Metabolic Profiling for the Discovery of Structurally Diverse Gibberellins and Their Precursors from the Endophytic Fungus Fusarium sp. NJ-F5

Abstract

Gibberellins (GAs) are well-known tetracyclic diterpenoid phytohormones since the 1950s. In this work, eight skeletally diverse GAs (1–8) including four new compounds (1–4), and three known ent-kaurene diterpenoids (9–11), were isolated from the endophytic fungus Fusarium sp. NJ-F5 by integrating mass spectrometry (MS)- and nuclear magnetic resonance (NMR)-based metabolic profiling. Their planar structures and stereochemistry were determined by extensive spectroscopic analyses including MS, NMR, as well as electronic circular dichroism and their calculations, together with single-crystal X-ray diffraction studies. As far as we know, this is a rare report of naturally occurring GAs and their detailed spectroscopic data including MS and NMR in recent decades. Compound 1, as a new member of GAs family, showed an obvious promoting effect on the seedling’s growth ofArabidopsis thaliana.

Introduction

Gibberellins (gibberellic acids; GAs), an important group of natural plant growth hormones, have attracted considerable attention owing to their diterpenoid structures and agricultural uses.¹ Since the structural identification of the first member GA₃ in the fungus Fusarium fujikuroi in 1959,² about 136 GAs have been discovered from fungi, bacteria, and plants.^3,4 The members of GAs are typically characterized by the presence of a 6/5/6/5-fused tetracyclic diterpenoid framework, which is biosynthetically derived from the precursor, ent-kaurene with a 6/6/6/5 carbon skeleton.⁵ The variations in oxidation, reduction, cyclization, and substituents, especially the number and position of carboxylic acid groups, greatly contribute to the structural diversity of GAs.⁵ GAs, especially the well-known GA₁, GA₃, GA₄, and GA₇, can regulate plant growth and development processes in agriculture, such as seed germination, shoot elongation, flower initiation, and fruit formation.⁶⁻⁹ They also showed antitumor activity^10,11 and were recently found to have antimicrobial properties.¹²

Even though the intriguing structures and biological activities of GAs, the isolation and identification of new naturally occurring GAs were scarcely mentioned in recent decades, and the NMR and related spectroscopic data of known GAs were not fully recorded in databases and references. During our continuous search for novel bioactive secondary metabolites from endophytic fungi,¹³⁻¹⁵Fusarium sp. NJ-F5 was found to produce comparatively more secondary metabolites than its counterparts during our preliminary screening of 50 selected fungal endophytes by HPLC–DAD profiling. This indicated that NJ-F5 was a prolific producer of compounds in our laboratory conditions. To maximally understand the structural classification and novelty of secondary metabolites,¹⁶ mass spectrometry (MS)- and nuclear magnetic resonance (NMR)-based metabolomics were explored for the crude extract of NJ-F5, suggesting the presence of diverse GAs. As a result, the search for its secondary metabolites afforded eight structurally diverse GAs (1–8) including four new ones (1–4), and three biosynthetically related ent-kaurene diterpenoids (9–11). Herein, the discovery, structures, and biological activities of these isolates are reported.

Materials and Methods

General Experimental Procedures

Optical rotations were recorded on a GYROMAT-HP polarimeter (Kernchen, Germany), and electronic circular dichroism (ECD) measurements were performed using a Chirascan spectropolarimeter (Applied Photophysics, UK). Infrared (IR) spectra were obtained using a Nicolet iN10 spectrometer (Thermo Fisher). Nuclear magnetic resonance (NMR) spectra were recorded on a JNM-ECZ600R/S1 spectrometer operating at 600 (¹H) and 150 (¹³C) MHz with TMS as an internal standard (JEOL, Japan). The high-resolution mass spectrometry (HRMS) experiment was carried out on an Orbitrap Exploris 480 spectrometer (Thermo Fisher, Germany) equipped with an ESI source. Liquid chromatography (LC)-HRMS data for the crude extracts were obtained from an Agilent 6530 TOF spectrometer equipped with an Agilent Porshell 120 EC-C18 column (4.5 mm × 50 mm, 2.7 μm), eluting with an 18.0 min acetonitrile/water gradient (10–90% acetonitrile, 13.0 min; 100–100% acetonitrile, 5.0 min). Semipreparative high-performance liquid chromatography (HPLC) was performed on Agilent 1260 HPLC with an Agilent Zorbax SB-C18 column (5 μm, 150 mm × 9.4 mm). Medium-pressure liquid chromatography (MPLC) was achieved by a CombiFlash RF+ UV instrument (Teledyne Isco, USA). Silica gel (200–300 mesh; Qingdao Haiyang Chemical Company, China) and octadecyl carbon chain-bonded silica (ODS, 40–60 μm, 100 Å) were employed for column chromatography (CC). Thin-layer chromatography (TLC) was carried out with glass-precoated silica gel GF₂₅₄ plates (Qingdao Haiyang Chemical Company, China). Compounds were visualized under UV light and by spraying with H₂SO₄-EtOH (1:9, v/v) followed by heating. All solvents were of analytical grade.

Fungus and Fermentation

The endophytic fungus Fusarium sp. NJ-F5 was isolated from the roots of Mahonia fortunei collected from Nanjing, Jiangsu Province, People’s Republic of China. The isolation of endophytes was done following our previously reported method.¹⁷ The fungal strain was identified by internal transcribed spacer (ITS) sequencing (GenBank Accession no. OP159849; Figure S123). The fungus was assigned the strain designation NJ-F5 and was deposited at the School of Pharmacy, Qingdao University, Qingdao, People’s Republic of China. For large-scale fermentation, Fusarium sp. NJ-F5 was cultured on potato dextrose agar (PDA) media at 28 °C for 2 weeks. Then, agar plugs (around 0.5 cm × 0.5 cm) were inoculated in 75 flasks (1.0 L) each containing 160 g of rice, 240 mL of water, and 0.48 g peptone. The cultures were further incubated in an incubator at 28 °C for 60 days.

