New Cassane-Type Diterpenoids of Caesalpinia echinata (Leguminosae) Exhibiting NF-κB Inhibitory Activities

Abstract

Seven new cassane-type diterpenoids, echinalides A–G (1–7), were isolated from the stem of Caesalpinia echinata LAM. (Leguminosae). The structures were established on the basis of their chemical properties and spectroscopic evidence, including two dimensional (2D)-NMR analysis. These compounds were assessed for inhibitory activity against nuclear factor κB (NF-κB). Echinalides C and D, in particular, significantly inhibited NF-κB-responsive reporter gene expression at 5.0 µM, an effect almost equivalent to that of parthenolide, a known potent inhibitor of NF-κB.

Caesalpinia echinata LAM. (Leguminosae family), native to the Brazilian Atlantic coast, is locally called Pau brasil (brazil wood), and the name of Brazil refers to the red color of the trunk which resembles burning wood.^1,2) The plant has been used to manufacture violin bows^3,4) and as a source of a red pigment composed of brazilin, a flavonoid derivative.^5–7) Although the Caesalpinia genus has been previously reported to possess interesting biological activities,^8–10) there are only a few published chemical and pharmacological studies concerning this particular species.^11,12)

Nuclear factor κB (NF-κB) is a transcription factor and functions as a key regulatory element in inflammatory and immune responses. In unstimulated cells, NF-κB is retained in the cytoplasm via interaction with its inhibitor, IκB.¹³⁾ In response to pro-inflammatory stimuli, IκB is phosphorylated by IκB kinase complex, leading to its ubiquitination and subsequent proteasome-mediated degradation, allowing NF-κB to homodimerize and enter the nucleus. NF-κB then binds to its recognition sequences in promoters to induce transcription of target genes, including mediators of the immune and inflammatory response such as the inflammatory cytokines and tumor necrosis factor (TNF)-α.¹⁴⁾ Therefore, inhibitors of NF-κB are interesting as lead compounds for the treatment of acute and chronic inflammation. In the course of a search for a new class of inhibitors of NF-κB from plant extracts, we previously reported that culcitiolides, new eremophilane-type sesquiterpenes from Senecio culcitioides (Asteraceae), are significant NF-κB inhibitors.^15,16) This finding prompted to search for more potent NF-κB inhibitors from plant extracts. Consequently, in the stem of Caesalpinia echinata LAM., we discovered new cassane-type diterpenoids, some of which were potent NF-κB inhibitors. Here, we describe their isolation, structural characterization and biological activity.

Results and Discussion

Echinalide A (1) was obtained as a yellow oil and was determined to have a molecular formula of C₂₉H₃₄O₅ from its high resolution (HR)-FAB-MS (m/z 463.2479 [M+H]⁺; Calcd: 463.2484). The IR spectrum exhibited absorptions for carbonyl ester (1708 cm⁻¹) and conjugated double bond (1627 cm⁻¹) functional groups. The UV spectrum showed absorption bands at λ_max 217 and 278 nm. In addition, 1 showed a red-pink color in the Ehrlich test, indicating a furan chromophore.¹⁷⁾ The ¹H-NMR spectrum showed a secondary methyl, two tertiary methyls, an oxymethine, and two mutually coupled vinyl proton signals attributed to a 2,3-disubstituted furan ring.^18,19) Furthermore, the presence of a trans-cinnamoyl moiety was confirmed by two coupled doublets at δ_H 6.38 and 7.66 as well as multiplet signals of aromatic protons between δ_H 7.37 and 7.52. In the ¹³C-NMR spectrum, twenty-nine carbon signals were present, including signals for a carboxylic carbonyl carbon and a cinnamoyl carbonyl carbon. These spectral data showed the characteristics of a 2, 3-dihydrofuranocassane-type skeleton containing a cinnamoyl moiety.^12,17,19) The full connectivities of the two dimensional (2D)-NMR signals allowed the assignment of all proton and carbon signals, as shown in Tables 1 and 2, and the planar structure was deduced mainly from the heteronuclear multiple bond correlation (HMBC) spectrum. The HMBC correlations of H-6/C-1′ and H-19/C-18 indicated that the cinnamoyl ester and carboxylic acid groups were located at the C-6 and C-4 positions, respectively. The position of the furan ring was elucidated from the H-14/C-12, C-13, C-8 and H-11/C-8, C-9, C-12 correlations (Fig. 2). Thus, the planar structure of 1 was determined to be that shown in Fig. 1. The relative stereostructure was assigned based on the results of nuclear Overhauser enhancement and exchange spectroscopy (NOESY) experiments. NOE enhancements were observed between H-20/H-19, H-11β, H-8/H-11β, H-14 and H-9/H-5, H-11α, H-17, suggesting that the configurations of the ring A, B and C junctions were trans/anti/trans and that H-14 was in a β-equatorial position (Fig. 3). An NOE interaction between H-5 and H-6 indicated that the cinnamoyl moiety was β-axial, a finding supported by the small vicinal coupling constant between H-6 and H-5. Based on the above results, the relative structure of 1 was determined to be that in Fig. 1. Compound 1 was treated with m-chloroperbenzoic acid (mCPBA) to afford a known compound, 6β-O-cinnamoyloxy-12-hydroxy-(13) 15-en-16, 12-olide-18-cassaneoic acid (1a)^12,20) which had been simultaneously isolated from this plant during this work. The relative configuration of 1a was deduced from NOESY spectra and agreed with the reported relative stereochemistry.¹²⁾ Because the absolute configuration of known compound 1a was unresolved, the absolute stereochemistry was determined from the circular dichroism (CD) curves for the γ-lactone chromophore.^20–22) The sign of the CD curve in EtOH (λ_max 225 nm) was negative, indicating that the chirality at C-12 in compound 1a was R, as shown in Fig. 1. Thus, the absolute configuration of 1 was determined to be 6β-O-cinnamoyloxy-18-vouacapaneoic acid, as shown in Fig. 1. The co-occurrence of 1 and 1a suggested that 1a may be derived biogenetically from 1 through the plausible route shown in Fig. 4.²³⁾

