Anti-Complementary Components of Helicteres angustifolia

A first phenalenon derivative with an acetyl side chain at C-8, 8-acetyl-9-hydroxy-3-methoxy-7-methyl-1-phenalenon (compound 1), and a pair of new sesquilignan epimers at C-7″ of hedyotol C and hedyotol D analogs, hedyotol C 7″-O-β-d-glucopyranoside (compound 2) and hedyotol D 7″-O-β-d-glucopyranoside (compound 3) were isolated from the aerial parts of Helicteres angustifolia together with nine known compounds (4–12). Their structures were elucidated on the basis of spectroscopic methods, including mass spectroscopy, and 1D and 2D nuclear magnetic resonance. Eleven isolates exhibited anti-complementary activity. In particular, compounds 4 and 5 exhibited potent anti-complementary activities against the classical and alternative pathways with CH50 values of 0.040 ± 0.009 and 0.009 ± 0.002 mM, and AP50 values of 0.105 ± 0.015 and 0.021 ± 0.003 mM, respectively. The targets of compounds 4 and 5 in the complement activation cascade were also identified. In conclusion, the anti-complementary components of H. angustifolia possessed chemical diversity and consisted mostly of flavonoids and lignans in this study.


Introduction
The complement system, as a chief component of innate immunity, plays a significant role in host immune defense against infection and in the clearance of antigen-antibody complexes from the bloodstream. It can be activated by a cascade mechanism through the classical pathway (CP), alternative pathway (AP) or the lectin pathway [1]. Generally, the normal activation of the complement system leads to elimination of invading pathogens and optimal host response. However, excessive activation of the complement system may induce tissue damage, inflammation and a number of pathological situations such as systemic lupus erythematosus, rheumatoid arthritis and acute respiratory distress syndrome [2,3]. Thus, inhibition of the complement system is potentially therapeutic in diseases resulting from uncontrolled or overshooting complement activation. The desirable properties of a valuable anti-complementary therapeutic agent are that it should be inexpensive and highly specific, either having a long plasma half-life or being active orally [2]. However, none of the currently available agents meet all these criteria. A wide array of specific complement inhibitors, for instance cobra venom factor (CVF), synthetic nafamastat mesilate (FUT-175), a recombinant humanized antibody fragment, and a monoclonal antibody, Eculizumab, have been developed to target various components of the complement cascade [4][5][6][7][8]. However, the clinical development of this plethora of complement therapeutics has exhibited some side effects. Therefore, there is an urgent need to search for naturally occurring and low-toxicity anti-complementary agents from traditional Chinese medicines.
The dried roots or whole plants of Helicteres angustifolia L. (Sterculiaceae) have been used in Chinese folk medicine as anti-inflammatory, antidotal, analgesic, anti-bacterial and anti-cancer agents [9]. Previous phytochemical studies revealed that this plant mainly contains triterpenoids, penylpropanoids, quinones, lignans and flavonoids [10][11][12]. It is commonly accepted that the complement system is a crucial trigger for inflammation [13,14]. Therefore, one may hypothesize that complement inhibition intercepts the process of complement-dependent inflammation diseases. In our effort to search for anti-complementary agents from Chinese herb medicines and to clarify their anti-complementary constituents [3,15,16], an ethanolic extract of the aerial parts of H. angustifolia was found to show a potent anti-complementary activity, which encouraged us to investigate its anti-complementary constituents systematically. In this study, one new skeleton compound (1) and two new compounds (2 and 3) were isolated from the aerial parts of this plant, together with nine known compounds, on the basis of MS, 1D and 2D NMR, as well as comparison with the literature. Herein, the isolation, characterization, and anti-complementary activity of these compounds are reported. In addition, the target identification in the complement activation cascade has been investigated for anti-complementary constituents.