Extraction and Isolation

The fermented cultures were extracted three times with ethyl acetate (EtOAc). The organic solvent was combined and evaporated to dryness under reduced pressure to give a crude extract (23.3 g). The extract was fractionated by CC on silica gel eluting with a gradient of petroleum ether–EtOAc (from 100:0 to 0:100; v/v). Based on the TLC and HPLC analyses, all of the eluents were combined to give 12 fractions (fractions 1–12). The fraction 6 (0.9 g) was purified by semipreparative HPLC (MeOH–H₂O, 85/15, 2.0 mL/min) to afford compounds 2 (15.8 mg, t_R = 12.0 min), 3 (22.2 mg, t_R = 16.5 min), 9 (22.0 mg, t_R = 10.8 min), and 10 (11.3 mg, t_R = 7.2 min). Fraction 9 (3.6 g) was chromatographed over an ODS column with a gradient of MeOH–H₂O (from 10:90 to 100:0; v/v) to yield five subfractions (9-1–9-5). Fraction 9-3 (572 mg) was further separated by MPLC eluting with a gradient of MeOH–H₂O (from 10:90 to 100:0; v/v), followed by HPLC (MeOH–H₂O, 45/55, 2.0 mL/min) to provide compound 4 (17.9 mg, t_R = 24.2 min). Fraction 9-5 (806 mg) was directly purified by HPLC (MeOH–H₂O, 68/32, 2.0 mL/min) to afford compounds 5 (29.5 mg, t_R = 18.6 min), 7 (18.4 mg, t_R = 22.2 min), 8 (13.3 mg, t_R = 36.6 min), and 11 (116.0 mg, t_R = 25.0 min). Fraction 11 (5.6 g) was partly purified by HPLC (MeOH–H₂O, 75/25, 2.0 mL/min) to afford compounds 1 (118.9 mg, t_R = 8.7 min), and 6 (33.7 mg, t_R = 9.2 min).

3β,16α-Dihydroxy-9,15-cyclo-gibberellin A₉ (1)

Pale yellow oil; [α]_D²²: −11.2 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max: 196 nm; IR v_max: 3420, 2938, 2869, 1752, 1712, 1458, 1362, 1343, 1274, 1195, 1053, 1026, 990, 911 cm^–1; CD (MeOH) λ (Δε): 216 (−2.33) nm; ¹H NMR and ¹³C NMR, see Tables 1 and 3; Negative ESI-HRMS m/z: 347.1482 [M – H]⁻ (calcd. for C₁₉H₂₃O₆, 347.1489, Δ −1.9905).

Table 1. ¹H NMR Data of Compounds 1–5 (in DMSO-d₆)^a.

no.	1	2	3	4	5
1α	1.80, mc	2.04, mc	2.02, dt (13.2, 3.6)	2.01–2.10, m	1.28, mc
1β	1.62, m	0.97, td (13.2, 4.2)	0.95, td (13.2, 3.6)		1.39, mc
2α	1.51, m	1.91, m	1.95, m	1.52, mb	1.93, td (13.8, 1.8)
2β	1.68, m	1.35, m	1.33, m	1.62, mb,c	1.48, mc
3α	3.50, d (4.2)	1.69, dt (13.8, 3.6)	1.68, mc	3.78, t (3.0)	3.81, br s
3β		1.21, td (13.8, 3.6)	1.17, td (13.2, 3.6)
5	2.58, d (9.0)	2.93, s	3.32, s	3.05, m	2.17, d (12.6)
6	2.53, d (9.0)			3.03, d (6.6)	3.18, d (12.0)
7		5.57, s	4.68, s
9		1.64, mc	1.74, mc		1.23, mc
11α	1.86, td (12.0, 3.0)	1.80, mc	1.79, mc	2.22, mc	1.32–1.37, mc
11β	1.78, mc	1.62, m	1.54, m	1.88, m
12α	1.32, mc	1.63, mc	1.66, mc	1.29, mc	1.34, mb,c
12β	1.49, mc	1.41, m	1.41, m	1.57, m	1.82, mb
13	1.56, m	2.60, m	2.64, m	1.79, m	2.52, mc
14α	1.37, mc	1.54–1.59, mc	1.29, mc	2.24, mb,c	1.59, d (11.4)
14β	2.28, dd (11.4, 5.4)		1.77, mc	1.42, d (10.2)b	1.46, mc
15α	1.33, mc	2.70, dt (16.8, 3.0)	2.27, dt (17.4, 2.4)	1.65, d (13.2)	2.02, d (15.6)
15β		2.04, mc	2.20, d (17.4)	1.32, d (13.2)	2.07, d (15.6)
17a	1.22, s	4.72, s	4.75, s	1.25, s	4.80, br s
17b		4.81, s	4.84, s		4.86, br s
18	0.96, s	1.14, s	1.09, s	1.04, s	1.08, s
20		1.24, s	1.25, s		0.73, s
21		3.39, s	3.43, s

^aRecorded at 600 MHz.

^bProtons may be interchangeable for the corresponding methylenes.

^cOverlapped.

Table 3. ¹³C NMR Data of Compounds 1–11 (in DMSO-d₆)^a.