Table 1. ¹H-NMR Data of Echinalides A–G (1–7) (400 MHz) (δ in ppm, Multiplicities, J in Hz)

	1a)	2a)	3b)	4b)	5a)	6a)	7a)
1	1.23 (m)	1.56 (m)	1.10 (m)	1.01 (m)	1.10 (m)	1.17 (m)	1.01 (m)
	1.82 (dd, 13.2, 3.6)	1.79 (m)	1.72 (m)	1.71 (m)	1.79 (m)	1.84 (m)	1.75 (m)
2	1.65 (m)	1.59 (m)	1.46 (m)	1.33 (m)	1.54 (m)	1.62 (m)	1.55 (m)
	1.72 (m)	1.70 (m)	1.68 (m)	1.47 (m)	1.62 (m)	1.86 (m)	1.63 (m)
3	1.68 (m)	1.64 (m)	1.28 (m)	1.24 (m)	1.70 (m)	1.64 (m)	1.16 (m)
	1.75 (m)	1.77 (m)	2.01 (m)	1.97 (m)	1.74 (m)	1.72 (m)	1.20 (m)
5	2.11 (br s)	2.01 (br s)	2.13 (br s)	2.02 (m)	1.92 (br s)	2.05 (br s)	1.51 (m)
6	5.32 (br s)	5.14 (br s)	6.17 (br s)	5.96 (br s)	5.11 (br s)	5.32 (br s)	5.43 (br s)
7	1.73 (m)	1.64 (m)	1.75 (m)	1.61 (m)	1.56 (m)	1.69 (m)	1.54 (m)
	1.96 (dt, 7.6, 3.6)	1.77 (m)	1.89 (m)	1.68 (m)	1.59 (m)	1.82 (m)	1.58 (m)
8	2.10 (m)	1.79 (m)	2.07 (m)	1.55 (m)	1.53 (m)	1.94 (m)	1.53 (m)
9	1.69 (m)	1.62 (m)	1.91 (m)	1.83 (m)	1.33 (m)	1.44 (m)	1.32 (m)
11	2.55 (m)	2.44 (m)	1.61 (m)	1.52 (m)	0.99 (m)	1.14 (m)	1.04 (m)
	2.65 (m)	2.59 (m)	2.77 (dd, 12.8, 3.2)	2.74 (m)	2.48 (m)	2.55 (m)	2.49 (m)
12					4.83 (dd, 11.6, 6.4)	4.88 (dd, 11.6, 6.0)	4.83 (dd, 11.6, 6.0)
14	2.64 (m)	2.51 (m)	2.82 (dd, 7.2, 4.8)	2.71 (m)	2.76 (dd, 7.2, 4.0)	2.92 (m)	2.75 (m)
15	6.18 (d, 1.2)	6.19 (s)	5.91 (s)	5.94 (s)	5.70 (d, 1.2)	5.67 (s)	5.69 (d, 1.2)
16	7.23 (d, 1.2)	7.24 (m)
17	0.99 (d, 7.2)	0.94 (d, 7.0)	1.17 (d, 7.2)	1.13 (d, 7.2)	1.03 (d, 7.2)	1.07 (d, 7.6)	1.01 (d, 7.2)
18			3.38 (d, 11.2)	3.33 (d, 11.2)			3.09 (d, 10.8)
			4.03 (d, 11.2)	3.93 (d, 10.8)			3.54 (d, 10.8)
19	1.43 (s)	1.35 (s)	1.11 (s)	1.01 (s)	1.31 (s)	1.36 (s)	0.84 (s)
20	1.38 (s)	1.21 (s)	1.33 (s)	1.18 (s)	1.08 (s)	1.25 (s)	1.07 (s)
2′	6.38 (d, 16.0)	2.61 (t, 6.3)	6.92 (d, 16.0)	2.82 (t, 8.0)	2.64 (td, 7.2, 3.2)	6.37 (d, 16.0)	2.64 (t, 7.2)
3′	7.66 (d, 16.0)	2.95 (m)	8.04 (d, 16.0)	3.10 (m)	2.96 (td, 7.2, 3.2)	7.66 (d, 16.0)	2.96 (m)
5′, 9′	7.50 (m)	7.16 (m)	7.73 (dd, 7.6, 1.6)	7.36 (m)	7.18 (m)	7.51 (m)	7.53 (m)
6′, 8′	7.38 (m)	7.26 (m)	7.38 (m)	7.36 (m)	7.26 (m)	7.39 (m)	7.40 (m)
7′	7.38 (m)	7.18 (m)	7.40 (m)	7.34 (m)	7.19 (m)	7.39 (m)	7.39 (m)

a) Measured in CDCl₃. b) Measured in C₅D₅N.

