Download full-text PDF - ACS Publications - American Chemical Society

Article pubs.acs.org/jnp

Gardenifolins A−H, Scalemic Neolignans from Gardenia ternifolia: Chiral Resolution, Configurational Assignment, and Cytotoxic Activities against the HeLa Cancer Cell Line Dieudonné Tshitenge Tshitenge,†,‡ Doris Feineis,† Suresh Awale,§ and Gerhard Bringmann*,† †

Institute of Organic Chemistry, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany Faculty of Pharmaceutical Sciences, University of Kinshasa, B.P. 212, Kinshasa XI, Democratic Republic of the Congo § Division of Natural Drug Discovery, Institute of Natural Medicine, University of Toyama, 2630 Sugitani, Toyama 930-0194, Japan ‡

S Supporting Information *

ABSTRACT: From the tropical plant Gardenia ternifolia Schumach. and Thonn. (Rubiaceae), eight stereoisomeric 2,3-dihydrobenzo[b]furan neolignans, named gardenifolins A− H (1a−d and 2a−d), were isolated and fully structurally characterized. Reversed-phase chromatography of a stem bark extract afforded two peaks, viz. mixtures I and II, each one consisting of two diastereomers and their respective enantiomers. They were resolved and stereochemically analyzed by HPLC on a chiral phase coupled to electronic circular dichroism (ECD) spectroscopy, giving single ECD spectra of all eight stereoisomers. The double-bond geometries (E or Z) of the gardenifolins A−H and their relative configurations (cis or trans) at the stereogenic centers C-7 and C-8 in the dihydrofuran ring system were assigned by 1D and 2D NMR methods, in particular, using NOE difference experiments, whereas the absolute configurations of the isolated enantiomers were established by ECD spectroscopy by applying the reversed helicity rule. The individual pure gardenifolin isomers A−H showed the most different cytotoxic effects against the human cancer HeLa cell line, with 1d and 2a displaying the highest activities, with IC50 values of 21.0 and 32.5 μM, respectively. Morphological experiments indicated that gardenifolin D (1d) induces apoptosis of HeLa cells at 25 μM.

L

purity. Given the widespread occurrence of neolignans in the plant kingdom and their potential role as pharmaceutical lead compounds, it was essential to develop analytical methods for the reliable assessment of the stereochemical purities of these compounds and their absolute configurations. The pertinence of performing such a test is due to the fact that a scalemic (i.e., non-1:1) mixture of enantiomers will be isolated as a single chromatographic peak on the usual achiral phases and with only one set of NMR signals.20 Whether enantiomerically pure or “only” a scalemic mixture, it will display an optical rotation and an ECD spectrum. Purification of enantiomers has thus been reported mostly for fully racemic mixtures, as in that case there was no ECD spectrum and no specific rotation ([α]D) to be seen.20−25 For scalemic mixtures, the measured ECD spectrum and/or [α]D value is dictated by the preponderance of one enantiomer.26 In view of the sometimes drastically different bioactivities of enantiomers27 and the possible phylogenetic significance of stereoisomeric mixtures,28 it is mandatory to resolve them and provide pure stereoisomers. This should be particularly considered for herbal medicines.

ignans and neolignans are widely distributed plant secondary metabolites, showing a large structural diversity. Biosynthetically, they arise from the shikimic acid pathway, formed by dimerization of two phenylpropenoid units.1 They show remarkable pharmaceutical properties.2 Thus, some 2,3dihydrobenzo[b]furan neolignans (2,3-DBFs) exhibit antioxidant,3,4 cytotoxic,4−6 antiviral,6 antiparasitic,7 PGI2 release inducing,8 or antiplatelet9 activities. Despite the plethora of known neolignans,1,2 the unambiguous determination of the absolute configuration of 2,3-DBFs remains a thrilling challenge.10−12 In the literature, there have been several cases of misleading assignments and structural revisions,11−13 and some natural products have remained without a reliable assessment of the absolute stereostructures.8,14,15 For the determination of the complete threedimensional structures of neolignans, several techniques have been used, among them NMR spectroscopy, X-ray crystallography, electronic circular dichroism (ECD) spectroscopy, optical rotatory dispersion (ORD), and their interpretation by chiroptical rules and quantum-chemical calculations.8,10−18 To our surprise, such naturally occurring 2,3-DBFs, despite their biosynthetic origin through free-radical processes,1,19 have rarely been analyzed systematically for their enantiomeric © 2017 American Chemical Society and American Society of Pharmacognosy

Received: March 2, 2017 Published: May 10, 2017 1604

DOI: 10.1021/acs.jnatprod.7b00180 J. Nat. Prod. 2017, 80, 1604−1614

Journal of Natural Products

Article

Figure 1. Structures of gardenifolins A−D (1a−1d) and E−H (2a−2d), stereoisomeric neolignans from the stem bark of Gardenia ternifolia eluting as two peaks on an achiral reversed-phase column (Figure 2), mixtures I (consisting of 1a−1d) and II (containing 2a−2d), and then further resolved by HPLC on a chiral phase (Figure 7).

One important medical plant is Gardenia ternifolia Schumach. and Thonn. (Rubiaceae), an evergreen shrub widely used in African traditional medicine against several infectious diseases.29 In the course of our phytochemical analysis of this plant, we have developed a high-performance liquid chromatography (HPLC) method enabling the simultaneous resolution of diastereomeric and enantiomeric mixtures of 2,3-DBF neolignans both at an analytical and a preparative scale. We here describe the isolation, chiral resolution, and first assignment of the full absolute stereostructures of the complete series of all eight possible stereoisomers of a 2,3-DBF neolignan from G. ternifolia. These metabolites, named gardenifolins A−H (1a−d and 2a−d) (Figure 1), were found to occur in the plant as a mixture of its four diastereomers, each in a scalemic form. Their constitution and relative configurations were elucidated by spectroscopic techniques such as NMR and HRESIMS, whereas their absolute configurations were assigned by online ECD analysis in combination with their chiral resolution by HPLC. The ECD spectra were interpreted by application of the well-established reversed helicity rule.13,21 Prior to the study presented here, an endogenous prostacyclin (PGI2) inducer isolated from Zizyphus jujuba (Rhamnaceae) had been reported to possess the same constitution as that of the gardenifolins.8 This plant metabolite was shown to have E-configuration in the methyl acrylate unit and a relative 7,8-trans-configuration, thus resembling the structures of the two gardenifolins C (1c) and D (1d). The absolute configuration of this PGI2 inducer and its enantiomeric purity, however, were not determined. Even now, with our unambiguous structural assignments of all stereoisomers, the stereostructure of the Z. jujuba analogue cannot be assigned because its optical or chiroptical data had not been published. Thus, this is the first report on the full structural

characterization of all eight possible stereoisomers of those neolignan plant metabolites, particularly focusing on a reliable assessment of their relative and absolute stereostructures.

