Can Stereoclusters Separated by Two Methylene Groups Be Related

Article Cite This: J. Nat. Prod. 2018, 81, 343−348

pubs.acs.org/jnp

Can Stereoclusters Separated by Two Methylene Groups Be Related by DFT Studies? The Case of the Cytotoxic Meroditerpenes Halioxepines Guillermo Tarazona,† Gonzalo Benedit,† Rogelio Fernández,*,† Marta Pérez,† Jaime Rodríguez,*,‡ Carlos Jiménez,‡ and Carmen Cuevas† †

Medicinal Chemistry Department, PharmaMar S. A., Pol. Ind. La Mina Norte, Avenida de los Reyes 1, 28770, Colmenar Viejo (Madrid), Spain ‡ Departamento de Química, Facultade de Ciencias e Centro de Investigacións Científicas Avanzadas (CICA), Universidade da Coruña, 15071, A Coruña, Spain S Supporting Information *

ABSTRACT: QM/NMR-DFT (quantum mechanics combined with nuclear magnetic resonance parameters calculated by density functional theory approximations) studies allowed us to link two stereoclusters separated by two methylene groups present in the new meroditerpenes halioxepine B (2) and halioxepine C (3) and the known halioxepine (1), isolated from two Indonesian sponges of the genus Haliclona (Reniera). DP4 and DP4+ probabilities were used to discriminate the two diastereotopic arrangements of the two stereoclusters, whose unconnected relative configurations were determined by ROESY and J-based configurational analysis. To confirm the DFT studies, the full relative configuration of 1 was deduced using a mixture of benzene-d6 and pyridine-d5 as the NMR solvent. ROESY measurements connected the two stereoclusters and demonstrated that DFT calculations accurately predict the configuration when two methylenes separate the two stereoclusters. The different arrangements of the distant stereoclusters C-1/C-2/C-7 and C-10/C-15 for compounds 2 and 3 were deduced by DFT calculations and explained the opposite optical rotations observed for the two compounds. Halioxepines B (2) and C (3) display moderate cytotoxicity against different human cancer cell lines.

D

calculations7 have provided a remarkable contribution to the study of the relative configuration of new chemical entities, beyond the limits of X-ray and neutron diffraction methods.8 Although this approach has been commonly used in rigid systems and for close stereogenic centers, it becomes particularly challenging when the two stereoarrangements are separated, for instance, by either contiguous methylenes ([CH2]n), methynes [(CHCH)n], or nonstereogenic quaternary carbons.9 This is the case for the meroditerpene halioxepine (1), where the two stereoclusters A and B are separated by two methylenes, and the lack of spectroscopic relationships between them complicates establishing the full relative configuration (Figure 1). As part of our search for new cytotoxic metabolites from marine sources,10 we describe the isolation of halioxepines B (2) and C (3) as new derivatives of the known meroditerpene halioxepine (1),11 also isolated in this study, from two different samples, collected separately in Indonesia, of a sponge belonging to the genus Haliclona (Reniera).12

uring the structure determination of a new molecular entity, the assignment of its stereochemical features is a crucial part of the process.1 Specifically, the determination of the relative configuration of all of the stereogenic centers in a new compound can be troublesome and particularly problematic if using only NMR spatial correlations (NOE) or by measuring proton−proton and long-distance heteronuclear coupling constants in the well-known J-based configurational analysis (JBCA).2 Other methods that can be effective include chiral derivatization,3 comparison with bibliographic chemical shifts in the Kishi database,4 or the use of molecular alignment media to calculate residual dipolar couplings (RDCs) or residual chemical shift anisotropic data.5 Unfortunately these techniques may not be enough to elucidate the overall relative configuration, especially when the compounds have clusters of stereocenters separated by two or more methylene groups. Over recent years, the use of computational modeling has emerged as a supplementary method to determine the arrangements of these molecules, being able to calculate not only NMR chemical shifts6 but also other physical properties such as circular dichroism or infrared data. Nowadays, the 1H and 13C NMR chemical shift values predicted based on DFT © 2018 American Chemical Society and American Society of Pharmacognosy

Received: September 22, 2017 Published: February 2, 2018 343

DOI: 10.1021/acs.jnatprod.7b00807 J. Nat. Prod. 2018, 81, 343−348

Journal of Natural Products

Article

Table 1. NMR Data of 2 and 3 in CD3OD (500 MHz for 1H and 125 MHz for 13C) 2 pos.