Table 2. ¹³C-NMR Data of Echinalides A–G (1–7) (100 MHz)

	1a)	2a)	3b)	4b)	5a)	6a)	7a)
1	41.4	41.4	41.7	41.7	41.1	41.2	41.4
2	18.0	17.9	18.8	18.8	18.0	18.1	18.1
3	39.0	38.9	37.6	37.6	39.0	39.0	37.0
4	47.4	47.4	39.9	38.8	47.4	47.6	38.1
5	49.7	49.3	47.9	47.6	49.2	49.6	47.6
6	72.5	72.6	70.0	69.9	72.0	72.1	69.3
7	36.0	35.8	35.9	35.7	34.9	35.2	35.1
8	31.4	31.2	36.4	36.4	35.2	35.8c)	35.8
9	45.8	45.7	45.8	45.6	45.4	45.4	45.2
10	37.5	37.3	37.7	37.5	37.3	37.5	37.6
11	21.7	21.6	39.0	38.9	33.7	33.8	33.9
12	149.1	149.1	107.1	106.9	79.0	79.0	79.2
13	122.2	122.1	174.1	174.2	176.3	176.3	176.5
14	31.0	31.0	36.4	36.1	35.7	35.6c)	35.0
15	109.5	109.4	113.5	113.4	110.7	110.8	110.6
16	140.5	140.5	171.5	171.4	173.6	173.6	173.7
17	17.7	17.6	13.2	13.1	14.1	14.1	14.1
18	182.8	184.4	70.9	70.8	183.4	184.0	71.1
19	18.4	17.8	20.1	20.0	18.3	18.4	19.4
20	18.0	18.3	18.3	18.1	17.5	18.4	17.6
1′	166.4	172.2	166.6	172.4	172.1	166.4	172.4
2′	118.7	36.6	119.8	36.6	36.3	118.2	36.3
3′	144.8	30.9	145.0	31.2	30.8	145.4	30.8
4′	134.3	140.2	134.1	141.1	140.1	134.1	140.1
5′, 9′	128.1	128.2	128.7	128.7	128.2	128.2	128.2
6′, 8′	128.9	128.5	129.4	128.4	128.5	128.9	128.5
7′	130.3	126.3	130.7	126.7	126.4	130.6	126.4

a) Measured in CDCl₃. b) Measured in C₅D₅N. c) Assignment will be interchangeable.

Fig. 1. Structures of Echinalides A–G (1–7) from Caesalpinia echinata LAM.

Fig. 2. ¹H–¹H COSY and HMBC Correlations of Echinalides A (1), C (3), and E (5)

Arrows indicate HMBC correlations. Bold lines represent the connections in ¹H–¹H COSY.

Fig. 3. Key NOESY Correlations of Echinalides A (1) and C (3)

Fig. 4. Plausible Biogenetic Pathways of Compound 1a

Echinalide B (2) was obtained as a yellow oil, and its molecular formula of C₂₉H₃₆O₅ was determined by HR-FAB-MS (m/z 465.2657 [M+H]⁺; Calcd: 465.2641). The IR spectrum showed an absorption band at 1718 cm⁻¹ from a carbonyl ester, and the UV spectrum exhibited absorption bands at λ_max 212 and 279 nm. The ¹H- and ¹³C-NMR spectroscopic data were closely related to those of 1 except for the appearance of signals attributable to a dihydrocinnamoyl ester instead of the cinnamoyl moiety observed in 1. The location of the dihydrocinnamoyl group was confirmed by HMBC correlations of H-6/C-1′ and H-3′/C-9′(C-5′), and its configuration was the same as 1, as supported by NOESY experiments. Thus, the identity of 2 was established as 6β-O-2′,3′-dihydrocinnamoyloxy-18-vouacapaneoic acid.