■

RESULTS AND DISCUSSION Isolation and Structural Elucidation of Neolignans. Resolution of the stem bark extract of G. ternifolia by preparative HPLC on a SymmetryPrep-C18 column led to the isolation of two peaks, denoted as I and II (Figure 2). All of the components contained in I and II corresponded to the same molecular formula of C21H22O7, as deduced by HRESIMS. The NMR data (Figure 3) of I and II were similar, too. The 13C NMR spectrum exhibited 21 carbon signals, among them one ester carbonyl (C-9′) and two methoxy group carbons belonging to two different aromatic moieties. The DEPT-135 spectrum indicated one oxymethylene carbon (C-9), two vinylic carbons (C-7′ and C-8′), and five aromatic methine carbons (C-2, C-5′, C-5, C-3′, and C-6). The integration of the signals in the 1H NMR spectrum indicated the presence of 20 protons, including five aromatic hydrogens comprising two spin systems. The signals of two adjacent vinylic protons were detected as doublets, confirmed by COSY interactions. Two geminal protons, H-9a,b, were shown by TOCSY data to belong to a methylene group attached to an sp3 methine. Two sp3 methine groups, H-8 and H-7, were deduced to be part of a dihydrofuran moiety. In addition, three methoxy groups were observed, one belonging to a methyl ester residue (CH3O-9′), and the other two, CH3O-3 and CH3O-6′, were attached to an aromatic system. A detailed analysis of the NMR data revealed the presence of a π-system consisting of an O-substituted orthomethoxyphenol moiety. Taken together, the constitutions of I and II were assigned to be identical, representing stereo1605



Article

and E-II (6.51) were too similar to permit differentiation between the relative cis- and trans-orientations at C-7 versus C8 in the dihydrofuran subunit in these four samples. NOESY measurements did not permit unambiguous assignment of the relative configurations of the two stereocenters for any of the Eor Z-isomers of I and II. The distinction was facilitated by NOE difference (NOEDIFF) experiments14 (Figure 4). Irradiation of

Figure 2. Reversed-phase chromatography on the extract of the stem bark of Gardenia ternifolia containing peak I (all 7,8-trans) and peak II (all 7,8-cis), consisting of two pairs of diastereomeric neolignans 1a, 1c, 2a, and 2c and their enantiomers 1b, 1d, 2b, and 2d (Figure 1), as nonracemic mixtures.

Figure 4. NOEDIFF enhancements upon irradiation exemplarily illustrated (A) for Z-I (7,8-trans) and (B) for Z-II (7,8-cis). Similar results were also obtained for the samples E-I (7,8-trans) and E-II (7,8cis). Weak interactions are given as dotted lines.

H-7 led to a significant enhancement of the signals of H-9a,b in the NOEDIFF spectra of E-I and Z-I, indicating a relative transconfiguration of the two stereocenters at C-7 and C-8. By contrast, the H-8 signal was distinctly enhanced in the NOEDIFF spectra of E-II and Z-II when H-7 was irradiated, thus verifying a relative cis-configuration of the isomers of these two subfractions (see Supporting Information). Moreover, all four samples, Z-I, E-I, Z-II, and E-II, were found to be optically active. Chiral Resolution of Neolignan Stereoisomers. Despite the optical and chiroptical activities of the four samples, Z-I, EI, Z-II, and E-II, chromatography on a chiral Lux Cellulose-1 phase, with acetonitrile and water as the mobile phase, afforded two peaks in each case, thus hinting at the presence of enantiomeric mixtures. This finding was further corroborated by the hyphenation of HPLC on that chiral phase with ECD spectroscopy, permitting acquisition of two pairs of mirrorimaged ECD spectra, for each of the mixtures I and II, by HPLC-DAD-ECD hyphenation.30 HPLC resolution on a preparative scale, directly starting from each of the mixtures I and II, gave rise to four baseline-separated peaks each (Figure 5), thus corroborating that there were, in total, eight stereoisomers (see also Supporting Information). Thus, the samples Z-I, E-I, Z-II, and E-II consisted each of scalemic mixtures of 1a and 1b (Z-I), 1c and 1d (E-I), 2a and 2b (Z-II), and 2c and 2d (E-II). The eight stereoisomers were named gardenifolins A−H (1a−1d and 2a−2d), based on their occurrence in Gardenia ternifolia. Stereochemical Assignment of Gardenifolins A−H. With the eight pure stereoisomers in hand, we focused on their reliable stereochemical assignment. For each isomer, the constitution was confirmed by NMR data (Table 1), whereas

Figure 3. (A) 1H and 13C NMR spectroscopic data (in ppm) for the Z-isomeric forms contained in fractions I and II. Values of the Eisomers of I and II, as far as they are different from those of the corresponding Z-isomers, are denoted in {}. (B) HMBC (single red arrows) and NOESY (double blue arrows) interactions for evidencing the constitution of the neolignans of I and II.

isomeric neolignans. The NMR data of I and II (Figure 3) furthermore indicated that these two samples were, in fact, mixtures of E- and Z-diastereomeric 2,3-dihydrobenzo[b]furans. Further resolution of I and II by repeated preparative HPLC on an achiral reversed-phase C18 column afforded four samples, which, each of them, gave the same molecular formula by HRESIMS as I and II in total, but the 1H NMR spectra suggested the presence of mixtures of Z- and E-isomers in I and II (Figure 3), as evidenced by the signals of the vinylic doublets of H-7′ and H-8′ (J7′,8′ ≈ 12.9 Hz for Z) and (J7′,8′ ≈ 15.9 Hz for E). The 1H NNR coupling constants J7,8 (6.65 vs 6.51 Hz) of subfractions Z-I (6.65 Hz), E-I (6.51 Hz), Z-II (6.65 Hz), 1606



Article

and had displayed clear ECD spectra (Figure S3, Supporting Information). Assignment of the Stereostructures of the Gardenifolins A−H. To assign the absolute configurations of these 2,3-DBFs, the conformation of the dihydrofuran moiety had to be deduced from the sign of the 1Lb band in the ECD spectrum indicating its helicity.13,14,20,21 This helicity describes the chirality of the heterocycle, which adopts an envelope-like half-chair conformation, with a defined dihedral angle of bond a versus c (Figure 6).13 The aryl substituent at C-7 is decisive for

Figure 5. Online ECD spectra of four pure stereoisomers in peak I, showing two similar pairs of mirror-imaged spectra. For the four ECD spectra arising from peak II, see Supporting Information.

the relative configurations were corroborated by NOEDIFF measurements (vide supra) for the scalemic mixtures (for the detailed NMR assignments, see Supporting Information). For the absolute configurations, the online ECD spectra of the pure enantiomers were used. Each pair of compounds displayed opposite ECD spectra, hinting at enantiomeric mixtures. The four diastereomers, E-I, Z-I, E-II, and Z-II, were, in fact, pairs of enantiomers in unequal amounts and were thus not racemic but scalemic, which was unprecedented in this subgroup of natural products. This explained why each of the four enantiomeric mixtures (e.g., E-I) had been optically active

Figure 6. Assignment of the helicity of the dihydrofuran unit based on the reversed helicity rule13 and the ECD sign of the Cotton effect at the 1Lb band for substituted 2,3-dihydrobenzo[b]furans, here for derivatives with a methoxy function at C-6′ (according to the nomenclature applied in this paper), exemplarily illustrated for gardenifolin E (2a).