δC, type

1 2 3 4 5

72.0, CH 75.4, CH 138.7, C 129.6, CH 24.4, CH2

6

40.0, CH2

7 8

78.0, C 44.8, CH2

9

21.1, CH2

10 11 12

55.7, CH 74.3, C 42.0, CH2

13

19.5, CH2

14

43.2, CH2

15 16 17 18 19 20 1′ 2′ 3′ 4′ 5′ 6′

36.0, C 21.2, CH3 21.3, CH3 31.6, CH3 22.0, CH3 32.5, CH3 150.8, C 131.3, C 116.3, CH 148.2, C 114.8, CH 116.4, CH

Figure 1. Two separated stereoclusters in halioxepine (1).11

These compounds provide an excellent example of how the stereochemical relationship between distant stereoclusters can be established with a remarkable level of confidence by combining affordable quantum chemical (QM) calculations with sophisticated data processing such as DP413 or DP4+.14

■

RESULTS AND DISCUSSION The (+)-HRESITOFMS of 2 provided the [M + Na]+ ion peak at m/z 455.2742, which agrees with the molecular formula of C26H40O5. The 1H, 13C, and HSQC-DEPT NMR data (Table 1) showed the presence of four sp2 nonprotonated carbons (δC 150.8, 148.2, 138.7, and 131.3), four sp2 methines (δH 6.86/δC 116.3, δH 6.56/δC 116.4, δH 6.52/δC 114.8) characteristic of a 1,2,4-trisubstituted aromatic ring and δH 5.59 /δC 129.6, three sp3 nonprotonated carbons (two of them linked to an oxygen atom at δC 78.0 and 74.3 and another at δC 36.0), seven sp3 methylenes, three sp3 methines (two oxymethines at δH 5.25/ δC 72.0 and δH 4.41/δC 75.4 and another at δH 0.75/δC 55.7), and five methyl groups (δH 1.21/δC 31.6, δH 0.96/δC 22.0, δH 0.79/δC 32.5, δH 0.55/δC 21.3 and the allylic methyl at δH 1.88/ δC 21.2). Five spin systems were determined from the COSY spectrum of 2 that were connected through an HMBC experiment (Figure 2). This information allowed us to complete the planar structure composed of three rings including a 1,2,4-trisubstituted benzene, a tetrahydro-oxepine, and a cyclohexane ring linked through an aliphatic chain, as show in Figure 2. This structure displays the presence of two stereoclusters separated by two methylene groups: stereocluster A, having a substituted cyclohexane ring, and stereocluster B, bearing a tetrahydro-oxepine ring and a benzyl moiety, which were similar to halioxepine (1). Once the planar structure of 2 was completed, the relative configuration of each stereocluster was addressed independently. The 10S*11R* relative configuration in stereocluster A was determined by comparison with similar reported structures.15 Thus, the equatorial disposition of the C-18 methyl group in 2 was deduced from its 13C NMR chemical shift at 31.6 ppm because of the match with literature data for equatorial (i.e., δC 30.7) instead of axial (i.e., δC 23.2) methyl groups in similar cyclohexane rings as shown in Figure 3.16

3

δH, mult (J in Hz) 5.25, d (1.7) 4.41, br s 5.59, d (8.0) 2.39, bt 1.80, m 1.67, m 1.58, m 1.69, ovl 1.29, m 1.60, m 1.07, m 0.75, dd (5.0, 2.4) 1.67, 1.34, 1.81, 1.34, 1.39, 1.18,

m m m m m m

1.88, 0.55, 1.21, 0.96, 0.79,

s s s s s

6.86, d (2.8) 6.52, dd (8.6, 2.8) 6.56, d (8.5)