Echinalide C (3), a yellow oil, gave a quasi-molecular ion peak [M+H]⁺ at m/z 481.2579 (Calcd: 481.2590) from the HR-FAB-MS, corresponding to a molecular formula of C₂₉H₃₆O₆. The IR spectrum implied the presence of a hydroxyl group (3693 cm⁻¹), an α,β-unsaturated γ-lactone (1739 cm⁻¹) and a conjugated double bond (1692 cm⁻¹). The UV spectrum exhibited absorption bands at 217 and 278 nm, and the former band as well as the corresponding IR absorption band suggested the presence of an α,β-butenolide ring.^21,24) The ¹H-NMR spectrum of 3 showed the signals for two tertiary methyls, a secondary methyl, an oxymethine proton, an oxymethylene proton, and an olefinic proton attributable to the α,β-unsaturated γ-lactone. Furthermore, two coupled doublet signals (δ_Η 6.92, d, J=16.0 Hz, H-2′ and δ_Η 8.04, d, J=16.0 Hz, H-3′) assigned to olefinic protons, and the monosubstituted benzene proton signals confirmed the presence of a cinnamoyl moiety. Detailed analyses of the ¹³C-NMR, heteronuclear multiple quantum coherence (HMQC) and HMBC spectra showed twenty-nine carbons in total including a hemiketal carbon, and the full assignments are shown in Table 2. These data indicated that 3 was a cassane diterpene lactone with a cinnamoyl moiety. The position of the α,β-butenolide hemiketal ring was deduced from the HMBC correlations of H-14/C-12, C-13, H-17/C-8, C-13, C-14 and H-15/C-12, C-13, C-14, C-16, and the oxymethylene group was assigned to C-4 based on the correlations of H-19/C-3, C-4, C-5, C-18 and H-18/C-3, C-4, C-5, C-19. In addition to the HMBC correlations of H-6/C-7, C-8 and H-5/C-7, C-20, the deshielded carbon and proton at δ_C 70.0 and δ_H 6.17, respectively, suggested that the cinnamoyl moiety was attached to C-6, although the HMBC connectivity between C-6 and C-1′ was not clearly detected. In the NOESY spectrum, the correlations of H-18/H-5, H-6, H-19 and H-5/H-6 indicated that the oxymethylene group was α-equatorial and that the cinnamoyl moiety was β-axial. The spectrum also revealed that H-14 was correlated with H-8 and H-15, and H-9 displayed a cross peak with Me-17, supporting an α-orientation for both the 12-hydroxy group and Me-17. The CD curve of 3 also showed a negative Cotton effect for the γ-lactone chromophore at λ_max 226 nm in EtOH, indicating that the chirality at C-12 was R.²¹⁾ Combined with NOESY data (Fig. 3), the absolute configuration of 3 was determined to be 6β-O-cinnamoyloxy-12α-hydroxy-(13) 15-en-16,12-olide-18-cassaneol, as shown in Fig. 1.

Echinalide D (4) was isolated as a yellow oil, and its molecular formula was assigned as C₂₉H₃₈O₆ based on HR-FAB-MS (m/z 483.2724 [M+H]⁺; Calcd: 483.2746). The IR spectrum indicated the presence of a hydroxyl group (3855 cm⁻¹) and an α,β-unsaturated γ-lactone (1719 cm⁻¹). The UV absorption at 212 nm along with the relevant IR absorption band suggested the presence of an α,β-butenolide ring.²⁵⁾ The ¹H- and ¹³C-NMR spectra were similar to those of 3 except for C-6, where a cinnamoyl ester was replaced with a dihydrocinnamoyl moiety [δ_C 31.2, 36.6; δ_H 2.82 (2H, t, J=8.0 Hz), 3.10 (2H, m)]. These data were supported by ¹H–¹H COSY and HMBC spectra. The configurations of C-4, C-6, C-12 and C-14 were also the same as 3, as confirmed by NOESY experiments. Although the CD curve of 4 did not clearly show a negative Cotton effect, 4 was presumed to be 6β-O-2′,3′-dihydrocinnamoyloxy-12α-hydroxy-(13) 15-en-16,12-olide-18-cassaneol, with the same absolute configuration as 3, based on the co-occurrence of 3 in this plant.

Echinalide E (5) was obtained as a yellow oil with the molecular formula of C₂₉H₃₆O₆ on the basis of HR-FAB-MS (m/z 481.2567 [M+H]⁺; Calcd: 481.2590). The UV absorption at 220 nm in combination with an IR band at 1717 cm⁻¹ suggested the presence of an α,β-butenolide ring. The ¹H- and ¹³C-NMR spectra displayed characteristic signals attributable to an α,β-unsaturated γ-lactone [δ_C 173.6, 110.7, 176.3, 79.0 and δ_H 5.70 (1H, d, J=1.2 Hz)] and a dihydrocinnamoyl moiety [δ_C 172.1, 36.3, 30.8, 140.1, 128.2 (×2), 128.5 (×2), 126.4 and δ_H 2.64, 2.96, each 2H, td, J=7.2, 3.2 Hz] in addition to signals corresponding to a secondary methyl, two tertiary methyls and two oxymethines. A total of twenty-nine carbon signals were present in the ¹³C-NMR spectrum, including a signal for a carboxylic carbonyl carbon as along with the afore mentioned signals. Thus, 5 was considered to be cassane diterpene lactone similar to 4 with a dihydrocinnamoyl group, except for the presence of an oxymethine at C-12 [δ_H 4.83 (1H, dd, J=11.6, 6.4 Hz), δ_C 79.0]. The 2D-NMR spectra allowed the assignment of all proton and carbon signals, as shown in Tables 1 and 2, and the location of the substituents was determined from the HMBC spectrum (Fig. 2). The carboxylic acid group was located at C-4 based on the correlation of H-19/C-3, C-4, C-5, C-18 and H-5/C-4, C-18, C-19 and the dihydrocinnamoyl group was located at C-6 based on the correlations of H-6/C-1′, C-8, C-10. The position of the α,β-unsaturated γ-lactone was deduced from the correlations of H-15/C-12, C-13, C-14, C-16 and H-12/C-11, C-13, and the position of the secondary methyl (Me-17) was deduced from the correlations of H-17/C-8, C-13, C-14. Regarding the relative configuration, NOESY cross peaks between H-20/H-8, H-19 and H-5/H-6 suggested that the carboxylic acid group was α-equatorial and the dihydrocinnamoyl group was β-axial, whereas the correlations of H-14/H-8, H-15 and H-12/H-9, H-17 indicated that H-12 and Me-17 were in the α-orientation. The absolute configuration of 5 was deduced from the CD curve.^21,24) The strong negative Cotton effect for the γ-lactone chromophore at λ_max 225 nm in EtOH indicated that the chirality at C-12 was R. Therefore, the structure of 5 was established to be 6β-O-2′,3′-dihydrocinnamoyloxy-(13) 15-en-16, 12-olide-18-cassaneoic acid. Cassane diterpene lactones deoxygenated at C-12 appear to be rare in nature.