Table 1. 1H (600 MHz) and 13C (151 MHz) Data of Gardenifolins A−H in Acetone-d6 (δ in ppm) gardenifolins A (1a) and B (1b) no. 1 2 3 4 5 6 7 8 9a 9b 1′ 2′ 3′ 4′ 5′ 6′ 7′ 8′ 9′ 3-OCH3 6′-OCH3 9′-OCH3 a

δH, mult. (J in Hz) 7.04, d (2.00)

6.81, d (8.12) 6.88, dd (1.79, 8.15) 5.63, d (6.66) 3.58, pqa (6.36) 3.88, m 3.84, m

7.34, br s 7.72, d (1.50) 6.90, d (12.93) 5.80, d (12.93)

3.82, s 3.86, s 3.69, s

δC, type 134.3, 110.6, 148.9, 147.5, 116.2, 119.7,

C CH C C CH CH

89.2, CHOC 54.5, CH 64.6, CH2OH 64.6, CH2O 150.9, C 131.0, C 121.9, CH 129.2, C 115.8, C 144.6, C 144.7, CH 116.5, CH 167.5, CO2CH3 56.3, CH3 56.4, CH3 51.6, CH3

gardenifolins C (1c) and D (1d) δH, mult. (J in Hz) 7.04, d (1.92)

6.81, d (8.12) 6.87, dd (1.99, 8.15) 5.63, d (6.66) 3.58, pqa (6.36) 3.89, m 3.84, m

7.25, br s 7.24, br s 7.60, d (15.90) 6.40, d (15.90)

3.81, s 3.90, s 3.71, s

δC, type 133.9, 110.6, 148.5, 147.6, 115.8, 119.7,

C CH C C CH CH


gardenifolins E (2a) and F (2b) δH, mult. (J in Hz) 7.04, d (2.00)

6.81, d (8.11) 6.88, dd (1.79, 8.15) 5.62, d (6.51) 3.58, pqa (6.20) 3.88, m 3.84, m

7.34, br s 7.72, d (1.50) 6.90, d (12.93) 5.80, d (12.93)

3.82, s 3.86, s 3.70, s

δC, type 134.3, 110.6, 148.9, 147.5, 116.2, 119.7,

C CH C C CH CH


gardenifolins G (2c) and F (2d) δH, mult. (J in Hz) 7.04, d (1.92)

6.81, d (8.12) 6.88, dd (1.95, 8.10) 5.61, d (6.52) 3.59, pqa (6.31) 3.89, m 3.85, m

7.25, br s 7.24, br s 7.60, d (15.90) 6.40, d (15.90)

3.82, s 3.91, s 3.72, s

δC, type 133.9, 110.6, 148.5, 147.6, 115.8, 119.7,

C CH C C CH CH


pq = pseudo-quartet. 1607



Article

the respective conformations of the dihydrofuran portion as it tends to adopt an equatorial position, regardless of the configuration at C-8, where the “slender” CH2OH substituent thus adopts the resulting axial or equatorial position. As a consequence, S-configuration at C-7 leads to P-helicity (as shown in Figure 6), whereas 7R implies M-helicity. A chiroptical rule established by Snatzke et al.31 permits deduction of this helicityand thus the absolute configuration at C-7from the sign of the Cotton effect within the 1Lb band (270−300 nm) in the ECD spectrum.13,21 It is also known13 that the substitution at C-6′ of the benzene unit by an achiral group having a large spectroscopic moment such as hydroxy, alkoxy, or alkenyl residues results in an inversion of the sign of the Cotton effect in the 1Lb band, even though the absolute configuration and conformation of the furan portion remain unchanged. With the presence of a methoxy group at C-6′ in all eight isolated 2,3-dihydrobenzo[b]neolignan stereoisomers, this reversed helicity rule13,21 should be applicable here. According to this modified chiroptical rule, P-helicity of the dihydrofuran unit should lead to a positive sign of the 1Lb band in the ECD spectrum (Figure 6), and M-helicity should result in a negative Cotton effect.13,21 On the basis of the reversed helicity rule for the 1Lb band ECD of 7-methoxy-2,3-dihydrobenzo[b]furan chromophores,13 the positive 1Lb Cotton effect (CE) of 1a around 308 nm (Δε +1.4) evidenced P-helicity and, consequently, S-configuration at C-7. The negative CE of 1b around 309 nm (Δε −1.2), in turn, indicated 7R-configuration for this enantiomer. (+)-Gardenifolin A (1a) was thus assigned as (7′Z,7S,8R)-configured and (−)-gardenifolin B (1b = ent-1a) as (7′Z,7R,8S). As shown in Figure 5, the neolignans 1c and 1d exhibited mirror-image-like ECD curves, too. The ECD spectrum of 1c showed a positive CE around 317 nm (Δε +3.0) and, thus, P-helicity, whereas Mhelicity was deduced for 1d according to a negative CE around 314 nm (Δε −3.5). Therefore, the absolute configurations of these two enantiomers were established as (7′E,7S,8R) for (+)-gardenifolin C (1c) and (7′E,7R,8S) for (−)-gardenifolin D (1d = ent-1c) (Figure 1). In a similar way, resolution of the stereoisomers of mixture II by HPLC on a chiral Lux Cellulose-1 phase provided two pairs of enantiomers, 2a/2b and 2c/2d (Figure 7). The gardenifolins E (2a) and G (2c) displayed positive CEs and, thus, P-helicity and hence a 7S-configuration, whereas M-helicities and thus 7Rconfigurations were assigned for their corresponding enantiomers, gardenifolins F (2b) and H (2d). Based on the optical, chiroptical, and spectroscopic results obtained, the absolute configurations of these four stereoisomers were assigned as follows: (+)-7′Z,7S,8S for gardenifolin E (2a), (−)-7′Z,7R,8R for gardenifolin F (2b = ent-2a), (+)-7′E,7S,8S for gardenifolin G (2c), and (−)-7′E,7R,8R for gardenifolin H (2d = ent-2c). Thus, all eight possible stereoisomers for the 2D structure of gardenifolins were discovered in G. ternifolia (Figure 1). Resolution of the E- and Z-diastereomers in both mixtures I and II by HPLC on a chiral phase provided the corresponding four pairs of enantiomers in ratios distinctly differing from 1:1 (Figure 7). All thesehence nonracemicpairs of neolignan enantiomers were found to be mixtures occurring in ratios of about 1.5:1 (for 1a and 1b), 1.4:1 (for 1c and 1d), 1.9:1 (for 2a and 2b), and 2:1 (for 2c and 2d). The scalemic nature of the diastereomers in mixtures I and II in G. ternifolia may hint at a lack of stereocontrol and, thus, at a weak enzymatic assistance by dirigent proteins32 in the biosynthesis of these compounds.

Figure 7. Chromatograms of the mixtures I (1a−d) and II (2a−d) resolved by preparative HPLC on a chiral Lux Cellulose-1 phase and stereochemical assignment of the peaks (as achieved by ECD).

This is the first report on the discovery and successful resolution of two 2,3-dihydrobenzo[b]furan neolignan mixtures consisting of all possible stereoisomers, that is, four diastereomers and their four enantiomers and their configurational assignment. This separation was accomplished by two HPLC runs, first on an achiral column for the isolation of the diastereomers, followed by resolution of the enantiomers on a chiral phase. Only one similar literature example20 is known regarding the separation and characterization of four pairs of enantiomeric sesquineolignans isolated from the Chinese shrub Phyllanthus glaucus (Euphorbiaceae).20 These sesquineolignans consist of a 2-phenyl-2,3-dihydrobenzo[b]furanpropanol moiety and a 1-phenylpropan-1,3-diol unit possessing two stereogenic centers. Similar to the gardenifolins A−H, a series of eight sesquineolignan enantiomers had been isolated by HPLC on a chiral phase from two stereoisomeric mixtures, each of them consisting of four diastereomers and their four enantiomers. However, different from the neolignans from G. ternifolia described here, only eight out of a total of 16 possible stereoisomers had been identified in the case of those sesquineolignans from P. glaucus.20 The discovery of the gardenifolins E−H (2a−d), comprising a complete series of cis-configured isomers, was remarkable, also in view of the fact thatapart from those sesquineolignans from P. glaucus20no further reliable similar example of the natural occurrence of four cis-configured isomers was known from the literature. Only a few examples of naturally occurring pairs of neolignan enantiomers with a 2,3-dihydrobenzo[b]furan motif have so far been described in the literature,1,4,11,17,20,33 and even less frequently, are reports on investigations concerning the enantiomeric purity of naturally occurring 2,3-DBFs and related neolignans.20,33,34 Thus, little is known about the presence of these compounds as scalemic mixtures in nature. Only recently, 1608



