Six New Phragmalin Limonoids from the Stems of Chukrasia tabularis A. Juss

Six new phragmalin limonoids, named moluccensin Z1 (1), moluccensin Z2 (2), carapanolide Y (3), tabulalin N (4), chukvelutilide A1 (5), and velutinasin J (6), as well as two known compounds, chukvelutilide A (7) and velutinasin D (8) were isolated from the stems of Chukrasia tabularis A. Juss. The structures of the new compounds 1–6 were confirmed by spectroscopic methods, including IR and HRESIMS, as well as 1D and 2D NMR, and by comparisons with the data of known analogues. All compounds were tested for α-glucosidase and acetylcholinesterase inhibitory activities. However, none of the compounds was active against α-glucosidase and acetylcholinesterase in vitro.

In our previous study, some phragmalin limonoids such as chukbularisin B-E isolated from the big polar part of EtOAc-soluble extract of C. tabularis significantly inhibited the α-glucosidase [19]. As part of our investigation towards limonoids with novel structures, we continued to study on the small polar part of EtOAc-soluble extract of Chukrasia tabularis A. Juss, which afforded six new phragmalin limonoids, named moluccensin Z1 (1), moluccensin Z2 (2), carapanolide Y (3), tabulalin N (4), chukvelutilide A1 (5), and velutinasin J (6), together with two known compounds chukvelutilide A (7) and velutinasin D (8) (Figure 1). Compounds 1-8 were evaluated for the inhibitory effects on α-glucosidase and acetylcholinesterase. In this paper, the isolation, structural elucidation as well as the evaluations focused on the α-glucosidase and acetylcholinesterase inhibitory effects of eight limonoids from the stems of C. tabularis are described. elucidation as well as the evaluations focused on the α-glucosidase and acetylcholinesterase inhibitory effects of eight limonoids from the stems of C. tabularis are described.

Results and Discussion
Compound 1, a white amorphous powder, had a molecular formula of C34H40O15 as determined by the HRESIMS ion at m/z 711.2274 ([M + Na] + calcd. 711.2259), corresponding to 15 degrees of unsaturation. The IR absorptions showed the presence of hydroxy group (3528 cm −1 ) and carbonyl group (1731 cm −1 ). The 1 H-NMR (Table 1), 13 C-NMR (Table 2) along with the HSQC data of 1 revealed the presence of two methoxy groups, two acetoxy groups, three ester carbons, four methyls, four methylenes, seven methines with four oxygenated, and ten quarternary carbons (two olefinic and four oxygenated). These data were similar to those of moluccensin Y [20], suggesting that compound 1 was also an 8,9,30-phragmalin ortho ester. The main differences between them were the presence of a lactone carbonyl (δC 169.0), a methoxy (δH 3.56; δC 57.3) and an acetal methine (δH 5.85; δC 103.8) signals and the absence of two olefinic methine signals in 1 compared to moluccensin Y. HMBC correlations between 21-OMe/C-21, H-21/C-20, H-21/C-22, H-21/C-23, H-17/C-20, H-17/C-21, and H-17/C-22 indicated that a β-furyl ring moiety located at C-17 in moluccensin Y was replaced by a 21methoxy-20(22)-en-21,23-γ-lactone moiety in 1. The remaining substructure was determined to be the same as moluccensin Y based on the 2D NMR data as shown in Figure 2. The nearly identical chemical shifts and J-values suggested that compound 1 and moluccensin Y shared the same relative configuration.  Figure 3). Therefore, the structure of 1, named moluccensin Z1, was established as shown.   (Table 2) along with the HSQC data of 1 revealed the presence of two methoxy groups, two acetoxy groups, three ester carbons, four methyls, four methylenes, seven methines with four oxygenated, and ten quarternary carbons (two olefinic and four oxygenated). These data were similar to those of moluccensin Y [20], suggesting that compound 1 was also an 8,9,30-phragmalin ortho ester. The main differences between them were the presence of a lactone carbonyl (δ C 169.0), a methoxy (δ H 3.56; δ C 57.3) and an acetal methine (δ H 5.85; δ C 103.8) signals and the absence of two olefinic methine signals in 1 compared to moluccensin Y.    Figure 2). The relative configuration of 2 was assigned to be the same as that of 1 based on the explanation of ROESY correlations ( Figure 3). Thus, the structure of 2, named moluccensin Z2, was elucidated as shown.    Figure 2). The relative configuration of 2 was assigned to be the same as that of 1 based on the explanation of ROESY correlations ( Figure 3). Thus, the structure of 2, named moluccensin Z2, was elucidated as shown.   Figure 2). The relative configuration of 2 was assigned to be the same as that of 1 based on the explanation of ROESY correlations ( Figure 3). Thus, the structure of 2, named moluccensin Z2, was elucidated as shown.   (Tables 2 and 3) were similar to those of tabulalin C [22]. Compared with tabulalin C, 4 had three acetoxy groups, which replaced 2-OH, 3-OH and H-11 in tabulalin C, respectively, and lacked an acetoxy group at C-19. The methyl at C-19 was confirmed by the HMBC correlations between H-19/C-10, H-19/C-5 and H-19/C-9. The acetoxy groups at C-2, C-3 and C-11 were revealed by the HMBC correlations from H-2, H-3 and H-11 to the corresponding carbonyl of the acetoxy group (Figure 2). The relative configuration of 4 was established to be the same as these of tabulalin C based on the explanation of ROESY correlations (Figure 3). Thus, the structure of 4 was assigned as depicted and it was named tabulalin N.  (Tables 2 and 3) showed great similarity to those of chukvelutilide A [11]. The only difference was the replacement of the 12-O-acetyl group in chukvelutilide A by the 12-O-propionyl group in 5, which was further confirmed by HMBC and 1 H-1 H COSY correlations as depicted in Figure 2. The relative configuration of 5 was established to be the same as that of chukvelutilide A by the ROESY spectrum ( Figure 3). Therefore, the structure of 5 was elucidated and it was named chukvelutilide.  (Tables 2 and 3) were similar to those of velutinasin D [23], except for the replacements of the