Echinalide F (6) was isolated as a yellow oil, and its molecular formula of C₂₉H₃₄O₆ was established by HR-FAB-MS (m/z 479.2420 [M+H]⁺; Calcd: 479.2433). The IR spectrum supported the presence of a conjugated double bond and an α,β-unsaturated γ-lactone; the presence of the latter functional group was also supported by the UV absorption at 216 nm. The ¹H- and ¹³C-NMR spectra were similar to those of 5 except that the dihydrocinnamoyl ester was replaced with a cinnamoyl moiety [δ_H 6.37, 7.66 (each 1H, d, J=16.0 Hz), H-2′, H-3′; δ_C 118.2, 145.4] at the C-6 position. These data were fully supported by the HMBC spectrum, and the complete assignment of all protons and carbons is shown in Tables 1 and 2. The configurations of C-4, C-6, C-12 and C-14 were the same as 5, as confirmed by NOESY experiments. Thus, compound 6 was determined to be 6β-O-cinnamoyloxy-(13) 15-en-16, 12-olide-18-cassaneoic acid.

Echinalide G (7), a yellow oil, had a molecular formula of C₂₉H₃₈O₅ as determined by HR-FAB-MS (m/z 467.2820 [M+H]⁺; Calcd: 467.2797). The IR spectrum displayed absorption bands attributed to a hydroxyl group (3398 cm⁻¹) and an α,β-unsaturated γ-lactone (1733 cm⁻¹) group, and the UV absorption at 218 nm suggested the presence of an α,β-butenolide ring. The ¹H- and ¹³C-NMR spectra revealed the same cassane-butenolide skeleton as 5, which contained a dihydrocinnamoyl moiety at the C-6 position. The major difference was the replacement of the carboxylic group by a hydroxymethylene group, as shown by two coupled doublet signals at δ_H 3.09 (J=10.8 Hz) and 3.54 (J=10.8 Hz) and an oxygenated carbon signal at δ_C 71.1. The hydroxymethylene group was located at C-4 as supported by the HMBC correlations of H-18 with C-3, C-4, C-5 and C-19, and its configuration was α-equatorial on the basis of the observed NOESY correlations between H-6/H-5, H-18 and H-19/H-20. Thus, 7 was determined to be 6β-O-2′,3′-dihydrocinnamoyloxy-(13) 15-en-16,12-olide-18-cassaneol.

The absolute stereochemistries of 6 and 7 were also deduced from the CD curves of their γ-lactone chromophores. The signals of both CD curves in EtOH [6: λ_max 225 nm; 7: λ_max 223 nm] were negative, indicating that the chiralities at C-12 for 6 and 7 were those shown in Fig. 1.

The NF-κB inhibitory effects of echinalides A–G (1–7) on TNF-α regulated gene expression were evaluated using HeLa_NF-κB-3 cells that expressed the secreted alkaline phosphatase (SEAP) reporter enzyme under the control of the NF-κB responsive promoter. Echinalides A–G (1–7) were subjected to this assay system at 20 µM and 5 µM, and the value was determined from at least three individual experiments. Echinalides C (3) and D (4) showed potent activities [for C (3), 20 µM: 100%, 5 µM: 70% inhibition and for D (4), 20 µM: 100%, 5 µM: 78% inhibition] nearly equivalent to that of parthenolide²⁶⁾ [5 µM: 98%], a known potent inhibitor of NF-κB; the inhibitory activities of echinalides E (5), F (6) and G (7) [for E (5), 20 µM: 100%, 5 µM: 32%, for F (6), 20 µM: 100%, 5 µM: 16%, for G (7), 20 µM: 100%, 5.0 µM: 27% inhibition] were weaker than that of parthenolide. In contrast, inhibitory activity was not observed for echinalides A (1) and B (2) [for A (1); 20 µM: 0%, for B (2), 20 µM: 23% inhibition]. The active compounds (C, D, E, F and G) have an α,β-unsaturated γ-lactone moiety as the common substructure, suggesting that the presence of an α,β-unsaturated γ-lactone plays a crucial role in the inhibitory activities of these compounds, similar to observations with parthenolide. These results are also consistent with previous reports of culcitiolide derivatives.^15,16) In particular, the γ-hemiketal structures of culcitiolides C and D were promoting its inhibitory activities.