Article

Scheme 1. Possible Chemically Induced Predominant Formation of trans-Configured Gardenifolins, Illustrated Here for the Interconversion of the cis-Configured Gardenifolins F (2b) and H (2d) into the trans-Isomers Gardenifolins A (1a) and C (1c) (*Configurationally Stable Stereogenic Center)

In 1995, Wahl et al.35 reported the isolation of a neolignan named xylobuxin from the stem bark of Xylopia buxifolia (Annonaceae), yet without mentioning the PGI2 inducer isolated earlier from Zizyphus jujuba.8 The compound was described to have the same constitution and was attributed the novelty of a relative cis-configuration in the furan ring based on the coupling constant of J = 6.0 Hz, which was erroneously expected to have a value of J = 10.5 Hz for a trans-configured 2,3-DBF neolignan.35 Virtually the same coupling constant (J = 6.7 Hz), however, had already been reported for that transconfigured PGI2 inducer.8 As discussed above for Z-I (7,8trans) and Z-II (7,8-cis) and also demonstrated for related neolignans, coupling constants for cis- and trans-configured 2,3DBFs are too similar14,18,36 to permit a reliable assignment of the relative configuration of the 2,3-dihydrobenzofuran ring system, an exercise that can only be achieved by comprehensive 2D NMR measurements in combination with NOEDIFF experiments.4,7,14,33 The co-occurrence of E- and Z-isomers of the cinnamoyl residues in the gardenifolins A−H (1a−d and 2a−2d) is welldocumented for many other natural products.37 It is also known that these moieties are prone to E/Z interconversion upon light exposure, even in the NMR tube.38,39 This isomerizationeven in combination with a cis/trans interconversion by epimerization at C-7might, however, also occur by an acid (or base)-catalyzed ring opening reaction (Scheme 1). From previous studies,12,40−42 it is well-known that thermodynamically less stable cis-7-(4-hydroxyphenyl)-7,8dihydrobenzofurans can undergo an epimerization reaction at C-7 to give the respective more stable trans-isomers by

some lignan, norlignan, and neolignan enantiomers have been isolated from two traditional Chinese medicinal plants,20,33,34 some of which weresimilar to the gardenifolin enantiomers presented in this paperscalemic mixtures, as well. That such cases are very recent discoveries might hint at a more widespread presence of scalemic mixtures of neolignan enantiomers in plants, in general, because enantiomeric mixtures with variable compositions might have been erroneously considered as stereochemically homogeneous due to their reduced but definitive optical and chiroptical activities.33 A 2,3-dihydrobenzo[b]furan neolignan8 isolated from the leaves of Zizyphus jujuba (Rhamnaceae) in 1986 and possessing the same constitution as the gardenifolins discussed here was described as an endogenous prostacyclin (PGI2) inducer. Although, as for 1c/d, an E-configuration at the methyl acrylate unit and a relative 7,8-trans-configuration in the dihydrofuran portion had been established for that neolignan, its absolute configuration remained unknown. Moreover, neither an ECD spectrum nor any specific rotation values had been reported for that compound, rendering it impossible to assign the complete stereostructure of that plant metabolite from the literature data.8 Back in 1998, Yuen et al.15 faced the same dilemma: the group had synthesized the two respective enantiomers, but due to the insufficient information published on the absolute configuration and the enantiomeric purity of that natural PGI2 inducer, a stereochemical assignment of the natural product from Z. jujuba was not possible. Thus, our phytochemical work on G. ternifolia is the first reliable report on the natural occurrence of 1c and 1d. 1609



Article

treatment with acid (e.g., TFA in CH2Cl2)40,41 or base (e.g., Na2CO3 in MeOH).12,41 Such cis/trans isomerization processes may possibly interconvert any cis-gardenifolins into the respective trans-isomers, yet always within the same enantiomeric series, as the configuration at C-8 should remain unaffected.12,40−42 As illustrated in Scheme 1 exemplarily for the cis-configured gardenifolin 2d (7R,8R,7′E), ring openipng will give intermediate A, which will cyclize preferentially to the thermodynamically more stable trans-isomer 1c (7S,8R,7′E). Alternatively, A may cyclize back to 2d. As a third possibility, A may also be transformed into intermediate B by rotation about the C-7′−C-8′ bond. Subsequent cyclization of B will again mainly lead to the trans-isomer 1a (7S,8R,7′Z), that is, to an overall E/Z isomerization in combination with epimerization at C-7 or (to a smaller degree) to the cis-compound 2b (7R,8R,7′Z). The 2,3-dihydrobenzo[b]furan neolignans (Figure 1) described in this paper, however, were unexpectedly stable in the presence of acid as long as light was excluded. We also noticed that the E/Z isomerization of the gardenifolins A−H was significantly minimized when replacing methanol by the aprotic solvent acetone and performing the entire isolation process under light-reduced conditions, for example, when using brown flasks and acetone-d6 for NMR data acquisition. This might also be the reason why we surprisingly succeeded to isolate the complete series of all possible cis-configured gardenifolin isomers E−H (2a−d) from G. ternifolia, although thermodynamically the formation of the trans-isomers A−D (1a−d) should be favored. The configurational stability of the gardenifolins A−H under the mild isolation conditions provided evidence that all eight stereoisomers are true natural products. Cytotoxic Activities of Gardenifolins A−H. Many lignans and neolignans are constituents of traditional herbal medicines and are, thus, in the focus of biological evaluations regarding their cytotoxic potential toward various cancer cell lines,1,2 showing that some of them exhibited promising antiproliferative effects.1,2,4−6,43 The availability of the eight possible gardenifolin isomers A−H (1a−d and 2a−d) in a stereochemically pure form from the African medicinal plant G. ternifolia now permitted assessment of their individual cytotoxic activities against the human cervical HeLa cell line. As shown in Table 2, all of the gardenifolins reduced the growth of the tumor cells. While most isomers displayed only weak to moderate effects with IC50 values of 78.8−105.0 μM, two of

