Search for Cell Motility and Angiogenesis Inhibitors with Potential Anticancer Activity: Beauvericin and Other Constituents of Two Endophytic Strains of Fusarium oxysporum

Abstract

Wound healing assay-guided fractionation of an EtOAc extract of the fungal strain, Fusarium oxysporum EPH2R_AA endophytic in Ephedra fasciculata afforded beauvericin (1), (−)-oxysporidinone (2), and two new N-methyl-2-pyridones, (−)-4,6′-anhydrooxysporidinone (3) and (−)-6-deoxyoxysporidinone (4). Beauvericin (1) inhibited migration of the metastatic prostate cancer (PC-3M) and breast cancer (MDA-MB-231) cells and showed anti-angiogenic activity in HUVEC-2 cells at sublethal concentrations. Cytotoxicity-guided fractionation of an EtOAc extract of F. oxysporum strain CECIS occurring in Cylindropuntia echinocarpus afforded rhodolamprometrin (5), bikaverin (6), and the new natural product 6-deoxybikaverin (7). All compounds were evaluated for cytotoxicity in a panel of four sentinel cancer cell lines, NCI-H460 (non-small cell lung), MIA Pa Ca-2 (pancreatic carcinoma), MCF-7 (breast), and SF-268 (CNS glioma), and only beauvericin (1) and bikaverin (6) were active with 1 and 6 showing selective toxicity toward NCI-H460 and MIA Pa Ca-2, respectively. Interestingly, 6-deoxybikaverin (7) was completely devoid of activity suggesting the requirement of the C−6 hydroxy group of bikaverin for its cytotoxic activity.

Cancer chemotherapy has relied mostly on cytotoxic drugs which inhibit tumor cell proliferation and cause cell death. In recent years targeting cell motility has attracted attention as one of the alternative strategies for the development of anticancer therapies.² Metastasis is reported to be responsible for over 90% of cancer deaths.³ Cell motility (migration) is a critical cause of tissue invasion allowing primary tumors to disseminate and metastasize. Cell migration is also involved in a number of physiological processes including ovulation, embryonic development, tissue regeneration (wound healing), and inflammation. These migration activities of cells in vitro are thought to be related to many in vivo cellular behaviors such as tumor angiogenesis, cancer cell invasion, and metastasis.⁴ It is known that primary solid tumors depend on angiogenesis (formation of new blood vessels) to obtain the necessary oxygen and nutrients for growth beyond a certain size (ca. 1–2 mm). The transition from a pre-angiogenic condition to tumor angiogenesis, often referred to as the angiogenic switch, is followed by tumor growth, cancer cell invasion, and metastasis. It should therefore be possible to halt (or retard) this process at different stages with the help of cell motility inhibitors. Cell motility under ordinary physiological conditions in an adult is rather infrequent; its repression might therefore be accompanied by manageable toxicity.⁵ Small molecule inhibitors have been used to study the migration of living cells, and it was found that the most readily available compounds directly disrupt actin microfilaments or microtubules.⁶ Several small-molecule natural products such as withaferin A,⁷ prodigiosin,⁸ and migrastatin⁹ have been reported to inhibit in cell motility. Of the methods available to screen natural product libraries/extracts for cell motility inhibitors, a moderate throughput assay known as the wound healing assay (WHA) employing metastatic cancer cell lines has gained popularity due to its simplicity. Utilization of this assay has led to the discovery of motuporamines capable of inhibiting angiogenesis,¹⁰ and the Rho-kinase inhibitor, 3-(4-pyridyl)indole (Rockout).¹¹