δC, type 69.8, CH 75.1, CH 137.3, C 130.4, CH 23.8, CH2

41.5, CH2 77.4, C 36.0, CH2 31.4, CH2

δH, mult (J in Hz) 5.00, dd (2.2, 2.0) 4.40, d (2.2) 5.65, br d (8.5) 2.38, ddd (12.1, 8.5, 2.8) 1.85 m 1.86 m 1.64 m 1.31, 1.18, 1.41, 1.17,

m m m m

41.3, C 141.1, C 124.9, CH

5.38, dd (5.5, 1.8)

26.7, CH2

1.92, m

28.2, CH2

1.41, m

34.5, CH 21.5, CH3 22.8, CH3 21.5, CH3 19.8, CH3 16.2, CH3 189.1, C 151.2, C 134.5, CH 189.1, C 137.5, CH 137.3, CH

1.64, 1.83, 0.84, 0.79, 1.60, 0.79,

m d (1.9) s s d (1.8) d (6.6)

6.87, dd (1.9, 1.9) 6.80, m 6.80, m

Figure 2. Key COSY and HMBC correlations of 2.

Figure 3. Bibliographic comparison of δC values in the cyclohexane ring (stereocluster A) in compound 2. 344



Article

carbon and proton shifts were included. The comparison of both sets of NMR chemical shifts using the improved modified method DP4+ recently published by Sarotti and co-workers14 showed again 99.62% (1H NMR), 99.98% (13C NMR), and 100% (both 1H and 13C NMR) probabilities for the 2(7R*10S*) isomer, which confirmed the earlier statistical results. Therefore, the relative configuration of 2 is proposed as 1S*2S*7R*10S*11R*. A similar strategy to that described for 2 was then used to relate the two separated stereoclusters in the known halioxepine (1), whose relative configuration had been proposed as either 1S*,2S*,7R*,10S*,15S* 1(7R*10S*) or 1S*,2S*,7R*,10R*,15R* 1(7R*10R*) (Figure 6). Both stereo-

For the second stereocluster, JBCA methodology and NOE correlations for 2 showed that the deduced 1S*2S*7R* relative configuration of the stereogenic centers in halioxepine B (2) was the same as that for halioxepine (1).11 Thus, the strong NOE correlation between H-2 at 4.41 ppm and H3-17 at 0.55 ppm confirmed the cis configuration between these protons. Furthermore, the relative relationship between stereocenters at C-1 and C-2 was solved using the JBCA method (Figure 4).

Figure 4. Key NOE correlations and JBCA used to determine the relative configuration of stereocluster B in compound 2.

The homonuclear coupling constant between the H-1 and H-2 protons of 1.7 Hz, indicating a gauche arrangement, and the small two-bond heteronuclear coupling constants, measured through HSQC-HECADE and J-HMBC bidimensional experiments corroborated the syn disposition of these stereocenters. At this point, separation of stereoclusters A and B by two methylene groups (C-8 and C-9) prevented us from establishing the stereorelationship between them, with two relative configurations being possible for 2: 1S*,2S*,7R*,10R*,11S* or 1S*,2S*,7R*,10S*,11R*. A similar situation was found in the description of the known halioxepine (1), also isolated from this sponge. In order to solve this problem, DFT computational calculations were carried out on the two possible diastereoisomers, named as 2(7R*10R*) and 2(7R*10S*), to explore the connection between the two fragments A and B (Figure 5).

Figure 6. DFT studies and statistical DP4 and DP4+ parameters found for 1.