these neolignans, gardenifolin D (1d) and gardenifolin E (2a), strongly inhibited HeLa cell proliferation in a concentration range similar to that of 5-fluorouracil (5-FU) used as the positive reference. Dose−response curves (Figure 8) revealed compounds 1d and 2a to exhibit cytotoxicity toward the HeLa cell line with IC50 values of 21.0 μM (for 1d) and 32.5 μM (for 2a). The IC50 values of the gardenifolins 1a (IC50 = 78.8 μM) and 2c (IC50 = 105.0 μM) were found to be nearly the same as those of their enantiomers 1b (IC50 = 87.3 μM) and 2d (IC50 = 96.1 μM), whereas for the other two pairs of gardenifolin enantiomers, the individual IC50 values varied significantly: Compound 2a (IC50 = 32.5 μM) was more active against HeLa cells by a factor of 3.2 than its enantiomer 2b (IC50 = 103.0 μM), and 1d (IC50 = 21.0 μM) was more active than 1c (IC50 = 90 μM) by a factor of 4.1. These findings underline the strong impact of chirality on the bioactivities of the new neolignans described in this paper. Gardenifolin D (1d), as the most potent isomer within this series of 2,3-dihydrobenzo[b]furan neolignans, was further studied for its effects on cell morphology and apoptosis using an acridine orange (AO) and Hoechst 33342 double staining assay (AO/Hoechst 33342) (Figure 9). AO and Hoechst 33342 are cell-permeable dyes. AO emits bright green fluorescence in live cells, whereas Hoechst 33342 is a nuclear counterstain that emits blue fluorescence when bound to DNA. In living cells, cellular morphology and nuclei remain intact. Apoptotic cells, by contrast, are characterized by the loss of cytoplasmic and nuclear integrity, leading to irregular cellular morphology and fragmented nuclei. Untreated HeLa cells showed intact cell morphology with bright green AO fluorescence, counter stained with flattened blue nuclei (Figures 9A−D). Treatment of the tumor cells with 25 μM of 1d (Figure 9E−H) induced apoptosis, as indicated by the dramatic alteration of the morphology of the HeLa cells (Figure 9E, black arrows) and by fragmented nuclei (Figure 9G, white arrows). In the past decade, numerous new neolignans,1 including compounds with unusual carbon skeletons,20,25,33,34 were isolated from diverse plant sources, but the discovery of a complete series of all possible stereoisomers of such chiral plant metabolites, as presented here for the first time for the 2,3dihydrobenzo[b]furan neolignans gardenifolins A−H (1a−d and 2a−d), is unprecedented. All eight stereoisomers were isolated and fully characterized, with the four diastereomers occurring in the African medicinal plant Gardenia ternifolia as nonracemic mixtures. Similarly unusual was the identification of a complete series of all four possible cis-isomers. Thus, cis-2-(4hydroxyphenyl)-2,3-dihydrobenzo[b]furans neolignans11,20,44 such as the gardenifolins E−H (2a−d) have less frequently been discovered in nature because they may epimerize under acidic or basic conditions to give, thermodynamically driven, the corresponding trans-compounds. The resolution of 1a−d and 2a−d was achieved on a chiral HPLC column, followed by stereochemical assignment, which succeeded by a combination of NOEDIFF measurements for the relative configurations and ECD spectroscopy, with application of the reversed helicity rule, for the absolute stereostructures. The work described here will be of particular value for the future structural characterization of other natural 2,3-DBF neolignans.

Table 2. Activities (IC50 Values in μM) of the Gardenifolins A−D (1a−d) and E−H (2a−d) and 5-Fluorouracil against Human Cervical HeLa Cancer Cells compound 5-fluorouracil (5-FU) 1a 1b 1c 1d 2a 2b 2c 2d a

HeLa cells

stereochemical features

13.9 78.8 87.3 90.0 21.0 32.5 103.2 105.0 96.1

(+)-7S,8R,7′Z (−)-7R,8S,7′Z (+)-7S,8R,7′Z (−)-7R,8S,7′Z (+)-7S,8S,7′E (−)-7R,8R,7′E (+)-7S,8S,7′E (−)-7R,8R,7′E

a

■

EXPERIMENTAL SECTION

General Experimental Procedures. Optical rotations were determined on a Jasco P-1020 polarimeter operating with a sodium

Used as a reference compound. 1610



Article

Figure 8. Inhibition of HeLa cells induced by the two most active of the new neolignans, (A) gardenifolin D (1d) and (B) gardenifolin E (2a), and by (C) 5-fluorouracil (5-FU) as the positive reference.

Figure 9. Morphological changes of HeLa cells induced by gardenifolin D (1d) (E−H) in comparison to untreated cells (A−D): HeLa cells (2 × 104/well/500 μL media) were treated with 25 μM of 1d in a 24-well plate and incubated for 72 h. Cells were stained with acridine orange (AO) and Hoechst 33342 and further incubated for 10 min. The cells were photographed under fluorescence and phase contrast modes using an EVOS FL digital inverted microscope. (A,E) Phase contrast. (B,F) Cells stained with AO (green). (C,G) Nuclei stained with Hoechst 33342 (blue). (D,H) Merged images. (E−H) Cells treated with 1d showed morphological alterations (black arrows) and nuclear fragmentation (white arrows) indicating apoptosis. light source (λ = 589 nm). UV spectra were recorded on a Shimadzu UV-1800 spectrophotometer, and ECD measurements were performed under nitrogen on a Jasco J-715 spectrometer. For the ECD spectra measured online, by hyphenation with HPLC-DAD, a baseline correction was performed. The ECD spectra of the eluent at the retention time of each peak were subtracted from the ECD curves of the compounds analyzed. 1D and 2D NMR spectra were monitored on a Bruker AMX 400 and on a Bruker DMX 600 instrument using acetone-d6 (δH 2.09 and δC 29.9/206.7) as the solvent. Chemical shifts (δ) are reported in parts per million (ppm), and coupling constants (J) are given in hertz (Hz). NMR signal multiplicities are given as singlet (s), doublet (d), doublet of doublets (dd), quartet (q), pseudo-quartet (pq), or multiplet (m). Spectra were acquired and processed using an ACD/NMR processor (version 12.01) and the Topspin 3.2 software (Bruker Daltonics). HRESIMS measurements were performed in positive mode on a Bruker Daltonics micrOTOF-focus mass instrument. HPLC-DAD investigations were conducted on a Jasco LC-2000 Plus series system. For LC-MS analyses, an Agilent 1100 series system was used. Preparative HPLC was carried out on a Jasco system (PU-1580 Plus) in combination with UV/vis detection at 200− 680 nm (Jasco MD-2010 Plus diode array detector) at room temperature. For maceration, a mechanical shaker was used at 160 rpm (rotations per minute). All organic solvents were of analytical grade quality. Ultrapure water was obtained from an Elga Purelab Classic system. Plant Material. Stem barks of Gardenia ternifolia Schumach. and Thonn. (Rubiaceae) were harvested in February 2012 in the area of Kimwenza near Kinshasa (Democratic Republic of the Congo). The

plant material was identified by Mr. Nlandu Lukebakio of the Institut National pour l’Etude et la Recherche Agronomiques (INERA), of the University of Kinshasa, where a voucher specimen was deposited under the number 843. Extraction and Isolation. The air-dried powder of stem barks of G. ternifolia (650 g) was macerated in MeOH for 48 h under lightreduced conditions, assisted by mechanical shaking, and then filtered. The procedure was repeated several times for exhaustive extraction. The combined filtrates were evaporated to a small volume, affording a viscous solution, which was partitioned between water and n-hexane to remove traces of chlorophyll. The aqueous phase was exhaustively extracted with CH2Cl2. The organic layers were evaporated to dryness to yield 202 mg of extract A. The remaining aqueous phase was exhaustively extracted with EtOAc. The combined EtOAc fractions were evaporated to dryness to give 101 mg of extract B. Throughout the whole process, precautions were taken to minimize the light exposure of the samples. The extracts A and B were each dissolved in MeOH and separately subjected to column chromatography on silica gel and successively eluted with 10:0, 10:1, 10:2, 8:3, 7:5, 5:5, and 0:5 CH2Cl2/MeOH to afford 10 fractions for each of the two extracts (fractions F1−F10). Fraction F3 from the extracts A and B was combined and directly subjected to preparative HPLC on a SymmetryPrep-C18 column (Waters, 300 × 19 mm, 7 μm) at room temperature; mobile phase: (S1) H2O (0.05% TFA), (S2) CH3CN (0.05% TFA). The flow rate was set to 8 mL/min, using the following gradient: 0−4 min, 25−30% S2; 4−28 min, 30−60% S2; 28−30 min, 60−98% S2, providing two subfractions. Peak I and peak II (Figure 2) were further resolved and purified by renewed chromatography under 1611