Recent studies have demonstrated that plant-associated microorganisms are prolific producers of novel and pharmacologically active secondary metabolites.¹² In the course of our ongoing efforts to discover potential anticancer agents from Sonoran desert plants and their associated microorganisms,¹ EtOAc extracts derived from two endophytic strains of Fusarium oxyporum (mitosporic Hypocreales) inhabiting the root tissue of Ephedra fasciculata (Mormon tea; Ephedraceae) and the stem tissue of Cylindropuntia echinocarpus (silver cholla; Cactaceae) were found to have activity in WHA and the MTT (tetrazolium-based colorometric) assay ¹³ for cancer cell proliferation/survival. WHA-guided fractionation of an EtOAc extract of a solid culture of F. oxysporum strain EPH2R_AA led to the isolation of beauvercin (1), along with (−)-oxysporidinone (2) and two of its new analogues, (−)-4,6′-anhydrooxysporidinone (3) and (−)-6-deoxyoxysporidinone (4). MTT assay-guided fractionation of an EtOAc extract of a liquid culture of F. oxysporum strain CECIS led to the isolation of rhodolamprometrin (5), bikaverin (6), and the new natural product, 6-deoxybikaverin (7). Known to be a plant pathogen, F. oxysporum has been previously subjected to a number of investigations that has resulted in the isolation of several bioactive metabolites including fumonisins, oxysporidinone, sambutoxin, and cyclosporine A.¹⁴ Beauvericin (1) has previously been encountered in a number of fungal strains including Beauveria bassiana,^15a Paecilomyces fumosoroseus, ^15b Polyporus sulphureus,^15c Fusarium semitectum,^16a and F. moniliforme var. subglutinans;^16a beauvericin has been detected in fruit tissues of two cultivars of melon fruits artificially infected with a strain of F. oxysporum f. sp. melonis.^16b In addition to its well characterized entomopathogenic activity,¹⁶ 1 has been reported to be toxic to the brine shrimp, Artemia salina,^15a and two human cell lines, Hep G2 (hepatocellular carcinoma) and MRC-5 (fibroblast-like fetal lung cells),¹⁷ to possess anti-convulsion and anti-arrhythmic activities, induce apoptosis in several cancer cell lines, and to inhibit acyl-CoA: cholesterol acyltransferase activity.¹⁸ It has recently been suggested that the induction of apoptosis by beauvericin (1) involves multiple cellular/molecular pathways including pro- and anti-apoptotic Bcl-2 family proteins, mitochondrial membrane potential, mitochondrial cytochrome c, and caspase 3.¹⁹ Bikaverin (6) has been reported to be cytotoxic to Erlich ascites carcinoma, leukemia L5178, and sarcoma 37 cells probably by inhibiting nucleic acid and protein synthesis.²⁰

Results and Discussion

The EtOAc extract of F. oxysporum strain EPH2R_AA exhibiting activity in wound healing assay (WHA), on bioassay-guided fractionation involving solvent-solvent partitioning followed by Sephadex LH-20 size-exclusion and silica gel chromatography afforded 1 – 4. Compound 1 was identified as beauvericin by comparison of its mass and NMR spectral data with those reported¹⁵ for this cyclic hexadepsipeptide. ¹H and ¹³C NMR data (Tables 1 and 2) together with its mass spectrum suggested compound 2 to be identical with (+)- or (−)-oxysporidinone previously encountered in F. oxysporum strains CBS 330.95^14d and N17B,^14f respectively. However, the optical rotation observed for 2 ([α]_D -101) confirmed it to be (−)-oxysporidinone.^14f Although the application of ¹H−¹H coupling constants and NOE data has aided the determination of relative orientation of groups on cyclohexanone and pyran rings of (+)-oxysporidinone, absence of NOE interactions between protons of these two moieties have precluded determination of complete relative stereochemistry for this compound.^14d Compound 3 was determined to have the molecular formula C₂₈H₄₁NO₅. Its ¹H and ¹³C NMR data (Tables 1 and 2) were similar to those of 2 except for the chemical shifts in the neighborhood of C-4 and C-6'. In ¹³C NMR spectrum of 3, compared with compound 2, the signal due to C-4 was shifted from δ 163.7 to δ 168.5 and that for C-6' from δ 71.5 to δ 92.0, indicating that a molecule of H₂O may have been lost from the two hydroxyl groups at these positions in 2 to form an ether linkage. The coupling constant (4.3 Hz) observed for H-6′ suggests that this proton is equatorial.^14f The ¹H NMR spectrum of 3 when recorded in DMSO-d₆ showed, in addition to other signals, only one D₂O exchangeable signal which appeared as a sharp singlet at δ 6.00 and this was assigned to 1'-OH. The absence of a signal around δ 10.0 due to 4-OH in its ¹H NMR spectrum in DMSO-d₆ further supported that this OH had participated in the formation of an ether linkage. The HMBC correlations of H-6′ with C-4 and C-4′ (Figure 1) confirmed the presence of the C-4−O−C-6′ linkage in 3. On the basis of COSY, HSQC and HMBC spectra, the structure of 3 was established as (−)-4,6′-anhydrooxysporidinone. Spectroscopic data of compound 4 indicated that it is structurally related to 2 and 3. The molecular formula of 4 was determined to be C₂₈H₄₃NO₅ from HRFABMS, suggesting that it has one oxygen atom less than oxysporidinone (2). However, its ¹³C NMR spectrum exhibited only 27 signals with a strong signal at δ 36.8 due to the overlapping signals of two CH₂ groups as indicated by DEPT, ¹H NMR and HSQC spectra. Compared with 2 the ¹H NMR spectrum of 4 also exhibited an overlapping signal due to two CH₂ groups which had cross peaks to two methylene carbon signals at δ 36.2 and 36.3 in its HSQC spectrum. These two CH₂ groups were found to be attached to each other as indicated by its COSY and HSQC spectra. Comparison of the NMR data of 4 with those of 2 suggested that in 4 the hydroxymethine group is replaced by a methylene group. Thus, compound 4 was identified as (−)-6'-deoxyoxysporidinone. Assignment of ¹H and ¹³C NMR signals of 4 (Tables 1 and 2) have been made with the help of its 2D spectra, including the HMBC spectrum (Figure 1).