no.	1	2	3	4	5	6	7	8	9	10	11
1	23.6, CH2	40.2, CH2	40.5, CH2	18.5, CH2	33.5, CH2	30.3, CH2	36.3, CH2	31.9, CH2	40.4, CH2	36.8, CH2	36.2, CH2
2	28.0, CH2	17.6, CH2	17.7, CH2	29.0, CH2	26.9, CH2	29.2, CH2	21.3, CH2	17.3, CH2	18.8, CH2	16.9, CH2	16.6, CH2
3	68.9, CH	35.7, CH2	36.2, CH2	70.5, CH	69.3, CH	69.3, CH	37.6, CH2	27.6, CH2	39.4, CH2	27.8, CH2	22.5, CH2
4	53.2, C	43.4, C	43.4, C	51.2, C	47.8, C	48.8, C	44.3, C	44.3, C	43.4, C	40.6, C	47.9, C
5	48.1, CH	57.1, CH	59.1, CH	50.6, CH	48.8, CH	49.1, CH	55.6, CH	54.9, CH	50.6, CH	50.4, CH	46.9, CH
6	45.9, CH	171.9, C	172.8, C	50.9, CH	50.3, CH	50.2, CH	50.6, CH	172.5, C	70.3, CH	83.6, CH	84.2, CH
7	173.5, C	105.2, CH	109.6, CH	176.9, C	176.3, C	176.3, C	176.3, C	204.3, CH	80.4, CH	69.9, CH	69.8, CH
8	38.6, C	52.1, C	50.6, C	58.1, C	49.0, C	49.5, C	49.4, C	58.3, C	48.0, C	45.3, C	45.0, C
9	35.4, C	56.5, CH	52.8, CH	134.9, C	56.8, CH	55.8, CH	55.6, CH	46.0, CH	47.9, CH	58.3, CH	54.9, CH
10	94.2, C	40.1, C	39.1, C	126.5, C	43.6, C	56.0, C	55.9, C	41.3, C	40.2, C	34.6, C	33.5, C
11	13.8, CH2	19.2, CH2	18.9, CH2	19.6, CH2	16.5, CH2	18.2, CH2	18.2, CH2	18.8, CH2	17.7, CH2	62.7, CH	16.7, CH2
12	22.4, CH2	32.5, CH2	32.5, CH2	26.2, CH2	31.8, CH2	31.3, CH2	31.2, CH2	31.9, CH2	33.2, CH2	44.8, CH2	32.4, CH2
13	41.8, CH	42.4, CH	42.9, CH	47.5, CH	40.1, CH	39.9, CH	39.3, CH	43.4, CH	43.1, CH	37.5, CH	37.3, CH
14	32.1, CH2	31.4, CH2	36.5, CH2	44.6, CH2	39.0, CH2	36.7, CH2	36.7, CH2	31.6, CH2	37.9, CH2	33.8, CH2	33.5, CH2
15	33.2, CH	42.0, CH2	45.1, CH2	49.4, CH2	46.0, CH2	46.0, CH2	45.8, CH2	42.3, CH2	45.6, CH2	42.4, CH2	41.8, CH2
16	76.2, C	154.0, C	153.8, C	78.0, C	156.2, C	156.6, C	156.5, C	151.5, C	154.8, C	159.6, C	159.8, C
17	23.3, CH3	103.4, CH2	103.8, CH2	25.1, CH3	105.7, CH2	105.8, CH2	105.7, CH2	104.8, CH2	103.3, CH2	106.4, CH2	106.7, CH2
18	14.1, CH3	28.9, CH3	29.6, CH3	21.6, CH3	24.1, CH3	23.8, CH3	29.1, CH3	28.5, CH3	32.1, CH3	25.3, CH3	66.8, CH2
19	177.8, C	176.5, C	176.6, C	176.4, C	179.0, C	176.5, C	176.6, C	176.6, C	179.4, C	181.9, C	181.1, C
20		19.1, CH3	19.3, CH3		14.7, CH3	175.4, C	175.1, C	22.4, CH3	16.7, CH3	22.9, CH3	20.7, CH3
21		56.3, CH3	56.2, CH3

^aRecorded at 150 MHz; ¹³C multiplicities were determined by HSQC experiment.

7α-Methoxy-6,7-lactone-gibberellin A₁₂ (2)

Pale yellow oil; [α]_D²²: −103.8 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 194 nm; IR v_max 2978, 2932, 2866, 1719, 1692, 1468, 1442, 1369, 1307, 1201, 1171, 1007, 951, 881, 802 cm^–1; CD (MeOH) λ (Δε): 204 (−1.49), 210 (−0.60), 222 (−0.99) nm; ¹H NMR and ¹³C NMR, see Tables 1 and 3; Negative ESI-HRMS m/z: 347.1856 [M – H – CH₂]⁻ (calcd. for C₂₀H₂₇O₅, 347.1853, Δ 0.8857 ppm); Positive ESI-HRMS m/z: 385.1959 [M + Na]⁺ (calcd. for C₂₁H₃₀O₅Na, 385.1985, Δ −6.9663 ppm).

7β-Methoxy-6,7-lactone-gibberellin A₁₂ (3)

Pale yellow oil; [α]_D²²: −30.4 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 194 nm; IR v_max 3347, 1633, 1471, 1175 cm^–1; CD (MeOH) λ (Δε): 204 (−1.18), 210 (−0.56), 222 (−0.97) nm; ¹H NMR and ¹³C NMR, see Tables 1 and 3; Negative ESI-HRMS m/z: 347.1856 [M – H – CH₂]⁻ (calcd. for C₂₀H₂₇O₅, 347.1853, Δ 0.9736 ppm); Positive ESI-HRMS m/z: 385.1965 [M + Na]⁺ (calcd. for C₂₁H₃₀O₅Na, 385.1985, Δ – 5.3025 ppm).

16α-Hydroxy-9-ene-gibberellin A₁₄ (4)

Pale yellow oil; [α]_D²²: −22.2 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 198 nm; IR v_max 3420, 2932, 2853, 1745, 1171, 1101 cm^–1; CD (MeOH) λ (Δε): 201 (−0.69), 203 (−1.83), 231 (+0.67), 268 (−0.075) nm; ¹H NMR and ¹³C NMR, see Tables 1 and 3; Negative ESI-HRMS m/z: 349.1638 [M – H]⁻ (calcd. for C₁₉H₂₅O₆, 349.1646, Δ −2.0508 ppm).

Gibberellin A₁₄ (5)

Pale yellow oil; [α]_D²²: −36.0 (c 0.05 MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 200 nm; IR v_max 2934, 2867, 1713, 1454, 1226, 1138, 1056, 1028, 879, 827 cm^–1; CD (MeOH) λ (Δε): 201 (+1.28), 207 (−0.87), 227 (+0.19) nm; IR v_max 2934.2, 1713.2, 1226.1, 1028.2, 879.0 cm-1; ¹H NMR and ¹³C NMR, see Tables 1 and 3; Negative ESI-HRMS m/z: 347.1847 [M – H]⁻ (calcd. for C₂₀H₂₇O₅, 347.1853, Δ −1.7513 ppm).

Gibberellin A₁₃ (6)

Pale yellow oil; [α]_D²²: −25.4 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 204 nm; IR v_max 2939, 2873, 1709, 1264, 1224, 1191, 1030, 970 cm^–1; CD (MeOH) λ (Δε): 201 (+0.24), 202 (−6.13), 207 (−2.06), 253 (−0.40) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Negative ESI-HRMS m/z: 377.1585 [M – H]⁻ (calcd. for C₂₀H₂₅O₇, 377.1595, Δ −2.6561 ppm).