arrangements were submitted to computational studies by a standard conformational search using Maestro software, providing 18 conformers for 1(7R*10S*) and 13 for 1(7R*10R*). Then, theoretical calculations of chemical shifts using B3LYP/6-31G+(d,p)//GIAO MPW1PW91/6-311++G (2d,p) energy levels and DP4+ and DP4 probabilities were applied to both diasteroisomers. As a result, we found that isomer 1(7R*10R*) was the most likely structure using 1H and 13 C NMR data, with 98.9% probability. Isomer 1(7R*10R*) was also predicted as the most likely structure with 100% probability when using both carbon and proton data and Sarotti’s DP4+ parameter, including both 1H and 13C NMR data being 100% and 96.86%, respectively. It is worth noting that the DP4 parameter for 1H points out an 82.8% probability in favor of the 1(7R*10S*) diasteroisomer, although when both NMR nuclei are taken into consideration, a 98.9% probability is found in favor of 1(7R*10R*) (Supporting Information). Therefore, we propose 1S*,2S*,7R*,10R*,15R* as the full relative stereostructure for the known halioxepine (1). In order to corroborate the proposed relative configuration for 1, we tried to achieve a diastereotopic differentiation of the H-8 and H-9 methylene protons using a mixture of deuterated NMR solvents and then performing a spatial correlation with positions C-7 and C-10. In this way, a mixture of benzene-d6 and pyridine-d5 in a 9:1 ratio allowed us to distinguish H-8a, H8b, H-9a, and H-9b, which were unambiguously assigned by COSY, HMBC, and HSQC experiments and proton−proton coupling constants. In addition, various NOE correlations among the protons shown in Figure 7 agreed with the connection of the A and B stereoclusters in the 1(7R*10R) model. This experimental fact confirmed that the DFT-NMR calculations made for this compound are indeed able to distinguish two diastereoisomeric dispositions separated by two methylene moieties, and therefore relate two stereoclusters. From a second re-collection of the same sponge, we were able to isolate two meroterpenes. One of them showed the same NMR chemical shifts, MS, and specific rotation as

Figure 5. DFT studies and statistical DP4 and DP4+ parameters found for 2.

Conformational searches using Maestro software with an energy window of 3 kcal/mol afforded 12 conformers for 2(7R*10R*) and nine for 2(7R*10S*). Minimization of all of these conformers was done using DFT calculations at the B3LYP/6-31G+(d,p) level. A second step in the QM calculation included an energy selection of the minimized conformers in a 3 kcal/mol window, followed by application of a Boltzmann distribution. Finally, 1H and 13C NMR chemical shifts were predicted using the GIAO method in MeOH as solvent by DFT calculations at the MPW1PW91/6-311++G(2d,p) level of theory. Statistical analyses of the calculated and experimental chemical shifts (in CD3OD, Supporting Information) by using DP4 probability13 determined that the most probable structure was the 2(7R*10S*) isomer with a probability of 80.2% based on 1H NMR, 99.9% based on 13C NMR, and 100% when 345



Article

between quinone 3 and hydroquinone 1 can be explained by the different relative arrangement between the C-1/C-2/C-7 and the C-10/C-15 stereoclusters. Finally the modified Mosher’s method was performed using quinone 3 to establish the absolute configuration at C-1. Thus, compound 3 was treated with (R)- and (S)-MTPA-Cl separately, and analysis of the 1H NMR spectra of the corresponding esters allowed us to conclude that the absolute configuration of C-1 is S (Figure 8). Therefore, the absolute

Figure 8. MTPA ester analysis (values are ΔδH = δS − δR) and the resulting absolute configuration of 3.

Figure 7. Connecting the two isolated stereoclusters of 1 by ROESY.