Table 1.

¹H NMR Data (500 MHz) for Compounds 2–4

position	δH multiplicity (J in Hz)
	2a	3a	4b
4-OH			10.5 s
6	7.68 s	7.68 s	7.06 s
7	4.90 dd (11.2, 2.0)	4.76 dd (11.7, 2.3)	4.95 dd (11.1, 2.0)
8	1.94, 1.65 m	2.18, 1.53 m	2.06 m
9	1.93, 1.42 m	1.90, 1.31 m	1.90, 1.42 m
10	1.75 m	1.65 m	1.70 m
11	3.54 m	3.37 m	3.50 m
13	5.24 d (9.4)	5.11 dd (9.5, 1.1)	5.21 d (9.0)
14	2.55 m	2.51 m	2.49 m
15	1.24, 1.09 m	1.22, 1.08 m	1.21, 1.04 m
16	1.33 m	1.36 m	1.30 m
17	1.42, 1.06 m	1.43, 1.08 m	1.34, 1.04 m
18	0.85 mc	0.85 md	0.83 me
19	0.78 d (6.6)	0.73 d (6.6)	0.75 d (6.6)
20	1.66 s	1.57 d (1.1)	1.65 s
21	0.93 d (6.6)	0.90 d (6.6)	0.92 d (6.3)
22	0.85 mc	0.85 md	0.83 me
23	3.51 s	3.51 s	3.44 s
2'	2.60, 1.85 m	2.33, 2.25 m	2.38, 2.13 m
3'	2.78, 2.15 m	2.38, 2.03 m	2.90, 2.28 m
5'	2.83, 2.47 m	3.48 brd (11.0) 3.37 brd (11.0)	2.90, 2.28 m
6'	4.54 dd (11.4, 5.3)	4.92 d (4.3)	2.38, 2.13 m

^aIn CD₃OD.

^bIn CDCl₃.

^c–eOverlapping signals.

Table 2.

