Two New Phragmalin-Type Limonoids from Chukrasia tabularis var. velutina

Two new phragmalin-type limonoids with different structural skeletons, chuktabrin K (1) and tabulalin J (2), were isolated from the stem barks of Chukrasia tabularis var. velutina in the course of our ongoing research work in this area. Compound 1 was a 16-norphragmalin with an enolic alkyl appendage at C-15, and the carbonate moiety in 1 was also rare in natural organic molecules. The basic skeleton of compound 2 was a D-ring-opened phragmalin. Their structures were elucidated on HR-ESI-MS, 1H and 13C-NMR, HSQC, HMBC, and ROESY experiments.


Results and Discussion
Chuktabrin K (1) was isolated as a white amorphous powder, and its molecular formula was determined as C 31  Characteristic 1D-NMR spectra, i.e., three upfield proton signals at δ H 6.42, 7.59, and 7.68 and a set of double proton signals at δ H 1.92 and 1.46 with an 11.0 Hz coupling constant in the 1 H-NMR (Table 1), and four olefinic carbons at δ C 109.9, 121.5, 141.1, and 143.7 in the 13 C-NMR (Table 1), indicated that compound 1 was a phragmalin-type limonoid possessing an α,β-substituted furan ring and a 4,29,1-bridge moiety [8]. The presence of a carbon signal at  C 92.4 showing a HSQC correlation with the proton signal at  H 4.67 and two down-field carbon signals at  C 152.0 and 152.8 suggested that compound 1 was a 16-norphragmalin with an enolic alkyl appendage at C-15 and a characteristic carbonate moiety like chuktabrins C-H [8]. The obvious HMBC correlations ( Figure 2a) from a set of ethyl proton signals [ H 2.10, q, (7.5), 2H;  H 1.04, t (7.5), 3H] and proton signal of H-15 ( H 4.67) to a carbon signal at  C 152.8 (C-1') indicated that a propionyl group was attached at C-15 biosynthetically as in the chuktabrins C-H [8].
Comparison of the NMR data between 1 and chuktabrin H indicated that the former was a deacetyl derivative of the latter, which was also confirmed from the molecular formula by the absence of one C 2 H 2 O unit. An obvious HMBC correlation (Figure 2a) from H-17 (δ H 5.79) to the 13 C signal for an acetoxyl group (MeCOOR) at δ C 167.9 suggested that the only acetyl group was located at OH-17. Hitherto, the planar structure of compound 1 was determined, except for the location of the carbonate moiety and the ether linkage of C-1' due to a lack of direct HMBC evidence. Hydroxyl groups must be connected at C-1, C-2, C-3, C-11 and C-12 due to the observed correlations between the 1-OH signal (δ H 5.35) to the 13 C signal for C-1 at δ C 82.3 and C-2 at δ C 74.0; 2-OH (δ H 4.23) to C-1 at δ C 82.3, C-3 at δ C 85.7 and C-30 at δ C 65.7; 3-OH (δ H 5.92) to C-2 at δ C 74.0, C-3 at δ C 85.7 and C-4 at δ C 43.1; 11-OH (δ H 5.67) to C-11 at δ C 67.3 and C-9 at δ C 86.6; 12-OH (δ H 5.14) to C-11 at δ C 67.3, C-12 at δ C 73.4 and C-13 at δ C 43.9. The 1 H signal for H-30 at δ H 4.57 (δ C 65.7) showed correlation to the 13 C signal at δ C 83.2 (C-8). Thus, these correlation required the presence of -OR substituent at C-8, C-9 and C-30. The 13 C-NMR data for C-30, C-9 and C-8 were similar to those for chuktabrin C, which was determined by single-crystal X-ray diffraction [8], determining the position of enol ether at C-30 and the carbonate moiety at C-8 and C-9. , and H-29b with H-19b, indicated that the relative stereochemistry of the key asymmetric carbons of 1 was well matched with those of chuktabrin C obtained by X-ray crystallography [8]. Thus, the structure of 1 was established as shown in Figure 1, namely as a 12-deacetyl derivate of chuktabrin H [8].  The similarity of the 1 H and 13 C-NMR spectroscopic data of 2 ( Table 1) to those of tabulalin A, a phragmalin-type limonoid isolated in our previous research [13], indicated that these two molecules possessed the same carbon framework. Obvious HMBC correlations (Figure 4a) from H-17 (δ H 6.00) to the 13 C signal for the acetoxyl group (MeCOOR) at  C 168.6 and H-30 (δ H 4.98) to C-16 ( C 168.9) indicated that compound 2 possesses the same phragmalin skeleton with a C-16/C-30 δ-lactone ring as tabulalin A [13]. Comparison of the NMR data and molecular formula suggested that 2 was a bisacetyl derivative of tabulalin A [13]. A significant downfield shift for H-11 ( H 5.11) and H-12 (δ H 4.96), when compared with tabulalin A, determined the position of the acetoxyl groups at C-11 and C-12. In comparison with the 13 C-NMR spectrum of the parent compound tabulalin A (C-9  C 79.3, C-11  C 67.7, C-12  C 76.0), the acetylation of hydroxyl groups at C-11 and C-12 resulted in 2.9 ppm upfield shifts for the C-9 and 5.2 ppm for C-12 resonances, and 3.5 ppm downfield ships for C-11 position, confirming the acetoxyl groups at C-11 and C-12. Thus, the planar structure of 2 was determined. The relative configuration of 2 was determined to be the same as tabulalide A [13]

Plant Material
The

Extraction and Isolation
The air-dried stem bark (10 kg) was extracted by refluxing with 95% ethanol (40 L) three times. The EtOH extract was concentrated under reduced pressure (2,000 g) and then extracted with CHCl 3 to give a chloroform extract (300 g). The oily chloroform extract was dissolved in 2 L MeOH/H 2 O (50:50, v/v) and then extracted with petroleum ether (6 L, 60-90 °C, ×3). After removal of the fatty components, 210 g of extract were obtained, which was subjected to silica gel column chromatography eluting with CHCl 3 /MeOH in a gradient from 1:0 to 1:2 to afford eight fractions (Fractions A-H). Fraction E (20 g) was chromatographed on a column of reversed-phase C 18 silica gel eluted with MeOH/H 2 O (4:6 to 7:3) to give six sub-fractions (Fractions E1-6). Fraction E6 (7 g) was chromatographed on a column of reversed-phase C 18 silica gel eluted with MeOH/H 2 O (2:3 to 7:3) to give four sub-fractions (Fractions E6a-d), Fraction E6c was separated by prep-HPLC using MeOH/H 2 O (55:45) as the mobile phase to give compound 2 (3 mg). Fraction F (13 g) was chromatographed on a column of silica gel eluted successively with a gradient of petroleum ether/EtOAc (1:1 to 1:4) to give four sub-fractions (Fractions F1-4). Fraction F3 was chromatographed on a column of reversed-phase C 18 silica gel eluted with MeOH/H 2 O (2:3 to 7:3) to give four sub-fractions (Fractions F3a-d). Fraction F2a was separated by preparative HPLC using CH 3 OH/H 2 O (52:48, 10 mL/min) as the mobile phase to give 1 (4 mg).

Conclusions
Two new phragmalin-type limonoids with different structure skeletons, chuktabrin K (1) with a 16-norphragmalin skeleton and tabulalin J (2) with a normal phragmalin skeleton, were isolated from the stem barks of Chukrasia tabularis var. velutina. The complex and new structures of these two phragmalins represent new additions to the molecular diversity of natural organic limonoid molecules.