Table 2. ¹H NMR Data of Compounds 6–11 (in DMSO-d₆)^a.

no.	6	7	8	9	10	11
1α	2.01, mc	2.28, d (13.8)	1.80, mc	1.74, mc	1.61, m	1.45, mc
1β	1.32, mc	0.97, mc	1.56, mc	0.77, t (13.2)	1.00, m	0.90, m
2α	1.53, mb	1.42, mc	1.34, mc	1.70, mc	1.42–1.50, m	1.41–1.49, m
2β	2.07, mb,c	1.82, mc	1.52, mc	1.31, mc
3α	3.67, t (3.0)	1.03, mb,c	1.34, mb,c	1.98, d (13.2)	1.92, m	1.66, m
3β		1.92, mb,c	2.07, mb,c	0.97, t (13.2)	1.30, m	1.45, mc
5	2.34, d (12.6)	1.94, d (12.0)	2.82, mc	1.68, mc	1.85, d (6.6)	2.04, dd (4.2, 6.6)
6	3.70, d (12.6)	3.72, d (12.0)		4.14, d (10.8)	4.52, t (6.6)	4.53, m
7			9.61, s	3.28, s	4.27, d (6.6)	4.05, m
9	1.43, mc	1.45, mc	2.54, mc	1.34, mc	0.95, d (3.0)	1.05, m
11α	1.45, mb,c	1.49, mb,c	1.87, mc	1.48, mb,c	4.03, dd (3.0, 6.0)	1.28, mc
11β	1.26, mb,c	1.27, mb,c	1.72, mc	1.39, mb,c
12α	1.83, m	1.85, mc	1.80, mc	1.56, mb	2.06, dd (9.0, 14.4)	2.07, m
12β	1.22, mc	1.28, mc	1.56, mc	1.37, mb,c	1.56, dd (6.0, 14.4)	1.30, mc
13	2.49, mc	2.52, mc	2.78, mc	2.60, s	2.59, m	2.48, m
14α	1.65, dd (1.8, 12.0)b	1.65, d (11.4)b	1.80, mb,c	1.77, mc	2.22, dd (3.0, 10.2)	1.41, mc
14β	1.34, dd (4.8, 12.0)b	1.36, mb,c	1.56, mb,c	1.15, dd (3.6, 10.2)	1.48, mc	1.55, dd (11.4, 5.4)
15α	2.01, mc	2.01, d (16.8)	2.10, mb,c	2.18, d (17.4)	2.53, mc	2.52, mc
15β	2.08, mc	2.08, d (16.8)	2.42, mb,c	2.13, d (17.4)	1.76, d (15.0)	1.72, d (15.6)
17a	4.77, s	4.77, s	4.79, s	4.73, s	4.78, br s	4.81, s
17b	4.85, s	4.86, s	4.90, s	4.78, s	4.92, br s	4.95, s
18	1.11, s	1.09, s	1.34, s	1.35, s	1.19, s	3.45, mc
20			0.71, s	0.89, s	1.03, s	0.80, d (3.6)

^aRecorded at 600 MHz.

^bProtons may be interchangeable for the corresponding methylenes.

^cOverlapped.

Gibberellin A₂₅ (7)

Pale yellow oil; [α]_D²²: −27.0 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 198 nm; IR v_max 2934, 2873, 1710, 1202, 879 cm^-1; CD (MeOH) λ (Δε): 202 (−1.45), 230 (+0.11) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Negative ESI-HRMS m/z: 361.1641 [M – H]⁻ (calcd. for C₂₀H₂₅O₆, 361.1646, Δ −1.2222 ppm).

Fujenoic Diacid (8)

Pale yellow oil; [α]_D²²: −12.9 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 198 nm; IR v_max 2945, 1781, 1722, 1462, 1224, 1010, 928, 898 cm^-1; CD (MeOH) λ (Δε): 202 (−1.05), 226 (+0.003) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Negative ESI-HRMS m/z: 347.1860 [M – H]⁻ (calcd. for C₂₀H₂₇O₅, 347.1853, Δ +2.0284 ppm).

6β,7β-Dihydroxykaurenoic Acid (9)

Pale yellow oil; [α]_D²²: −66.2 (c 0.05, MeOH); LC-UV [(Acetonitrile in H₂O/0.1% FA)] λ_max 194 nm; IR v_max 3420, 2932, 2853, 1659, 1129, 1063, 878, 827 cm^–1; CD (MeOH) λ (Δε): 204 (−1.60), 247 (+0.003) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Negative ESI-HRMS m/z: 333.2059 [M – H]⁻ (calcd. for C₂₀H₂₉O₄, 333.2060, Δ −0.4015 ppm).

7β,11α-Dihydroxy-kaurenolide (10)

Pale yellow oil; [α]_D²²: −39.6 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 192 nm; IR v_max 3415, 2927, 1659, 1053, 1033, 865, 829 cm^–1; CD (MeOH) λ (Δε): 201 (+1.50), 233 (−0.10) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Negative ESI-HRMS m/z: 377.1959 [M – H + HCOOH]⁻ (calcd. for C₂₁H₂₉O₆, 377.1959, Δ +0.0723 ppm).

7β,18-Dihydroxy-kaurenolide (11)

Colorless crystals; [α]_D²²: −49.0 (c 0.05, MeOH); LC-UV [(MeOH in H₂O/0.1% FA)] λ_max 205 nm; IR v_max 3411, 2935, 2866, 1758, 1653, 1079, 1056, 878, 825 cm^-1; CD (MeOH) λ (Δε): 201 (−0.97), 205 (+0.63), 215 (−0.44), 225 (+0.62) nm; ¹H NMR and ¹³C NMR, see Tables 2 and 3; Positive ESI-HRMS m/z: 355.1878 [M + Na]⁺ (calcd. for C₂₀H₂₈O₄Na, 355.1880, Δ −0.4937), 687.3856 [2M + Na]⁺ (calcd. for C₄₀H₅₆O₈Na, 687.3867, Δ −1.6289).