¹³C NMR Data (125 MHz) for Compounds 2–4

position	δC, multiplicitya
	2b	3b	4c
2	163.3, C	165.3, C	161.1, C
3	111.2, C	109.5, C	111.2, C
4	163.7, C	168.5, C	162.0, C
5	118.1, C	119.1, C	117.2, C
6	138.4, CH	135.8, CH	133.0, CH
7	79.1, CH	73.7, CH	78.1, CH
8	31.5, CH2	30.3, CH2	30.6, CH2
9	33.2, CH2	34.1, CH2	32.0, CH2
10	33.7, CH	33.1, CH	32.5, CH
11	93.6, CH	93.2, CH	92.5, CH
12	132.0, C	133.9, C	130.0, C
13	138.7, CH	137.4, CH	138.1, CH
14	30.9, CH	30.9, CH	29.7, CH
15	46.1, CH2	46.4, CH2	44.7, CH2
16	33.4, CH	33.4, CH	32.0, CH
17	29.8, CH2	30.1, CH2	28.9, CH2
18	11.6, CH3	11.7, CH3	11.2, CH3
19	18.0, CH3	18.4, CH3	17.6, CH3
20	12.5, CH3	11.7, CH3	11.9, CH3
21	21.2, CH3	21.3, CH3	20.7, CH3
22	20.2, CH3	20.2, CH3	19.6, CH3
23	37.8, CH3	38.6, CH3	37.2, CH3
1'	74.8, C	77.2, C	70.2, C
2'	32.8, CH2	34.4, CH2	36.2d, CH2
3'	37.2, CH2	36.0, CH2	36.8, CH2
4'	212.1, C	210.3, C	211.4, C
5'	47.0, CH2	49.4, CH2	36.8, CH2
6'	71.5, CH	92.0, CH	36.3d, CH2

^aMultiplicity from DEPT.

^bIn CD₃OD.

^cIn CDCl₃.

^dInterchangeable.

Figure 1.

Selected HMBC correlations for 3 and 4.

MTT assay-guided fractionation of a cytotoxic EtOAc extract of F. oxysporum strain CEC1S resulted in the isolation compounds 5 – 7. Compound 5 was identified as rhodolamprometrin by comparison of its mass, ¹H and ¹³C NMR spectral data with those reported in the literature.²¹ Compounds 6 and 7 exhibited very similar NMR spectra except for the absence of an OH group in 7 compared with 6 as evident from their APCI-MS and ¹H NMR spectra. Spectroscopic data for 6 and 7 were found to be consistent with those reported for bikaverin²² and 6-deoxybikaverin,²³ respectively. Rhodolamprometrin (5) has previously been isolated from the feather star, Lamprometra klunzingeri,²⁴ and a mutant of the deuteromycete, Trichoderma viride.²¹ Bikaverin (6) has been encountered in a number of microorganisms including several Fusarium species²⁵ and Mycogone jaapii.²⁶ 6-Deoxybikaverin (7) has previously been obtained only as a by-product in the synthesis of bikaverin (6)²³ and this constitutes the first report of its natural occurrence.

Compounds 1 – 7 were evaluated for in vitro cytotoxic activity against a panel of four sentinel human cancer cell lines, NCI-H460 (non-small cell lung), MIA Pa Ca-2 (pancreatic carcinoma), MCF-7 (breast), and SF-268 (CNS glioma). Cells were exposed to serial dilutions of test compounds for 72 h in RPMI 1640 media supplemented with 10% fetal bovine serum, and when cell viability was evaluated by the MTT assay,¹³ only beauvericin (1) and bikaverin (6) were found to be cytotoxic. The concentrations resulting in 50% inhibition of cell proliferation/survival as measured by the MTT assay (IC₅₀) were found to range between 0.01 and 1.81 µM (Table 3). Interestingly, 6-deoxybikaverin (7) which lacks the OH group at C-6 did not exhibit cytotoxic activity towards any of the cell lines used even at concentrations of 4.0 and 2.0 µg/mL. As beauvericin (1) was isolated using WHA-guided fractionation, it was evaluated for cell migration inhibitory activity in two metastatic cancer cell lines, PC-3M (prostate cancer) and MDA-MB-231 (breast cancer). Beauvericin (1) inhibited migration of PC-3M and MDA-MB-231 cells at concentrations ranging from 2.0 – 2.5 and 3.0 – 4.0 µM, respectively (Figure 2). The extent of cell migration inhibition when determined using NIH ImageJ software²⁷ suggested that 1 caused ca. 50% inhibition of the migration of PC-3M and MDA-MB-231 cells at concentrations of ca. 3.0 and 5.0 µM, respectively (Figure 3). The concentrations of 1 required for 50% (IC₅₀) and 25% (IC₂₅) inhibition of PC-3M cell proliferation/survival as measured by the MTT assay at 20 h (time used for WHA with PC-3M) were determined to be 3.8 and 2.3 µM, respectively, and those for MDA-MB-231 at 40 h (time used for WHA with MDA-MB-231) were found to be 7.5 and 6.4 µM, respectively, suggesting that it is capable of inhibiting migration of both metastatic cancer cell lines at sublethal concentrations. Encouraged by this observation we next evaluated the antiangiogenic activity of beauvericin (1) by assessing its ability to interfere with endothelial cell network formation.²⁸ Endothelial (HUVEC-2) cells in suspension were placed on top of a thin layer of Matrigel,® allowing the cells to align and form tube like structures. Cells were then exposed to vehicle control (DMSO) or increasing concentrations of 1 for 16 h, fixed and observed under an inverted microscope. Beauvericin (1) clearly showed potent anti-angiogenic activity at sublethal concentrations, as demonstrated by the inhibition of endothelial cell network formation (Fig. 4); complete disruption of HUVEC-2 network formation was observed at a concentration (3.0 µM) below IC₅₀ (7.5 µM) and IC₂₅ (5.0 µM) of 1 for this cell type exposed to the same length of time (16 h). Studies to elucidate the molecular target(s) of beauvericin (1) by gene expression analysis and animal studies to evaluate its anticancer potential are currently in progress.