This is the first report to demonstrate that cassane diterpenoids show potent inhibitory activities against NF-κB. These findings thus provide new lead compounds for anti-inflammatory drugs. The mechanism of inhibition and the structure–activity relationships of these diterpenoids are now under investigation.

Experimental

General Procedures

Optical rotations were measured on a JASCO DIP-140 polarimeter. CD spectra were recorded on a JASCO J-820 spectropolarimeter. IR and UV spectra were recorded on HORIBA FT-720 and HITACHI U-2810 spectrometers, respectively. HR-FAB-MS spectra were taken on a JEOL JMS 700MS spectrometer. NMR spectra were recorded on a JEOL ECS400 NMR spectrometer using tetramethylsilane (TMS) as an internal standard. Silica gel, 230–400 mesh (Merck, silica gel 60), and octadecylsilica (ODS), 100–200 mesh (Fuji Silisia, ODS), were used for column chromatography and medium-pressure liquid chromatography, respectively. 3-Chloroperoxybenzoic acid (mCPBA) was purchased from Aldrich Chemical Co., Inc.

Plant Material

Samples of C. echinata LAM. were collected in Amazonas state, Brazil in March 2005 and were identified by Dr. Goro Hashimoto, Centro de Pesquica de Historia Natural, São Paulo, Brazil. A voucher specimen (No. OUS-1106) was deposited in the Herbarium of Okayama University of Science.

Extraction and Isolation

The shade-dried plant material (3.8 kg) of Caesalpinia echinata LAM. was extracted with acetone at room temperature for six months and then filtered. The filtrate was concentrated under vacuum to give 60 g of crude residue, which was subjected to silica gel column chromatography eluting with a gradient mixture of chloroform–methanol to afford six fractions (F1 to F6). F3 (22.0 g) was chromatographically separated using silica gel eluting with a mixture of n-hexane–acetone of increasing polarity to yield five fractions (F31 to F35). F32 (14.2 g) was fractionated by silica gel column chromatography with a mixture of benzene–methanol of increasing polarity to yield five fractions (F321 to 325). F323 (8.9 g) was further subjected to column chromatography on ODS-silica gel using a mixture of methanol–H₂O of decreasing polarity to give six fractions (F3231 to 3236). F3235 (4.1 g) was separated by column chromatography on silica gel and eluted with benzene–acetone (10 : 1) to give four fractions (F35351 to F32354). Further purification of F32352 (990.4 mg) was accomplished by column chromatography on ODS-silica gel using methanol–H₂O (18 : 1) to yield echinalides A (1) (164.1 mg) and B (2) (106.7 mg). F324 (1.65 g) was fractionated by silica gel column chromatography with mixture of chroloform–methanol (20 : 1) to give six fractions (F3241 to F3246). Further purification of F3244 (111.0 mg) was achieved with ODS-silica gel column chromatography using methanol–H₂O (4 : 1) to yield compound 1a (12.8 mg). F3233 (1.5 g) was separated by column chromatography on silica gel eluting with benzene–acetone (4 : 1) to give seven fractions (F32331 to F32337). Further purification of F32334 (86.6 mg) was achieved with ODS-silica gel column chromatography using methanol–H₂O (4 : 1) to give echinalides C (3) (6.2 mg) and D (4) (5.8 mg). F32332 (36.5 mg) was purified by column chromatography on ODS-silica gel using methanol–H₂O to yield echinalides E (5) (8.8 mg) and F (6) (8.6 mg). F33 (4.5 g) was separated by column chromatography on silica gel eluting with chloroform–acetone (10 : 1) to yield eight fractions (F331 to F338). F336 (221.0 mg) was, further subjected to column chromatography on ODS-silica gel using methanol–H₂O (3 : 1) to afford six fractions (F3361 to F3366). F3365 (19.6 mg) was purified by silica gel column chromatography eluting with benzene–acetone (7 : 1) to give echinalide G (7) (8.6 mg).

Assay for NF-κB Inhibitory Activity

The NF-κB inhibitory activities of echinalides A–G were evaluated using HeLa_NF-κB-3 cells as described previously.²⁶⁾ In brief, 0.5×10⁴ cells were seeded in a 96-well tissue culture plate. After 2 h, an aliquot of a test sample solution was added to each well. After 1 h, 10 µL of 800 ng/mL TNF-α, which induces inflammation, was added to each well. After 24 h, the medium from each well was removed and assayed for secreted alkaline phosphatase (SEAP) activity according to the manufacturer’s protocol. The NF-κB inhibitory activity was defined as the percent inhibition (%) of TNF-α-induced expression of the SEAP reporter gene under the control of NF-κB transcriptional activity. The inhibitory data are presented as the average (n=3) of three independent experiments. The S.D. of each population was lower than 5% of the average value. Parthenolide (1 µM), a known NF-κB inhibitor,²⁷⁾ was used as a positive control.