X-ray Diffraction Crystallographic Analysis

Single crystals for X-ray analysis were obtained by recrystallization from DMSO-acetonitrile-cyclohexane (0.1:1:4) for compound 1 and acetonitrile-cyclohexane (3:7) for compound 2. An XtaLAB Synergy R, DW system, HyPix diffractometer with Cu Kα radiation was used to perform X-ray crystallography. The crystal structures were solved with SHELXT¹⁸ and refined with SHELXL¹⁹ with the Olex2 software.²⁰ The Flack parameters were 0.05(3) and −0.13(18) for compounds 1 and 2, respectively. Molecular graphics were carried out using the software Ortep-3.²¹ Crystallographic data of compounds 1 and 2 were deposited in the Cambridge Crystallographic Data Center (CCDC) with deposition numbers 2195948 and 2195949, respectively. Copies of the detailed data can be obtained, free of charge, on application to CCDC, 12 Union Road, Cambridge, CB2 1EZ, United Kingdom (fax: + 44-(0)1223-336033 or email: data_request@ccdc.cam.ac.uk).

MS/MS Molecular Networking

The MS/MS data of the fungal crude extract was obtained by a Q Exactive Orbitrap mass spectrometer (Thermo Fisher) equipped with a HESI source. The spectrometer was connected with a Thermo UltiMate 3000 liquid chromatography (LC) system equipped with an Agilent Poroshell 120 EC-C18 (4.6 mm × 150 mm, 4 μm) column at 28 °C. The sample (0.1 mg/mL, 5 μL) was automatically injected into the LC system, eluting with H₂O (A)-MeCN (B) from 10–100% B (0–15 min) to 100-100% B (15–30 min). The compounds were analyzed in full scan/ddms2 method in both positive (+) and negative (−) ion modes. For the HESI source, the following parameters were set for experiments: spray voltage (+) 3800 V, spray voltage (−) 3200 V, capillary temperature (+ and −) 325 °C, sheath gas 40 arb, aux gas 20 arb, and probe heater temperature 350 °C. The full MS survey scan was carried out for a maximum inject time of 100 ms in the range of 70–1000 Da, followed by fragmentation of top 5 ions to obtain MS/MS spectra. Using MSConvert software, the MS/MS data were converted to the acceptable “.mzXML” format²² and then uploaded to the global natural products social molecular networking (GNPS) platform through the WinSCP client. The GNPS classical molecular networking workflow (METABOLOMICS–SNETS-V2) was employed to produce the MS/MS molecular network,²³ which was further visualized using Cytoscape software. The following parameters were used for molecular network generation: precursor mass tolerance m/z 0.02 Da, MS/MS fragment ion tolerance m/z 0.02 Da, minimum cosine score 0.6, minimum matched fragment ions 4, minimum cluster size 2, and network TopK 10.

Computational Details

ECD calculations for compounds 2 and 4 were performed as described previously.^13,15,17 The geometrical optimization and vibrational evaluation were achieved using density functional theory (DFT) calculations with the TZVP basis set and the B3PW91 exchange-correlation functional in Gaussian 09 program (Figures S125 and S126). The singlet electronic excitation energies and corresponding rotational strengths were obtained by employing the time-dependent DFT (TDDFT) calculations at the mPW1PW91/6-311G(d) level of theory. The solvation PCM model with methanol and the half-band of 0.3 eV were applied.

Antimicrobial Activity

The in vitro antimicrobial activity of compounds 1–11 was evaluated according to our previously described approaches.^13,17,24 For antimicrobial assay, two Gram-positive bacteria Staphylococcus aureus (ATCC 6538), and Bacillus subtilis (ATCC 9372), two Gram-negative bacteria Escherichia coli (ATCC 25922), and Pseudomonas aeruginosa (ATCC 27853), and two fungi Candida albicans CG3 and C. albicans CG4 were used in this study. Streptomycin and fluconazole were employed as the positive controls in the antibacterial and antifungal assays, respectively.

Cytotoxic Evaluation

Human prostate cancer cell line PC3, lung carcinoma cancer cell lines H460, and A549, and breast cancer MCF-7 cells were cultured in RPMI-1640 media supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. Liver cancer HepG2 cells were cultured in DMEM media with 10% FBS and antibiotics. All of these cells were incubated in a 5% CO₂ incubator at 37 °C. The above-mentioned PC3, H460, A549, MCF-7, and HepG2 cells were further seeded into 96-well plates at an initial density of 1 × 10⁴ cells/well and incubated at 37 °C overnight. The cells were then treated with compounds for 24 h. Cell viability was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay performed in 4 replicates for each concentration. The half-inhibitory concentration (IC₅₀) value was determined.^17,24 DMSO served as the blank control, and adriamycin was used as the positive control.

Anti-inflammatory Assay

PC3 Cells were seeded into six-well plates and treated with compounds at a concentration of 10 μM for 12 h. Total RNA was obtained by an RNA iso plus kit (TaKaRa, Japan). Complementary DNA was synthesized through reverse transcription using ReverTra Ace qPCR RT Kit (TOYOBO, Japan). Quantitative PCR analysis of cDNA was performed with SYBR Green reaction master mix (TOYOBO) on a QuantStudio 5 system (Thermo Fisher, Waltham, MA). Target mRNA levels were normalized to the level obtained for β-actin. Changes in the transcript level were calculated using 2^–ΔΔCt method. The forward and reverse primers for human IL-1β are GTCGGAGATTCGTAGCTGGAT and CTCGCCAGTGAAATGATGGCT, respectively. The forward and reverse primers for human IL-6 are TCAATGAGGAGACTTGCCTGG and GGCTGGCATTTGTGGTTGG, respectively.

Seedling Growth Assay

The seed germination regulation was evaluated using our previously established method.^25,26 The seeds of Arabidopsis thaliana were sterilized by 5% sodium hypochlorite for 4 min and then washed with distilled water three times. Compounds were dissolved in DMSO and added to 1/2 MS media (25 mL) containing 1.0% (w/v) agar to obtain plates with compounds (64 μg/mL). Seeds were then distributed on the top of each Petri dish, which was further incubated at 24 °C under light for 8 h and darkness for 6 h. The lengths of roots were measured after 7 days. The percentage of elongation growth of root average lengths of treatments was calculated in comparison with that of blank control. The 1/2 MS media with and without DMSO were used as the blank controls, and the well-known GA₃ was applied as the positive control.