Table 3.

Cytotoxicities of Compounds 1 and 6 against a Panel of Four Tumor Cell Lines^a

compound	cell lineb
	NCI-H460	MIA Pa Ca-2	MCF-7	SF-268
1	1.41	1.66	1.81	2.29
6	0.43	0.26	0.42	0.38
doxorubicin	0.01	0.05	0.07	0.04

^aResults are expressed as IC₅₀ values in µM.

^bKey: NCI-H460 = human non-small cell lung cancer; MIA Pa Ca-2 = human pancreatic carcinoma; MCF-7 = human breast cancer; SF-268 = human CNS cancer (glioma).

Figure 2.

Wound healing assay of beauvericin (1) with metastatic cancer cells PC-3M (A – D) and MDA-MB-231 (E – H). (A) and (E) DMSO control (negative). (B) and (F) LY294002 control (positive) at 7.5 µM. (C) Beauvericin at 2.0 µM. (D) Beauvericin at 2.5 µM. (G) Beauvericin at 3.0 µM. (H) Beauvericin at 4.0 µM.

Figure 3.

Quantitative wound healing assay data for beauvericin (1). (A) In metastatic prostate cancer (PC-3M) cells. (B) In metastatic breast cancer (MDA-MB-231) cells.

Figure 4.

In vitro antiangiogenesis and cell survival assay of beauvericin (1) in HUVEC-2 cells exposed for 16 h. (A) Negative (DMSO) control. (B) Positive (withaferin A) control at 1.0 µM. (C) Beauvericin at 2.0 µM. (D) Beauvericin at 3.0 µM. (E) MTT cell survival assay (n = 3).

Experimental Section

General Experimental Procedures

Melting points were determined with an Electrothermal micromelting point apparatus and are uncorrected. Optical rotations were measured with a Jasco DIP-370 digital polarimeter using MeOH as solvent. IR spectra were recorded on a Shimadzu FTIR-8300 spectrometer in KBr disks and UV spectra in MeOH on a Shimadzu UV-1601 spectrometer. 1D and 2D NMR spectra were recorded on a Bruker DRX-500 (500 MHz for ¹H NMR and 125 MHz for ¹³C NMR), and Varian Inova-600 (600 MHz for ¹H NMR and 150 MHz for ¹³C NMR) instruments. The Chemical shift values (δ) are given in parts per million (ppm), and the coupling constants are in Hz. Low resolution APCIMS were recorded on a Shimadzu LCMS-QP8000α HPLC-MS system, and high resolution FABMS were obtained with a JEOL HX110A mass spectrometer.