halioxepine (1). The molecular formula of the second meroterpene, compound 3, was determined to be C26H36O4 by (+)-HRESITOFMS. The 1H and 13C NMR chemical shifts of 3 were very similar to those reported for 1. The presence of two signals in the 13C NMR spectrum at δC 189.1 (nonprotonated carbons) in 3 was indicative of two carbonyl groups of a p-quinone instead of the hydroquinone moiety present in 1. A full analysis of 1H, 13C, edited-HSQC, and HMBC spectra of 3 (Table 1) corroborated the planar structure of 3. Surprisingly, the specific rotation determined for 3 ([α]D −23.6) had the opposite sign of that of 1 ([α]D +14.0), suggesting a possible difference in their relative configurations. A literature search for pairs of hydroquinone and quinone compounds yielded six examples, arenol−arenone,17 cacospongins B−cacospongin C,18 avarol−avarone,19 neoavarol−neoavarone,19 zonarol−zonarone,20 and isozonarol−isozonarone,20 and in all cases each pair of compounds had the same sign for their specific rotation values (Supporting Information). In order to compare 1 and 3, we attempted the conversion of the hydroquinone 1 to the corresponding quinone with DDQ oxidation in dioxane (10 °C), but unfortunately a mixture of unknown compounds was obtained and no information could be gathered. The relative configuration in each individual stereocluster of 3 was shown to be the same as 1. Indeed, the C-1/C-2/C-7 fragment displayed the same J value between H-1 and H-2 (2.0 Hz) and the NOE correlation between H-2 (δH 4.40)/H3-17 (δH 0.84). Furthermore, the C-10/C-15 fragment in 3 presented the following NOE correlations, H-15 (δH 1.64)/H3-18 (δH 0.79) and H-9 (δH 1.41;1.17)/H3-20 (δH 0.79), which were also detected in 1. Again, the complete relative configuration of compound 3 was established by computational methods that allowed us to link both stereoclusters. Thus, conformational analysis of the two possible diastereoisomers of 3 as in 1, 1S*,2S*,7R*,10S*,15S* and 1S*,2S*,7R*,10R*,15R*, followed by DFT-NMR calculations (energy//NMR B3LYP/6-31G +(d,p)//GIAO MPW1PW91/6-311++G(2d,p)) using the DP4 probability approximation was applied. As a result, the stereocluster combination 1S*,2S*,7R*,10S*,15S* showed the best DP4+ fitting (100% for 1H and 70% for 13C; 100% for both). Thus, the opposite sign of the specific rotation values

configuration for 3 can be defined as 1S,2S,7R,10S,15S. The new analogues 2 and 3 together with the known halioxepine (1) were tested for their potential cytotoxic activity against A549 human lung carcinoma cells, MDA-MB-231 human breast adenocarcinoma cells, PSN-1 human pancreatic adenocarcinoma cells, and HT-29 human colorectal carcinoma cells (Table 2). Halioxepine C (3) showed moderate but nonspecific cytotoxicity against all of the tested tumor cells.21 Table 2. Cytotoxic μM GI50 Values of the Isolated Halioxepines23,24 1 2 3

A549

HT29

MDA-MB-231

PSN-1

8.7 6.7 1.1

7.0 8.8 1.2

1.3 7.6 1.1

9.6 8.3 1.2

Two new analogues of the meroditerpene halioxepine (1) were isolated from two separate samples collected in Indonesia of a sponge belonging to the Haliclona genus, and their structures determined by mass spectrometry and NMR experiments. The relative configuration of each of the two stereoclusters present in compound 2 was solved by NOE correlations, bibliographic comparison, JBCA, and computational calculations. Additionally, the relationship between the two distant stereoclusters in halioxepine B (2), the known halioxepine (1), and halioxepine C (3) was resolved by an intensive study of their experimental chemical shifts and the theoretical chemical shifts obtained by DFT calculations. Finally, the absolute configuration of 3 was determined by application of the modified Mosher’s method. We believe that DFT-NMR analysis is a very useful tool in NMR-based structural studies. These DFT calculations generate conformational information and theoretical NMR chemical shift values that can be compared with the experimental data. Although it is well known how difficult it is to relate two distinct stereoclusters separated by more than one methylene group, we have shown that this approach can give accurate results for the halioxepines 1−3. Particularly, in halioxepine (1) we have been able to show that the isomer deduced from computational calculations can be corroborated by diasterotopic differentiation of the four protons belonging to 346