Results and Discussion

Discovery of GAs by MS- and NMR-Based Metabolomics

During our preliminary screening of 50 selected fungal endophytes, Fusarium sp. NJ-F5 was cultured in rice media, followed by extraction with ethyl acetate (EtOAc). HPLC–DAD analysis of the EtOAc crude extract showed a series of peaks with similar UV absorption, indicating that NJ-F5 strain was a prolific producer of fungal secondary metabolites with a similar substructure (Figure S1). To tentatively identify their structural types, LC-HRMS was employed to propose the possible molecular formulas, which were further analyzed with SciFinder database using the search mode of molecular formula (Figures S2–S5). It is clearly shown that most of candidate structures with a high number of references belong to GA family (Figures S2–S5). To reveal the chemical diversity of GAs in NJ-F5, the crude extract was further analyzed by LC-HRMS/MS. The MS/MS data was analyzed by molecular networking (MN) via the GNPS Web platform to cluster similar spectra as molecular families (Figures 1 and S8–S10). In the molecular network, a number of spectral nodes grouped together and were estimated to be the above-mentioned GAs on the basis of their precursor ions (Figures 1 and S2–S5). Unexpectedly, many GA structures had no spectroscopic data recorded in databases and literature, such as NMR data, deserving further chemical investigation.

Figure 1.

Metabolic profiling for the endophytic fungus Fusarium sp. NJ-F5. (A) MS/MS molecular networking for the EtOAc crude extract: Entire molecular network (a) and the identification of GAs and analogues within the selected molecular family (b). (B) ¹³C NMR spectra of crude extract, selected fractions, and GA₃, indicating the presence of carboxyl and/or lactone groups (a), a double bond between C-16 and C-17 (b, c), oxygenated methines (d), and saturated carbons (e).

Based on the above analysis, Fusarium sp. NJ-F5 was subjected to large-scale fermentation. The EtOAc crude extract was separated by CC on silica gel to afford 12 fractions (Fr. 1–12). Further isolation and purification were guided by ¹H and ¹³C NMR-based metabolic profiling. The known GA₃ was purchased and used as the standard for NMR comparison. The ¹H NMR spectra of Fr. 6, 9, and 11 showed some characteristic proton signals of GAs, such as the olefinic proton at C-17 (Figure S6). The ¹³C NMR spectra of Fr. 6, 9, and 11 had high similarities with that of GA₃ (Figures 1 and S7). The signals of carboxyl and/or lactone groups, the double bond between C-16 and C-17, and some saturated carbons were observed. Hence, we attempted to purify compounds from Fr. 6, 9, and 11, which contained a number of GAs. As expected, further chromatographic isolation and HPLC purification yielded 11 GAs and their biosynthetically related analogues with diverse skeletons (1–11, Figure 2), such as compound 11 from the nearly pure Fr. 11.

Figure 2.

Chemical structures of compounds 1–11.

Structural Elucidation of Compounds

Compound 1 was isolated as a pale yellow oil with the molecular formula of C₁₉H₂₄O₆ as derived from ESI-HRMS at m/z 347.1482 [M – H]⁻ (Figures 2 and S19). The ¹H NMR spectrum of 1 (Figure S11) indicated the presence of two singlet methyl signals (δ_H 0.96, and 1.22), and one oxygenated methine proton (δ_H 3.50, d, J = 4.2 Hz) (Table 1). Its ¹³C NMR spectrum (Figure S12) showed 19 carbons including three oxygenated carbons (δ_C 68.9, 76.2, and 94.2) and two carbonyl groups (δ_C 173.5, and 177.8) (Table 3).

The planar structure of compound 1 was further constructed by the ¹H-¹H COSY and HMBC spectra (Figures 3, S14, and S15). Based on the analysis of ¹H-¹H COSY spectrum, the connections from C-1 to C-3, from C-11 to C-14, and between C-5 and C-6 were determined. Key HMBC correlations from H₃-18 to C-3, C-4, and C-5, and from H₂-1 to C-5, and C-10 indicated a six-membered ring (A ring) in 1. Another six-membered carbon ring (C ring) was verified on the basis of the HMBC correlations (Figure 3). The above-mentioned A and C rings were fused to the same ring B, which was supported by the key HMBC correlations of H-1α/C-9, and H-6/C-8. By analyzing the HMBC correlations from H₃-17 to C-13, C-15, and C-16, and from H-14α to C-8, and C-15, the structure of five-membered ring D was constructed as shown. An unusual E ring, a cyclopropane, was fused to the C and D rings, which was verified by the key HMBC correlations of H-15/C-10 and H-11β/C-15. A carbonyl or lactone group was located at C-4 from the HMBC correlation of H₃-18/C-19, which contributed to the formation of a five-membered lactone (ring F) based on the evidence of the significant downfield shift of C-10 (δ_C 94.2).^27,28 The remaining carbonyl group was assigned as a free carboxyl group attached to C-6, supported by the HMBC correlations from H-5 and H-6 to C-7. Finally, two hydroxyl groups were located at C-3 and C-16, respectively, which was consistent with the molecular formula of 1. Accordingly, the planar structure of compound 1 is determined as depicted and it belongs to the class of GAs with a classic 6/5/6/5-fused tetracyclic core skeleton.

Figure 3.

Key ¹H-¹H COSY and HMBC correlations of new compounds 1–4.

The relative configuration of 1 was assigned by detailed analysis of its NOESY spectrum and coupling constants (Figures 4 and S16). Key NOESY correlations of H-5/H-1β, H-5/H-15, and H-5/H₃-18 indicated that they are on the same side. The NOESY correlations of H-3 with H₃-18 and the small coupling constant (J = 4.2 Hz) of H-3 with H₂-2 were indicative of H-3 in α-orientation. The assignment of the relative configurations of H-6, H-13, and H₃-17 was achieved using the correlations of H-6/H-14α, H₃-17/H-11β, H₃-17/H-12β, and H₃-17/H-13 in the NOESY spectrum.

Figure 4.

Key NOESY correlations of new compounds 1–4.

Considering the structural complexity of 1, a single-crystal X-ray diffraction technique was employed. After numerous attempts, crystals for X-ray diffraction were obtained from a mixed solvent of 10 μL DMSO, 100 μL acetonitrile, and 400 μL cyclohexane. Single-crystal diffraction analysis with Cu Kα radiation finally verified the planar structure, and stereochemistry (Figure 5). The absolute configuration of 1 was finally established as 3S,4S,5R,6S,8S,9S,10S,13R,15R,16S. Therefore, the structure of compound 1 was unambiguously confirmed and it was named 3β,16α-dihydroxy-9,15-cyclo-gibberellin A₉.