Fungal Isolation, Identification, and Cultivation

The two endophytic strains of F. oxyporum used in this study were isolated from the root tissue of Ephedra fasciculata and the stem tissue of Cylindropuntia echinocarpus growing in the Sonoran desert, and were deposited in the School of Life Sciences, Arizona State University under the accession numbers EPH2R_AA and CEC1S, respectively, and at the Southwest Center for Natural Products Research and Commercialization of the University of Arizona microbial culture collection under the accession numbers CS-36-33 and CS-69-50, respectively. Both strains were grown on small scale in solid (PDA) and liquid (PDB) media, processed as for large scale cultures described below, and the resulting EtOAc extracts tested in assays for cell proliferation/survival (MTT) and inhibition of metastatic cancer cell motility (WHA). F. oxysporum strain EPH2R_AA was found to produce bioactive metabolites when cultured on PDA whereas the strain CEC1S produced bioactive metabolites in PDB. Thus, for isolation of bioactive secondary metabolites, F. oxysporum EPH2R_AA was cultured on PDA in 80 T-flasks (135 mL of PDA each) for 14 days at 28°C, while the strain CEC1S was cultured in 12 L of PDB in six 4 L flasks (2 L of PDB each) on a shaker (150 rpm) for 14 days at 28 °C.

Extraction and Isolation of the Metabolites from F. oxysporum EPH2R_AA

After 14 days of cultivation, MeOH (200 mL/T-flask) was added to all 80 T-flasks and shaken in a rotary shaker for 8 h at 27 °C, and the resulting extract was filtered through Whatman No. 1 filter paper and a layer of Celite 545. The filtrate was concentrated to one-fourth of its original volume and extracted with EtOAc (3 × 1500 mL). Combined EtOAc extracts were evaporated under reduced pressure to afford a dark brown solid (696 mg) which showed activity in both WHA and MTT assays. A portion of this extract (595 mg) was partitioned between hexane and 80% aqueous MeOH. The WHA-active aqueous MeOH fraction was diluted to 50% aqueous MeOH with water and extracted with CHCl₃. Of the resulting fractions, only the CHCl₃ fraction (322 mg) was found to be active and this was subjected to size exclusion chromatography on a column of Sephadex LH-20 (12 g) made up in hexane/CH₂Cl₂ (1:1) and eluted with hexane/CH₂Cl₂ (1:4) (100 mL), of CH₂Cl₂/acetone (3:2) (100 mL), CH₂Cl₂/acetone (1:4) (100 mL), CH₂Cl₂/MeOH (1:1) (50 mL) and finally MeOH (50 mL). Twenty fractions were collected and combined into six fractions (F1–F6) based on their TLC profiles of which fractions F1 (110.9 mg) and F2 (71.2 mg) were found to be active. These two fractions were combined and the combined fraction (F7) was fractionated further on reversed phase silica gel (RP-18; 6.0 g) column by elution with increasing amounts of MeOH (66% – 100%) in H₂O, from which 11 fractions (F7A–F7K) were collected. Fractions F7E (23.4 mg), F7F (25.4 mg), and F7G (22.2 mg) were combined and further separated by reversed phase preparative TLC (5% H₂O in MeOH), leading to the isolation of 1 (62.3 mg) and 4 (1.3 mg). Fraction F7D (35.2 mg) was separated by silica gel preparative TLC (hexane/acetone, 2:5) to yield 2 (10.5 mg) and 3 (6.6 mg).

Beauvericin (1)

colorless needles (MeOH) mp 92–93 °C (lit.^15a mp 93–94 °C); ¹H NMR (500 MHz, CDCl₃): δ 7.14–7.26 (5H, m), 5.45 (1H, dd, J = 11.1, 4.1 Hz), 4.91 (1H, d, J = 8.6 Hz), 3.34 (1H, dd, J = 14.6, 5.0 Hz), 2.98 (3H, s), 2.97 (1H, dd, J = 14.6, 11.8 Hz), 2.00 (1H, m), 0.79 (3H, d, J = 6.7 Hz), 0.42 (3H, d, J = 6.8 Hz); APCIMS (+) mode m/z 784 [M+1]⁺.

(−)-Oxysporidinone (2)

colorless amorphous solid; ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{25}-101.1$ (c, 0.20, MeOH); ¹H and ¹³C NMR data, see Tables 1 and 2; APCIMS (+) mode m/z 490 [M+1]⁺; APCIMS (-) mode m/z 488 [M-1]⁺.