Article

(+3.0), 330 (+3.0), 354 (0.0) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12488 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin H (2d): white, amorphous powder; [α]23 D −20 (c 0.04, acetone); UV (acetone) λmax (log ε) 226 (2.3), 287 (1.7), 327 (4.3) nm; ECD (MeCN/H2O) λmax (log ε) 227 (+3.3), 236 (+2.9), 256 (−2.0), 277 (−0.1), 313 (−5.0), 329 (−1.8), 355 (0.0) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12486 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Biological Evaluation. The gardenifolins A−H (1a−d and 2a−d) were tested for their individual cytotoxic activities in vitro using the human cervical HeLa cell line.45 Cell viability in the presence or absence of tested compounds was determined using the Cell Counting Kit-8 (Dojindo Molecular Technologies, Inc., USA). The HeLa cell line was maintained in standard DMEM with 10% FBS supplement, 0.1% NaHCO3, and 1% antibiotic antimycotic solution. For cytotoxicity evaluation, exponentially growing cells were harvested and plated in 96-well plates (2 × 103/well) in DMEM at 37 °C under an atmosphere of humidified 5% CO2 and 95% air for 24 h. After the cells had been washed with PBS, the medium was changed to serially diluted test samples in DMEM, with the control and blank in each plate. After 72 h of incubation, the cells were washed twice with PBS, and a solution of 100 μL of DMEM containing 10% WST-8 cell counting kit was added in each well. After incubation for 3 h, the absorbance at 450 nm was measured (PerkinElmer EnSpire multilabel reader). Cell viability was calculated from the mean values from three wells using the following equation:

identical conditions as described above to give four samples (Z-I, E-I, Z-II, and E-II). The gardenifolins A−D (1a−1d) and E−H (2a−2d) were finally obtained in a stereochemically homogeneous form by preparative HPLC on a chiral Lux PrepCellulose-1 column (Phenomenex, 250 × 4.6 mm, 5 μm); mobile phase: (S1) H2O (0.05% TFA), (S2) CH3CN (0.05% TFA). The flow rate was set to 10 mL/min, using the following gradient: 0−7 min, 20−35% S2; 7−23 min, 35−55% S2; 23−28 min, 55−98% S2. From peak I, 1.5 mg of gardenifolin A (1a) (retention time 25.4 min), 1.4 mg of gardenifolin B (1b) (retention time 26.3 min), 3.5 mg of gardenifolin C (1c) (retention time 26.8 min), and 3.2 mg of gardenifolin D (1d) (retention time 27.6 min) were isolated. Peak II provided 1.8 mg of gardenifolin E (2a) (retention time 25.6 min), 1.0 mg of gardenifolin F (2b) (retention time 26.6 min), 1.7 mg of gardenifolin G (2c) (retention time 27.2 min), and 0.9 mg of gardenifolin H (2d) (retention time 27.8 min). Stereochemical Analysis of the Gardenifolins A−H. The resolution of the compounds at the analytical scale was achieved on a Phenomenex Lux Cellulose-1 column (stationary phase: cellulose tris(3,5-dimethylphenylcarbamate, 250 × 4.6 mm, 5 μm), using a dual mobile phase system consisting of 0.05% TFA in water (S1) and in CH3CN (S2). At a flow rate of 1 mL/min, the following gradient was used: 0−5 min, 10−40% S2; 5−25 min, 40−50% S2; 25−27 min, 50− 98% S2. These conditions were likewise used for the online HPLCDAD-ECD measurements in the stopped-flow mode. For these investigations, the system was equipped with a motor valve, thus giving rise to the acquisition of the ECD spectra of the compounds directly after identifying their UV spectra using a diode array detector. The peaks were analyzed at their highest absorbances. The ECD spectra were subtracted from the blank recorded at a solvent ratio corresponding to their retention time. Gardenifolin A (1a): white, amorphous powder; [α]23 D +20 (c 0.02, acetone); UV (acetone) λmax (log ε) 207 (4.9), 226 (4.4), 288 (3.5), 321 (4.1) nm; ECD (MeCN/H2O) λmax (log ε) 210 (3.2), 236 (−2.5), 256 (+1.3), 275 (−0.5), 308 (+1.4), 352 (0.0) nm; for 1H NMR and 13 C NMR, see Table 1; HRESIMS m/z 409.12485 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin B (1b): white, amorphous powder; [α]23 D −15 (c 0.02, acetone); UV (acetone) λmax (log ε) 207 (5.1), 226 (4.2), 288 (1.9), 321 (5.2) nm; ECD (MeCN/H2O) λmax (log ε) 209 (−1.9), 234 (+2.9), 255 (−1.2), 275 (+0.3), 309 (−1.2), 353 (0.0) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12485 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin C (1c): white, amorphous powder; [α]23 D +21 (c 0.04, acetone); UV (acetone) λmax (log ε) 226 (4.3), 287 (1.9), 327 (5.2) nm; ECD (MeCN/H2O) λmax (log ε) 234 (−5.0), 255 (+2.1), 317 (+3.0), 330 (+2.9), 353 (0.0) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12486 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin D (1d): white, amorphous powder; [α]23 D −21 (c 0.03, acetone); UV (acetone) λmax (log ε) 226 (2.3), 287 (1.7), 327 (4.3) nm; ECD (MeCN/H2O) λmax (log ε) 227 (+3.3), 236 (+2.9), 256 (−2.0), 277 (−0.1), 314 (−3.5), 330 (−2.9), 358 (−0.5) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12486 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin E (2a): white, amorphous powder; [α]23 D +27 (c 0.03, acetone); UV (acetone) λmax (log ε) 207 (5.3), 226 (4.3), 288 (3.6), 321 (4.5) nm; ECD (MeCN/H2O) λmax (log ε) 210 (4.5), 234 (−2.7), 255 (+2.1), 275 (−0.7), 308 (+1.9), 355 (0.0) nm; for 1H NMR and 13 C NMR, see Table 1; HRESIMS m/z 409.12487 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin F (2b): white, amorphous powder; [α]23 D −23 (c 0.03, acetone); UV (acetone) λmax (log ε) 207 (3.1), 226 (4.2), 288 (1.7), 321 (5.2) nm; ECD (MeCN/H2O) λmax (log ε) 209 (−1.9), 235 (+1.8), 254 (−1.1), 277 (+0.4), 310 (−1.4), 355 (0.0) nm; for 1H NMR and 13C NMR, see Table 1; HRESIMS m/z 409.12485 [M + Na]+ (calcd for C21H22O7Na, 409.12632). Gardenifolin G (2c): white, amorphous powder; [α]23 D +26 (c 0.05, acetone); UV (acetone) λmax (log ε) 226 (4.3), 287 (1.9), 327 (5.2) nm; ECD (MeCN/H2O) λmax (log ε) 234 (−5.2), 255 (+2.6), 318

cell viability (%) = [Abs(test sample) − Abs(blank)/Abs(control) − Abs(blank)] × 100% Morphological Assessment of Cancer Cells. HeLa cells were seeded in 24-well plates (1.0 × 104/well) and incubated in fresh DMEM at 37 °C under an atmosphere of 5% CO2 and 95% air for 24 h for the attachment. After the cells had been washed twice with PBS, the medium was changed with the test samples and incubated for 72 h. After incubation, acridine orange (30 μL, 100 mg/mL concentration) and Hoechst 33342 (NucBlue live readyprobes reagent, one drop) were added to each test well, and the sample was further incubated for 10 min. The cells were photographed using EVOSFL cell imaging system (40× objective) under fluorescent and phase contrast mode.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.7b00180. NMR (1H, 13C, COSY, HSQC, HMBC, NOESY, TOCSY), HRESIMS, and CD spectra of the gardenifolins A−D (1a−1d) and E−H (2a−2d) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Tel. +49-931-318-5323. Fax: +49-931-318-4755. E-mail: [email protected]. ORCID