Results and Discussion
In our search for anti-complementary activity compounds from plants, the 95% ethanolic extract of H. angustifolia possessed a potent anti-complementary activity. Thus, the EtOAc-soluble fraction was concentrated under reduced pressure to produce a residue that was subjected to multiple chromatographies. From this active fraction, 12 compounds (1-12) including one new skeleton compound (1) and two new compounds (2 and 3) were isolated and identified.
Compound 1 (Figure 1) was obtained as yellow crystals (CHCl 3 ), and its molecular formula was determined as C 17 (Table 1) showed two carbonyl carbon signals (δ C 204.1 and 181.1) and eight quaternary carbon signals. The 1 H-1 H COSY spectrum revealed the presence of a subunit C-4-C-5-C-6 ( Figure 2). These NMR signals indicated that compound 1 had a phenalenone skeleton as in the known compounds 7-methyl-1-phenalenon [17] and myeloconone A 2 [18]. The HMBC ( Figure 2) correlations from 7-CH 3 (δ H 2.58) to C-7 (δ C 144.1), C-8 (δ C 136.3) and C-6a (δ C 125.4) indicated that the methyl was attached to C-7. Different from 7-methyl-1-phenalenon, compound 1 possessed a methoxyl, an acetyl and a peri-hydroxy group in its structure. The HMBC ( Figure 2 and see Supplementary Materials) correlations from 3-OCH 3 (δ H 3.92) to C-3 (δ C 166.6), 8-COCH 3 (δ H 2.72) to 8-COCH 3 (δ C 204.1) and C-8 (δ C 136.3) further confirmed that the methoxyl was connected to C-3, and the acetyl group was connected to C-8. Consequently, the structure of 1 was deduced as 8-acetyl-9-hydroxy-3-methoxy-7-methyl-1-phenalenon. To our knowledge, 1 is the first phenalenon derivative with an acetyl side chain connected to C-8.   Compound 2 ( Figure 1) was obtained as white amorphous powder, and the molecular formula was determined as C37H46O16 by HR-ESI-MS (m/z 769.2662 [M + Na] + , calcd. for C37H46NaO16 + , 769.2678). The 1 H-and 13 C-NMR spectra ( Table 2) showed characteristic signals for a sesquilignan similar to hedyotol C [19], including eight aromatic protons (two sets of 1,2,4-trisubstituted phenyl rings and a 1,2,3,5-tetrasubstituted phenyl ring), four methoxyl groups, two phenolic and five alcoholic hydroxyl groups, one furan and one glycerol unit. In addition, an anomeric proton resonated at δH 4.67 (1H, d, J = 7.8 Hz) and the large coupling constant indicated β-glucosidic linkage. Detailed inspection of the NMR data revealed that the structure of compound 2 was similar to hedyotol C 4″-O-β-D-glucopyranoside [20], except that the glucosyl unit was connected to C-7″ in 2 rather than to C-4″, as indicated by the HMBC correlations from H-1″′ (δH 4.67) to C-7″ (δC 82.3) and H-7″ (δH 5.15) to C-1″′ (δC 104.4). The nature of sugar in compound 2 was further demonstrated to be D-glucose by GC analysis of an acid-treated hydrolysate, as well as comparison of their 13 C-NMR data with that in the literature [20]. The relative configuration of compound 2 was determined by interpretation of the NOESY spectrum and the chemical shift ratio of C-7″:C-8″ [21]. In the NOESY spectrum of compound 2 (See Figure S13   Compound 2 ( Figure 1) was obtained as white amorphous powder, and the molecular formula was determined as C37H46O16 by HR-ESI-MS (m/z 769.2662 [M + Na] + , calcd. for C37H46NaO16 + , 769.2678). The 1 H-and 13 C-NMR spectra ( Table 2) showed characteristic signals for a sesquilignan similar to hedyotol C [19], including eight aromatic protons (two sets of 1,2,4-trisubstituted phenyl rings and a 1,2,3,5-tetrasubstituted phenyl ring), four methoxyl groups, two phenolic and five alcoholic hydroxyl groups, one furan and one glycerol unit. In addition, an anomeric proton resonated at δH 4.67 (1H, d, J = 7.8 Hz) and the large coupling constant indicated β-glucosidic linkage. Detailed inspection of the NMR data revealed that the structure of compound 2 was similar to hedyotol C 4″-O-β-D-glucopyranoside [20], except that the glucosyl unit was connected to C-7″ in 2 rather than to C-4″, as indicated by the HMBC correlations from H-1″′ (δH 4.67) to C-7″ (δC 82.3) and H-7″ (δH 5.15) to C-1″′ (δC 104.4). The nature of sugar in compound 2 was further demonstrated to be D-glucose by GC analysis of an acid-treated hydrolysate, as well as comparison of their 13 C-NMR data with that in the literature [20]. The relative configuration of compound 2 was determined by interpretation of the NOESY spectrum and the chemical shift ratio of C-7″:C-8″ [21]. In the NOESY spectrum of compound 2 (See Figure S13  769.2678). The 1 H-and 13 C-NMR spectra ( Table 2) showed