(−)-4,6′-Anhydrooxysporidinone (3)

colorless crystalline solid (MeOH); mp 160–161 °C; ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{25}-{78.6}^{\mathrm{o}}$ (c, 0.30, MeOH); UV (MeOH) λ_max (log ε): 290.2 (3.70) nm; IR (KBr) ν_max: 3406, 3215, 2958, 2924, 2872, 1724, 1703, 1668, 1593, 1564, 1456, 1136, 1065, 1013 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; APCIMS (+) mode m/z 472 [M+1]⁺; APCIMS (-) mode m/z 470 [M-1]⁺; HRFABMS m/z 472.3062 [M+1]⁺ (calcd for C₂₈H₄₂NO₅, 472.3063).

(−)-6-Deoxyoxysporidinone (4)

colorless crystalline solid; mp 130–131 °C; ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{25}-72.0$ (c, 0.15, MeOH); UV (MeOH) λ_max (log ε): 219.8 (4.46), 296.0 (3.66) nm; IR (KBr) ν_max: 3418, 3234, 2959, 2924, 2872, 1705, 1651, 1564, 1445, 1383, 1065, 1013 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; APCIMS (+) mode m/z 474 [M+1]⁺; APCIMS (-) mode m/z 472 [M-1]⁺; HRFABMS m/z 474.3242 [M+1]⁺ (calcd for C₂₈H₄₄NO₅, 474.3219).

Extraction and Isolation of the Metabolites from F. oxysporum CEC1S

After 14 days of fermentation, the PDB culture was filtered through Whatman No. 1 filter paper, the filtrate (12 L; pH = 4.4) neutralized with 1 N NaOH, and extracted (x 3) with EtOAc (10 L). The EtOAc extract thus obtained was evaporated under reduced pressure to obtain a dark red solid (603 mg) which was found to be active in the MTT assay. Solvent-solvent partitioning of a portion (552 mg) of this EtOAc extract as for F. oxysporum EPH2R_AA above afforded hexane (20.1 mg), CHCl₃ (276 mg), and 50% aqueous MeOH (245 mg) fractions of which the latter two fractions were found to be active in the MTT assay. A portion (261 mg) of the CHCl₃ fraction was subjected to size exclusion chromatography on Sephadex LH-20 and eluted with hexane/CH₂Cl₂ (1:4) (100 mL), CH₂Cl₂/acetone (2:3) (100 mL), CH₂Cl₂/acetone (1:4) (100 mL), CH₂Cl₂/MeOH (1:1) (100 mL), and finally MeOH (100 mL) yielding five fractions, F1' (129.5 mg), F2' (71.9 mg), F3' (25.1 mg), F4' (16.4 mg), and F5' (14.7 mg). All five fractions were active in the MTT assay and were found to have common TLC spots including those due to a major red compound and a minor yellow compound. These fractions and the 50% aqueous MeOH fraction active in the MTT assay were separately subjected to silica gel preparative TLC (CH₂Cl₂/MeOH, 92:8), and the resulting fractions with the same R_f were combined affording 6 (58.1 mg; R_f = 0.6) and 7 (1.5 mg; R_f = 0.8). Similar fractionation of F3' provided 5 (2.9 mg; R_f = 0.2) together with small quantities of 6 and 7.

Rhodolamprometrin (5)

orange powder; ¹H NMR (500 MHz, DMSO-d₆): δ 12.4 (1H, s, 1-OH), 12.1 (1H, s, 8-OH), 7.04 (1H, d, J = 1.7 Hz, H-5), 6.59 (1H, d, J = 1.7 Hz, H-7), 6.63 (1H, s, H-2), 2.37 (3H, s, 4-COCH₃); APCIMS (−) mode m/z 314 [M]⁺. Its NMR and MS data were consistent with those reported for rhodolamprometrin.²¹

Bikaverin (6)

red powder; ¹H NMR (500 MHz, CDCl₃): δ 6.90 (1H, d, J = 2.5 Hz, H-4), 6.78 (1H, d, J = 2.5 Hz, H-2), 6.33 (1H, s, H-9), 3.94 (3H, s, 8-OCH₃), 3.91 (3H, s, 3-OCH₃), 2.84 (3H, s, 1-CH₃); APCIMS (+) mode m/z 383 [M+1]⁺. These NMR and MS data compared well with those reported for bikaverin.²²