Suresh Awale: 0000-0002-5299-193X Gerhard Bringmann: 0000-0002-3583-5935 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft (SFB 630 “Agents against Infectious Diseases”, project A2). D.T.T. was supported by a grant of the German Excellence 1612



Article

(18) Wang, Y.-H.; Sun, Q.-Y.; Yang, F.-M.; Long, C.-L.; Zhao, F.-W.; Tang, G.-H.; Niu, H.-M.; Wang, H.; Huang, Q.-Q.; Xu, J.-J.; Ma, L.-J. Helv. Chim. Acta 2010, 93, 2467−2477. (19) Suzuki, S.; Umezawa, T. J. Wood Sci. 2007, 53, 273−284. (20) Wu, Z.; Lai, Y.; Zhou, L.; Wu, Y.; Zhu, H.; Hu, Z.; Yang, J.; Zhang, J.; Wang, J.; Luo, Z.; Xue, Y.; Zhang, Y. Sci. Rep. 2016, 6, 24809. (21) Kurtán, T.; Antus, S.; Pescitelli, G. In Comprehensive Chiroptical Spectroscopy; Berova, N., Polavarapu, P. L., Nakanishi, K., Woody, R. W., Eds.; John Wiley & Sons: Hoboken, NJ, 2012; Vol. 2, Chapter 3 (Part II), pp 73−114. (22) Li, H.; Peng, S.-Y.; Yang, D.-P.; Bai, B.; Zhu, L.-P.; Mu, C.-Y.; Tian, Y.-J.; Wang, D.-M.; Zhao, Z.-M. Chirality 2016, 28, 259−263. (23) Pereira, A. C.; Magalhães, L. G.; Januário, A. H.; Pauletti, P. M.; Cunha, W. R.; Bastos, J. K.; Nanayakkara, D. N. P.; e Silva, M. L. A. J. Chromatogr. A 2011, 1218, 7051−7054. (24) Shi, Y.; Liu, Y.; Li, Y.; Li, L.; Qu, J.; Ma, S.; Yu, S. Org. Lett. 2014, 16, 5406−5409. (25) Cheng, Z.-B.; Lu, X.; Bao, J.-M.; Han, Q.-H.; Dong, Z.; Tang, G.-H.; Gan, L.-S.; Luo, H.-B.; Yin, S. J. Nat. Prod. 2014, 77, 2651− 2657. (26) Swinehart, D. F. J. Chem. Educ. 1962, 39, 333−335. (27) Agranat, I.; Caner, H.; Caldwell, J. Nat. Rev. Drug Discovery 2002, 1, 753−768. (28) Song, W.; Staudt, M.; Bourgeois, I.; Williams, J. Biogeosciences 2014, 11, 1435−1447. (29) (a) Tshitenge, D. T.; Ioset, K. N.; Lami, J. N.; Ndelo-di-Phanzu, J.; Mufusama, J.-P. K. S.; Bringmann, G. Fitoterapia 2016, 110, 189− 195. (b) Ochieng, C. O.; Ogweno Mid, J.; Okinda Owu, P. Res. J. Pharmacol. 2010, 4, 45−50. (c) Asase, A.; Oteng-Yeboah, A. A.; Odamtten, G. T.; Simmonds, M. S. J. J. Ethnopharmacol. 2005, 99, 273−279. (30) Bringmann, G.; Messer, K.; Wohlfarth, M.; Kraus, J.; Dumbuya, K.; Rückert, M. Anal. Chem. 1999, 71, 2678−2686. (31) (a) Snatzke, G.; Ho, P. C. Tetrahedron 1971, 27, 3645−3653. (b) Antus, S.; Snatzke, G.; Steinke, I. Liebigs Ann. Chem. 1983, 1983, 2247−2261. (c) Antus, S.; Baitz-Gács, E.; Snatzke, G.; Tóth, T. S. Liebigs Ann. Chem. 1991, 1991, 633−641. (32) (a) Davin, L. B.; Lewis, N. G. Curr. Opin. Biotechnol. 2005, 16, 398−406. (b) Moinuddin, S. G. A.; Hishiyama, S.; Cho, M. H.; Davin, L. B.; Lewis, N. G. Org. Biomol. Chem. 2003, 1, 2307−2313. (c) Fournand, D.; Cathala, B.; Lapierre, C. Phytochemistry 2003, 62, 139−146. (33) (a) Lu, Y.; Xue, Y.; Liu, J. J.; Yao, G.; Li, D.; Sun, B.; Zhang, J.; Liu, Y.; Qi, C.; Xiang, M.; Luo, Z.; Du, G.; Zhang, Y. J. Nat. Prod. 2015, 78, 2205−2214. (b) Lu, Y.; Xue, Y.; Chen, S.; Zhu, H.; Zhang, J.; Li, X.-N.; Wang, J.; Liu, J. J.; Qi, C.; Du, G.; Zhang, Y. Sci. Rep. 2016, 6, 22909. (34) Lai, Y.; Liu, T.; Sa, R.; Wei, X.; Xue, Y.; Wu, Z.; Luo, Z.; Xiang, M.; Zhang, Y.; Yao, G. J. Nat. Prod. 2015, 78, 1740−1744. (35) Wahl, A.; Roblot, F.; Cavé, A. J. Nat. Prod. 1995, 58, 786−789. (36) Garcia-Muñoz, S.; Á lvarez-Corral, M.; Jiménez-González, L.; López-Sánchez, C.; Rosales, A.; Muñoz-Dorado, M.; Rodríguez-García, I. Tetrahedron 2006, 62, 12182−12190. (37) (a) Veitch, N. C.; Grayer, R. J. Nat. Prod. Rep. 2011, 28, 1626− 1695. (b) Bienz, S.; Bisegger, P.; Guggisberg, A.; Hesse, M. Nat. Prod. Rep. 2005, 22, 647−658. (38) (a) Tawil, B. F.; Guggisberg, A.; Hesse, M. J. Photochem. Photobiol., A 1990, 54, 105−111. (b) Yoshida, K.; Okuno, R.; Kameda, K.; Mori, M.; Kondo, T. Biochem. Eng. J. 2003, 14, 163−169. (c) Yoshida, K.; Okuno, R.; Kameda, K.; Kondo, T. Tetrahedron Lett. 2002, 43, 6181−6184. (d) Silva, V. O.; Freitas, A. A.; Maçanita, A. L.; Quina, F. H. J. Phys. Org. Chem. 2016, 29, 594−599. (39) Talhi, O.; Lopes, G. R.; Santos, S. M.; Pinto, D. C. G. A.; Silva, A. M. S. J. Phys. Org. Chem. 2014, 27, 756−763. (40) Kurosawa, W.; Kan, T.; Fukuyama, T. Synlett 2003, 1028−1030. (41) Wada, H.; Kido, T.; Tanaka, N.; Murakami, T.; Saiki, Y.; Chen, C.-M. Chem. Pharm. Bull. 1992, 40, 2099−2101.