Echinalide A (1): Yellow oil. [α]_D²⁰ +41° (c=0.81, CHCl₃). IR (neat) cm⁻¹: 2924, 2853, 1720, 1708, 1692, 1627, 1202, 1148. UV λ_max (MeOH) nm (log ε): 278 (4.05), 217 (4.03). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₅O₅: 463.2484; Found 463.2479.

Echinalide B (2): Yellow oil. [α]_D²⁰ +3° (c=0.72, CHCl₃). IR (neat) cm⁻¹: 2899, 2359, 2342, 1718, 1698, 1599, 1213, 1152. UV λ_max (MeOH) nm (log ε): 279 (4.23), 212 (3.98). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₇O₅: 465.2641; Found 465.2657.

Echinalide C (3): Yellow oil. [α]_D²⁰ −5° (c=0.41, CHCl₃). IR (neat) cm⁻¹: 3693, 2925, 2844, 1739, 1692, 1638, 1210. UV λ_max (MeOH) nm (log ε): 278 (4.24), 217 (4.23). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. CD (EtOH) : [θ]₂₇₄=+8030° cm²/dmol, [θ]₂₂₆=−16480° cm²/dmol. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₇O₆: 481.2590; Found 481.2579.

Echinalide D (4): Yellow oil. [α]_D²⁰ −67° (c=0.17, CHCl₃). IR (neat) cm⁻¹: 3855, 2923, 2854, 1719, 1701, 1686, 1637, 1210. UV λ_max (MeOH) nm (log ε): 276 (3.63), 212 (4.06). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₉O₆: 483.2746; Found 483.2724.

Echinalide E (5): Yellow oil. [α]_D²⁰ −29° (c=0.97, CHCl₃). IR (neat) cm⁻¹: 3626, 2925, 2849, 1717, 1700, 1458, 1396. UV λ_max (MeOH) nm (log ε): 267 (3.55), 220 (2.62). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. CD (EtOH) : [θ]₂₂₅=−15273° cm²/dmol. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₇O₆: 481.2590; Found 481.2567.

Echinalide F (6): yellow oil. [α]_D²⁰ −63° (c=0.54, CHCl₃). IR (neat) cm⁻¹: 3630, 2929, 2852, 1709, 1687, 1459, 1395. UV λ_max (MeOH) nm (log ε): 277 (4.26), 216 (4.41). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. CD (EtOH) : [θ]₂₇₁=+4555° cm²/dmol, [θ]₂₂₅=−37531° cm²/dmol. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₅O₆: 479.2433; Found 479.2420.

Echinalide G (7): Yellow oil. [α]_D²⁰ −49° (c=0.50, CHCl₃). IR (neat) cm⁻¹: 3398, 2924, 2849, 1733, 1717, 1699, 1684, 1637, 1253. UV λ_max (MeOH) nm (log ε): 278 (3.18), 218 (3.81). ¹H-NMR: Table 1. ¹³C-NMR: Table 2. CD (EtOH) : [θ]₂₂₃=−17858° cm²/dmol. HR-FAB-MS: m/z [M+H]⁺ Calcd for C₂₉H₃₉O₅: 467.2797; Found 467.2820.

Preparation of 1a from 1

To a solution of 1 (4.0 mg) in benzene (1.0 mL) was added mCPBA (13.0 mg), and the mixture was stirred at room temperature for 4 h. The resulting solution was poured into saturated aqueous Na₂S₂O₃ and NaHCO₃ and extracted with ethyl acetate. The organic layer was washed with brine, dried over Na₂SO₄, and concentrated under reduced pressure. The residue was purified by silica gel column chromatography (benzene : ethyl acetate=5 : 1) to give 1a (2.1 mg, yield 34.5%) as a yellow oil. [α]_D²⁰ −63° (c=0.11, CHCl₃). ¹H-NMR (400 MHz in CDCl₃) δ: 1.20 (3H, d, J=7.2 Hz), 1.25 (1H, m), 1.27 (3H, s), 1.40 (3H, s), 1.48 (1H, m), 1.66 (1H, m), 1.70–1.77 (4H, m), 1.83–1.89 (2H, m), 1.96 (1H, m), 2.09 (1H, br s), 2.47 (1H, dd, J=13.0, 3.0 Hz), 2.93 (1H, dd, J=6.8, 5.2 Hz), 5.30 (1H, br s), 5.72 (1H, s), 6.38 (1H, d, J=16.0 Hz), 7.38–7.40 (3H, m), 7.51–7.53 (2H, m). ¹³C-NMR (100 MHz in CDCl₃) δ: 13.0, 17.9, 18.4, 35.2, 35.9, 36.1, 37.2, 37.7, 39.0, 41.1, 45.2, 47.4, 49.4, 72.1, 105.2, 113.7, 118.3, 128.1 (×2), 128.9 (×2), 130.5, 131.0, 134.2, 145.3, 166.3, 170.1, 172.2, 181.6. CD (EtOH): [θ]₂₂₅=−10891° cm²/dmol.