6-Deoxybikaverin (7)

yellow powder; ¹H NMR (500 MHz, CDCl₃): δ 7.61 (1H, s, H-6), 6.75 (2H, d, J = 2.5 Hz, H-2 and H-4), 6.16 (1H, s, H-9), 3.92 (3H, s, 8-OCH₃), 3.88 (3H, s, 3-OCH₃), 2.85 (3H, s, 1-CH₃). APCIMS (+) mode m/z 367 [M+1]⁺. These NMR and MS data were consistent with those reported in the literature.²³

Cytotoxicity Assay

The tetrazolium-based colorometric assay (MTT assay)¹³ was used for the in vitro assay of cytotoxicity to NCI-H460 (non-small cell lung cancer), MIA Pa Ca-2 (pancreatic carcinoma), MCF-7 (breast cancer) and SF-268 (CNS glioma). Doxorubicin and DMSO were used as positive and negative controls.

Wound Healing Assay (WHA)

For WHA the metastatic prostate cancer (PC-3M) cells were cultured in RPMI 1640 medium containing 10% fetal bovine serum, penicillin (100 IU/mL) and streptomycin (50 µg/mL), and the metastatic breast cancer (MDA-MB-231) cells were cultured in DMEM/Ham’s F-12 medium containing 10% fetal bovine serum and gentamycin (50 µg/mL). Both types of cells were grown in a 5% CO₂ atmosphere at 37 °C and the cells were harvested at or above 80% confluence. The cells were plated onto sterile 24-well plates at a density of 70,000 cells per well for PC-3M and 150,000 cells per well for MDA-MB-231, and were allowed to recover for 48 and 24 h, respectively, until a confluent cell monolayer has formed in each well (> 90% confluence). Wounds are then inflicted to each cell monolayer using a sterile toothpick, media were removed and the cell monolayers washed once with PBS, and fresh media were added to each well. Test samples (extracts, fractions or beauvericin) were prepared in DMSO at different concentrations and added to the plates, each in duplicate along with the two controls, phosphotidylinositol (PtdIns) 3-kinase inhibitor LY294002²⁹ at 7.5 µM (positive control), and DMSO (negative control). The plates were incubated for 40 h (PC-3M) or 20 h (MDA-MB-231) during which the wells treated with DMSO had healed entirely and the wells treated with LY294002 and samples containing cell motility inhibitors had wounds present. All treatments, including the controls, were documented photographically. A treatment was considered active if there was a wound present in duplicate wells at the completion of the assay.

Quantification of Wound Healing Inhibition

ImageJ software available from the NIH website (http://rsb.info.nih.gov/rj/) was used to quantify WHA data.²⁷ Three random pictures were taken for each wound using an inverted microscope at × 20 magnification; photos were taken immediately after a wound was inflicted to the cell monolayer, uploaded into the ImageJ software and the area of the wound was measured by using the rectangle area selection tool and the three areas per well averaged. After the DMSO wells have healed (usually 40 h after infliction of the wound for PC-3M and 20 h for MDA-MD-231), three random pictures were taken for each treated well and the areas of the wounds measured as above. The % of wound healed was then calculated using the formula 100 − (Final area/Initial area × 100%).²⁷

In Vitro Angiogenesis Assay

Matrigel® (BD Biosciences) was placed in sterile round-bottomed 96-well plates (45 µL of Matrigel® per well). Once Matrigel® has formed into a gel, HUVEC-2 cells cultured in EBM Basal Medium supplemented with EGM-MV SingleQuot kit (Cambrex Bio Science Walkersville, Inc) were plated onto Matrigel® at a density of 10,000 cells per well and were allowed to adhere to the gel (adherence occurs in ca. 1 h). After the cells have adhered to Matrigel,® DMSO (negative control), 1.0 µM withaferin A (positive control),⁷ and various concentrations of beauvericin (1) in the above media were added to separate wells (in triplicate). The plates were incubated overnight (ca. 16 h) and the results were documented photographically using an inverted microscope at × 4 magnification and were edited in Microsoft Office Picture Manager.

Figure 5.