characteristic signals for a sesquilignan similar to hedyotol C [19], including eight aromatic protons (two sets of 1,2,4-trisubstituted phenyl rings and a 1,2,3,5-tetrasubstituted phenyl ring), four methoxyl groups, two phenolic and five alcoholic hydroxyl groups, one furan and one glycerol unit. In addition, an anomeric proton resonated at δ H 4.67 (1H, d, J = 7.8 Hz) and the large coupling constant indicated β-glucosidic linkage. Detailed inspection of the NMR data revealed that the structure of compound 2 was similar to hedyotol C 4"-O-β-D-glucopyranoside [20], except that the glucosyl unit was connected to C-7" in 2 rather than to C-4", as indicated by the HMBC correlations from H-1" (δ H 4.67) to C-7" (δ C 82.3) and H-7" (δ H 5.15) to C-1" (δ C 104.4). The nature of sugar in compound 2 was further demonstrated to be D-glucose by GC analysis of an acid-treated hydrolysate, as well as comparison of their 13 C-NMR data with that in the literature [20]. The relative configuration of compound 2 was determined by interpretation of the NOESY spectrum and the chemical shift ratio of C-7":C-8" [21]. In the NOESY spectrum of compound 2 (See Figure S13), the correlation of H-7" (δ H 5.15) with H-8" (δ H 4.42-4.44) was observed, suggestive of the same α-orientation. Moreover, based on the similar chemical shift ratio of C-7":C-8" (δ C 82.3:86.6) to those of hedyotol C 4"-O-β-D-glucopyranoside [20], the relative configuration of compound 2 was proposed as an erythro configuration. Therefore, compound 2 was established as hedyotol C 7"-O-β-D-glucopyranoside. Compound 3 (Figure 1) was determined to have the same molecular formula C 37 H 46 O 16 as compound 2 by HR-ESI-MS (m/z 769.2662 [M + Na] + ). Their 1 H-and 13 C-NMR spectra closely resembled each other ( Table 2). The above evidence suggests that compounds 2 and 3 might be stereoisomeric, which could be clarified in accordance with the NOESY spectrum and the chemical shift ratio of C-7":C-8" [21]. The differences between compounds 2 and 3 were the absence of NOESY correlations between H-7" (δ H 5.28) and H-8" (δ H 4.26-4.27) (See Figure S20) and the similar chemical shift ratio of C-7":C-8" (δ C 77.7:86.8) as compared to those of hedyotol D 4"-O-β-D-glucopyranoside [20] in compound 3. This evidence confirmed compounds 2 and 3 were a pair of epimers at C-7". Accordingly, the relative configuration of compound 3 was presumed as a threo configuration. Consequently, compound 3 was identified as hedyotol D 7"-O-β-D-glucopyranoside.
All the isolated constituents (compounds 1-12) were evaluated for in vitro anti-complementary activity on CP and AP [16]. As shown in Figure 1 and Table 3, compounds 4 and 5 belonged to the benzofuran lignans and possessed the most potent anti-complementary activity with CH 50 values of 0.040 ± 0.009 and 0.009 ± 0.002 mM and AP 50 values of 0.105 ± 0.015 and 0.021 ± 0.003 mM, respectively. Seven flavonoids (compounds 6-12) showed anti-complementary activity with CH 50 and AP 50 values of 0.143-1 mM and 0.311-1 mM, respectively. Compounds 1-3 exhibited moderate activity on CP and weak activity on AP. It was found that flavonoids with 4 -OH (compounds 8 and 12) showed stronger anti-complementary activity than those with a -OCH 3 at C-4 (compounds 9-11). Thus, the 4 -OH appears to be essential for the complementary activity of flavonoids. 0.026 ± 0.005 0.055 ± 0.008 a Data were represented as mean ± SD of three independent measurements (n = 3), and described as 50% hemolytic inhibitory concentration (CH 50 for classical pathway and AP 50 for alternative pathway). ** ,b p < 0.01, compared to the positive control group. c Heparin was used as the positive control (mg/mL).
In order to illuminate the anti-complementary mechanism of compounds 4 and 5, identification of their targets in the complement activation cascade was conducted using complement-depleted (C-depleted) sera [16]. As shown in Figure 3, compound 4 regained the hemolytic capacity of C5-depleted serum, and compound 5 regained the hemolytic capacity of C4-and C5-depleted sera. These findings suggested that compound 4 probably acted on C1q, C2, C3, C4 and C9, while compound 5 interacted with the C1q, C2, C3 and C9 components of the complement. These results indicate that different compounds can act on different targets in the complement activation cascade. Thus, compounds 4 and 5 are promising candidates for development as anti-complementary agents from H. angustifolia. Moreover, the relevant pharmacology and toxicology of these compounds need further investigations.