Figure 5.

(a) ORTEP diagram of the crystal structure of compound 1. One molecule of DMSO was co-crystallized with compound 1 by forming a hydrogen bond at COOH-7. (b) ORTEP diagram of the crystal structure of compound 2. Two molecules of compound 2 were crystallized together and only one of them was shown herein. (c) Experimental and calculated ECD spectra of compound 2.

Compound 2 was isolated as a pale yellow oil. Its molecular formula was determined to be C₂₁H₃₀O₅ on the basis of ESI-HRMS data (Figure S29). In the ¹H NMR spectrum of 2 (Figure S21), three tertiary methyls (δ_H 1.14, 1.24, and 3.39) and three olefinic or oxygenated singlet protons (δ_H 4.72, 4.81, and 5.57) were present (Table 1). Its ¹³C NMR spectrum (Figure S22) indicated the presence of two carboxyl or lactone groups (δ_C 171.9, and 176.5), one double bond [δ_C 154.0, and 103.4 (or 105.2)], and an acetal or ketal group (δ_C 105.2 or 103.4) (Table 3). These NMR data revealed similar structural features to those of compound 1 and indicated that compound 2 was also a GA derivative.

The ¹H-¹H COSY and HMBC spectra were further employed to assemble the planar structure of compound 2 (Figures 3, S24, and S25). Interpretation of ¹H-¹H COSY spectrum identified the connections from C-1 to C-3, and from C-11 to C-14. HMBC correlations of H₃-18/C-3, C-4, and C-5, and of H₃-20/C-1, C-5, and C-10 constructed the A ring, while the C and D rings were verified by the HMBC correlations from H-11α, H₂-14, and H₂-15 to both C-8 and C-9, together with the HMBC correlations of H₂-17 with C-13, C-15, and C-16. The B ring, as a seven-membered lactone, was constructed by analysis of HMBC correlations from H₃-20 to C-9, from H-5 to C-6, and from H-7 to C-6 and C-8, together with the chemical shifts of C-6 (δ_C 171.9) and C-7 (δ_H 5.57; δ_C 105.2). A methoxy group was further attached to the ketal carbon C-7 as suggested by HMBC correlation of H₃-21/C-7. The remaining one degree of unsaturation came from a carboxyl group located at C-4, which was supported by the HMBC correlation of H₃-18/C-19 and the requirement of molecular formula. Accordingly, the planar structure of compound 2 was unambiguously established as depicted.

The key NOESY correlations of H-5/H₃-18, H-5/H-7, H-5/H-9, H₃-20/H-14, H₃-20/H-12α, and H-13/H-17b determined the relative configuration of 2 (Figures 4 and S26). Fortunately, the crystals of compound 2 were obtained from an organic solution containing acetonitrile-cyclohexane (3:7). A single-crystal diffraction analysis using Cu Kα radiation confirmed the gross structure and stereochemistry of 2 (Figure 5). ECD calculations were further carried out at the mPW1PW91/6-311G(d)//B3PW91/TZVP level of theory to confirm the absolute configuration of 2. As expected, the calculated ECD spectrum of 2 was in good accordance with the experimental curve (Figures 5 and S125). Therefore, the absolute configuration of 2 was identified as 4R,5S,7R,8R,9S,10S,13R, and compound 2 was named 7α-methoxy-6,7-lactone-gibberellin A₁₂.

Compound 3 was also obtained as a pale yellow oil and named 7β-methoxy-6,7-lactone-gibberellin A₁₂. ESI-HRMS spectrum of 3 showed that it had the same molecular formula as compound 2 (Figure S39). Further analysis of its ¹H, ¹³C, and HSQC NMR spectra (Figures S31, S32, and S33) revealed highly close structural features to those of 2, except for the significantly upfield shift of H-7 (δ_H 5.57 in 2; δ_H 4.68 in 3) and the downfield shift of C-7 (δ_C 105.2 in 2; δ_C 109.6 in 3) (Tables 1, and 3). The above analysis suggested that the configuration of C-7 chiral center should be different in compounds 2 and 3, and indicated an S-configuration of C-7 in 3 compared to the R-configuration in 2, which was strongly supported by comprehensive analysis of the NOESY data of 3 (Figures 4 and S36), especially the NOESY correlations of H-7/H-15α, and H-7/h-14β. Based on the above-mentioned analysis, and considering the same biosynthetic origin, and the similar NMR data and ECD curves of compounds 2 and 3 (Figures S30 and S40), the absolute configuration of 3 was determined as 4R,5S,7S,8R,9S,10S,13R.

For compound 4, a pale yellow oil, ESI-HRMS data determined its molecular formula as C₁₉H₂₆O₆ (Figure S49). The ¹H NMR spectrum of 4 (Figure S41) showed the presence of two singlet methyl groups (δ_H 1.04, and 1.25) and one oxygenated methine (δ_H 3.78, t, J = 3.0 Hz) (Table 1), and its ¹³C NMR spectrum (Figure S42) revealed 19 carbon signals including one double bond (δ_C 126.5, and 134.9), two oxygenated carbons (δ_C 70.5, and 78.0), and two carbonyl groups (δ_C 176.4, and 176.9) (Table 3). The 1D NMR data with the aid of HSQC spectrum (Figure S43) confirmed the presence of two methyls, six methylenes, four methines (one oxygenated), three quaternary carbons (one oxygenated), one double bond, as well as two carboxyl and/or lactone groups, indicating that compound 4 was also a GA derivative.