Article

437.2655 [M + Na]+ (calcd for C26H40O5Na, 437.2662); other spectroscopic data were in agreement with published values.11 Halioxepine B (2): pale purple glass; [α]D +2.0 (c 0.045, MeOH); UV (MeOH) λmax (log ε) 205 (5.97), 293 (2.35) nm; IR νmax 2924, 2853, 2361, 2341, 1735, 1653, 1463, 1204, 1123, 1073 cm−1; 1H NMR (500 MHz) and 13C NMR (125 MHz), Table 1; HREITOFMS m/z 455.2742 [M + Na]+ (calcd for C26H40O5Na, 455.2768). Halioxepine C (3): pale brown oil; [α]D −23.6 (c 0.053, MeOH); UV (MeOH) λmax (log ε) 247 (1.20), 301 (0.30) nm; IR νmax 3738, 3372, 2962, 2923, 2324, 1656, 1456, 1205, 947, 833 cm−1; 1H NMR (500 MHz) and 13C NMR (125 MHz), Table 1; HRESITOFMS m/z 395.2602 [M + H − H2O]+ (calcd for C26H35O3, 395.2581). Preparation of the (S)- and (R)-MTPA Ester Derivatives of 3 by Modified Mosher’s Reaction. Compound 3 (1.00 mg, 2.43 μmol) was dissolved in pyridine-d5 (500 μL) and transferred to an NMR tube. R-(−)-MTPA-Cl (10 μL, 45 μmol) was added to the NMR tube, and the mixture shaken carefully. After the reaction was completed (approximately 8 h), the 1H NMR spectrum was recorded. The preparation of the R-MTPA ester was done repeating the same process using S-(+)-MTPA-Cl instead of R-(−)-MTPA-Cl. (S)-MTPA ester of 3 (3a): 1H NMR (500 MHz, pyridine-d5) δ 7.22 (dd, J = 8.6, 2.6 Hz, H-5′), 7.01 (d, J = 8.7 Hz, H-6’), 6.97 (dd, J = 2.6, 1.0 Hz, H-3′), 6.69 (d, J = 1.6 Hz, H-1), 5.67 (m, H-4), 4.68 (br s, H2), 2.42 (m, H-5), 1.90 (s, H-16). (R)-MTPA ester of 3 (3b): 1H NMR (500 MHz, pyridine-d5) δ 7.36 (m, H-3′), 7.27 (dd, J = 8.6, 2.5 Hz, H-5′), 7.03 (d, J = 8.6 Hz, H-6′), 6.70 (d, J = 1.6 Hz, H-1), 5.51 (m, H-4), 4.57 (s, H-2), 2.34 (m, H-5), 1.86 (d, J = 1.6 Hz, H-16).

the two methylene groups. This spectroscopic fact helped in reaching a very reasonable result, although special care must be taken in the use of this QM approximation.

