dd, J = 4.8, 14.4 Hz, H_β−6), 2.09 (3H, s, H-10), 1.64 (2H, pent, J = 7.2 Hz, H-8), 0.84 (3H, t, J = 7.2 Hz, H-9); (+) HRESIMS m/z 399.1411 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.4.6. 11-(R)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.18 (1H, d, J = 9.0 Hz, H-3), 6.18 (1H, d, J = 9.0 Hz, H-2), 5.27 (1H, m, H-7), 2.58 (1H, dd, J = 9.0, 15.0 Hz, H_α−6), 2.47 (1H, dd, J = 4.2, 15.0 Hz, H_β−6), 1.88 (3H, s, H-10), 1.67 (2H, pent, J = 7.2 Hz, H-8), 0.93 (3H, t, J = 7.2 Hz, H-9); (+) HRESIMS m/z 399.1433 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.4.7. 12-(S)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.16 (1H, d, J = 9.6 Hz, H-3), 6.28 (1H, d, J = 9.0 Hz, H-2), 5.26 (1H, m, H-8), 2.30 (1H, m, H_α−6), 2.25 (1H, m, H_β−6), 2.04 (3H, s, H-10), 1.77 (1H, m, H_α−7), 1.67 (1H, m, H_β−7), 1.28 (3H, d, J = 6.0 Hz, H-9). (+) HRESIMS m/z 399.1417 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.4.8. 12-(R)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.05 (1H, d, J = 10.2 Hz, H-3), 6.25 (1H, d, J = 10.2 Hz, H-2), 5.26 (1H, m, H-8), 2.13 (1H, m, H_α−6), 2.12 (1H, m, H_β−6), 1.93 (3H, s, H-10), 1.68 (1H, m, H_α−7), 1.62 (1H, m, H_β−7), 1.35 (3H, d, J = 6.0 Hz, H-9). (+) HRESIMS m/z 399.1419 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.4.9. 13-(S)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.54 (1H, d, J = 9.6 Hz, H-3), 6.43 (1H, d, J = 9.6 Hz, H-2), 6.00 (1H, t, J = 7.2 Hz, H-6), 2.36 (3H, s, H-10), 1.85 (1H, m, H_α-7), 1.61 (1H, m, H_β-7), 1.23 (1H, m, H_α-8), 1.09 (1H, m, H_β-8), 0.79 (3H, t, J = 7.2 Hz, H-9); (+) HRESIMS m/z 399.1421 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.4.10. 13-(R)-MTPA ester

¹H NMR (pyridine-d₅, 600 MHz) δ_H 7.18 (1H, d, J = 9.6 Hz, H-3), 6.29 (1H, d, J = 9.6 Hz, H-2), 5.92 (1H, t, J = 7.2 Hz, H-6), 2.33 (3H, s, H-10), 1.87 (1H, m, H_α−7), 1.63 (1H, m, H_β−7), 1.31 (1H, m, H_α−8), 1.26 (1H, m, H_β−8), 0.85 (3H, t, J = 7.2 Hz, H-9); (+) HRESIMS m/z 399.1419 [M + H]⁺ (calcd for C₂₀H₂₂F₃O₅, 399.1414).

4.5. Computational Details

4.5.1. TDDFT ECD calculations for 1–7 and 10

ECD calculations of were performed at 298 K using Maestro and Gaussian 09. Molecular mechanics calculations were performed using Macromodel interfaced to the Maestro program (Version 2015.3, Schrödinger) (Schrödinger Release 2015–3: MacroModel). All conformational searches used the OPLS_2005 force field. Conformers having internal relative energies within 3 kcal/mol were subjected to geometry optimization on Gaussian 09 (Frisch et al., 2013) at the DFT level with the B3LYP functional and the 6–31G(d,p) basis set. Optimized conformers were then subjected to TDDFT calculations in MeOH on Gaussian 09 using the B3LYP functional, and the 6–31+G(d,p) basis set for 1, 3-7 and 10, and the 6–31G(d,p) basis set for 2. All calculations were performed in MeOH solvent. For each conformer, all of the resultant rotational strengths were converted into Gaussian distributions and summed to obtain the final calculated ECD spectrum based on the Boltzmann distribution of each conformer. ECD spectra were generated using the SpecDis program (Bruhn et al., 2015).

4.5.2. DFT NMR calculations for (3R, 8R)-5 and (3R, 8S)-5

Molecular mechanics and quantum chemical calculations for (3R, 8R)-5 and (3R, 8S)-5 were performed in DMSO with the same procedure as above. Single point calculations in DMSO with the B3LYP functional and the 6–311+G(d,p) basis set were then employed to provide the shielding constants of carbon and proton nuclei. Exchangeable protons were excluded in the calculation (Kwan and Liu, 2015). Meanwhile, the same procedure was applied on TMS (tetramethylsilane). Final ¹H and ¹³C chemical shifts were obtained as the result of the Boltzmann weighted average. The theoretical chemical shifts were calculated according to the equation

$${\delta }_{\mathit{\text{calc}}}^x={\sigma }_{TMS}-{\sigma }^x$$

where ${\delta }_{\text{calc}}^{\text{x}}$ is the calculated shift for nucleus × (in ppm); σ^x is the shielding constant for nucleus x; σ_TMS is the shielding constant of TMS computed at the same level of theory. The mean absolute error (MAE) was defined as ${\sum }_{i=1}^n\left|{\delta }_{\mathit{\text{calc}}.}-{\delta }_{exp.}\right|/n$ after removing systematic errors during the chemical shift calculations. DP4 parameters were calculated using the online applet at http://www-jmg.ch.cam.ac.uk/tools/nmr/.

4.6. MALT1 assay

Enzymatic expression and the MALT1 assay procedure were similar to those described previously (Tran et al., 2019).

Highlights.

• Twelve undescribed compounds were identified from the soil-derived fungus Dictyosporium digitatum.

• Absolute configurations of the isolated compounds were established using Mosher’s ester analysis and quantum chemical ECD calculations.

• Two sesquiterpenes, dictyosporins A (1) and B (2), have a unique carbon-oxygen structural framework.