Similarly, the A and C rings of 4 were determined based on the ¹H-¹H COSY correlations of H₂-1/H₂-2/H-3 and H₂-11/H₂-12/H-13/H₂-14, together with HMBC correlations from H₃-18 to C-3, C-4, and C-5, from H₂-1 and H-5 to C-10, from H₂-14 to C-8 and C-9, and from H₂-11 to C-9 (Figures 3, S44, and S45). Further analysis of HMBC correlations of H₂-15 with C-8 and C-9, and of H₃-17 with C-13, C-15, and C-16, and the chemical shift of C-16 (δ_C 78.0) allowed the determination of D ring with a hydroxyl group at C-16. HMBC correlations of H-5 to C-6, and C-9, and of H-6 to C-8, and C-15 established the connection of C-5/C-6/C-8 and the double bond between C-9 and C-10, generating the classic GA 6/5/6/5 tetracyclic system. Two carboxyl groups were found to be located at C-4 and C-6, respectively, as established by HMBC correlations of H-5 and H-6/C-7 and H₃-18/C-19 and their chemical shifts. One remaining hydroxyl group was attached to C-3, which was in accordance with the chemical shift of C-3 and molecular formula. Using the same strategy as above, the relative configuration of compound 4 was determined by the key NOESY correlations of H-5/H₃-18, H-3/H-2α and H-2β, H-6/H-14α and H-14β, and H-15β/H-11β and H₃-17 (Figures 4 and S46). From a biosynthetic standpoint, the absolute configuration of 4 was determined as 3S,4S,5S,6S,8S,13R,16R. Following the same ECD calculation method as 2, the calculated ECD curve of 4 matched well with the experimental data, further confirming the absolute configuration of 4 (Figures S50 and S126). Compound 4 was named 16α-hydroxy-9-ene-gibberellin A₁₄.

Compounds 1–4 are previously undescribed members of GA family (Figure 2). Among them, compound 1 features a complex 6/5/6/5/3/5 heptacyclic ring framework with a striking C-9/C-15 bond. As far as we know, until now, there are only six GA-like natural products with a 9,15-cyclo skeleton since the discovery of the first member, 1β-hydroxy-9,15-cyclogibberellin A₉ (Figure S124).²⁸⁻³³ However, all of the reported structures of 9,15-cyclogibberellins were recorded 20 years ago, and were determined by GC-MS, IR, ¹H NMR, ¹³C NMR, or synthetic investigation, and no detailed spectroscopic data were found.²⁸⁻³³ Compounds 2 and 3 are stereoisomers with a 6/7/6/5-fused tetracyclic skeleton, while compound 4 possesses a classic 6/5/6/5-fused tetracyclic system.

The skeleton of compound 4 is shared by three co-isolated known GAs, gibberellin A₁₄ (5), gibberellin A₁₃ (6), and gibberellin A₂₅ (7) (Figure 2). Furthermore, a known GA derivative, fujenoic diacid (8) with an opening B ring, together with three known ent-kaurene diterpenoids, 6β,7β-dihydroxykaurenoic acid (9), 7β,11α-dihydroxy-kaurenolide (10), and 7β,18-dihydroxy-kaurenolide (11) (Figure 2), were also isolated. Compounds 9–11 with a 6/6/6/5-fused framework should be the biosynthetic precursors of co-occurring GAs (1–8).⁵ Based on their structures, a possible biosynthetic pathway is proposed in Scheme 1. The general ent-kaurene diterpenoid biosynthesis provided 9 via ent-kaurenoic acid.⁵ Compound 9 was oxidized to afford compounds 8, 10, and 11. Further cyclization and methylation of 8 yielded 2 and 3. The formation of a C6 carbocation in 9 might be key for GA skeleton construction, guiding the oxidative extrusion of C-7 through a semipinacol rearrangement to provide GA₁₂-aldehyde.⁴ Further oxidation of GA₁₂-aldehyde afforded compounds 4–7. Finally, the oxidation and rearrangement of 4 could give compound 1.

Scheme 1. Proposed Biosynthetic Pathway of Compounds 1–11.

Notably, as far as we know, all known compounds 5–11 had no detailed spectroscopic data in the literature, especially NMR spectra. Herein, we have provided the MS, 1D and 2D NMR, IR, UV, optical rotation, and ECD data of all compounds 5–11 (Tables 1–3 and Figures S51–S120), facilitating future chemical and biological investigation. In addition, their key ¹H-¹H COSY, HMBC, and NOESY correlations were also provided in Figures S121 and S122.

Biological Activity

In this work, diverse biological assays were selected to evaluate the function of isolated GAs and analogues, enhancing our understanding of their pharmaceutical potential. Compounds 1–11 were tested for their antimicrobial activity against two Gram-positive bacteria, two Gram-negative bacteria, and two Candida strains. Only compound 1 exhibited weak antibacterial efficiency against S. aureus with an MIC value of 64 μg/mL. They were also evaluated for their cytotoxicity against five cancer cell lines, and none of them showed significant cytotoxic effects at the concentration of 20 μM. The inflammatory assay was also studied based on the relative changes of IL 1β and IL6 of PC3 cells in response to compounds. Unfortunately, all of the compounds 1–11 exhibited no anti-inflammatory activity at the concentration of 10 μM. These results indicated that the obtained GAs members have limited antimicrobial, cytotoxic, and anti-inflammatory potential.

Concerning the well-known regulation role of GAs in plant growth, especially GA₁, GA₃, GA₄, and GA₇,⁷⁻⁹ all of the compounds 1–11 were further evaluated for their effects on the germination of A. thaliana. Like the agriculturally used GA₃, compound 1 as the new member of GAs group, significantly promoted the seedling’s growth of A. thaliana, and stimulated the root elongation (Figure 6).

Figure 6.

Root elongation of A. thaliana on Petri dishes containing compounds.

Conclusions

GAs, especially GA₃, are well-known tetracyclic diterpenoid phytohormones since the 1950s. In this work, 11 skeletally diverse GAs and their biosynthetic precursors (1–11) including four new ones (1–4) were isolated and purified from the endophytic fungus Fusarium sp. NJ-F5 by integrating MS- and NMR-based metabolic profiling. The discovery of new GAs derivatives and the detailed spectroscopic data including MS, NMR, IR, UV, and CD of compounds 1–11, facilitate future chemical and biological investigation on this kind of secondary metabolites. Diverse biological assays revealed that compound 1, as a new member of the GAs family, showed comparable physiological effects on the seedling’s growth as GA₃. These results implied that plant endophytic fungus Fusarium sp. NJ-F5 and its secondary metabolites might have beneficial effects on the growth of host plant.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.2c06454.

• HPLC and LC-HRMS chromatograms; MS, IR, UV, ECD, and NMR spectra of compounds 1–11; and key ¹H-¹H COSY, HMBC, and NOESY correlations of compounds 5–11 (PDF)

• Crystallographic data and X-ray structures of compound 1 (CIF)

• Crystallographic data and X-ray structures of compound 2 (CIF)

The authors declare no competing financial interest.