■

EXPERIMENTAL SECTION

General Experimental Procedures. Optical rotations were determined using a Jasco P-1020 polarimeter. UV spectra were recorded using an Agilent 8453 UV−vis spectrometer. IR spectra were obtained with a PerkinElmer Spectrum 100 FT-IR spectrometer with ATR sampling. NMR spectra were recorded on a Varian “Unity 500” spectrometer at 500/125 MHz (1H/13C) and a Bruker Avance III 600 that was equipped with a TCI inverse-detection cryoprobe (5 mm). Chemical shifts were reported in ppm using residual CD3OD (δH 3.31 for 1H and δC 49.0 for 13C) as an internal reference. HRESITOFMS was performed on an Agilent 6230 TOF LC/MS chromatograph spectrometer. (+)-ESIMS spectra were recorded using an Agilent 1100 Series LC/MSD spectrometer. Conformational searches were performed by using the corresponding module implemented in the Maestro Quantum mechanical software. The OPLS 2005 force field with MeOH as solvent was used, and torsional enhanced sampling with 1000 or 10 000 steps was fixed using an energy window of 3 kcal/mol. Molecular geometry optimizations were performed at the DFT theoretical level using the Gaussian 09W package with B3LYP/6-31+G(d,p) for energy and frequency calculation and the mPW1PW91/6-311++G(d,2p) for proton and chemical shift calculations. Chemical shifts were calculated following Tantillo and co-workers’ recommendations,22 using the scaling factor protocol with the combination B3LYP/6-31+G(d,p) (gas phase)//mPW1PW91/6-311+G(2d,p)(giao, scrf) slope: −1.0754, intercept: 31.8463 for 1H and slope: −1.0399, intercept: 186.5993 for 13C. To calculate Boltzmann populations, saddle points, and final chemical shifts, Hoye et al. Phyton scripts were used.6 DP4 statistical parameters were obtained using Java script in http://wwwjmg.ch.cam.ac.uk/tools/nmr/DP4/. Biological Material. The two specimens of Haliclona (Reniera) sp. were collected in Indonesia. The first sponge (30 g) was collected by hand using rebreather diving near Tunumanu, (125°32.978′ E/08° 17.994′ S) at a depth of 50 m and frozen immediately after collection. The second sponge (57 g) was collected by hand using rebreather diving near Mapia anchorage, (124°45.965′ E/01°27.574′ N) at a depth of 40 m and frozen immediately after collection. Haliclona (Reniera) Schmidt, 1862: delicate sponge with an encrusting cushion shape; laterally spreading masses up to 12 cm in diameter and 1 cm thick; oscules at the top of elevations (oscular chimneys); color light pink, consistency soft, very fragile; surface smooth; ectosomal skeleton regular, tangential, unispicular, isotropic reticulation; choanosomal skeleton also regular, delicate, unispicular, isotropic reticulation; oxeas blunt-pointed of size 140−170 × 6−8 μm; spongin moderate, at the nodes of the spicula. The voucher specimens coded as ORMA142266 are deposited at PharmaMar. Extraction and Isolation. The sponges Haliclona (Reniera) sp. (30 and 57 g each) were triturated and exhaustively extracted with MeOH−CH2Cl2 (1:1, 3 × 300 mL). The combined extracts were concentrated to yield 2.4 and 3.0 g of organic extracts, respectively. The extracts were subjected to VLC on Lichroprep RP-18 with a stepped gradient of H2O−MeOH−CH2Cl2 mixtures. Fractions that eluted with MeOH (87.2 and 270 mg) were purified by semipreparative HPLC. In the case of the first collection the HPLC conditions were as follows: XTerra Prep RP18 10 μm, 10 × 150 mm, H2O + 0.04% TFA/CH3CN + 0.04% TFA gradient from 60% to 100% over 30 min, flow 3 mL/min, UV detection, to give halioxepine B (2, 3.3 mg, 0.011% wet wt) and halioxepine (1, 13.7 mg, 0.047% wet wt). From the extract coming from the second collection, the HPLC conditions were as follows: XBridge Prep C18 5 μm, 10 × 150 mm, H2O + 0.1% TFA/CH3CN + 0.1% TFA gradient from 50% to 100% over 25 min, flow 3 mL/min, UV detection, to give halioxepine C (3, 7.5 mg, 0.013% wet wt) and halioxepine (1, 2.0 mg, 0.004% wet wt). Halioxepine (1): from the second extract [α]D +14.0 (c 0.051, MeOH), literature +49.0 (c 0.5, MeOH); HRESITOFMS m/z

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.7b00807. 1 H, 13C, gHSQC, gCOSY, gHMBC, TOCSY, and ROESY NMR spectra and DFT computational data of halioxepines (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Guillermo Tarazona: 0000-0002-8724-3851 Jaime Rodríguez: 0000-0001-5348-6970 Carlos Jiménez: 0000-0003-2628-303X Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We gratefully acknowledge the help of our PharmaMar colleagues C. de Eguilior, for collecting the marine samples, S. Bueno, for determining the sponge taxonomy, B. de Castro, for the realization of the biological assays, S. Munt, for revision of the manuscript, and S. González, for performing the NMR experiments. The present research was financed by FEDER funds with a grant from Ministerio de Economi á y Competitividad of Spain (RTC-2016-4611-1). J.R. and C.J. acknowledge Xunta de Galicia and CESGA for the computational facilities.