Initiative to the Graduate School of Life Sciences, University of Würzburg, and by the Excellence Scholarship Program BEBUC (www.foerderverein-uni-kinshasa.de). The biological evaluation was supported by the Japanese Society for the Promotion of Science (JSPS), Japan, Kakenhi (16K08319) to S.A. Special thanks are due to Mr. Nlandu Lukebakio from INERA for the botanical identification of the plant. Furthermore, we are grateful to Dr. M. Grüne and Mrs. E. Ruckdeschel for the NMR experiments.

■

DEDICATION Dedicated to the memory of Professor Burchard Franck, deceased on February 21, 2017.

■

REFERENCES

(1) (a) Teponno, R. B.; Kusari, S.; Spiteller, M. Nat. Prod. Rep. 2016, 33, 1044−1092. (b) Pan, J.-Y.; Chen, S.-L.; Yang, M.-H.; Wu, J.; Sinkkonen, J.; Zou, K. Nat. Prod. Rep. 2009, 26, 1251−1292. (2) (a) Apers, S.; Vlietinck, A.; Pieters, L. Phytochem. Rev. 2003, 2, 201−217. (b) Saleem, M.; Kim, H. J.; Ali, M. S.; Lee, Y. S. Nat. Prod. Rep. 2005, 22, 696−716. (3) (a) Song, C.-W.; Wang, S.-M.; Zhou, L.-L.; Hou, F.-F.; Wang, K.J.; Han, Q.-B.; Li, N.; Cheng, Y.-X. J. Agric. Food Chem. 2011, 59, 1199−1204. (b) Mei, R.-O.; Wang, Y.-H.; Du, G.-H.; Liu, G.-M.; Zhang, L.; Cheng, Y.-X. J. Nat. Prod. 2009, 72, 621−625. (4) Huang, X.-X.; Zhou, C.-C.; Li, L.-Z.; Peng, Y.; Lou, L.-L.; Liu, S.; Li, D.-M.; Ikejima, T.; Song, S.-J. Fitoterapia 2013, 91, 217−223. (5) (a) Baek, S.; Xia, X.; Min, B. S.; Park, C.; Shim, S. H. Beilstein J. Org. Chem. 2014, 10, 2955−2962. (b) Suh, W. S.; Kim, K. H.; Kim, H. K.; Choi, S. U.; Lee, K. R. Helv. Chim. Acta 2015, 98, 1087−1094. (6) Sawasdee, K.; Chaowasku, T.; Lipipun, V.; Dufat, T.-H.; Michel, S.; Likhitwitayawuid, K. Fitoterapia 2013, 85, 49−56. (7) (a) Rakotondraibe, L. H.; Graupner, P. R.; Xiong, Q.; Olson, M.; Wiley, J. D.; Krai, P.; Brodie, P. J.; Callmander, M. W.; Rakotobe, E.; Ratovoson, F.; Rasamison, V. E.; Cassera, M. B.; Hahn, D. R.; Kingston, D. G. I.; Fotso, S. J. Nat. Prod. 2015, 78, 431−440. (b) De Castro Oliveira, L. G.; Brito, L. M.; de Moraes Alves, M. M.; Amorim, L. V.; Sobrinho-Júnior, E. P. C.; de Carvalho, E. S.; da Franca Rodrigues, K. A.; Arcanjo, D. D. R.; das Graças Lopes Citó, A. M.; de Amorim Carvalho, F. A. Basic Clin. Pharmacol. Toxicol. 2017, 120, 52− 58. (8) Fukuyama, Y.; Mizuta, K.; Nakagawa, K.; Wenjuan, Q.; Xiue, W. Planta Med. 1986, 52, 501−502. (9) Zhuo, J.-X.; Wang, Y.-H.; Su, X.-L.; Mei, R.-Q.; Yang, J.; Kong, Y.; Long, C.-L. Nat. Prod. Bioprospect. 2016, 6, 161−166. (10) Li, Y.-P.; Dong, L.-B.; Chen, D.-Z.; Li, H.-M.; Zhong, J.-D.; Li, F.; Liu, X.; Wang, B.; Li, R.-T. Phytochem. Lett. 2013, 6, 281−285. (11) Tang, G.-H.; Chen, Z.-W.; Lin, T.-T.; Tan, M.; Gao, X.-Y.; Bao, J.-M.; Cheng, Z.-B.; Sun, Z.-H.; Huang, G.; Yin, S. J. Nat. Prod. 2015, 78, 1894−1903. (12) Natori, Y.; Tsutsui, H.; Sato, N.; Nakamura, S.; Nambu, H.; Shiro, M.; Hashimoto, S. J. Org. Chem. 2009, 74, 4418−4421. (13) (a) Antus, S.; Kurtán, T.; Juhász, L.; Kiss, L.; Hollósi, M.; Májer, Z. S. Chirality 2001, 13, 493−506. (b) Kim, T. H.; Ito, H.; Hayashi, K.; Hasegawa, T.; Machiguchi, T.; Yoshida, T. Chem. Pharm. Bull. 2005, 53, 641−644. (14) Xiong, L.; Zhu, C.; Li, Y.; Tian, Y.; Lin, S.; Yuan, S.; Hu, J.; Hou, Q.; Chen, N.; Yang, Y.; Shi, J. J. Nat. Prod. 2011, 74, 1188−1200. (15) Yuen, M. S. M.; Xue, F.; Mak, T. C. W.; Wong, H. N. C. Tetrahedron 1998, 54, 12429−12444. (16) Nascimento, I. R.; Lopes, L. M. X.; Davin, L. B.; Lewis, N. G. Tetrahedron 2000, 56, 9181−9193. (17) Shi, Y.-S.; Liu, Y.-B.; Ma, S.-G.; Li, Y.; Qu, J.; Li, L.; Yuan, S.-P.; Hou, Q.; Li, Y.-H.; Jiang, J.-D.; Yu, S.-S. J. Nat. Prod. 2015, 78, 1526− 1535. 1613



Article

(42) López-Sánchez, C.; Á lvarez-Corral, M.; Jiménez-González, L.; Muñoz-Dorado, M.; Rodríguez-García, I. Tetrahedron 2013, 69, 5511− 5516. (43) (a) Matsuo, Y.; Mimaki, Y. Chem. Pharm. Bull. 2010, 58, 587− 590. (b) Chavez, K. J.; Feng, X.; Flanders, J. A.; Rodriguez, E.; Schroeder, F. C. J. Nat. Prod. 2011, 74, 1293−1297. (c) Huang, X.-X.; Zhou, C.-C.; Li, L.-Z.; Li, F.-F.; Lou, L.-L.; Li, D.-M.; Ikejima, T.; Peng, Y.; Song, S.-J. Bioorg. Med. Chem. Lett. 2013, 23, 5599−5604. (d) Gao, X.-M.; Wang, R.-R.; Niu, D.-Y.; Meng, C.-Y.; Yang, L.-M.; Zheng, Y.T.; Yang, G.-Y.; Hu, Q.-F.; Sun, H.-D.; Xiao, W.-L. J. Nat. Prod. 2013, 76, 1052−1057. (e) Yang, D.-T.; Lin, S.-S.; Chen, J.-H.; Yuan, S.-T.; Shi, J.-S.; Wang, J.-S.; Jia, A.-Q. Bioorg. Med. Chem. Lett. 2015, 25, 1976−1978. (44) Jin, H.-G.; Kim, A. R.; Ko, H. J.; Lee, S. K.; Woo, E.-R. Chem. Pharm. Bull. 2014, 62, 288−293. (45) Setiawati, A. Asian Pac. J. Cancer Prev. 2016, 17, 1655−1659.

1614


Download full-text PDF - ACS Publications - American Chemical Society

Recommend Documents