References

• 1)  Hashimoto G., Nishimoto S., “Illustrated Cyclopedia of Brazilian Medical Plant,” Aboc-sha & Co., Ltd., Kanagawa, 2002, pp. 645–646.
• 2)  de Oliveira L. F. C., Edwards H. G. M., Velozo E. S., Nesbitt M., Vib. Spectrosc., 28, 243–249 (2002).
• 3)  Matsunaga M., Sakai K., Kamitakahara H., Minato K., Nakatsubo F., J. Wood Sci., 46, 253–257 (2000).
• 4)  Matsunaga M., Sugiyama M., Minato K., Norimoto M., Holzforschung, 50, 511–517 (1996).
• 5)  Manhita A., Ferreira T., Candeias A., Barrocas Dias C., Anal. Bioanal. Chem., 400, 1501–1514 (2011).
• 6)  Ferreira E. S., Hulme A. N., McNab H., Quye A., Chem. Soc. Rev., 33, 329–336 (2004).
• 7)  Gilbert (nee stoker) K. G., Cooke D. T., Plant Growth Regul., 34, 57–69 (2001).
• 8)  Hikino H., Taguchi T., Fujimura H., Hiramatsu Y., Planta Med., 31, 214–220 (1977).
• 9)  Jiang R.-W., But P. P.-H., Ma S.-C., Ye W.-C., Chan S.-P., Mak T. C. W., Tetrahedron Lett., 43, 2415–2418 (2002).
• 10)  Kuria K. A., De Coster S., Muriuki G., Masengo W., Kibwage I., Hoogmartens J., Laekeman G. M., J. Ethnopharmacol., 74, 141–148 (2001).
• 11)  Cruz-Silva I., Gozzo A. J., Nunes V. A., Carmona A. K., Faljoni-Alario A., Oliva M. L., Sampaio M. U., Sampaio C. A., Araujo M. S., Biol. Chem., 385, 1083–1086 (2004).
• 12)  Cota B. B., de Oliveira D. M., de Siqueira E. P., Souza-Fagundes E. M., Pimenta A. M. C., Santos D. M., Rabello A., Zani C. L., Fitoterapia, 82, 969–975 (2011).
• 13)  Ghosh S., May M. J., Kopp E. B., Annu. Rev. Immunol., 16, 225–260 (1998).
• 14)  Baldwin A. S. Jr., Annu. Rev. Immunol., 14, 649–681 (1996).
• 15)  Nozaki H., Hayashi K., Kawai M., Mitsui T., Kido M., Tani H., Takaoka D., Uno H., Ohira S., Kuboki A., Matsuura N., Nat. Prod. Commun., 7, 427–430 (2012).
• 16)  Mitsui T., Hayashi K., Kawai M., Kido M., Tani H., Takaoka D., Matsuura N., Nozaki H., Chem. Pharm. Bull., 61, 816–822 (2013).
• 17)  Kuroda C., Ueshino T., Nagano H., Bull. Chem. Soc. Jpn., 77, 1737–1740 (2004).
• 18)  Kitagawa I., Simanjuntak P., Watano T., Shibuya H., Fujii S., Yamagata Y., Kobayashi M., Chem. Pharm. Bull., 42, 1798–1802 (1994).
• 19)  Jiang R. W., But P. P., Ma S. C., Mak T. C., Phytochemistry, 57, 517–521 (2001).
• 20)  Matsuno Y., Deguchi J., Hirasawa Y., Ohyama K., Toyoda H., Hirobe C., Ekasari W., Widyawaruyanti A., Zaini N. C., Morita H., Bioorg. Med. Chem. Lett., 18, 3774–3777 (2008).
• 21)  Stöcklin W., Waddell T. G., Geissman T. A., Tetrahedron, 26, 2397–2409 (1970).
• 22)  Bodensieck A., Fabian W. M., Kunert O., Belaj F., Jahangir S., Schuhly W., Bauer R., Chirality, 22, 308–319 (2010).
• 23)  Saito Y., Ichihara M., Okamoto Y., Gong X., Kuroda C., Tori M., Tetrahedron Lett., 52, 6388–6391 (2011).
• 24)  Pranithanchai W., Karalai C., Ponglimanont C., Subhadhirasakul S., Chantrapromma K., Phytochemistry, 70, 300–304 (2009).
• 25)  Kinoshita T., Haga Y., Narimatsu S., Shimada M., Goda Y., Chem. Pharm. Bull., 53, 717–720 (2005).
• 26)  Matsuura N., Murakami R., Katayama T., Hibino S., Choshi T., Yamada M., J. Health Sci., 55, 311–313 (2009).
• 27)  Bork P. M., Schmitz M. L., Kuhnt M., Escher C., Heinrich M., FEBS Lett., 402, 85–90 (1997).