12-O-isobutyryl group and 2-OH in velutinasin D by the 12-O-acetyl and 2-O-acetyl in 6.
The acetoxy at C-12 (δ C 69.6) was revealed by the HMBC correlations from H-12 (δ H 4.72) to the corresponding carbonyl of the acetoxy group. Similarly, the acetoxy at C-2 was confirmed by the HMBC correlations ( Figure 2). The relative configuration of 6 was established to be the same as that of velutinasin D by the ROESY spectrum (Figure 3). Thus, the structure of 6, named velutinasin J, was elucidated as shown. Two known compounds were identified as chukvelutilide A (7) [11] and velutinasin D (8) [23], respectively, by interpreting their NMR data and making comparisons with literature values. More details about the original spectra of NMR, IR and HRESIMS data for the new compounds 1-6 see Figures S1-S48 of the supplementary materials.
All the compounds were tested for the α-glucosidase and acetylcholinesterase inhibition activities according to the method of Li [24] and Xiang [25]. There was no obvious inhibition effect on α-glucosidase and acetylcholinesterase. Previous research showed that the EtOAc-soluble extract of C. tabularis and some phragmalin limonoids which were isolated from it had significant α-glucosidase inhibitory activity [19]. Compare the chemical structures between the previously isolated limonoids with significant α-glucosidase inhibitory activities and the newly isolated compounds, a quinary lactone ring instead of a β-furyl ring located at C-17 in compound 1 and 2, D-rings were opened and an acetoxy group was connected to C-17 in compounds 4-8. These differences of chemical structures might be the reason for missing the α-glucosidase inhibitory activity of the newly isolated compounds, and were consistent with the result of our previous study [19].

General Procedures
The NMR spectra were recorded with a Bruker AV III spectrometer (Bruker, Bremen, Germany) using TMS as an internal standard. Optical rotations were measured on an MCP 5100 polarimeter (Anton Paar, Graz, Austria). The infrared spectra were recorded with a Nicolet 380 FT-IR spectrometer (Thermo, Pittsburgh, PA, USA). UV spectra were recorded on a Shimadzu UV2550 spectrophotometer (Shimadzu, Kyoto, Japan). The mass spectrometric (HRESIMS) data were acquired using an API QSTAR Pulsar mass spectrometer (Bruker, Bremen, Germany). Melting points were obtained with an apparatus of Beijing Taike X-5 (Beijing Taike Instrument

Plant Material
The stems of C. tabularis were collected from Haikou, Hainan Province, P.R. China, in July 2014, and identified by Dr. Jun Wang. A voucher sample (No. 20140726) was deposited at the Institute of Tropical Bioscience and Biotechnology, Chinese Academy of Tropical Agriculture Science.

Extraction and Isolation
The dried stems of C. tabularis (110.0 kg) were pulverized and extracted three times with 95% ethanol (314 L) at room temperature. The extract was concentrated under reduced pressure to afford a crude extract (13.7 kg), followed by suspension in H 2 O and extraction with petroleum ether, EtOAc, and n-BuOH successively. Then, the extract solutions were evaporated to dryness under reduced pressure separately to get the petroleum ether extract (30.0 g), EtOAc extract (1700.0 g) and n-BuOH extract (800.0 g). The EtOAc extract (1700.0 g) was chromatographed on silica gel eluted with a petroleum ether-EtOAc system (20:1 to 0:1, v/v) to yield 18 fractions. Fr.15 (220.0 g) was further chromatographed on silica gel eluted with CHCl 3 -MeOH (50:1, v/v) to yield one fraction (90.