Extraction and Isolation
The extraction of the dried aerial parts of H. angustifolia with 95% ethanol was partitioned successively with petroleum ether, EtOAc, and n-BuOH. The EtOAc fraction was selected to further purification due to its significant anti-complementary activity with CH 50 value of 0.15 ± 0.02 and AP 50 value of 0.29 ± 0.03 mg/mL. The EtOAc extract (160 g) was subjected to a silica gel column (12 × 50

Acid Hydrolysis of Compounds 2 and 3
The acid hydrolysis was performed according to the modified method of Song et al. [29]. Compound 2 (2 mg) was dissolved in 2 mol/L HCl (1 mL), and heated at 95 • C for 2 h. After hydrolysis, the reaction mixture was extracted with EtOAc and H 2 O, and the aqueous layer was neutralized with silver carbonate (10 mg). Then the supernatant was concentrated to dryness and further evaporated to dryness under N 2 for 2 h. Subsequently, the residue was stirred with L-cysteine methyl ester hydrochloride (2 mg) in anhydrous pyridine (0.2 mL) at 60 • C for 1 h. The dried reactant was fractionalized with H 2 O and n-hexane (each 0.5 mL). The n-hexane layer was collected and subjected to GC-MS. The acid hydrolysis of compound 3 was conducted in the same procedure as compound 2. The GC conditions were: column temperature 150 • C, injector temperature 270 • C, carrier gas N 2. D-glucose was detected from compounds 2 and 3 (t R 15.523 and 15.524 min) by comparison with authentic samples: t R D-glucose 15.513 min, and L-glucose 15.808 min.

Anti-Complementary Activity Assay Against CP and AP
The anti-complementary activities against the CP and AP were investigated with the method of Xu et al. [16]. Heparin was used as positive control drug. The results of anti-complementary activity of compounds 1-12 are displayed in Table 3. Anti-complement activity was determined as the mean of triplicate measurements at each concentration and expressed as 50% inhibitory concentration (CH 50 and AP 50 values). All results were performed statistical analysis to compare the CH 50 and AP 50 values between each compound and positive control with unpaired t test.

Identification of the Targets in the Complement Activation Cascade
Assay to identify the targets in the complement activation cascade was performed according to the method of Di et al. [30]. The results of compounds 4 and 5 were displayed in Figure 3. Data were expressed as mean ± SD of triplicate measurements.

Conclusions
A first phenalenon derivative (compound 1) with an acetyl side chain at C-8 and a pair of new sesquilignan epimers (compounds 2 and 3) with a glucosyl unit at C-7" together with nine known compounds were isolated from the aerial parts of H. angustifolia. Chemical structures of the isolated compounds were identified on the basis of extensive spectroscopic data. Compounds 4, 8-12 were isolated for the first time from the aerial parts of H. angustifolia. Eleven compounds exhibited anti-complementary activity, and the targets of two compounds (4 and 5) in the complement activation cascade were also identified. In conclusion, the anti-complementary constituents of H. angustifolia possessed chemical diversity, and consisted mostly of flavonoids and lignans in this study.