■

REFERENCES

(1) Bifulco, G.; Dambruoso, P.; Gomez-Paloma, L.; Riccio, R. Chem. Rev. 2007, 107, 3744−3779. 347



Article

(2) Matsumori, N.; Kaneno, D.; Murata, M.; Nakamura, H.; Tachibana, K. J. Org. Chem. 1999, 64, 866−876. (3) Seco, M.; Quiñoá, E.; Riguera, R. Chem. Rev. 2004, 104, 17−117. (4) Lee, J.; Kobayashi, Y.; Tezuka, K.; Kishi, Y. Org. Lett. 1999, 1, 2181−2184. (5) Liu, Y.; Saurí, J.; Mevers, E.; Peczuh, M. W.; Hiemstra, H.; Clardy, J.; Martin, G. E.; Williamson, R. T. Science 2017, 356, eaam5349. (6) Willoughby, P. H.; Jansma, M. J.; Hoye, T. R. Nat. Protoc. 2014, 9, 643−660. (7) Lodewyk, M. W.; Siebert, M. R.; Tantillo, D. J. Chem. Rev. 2012, 112, 1839−1862. (8) Siskos, M. G.; Choudhary, M. I.; Gerothanassis, I. P. Org. Biomol. Chem. 2017, 15, 4655−4666. (9) (a) Cen-Pacheco, F.; Rodríguez, J.; Norte, M.; Fernández, J. J.; Hernández Daranas, A. Chem. - Eur. J. 2013, 19, 8525−8532. (b) Wang, C.-X.; Chen, G.-D.; Feng, C.-C.; He, R.-R.; Qin, S.-Y.; Hu, D.; Chen, H.-R.; Liu, X.-Z.; Yao, X. S.; Gao, H. Chem. Commun. 2016, 52, 1250−1253. (10) (a) Rodríguez, J.; Nieto, R. M.; Hunter, L. M.; Crews, P. Tetrahedron 1994, 50, 11079−11099. (b) Anta, C.; González, N.; Santafe, G.; Rodríguez, J.; Jiménez, C. J. Nat. Prod. 2002, 65, 766−768. (11) Trianto, A.; Hermawan, I.; de Voogd, N. J.; Tanaka, J. Chem. Pharm. Bull. 2011, 59, 1311−1313. (12) World Porifera Database: http://www.marinespecies.org/ porifera/porifera.php?p=sourcedetails&id=7515. (13) Smith, S. G.; Goodman, J. M. J. Am. Chem. Soc. 2010, 132, 12946−12959. (14) Grimblat, N.; Zanardi, M. M.; Sarotti, A. M. J. Org. Chem. 2015, 80, 12526−12534. (15) (a) Nishizawa, M.; Takenaka, H.; Hayashi, Y. J. Org. Chem. 1986, 51, 806−813. (b) Habtemariam, S.; Gray, A. I.; Waterman, P. G. J. Nat. Prod. 1993, 56, 140−143. (16) Crews, P.; Jiménez, C.; O’Neil Johnson, M. Tetrahedron 1991, 47, 3585−3600. (17) Schmitz, F. J.; Lakshmi, V.; Powell, D. R.; Van der Helm, D. J. Org. Chem. 1984, 49, 241−244. (18) Tasdemir, D.; Concepción, G. P.; Mangalindan, G. C.; Harper, M. K.; Hajdu, E.; Ireland, C. M. Tetrahedron 2000, 56, 9025−9030. (19) Iguchi, K.; Sahashi, A.; Yamada, Y.; Kohno, J. Chem. Pharm. Bull. 1990, 38, 1121−1123. (20) Laube, T.; Schröder, J.; Stehle, R.; Seifert, K. Tetrahedron 2002, 58, 4299−4309. (21) Vichai, V.; Kirtikara, K. Nat. Protoc. 2006, 1, 1112−1116. (22) (a) Lodewyk, M. W.; Siebert, M. R.; Tantillo, D. J. Chem. Rev. 2012, 112, 1839−1862. (b) Lodewyk, M. W.; Soldi, C.; Jones, P. B.; Olmstead, M. M.; Larrucea, J. R.; Shaw, J. T.; Tantillo, D. J. J. Am. Chem. Soc. 2012, 134, 18550−18553. (23) Skehan, P.; Storeng, R.; Scudeiro, D.; Monks, A.; McMahon, J.; Vistica, D.; Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82, 1107−1112. (24) Shoemaker, R. H. Nat. Rev. Cancer 2006, 6, 813−823.

348


Can Stereoclusters Separated by Two Methylene Groups Be Related

Recommend Documents