JOURNAL OF SEPARATION SCIENCE
J S S
ISSN 1615-9306 · JSSCCJ 38 (14) 2371–2558 (2015) · Vol. 38 · No. 14 · August 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
14 15
Vol. 38 (2015) · No. 14 · Pages 2371–2558
Methods Chromatography · Electroseparation
www.jss-journal.com
Applications Biomedicine · Foods · Environment
JSSC_38_14_cover.indd 1
10/07/15 4:54 PM
2545
J. Sep. Sci. 2015, 38, 2545–2550
Alain Krief1 Melissa Dunkle2 Masoud Bahar3 Patrick Bultinck4,5 Wouter Herrebout5,6 Pat Sandra2 1 Department
of Chemistry, University of Namur, Namur, Belgium 2 Research Institute for Chromatography (R.I.C.), Kortrijk, Belgium 3 Department of Plant Protection & Agricultural Biotechnology, College of Agriculture, Isfahan University of Technology, Isfahan, Iran 4 Department of Inorganic and Physical Chemistry, Ghent University, Krijgslaan 281 (S3), Ghent, Belgium 5 Scaldis Spectroscopy www.scaldis-spectroscopy.com 6 Department of Chemistry, University of Antwerp, Antwerp, Belgium
Research Article
Elucidation of the absolute configuration of rhizopine by chiral supercritical fluid chromatography and vibrational circular dichroism The absolute configuration of rhizopine, an opine-like natural product present in nitrogenfixing nodules of alfalfa infected by rhizobia, is elucidated using a combination of state-ofthe-art analytical and semi-preparative supercritical fluid chromatography and vibrational circular dichroism spectroscopy. A synthetic peracetylated racemate was fractionated into its enantiomers and subjected to absolute configuration analysis revealing that natural rhizopine exists as a single enantiomer. The stereochemistry of non-derivatized natural rhizopine corresponds to (1R,2S,3R,4R,5S,6R)-4-amino-6-methoxycyclohexane-1,2,3,5-tetraol. Keywords: Absolute configuration/Infrared spectroscopy/Rhizopine/Supercritical fluid chromatography/Vibrational circular dichroism DOI 10.1002/jssc.201500138
Received February 2, 2015 Revised April 13, 2015 Accepted April 15, 2015
1 Introduction Rhizopine is a natural compound produced in nitrogen-fixing nodules [1] resulting from infection of the root hairs of alfalfa (Medicago sativa) by rhizobia bacteria (Sinorhizobium meliloti and Rhizobium leguminosarum bv. Viciae) [2]. This infection occurs in soil with low bioactive nitrogen content and is known to favor the growth of alfalfa as well as the freeliving rhizobia, which are not only able to catabolize rhizopine but also use it as a source of carbon and nitrogen [3, 4]. The structure of rhizopine has been postulated as 4amino-6-methoxycyclohexane-1,2,3,5-tetraol (Scheme 1) more than 25 years ago but has not yet been firmly established. Moreover, although its relative stereochemistry has been proposed as the one shown in Scheme 1 [1], there remain some doubts whether rhizopine in nature exists as a racemate (1 + 1 ) or as a single enantiomer (1 or 1 ), hence no absolute configuration (AC) information is available nor shown in Scheme 1. Correspondence: Prof. Alain Krief, University of Namur, Rue de Bruxelles 61, 5000 Namur, Belgium E-mail: [email protected]
Abbreviations: AC, absolute configuration; LCPL, left circularly polarized light; RCPL, right circularly polarized light; VCD, vibrational circular dichroism
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ten years ago one of us described the synthesis of racemic 4-amino-6-methoxycyclohexane-1,2,3,5-tetraol [5] and some preliminary experiments indicated that natural rhizopine and its synthetic racemate exhibited similar biological properties. This, however, is no guarantee that natural rhizopine does indeed have exactly the same structure as the synthetic product and moreover that it would also occur as a racemate. We here report on the structure, relative and absolute configuration of natural rhizopine by means of analytical and semi-preparative supercritical fluid chromatography (SFC) and spectroscopic methods including vibrational circular dichroism (VCD) spectroscopy. State-of-the-art chromatography, allowing to effectively separate enantiomers, is followed by VCD that is used to establish the AC of the compound in the fractions of interest. The use of VCD as a technique to establish the stereochemistry of chirally separated molecules is not yet very widespread albeit very promising. VCD is a technique based on absorption of IR radiation although using circularly polarized light. For achiral molecules, the absorption of left circularly polarized light (LCPL) is the same as for right circularly polarized light (RCPL) and hence also for unpolarized light. IR spectra therefore consist of bands that all have the same sign. In the case of chiral molecules, the handedness of the molecule (e.g., the question whether a single chiral center has R or S configuration) causes a difference in interaction with the IR light of different handedness. For instance, the R enantiomer may interact slightly more strongly (or weakly) with RCPL than with LCPL. The VCD spectrum is a plot of the difference in absorption between RCPL and LCPL by the molecule of interest. For some bands, RCPL is slightly more absorbed than LCPL leading to a negative difference RCPL– LCPL whereas for a different band it may be opposite. VCD spectra therefore can have both signs for every band. The
www.jss-journal.com
2546
J. Sep. Sci. 2015, 38, 2545–2550
A. Krief et al.
Scheme 1. Structure of rhizopine (1) and its peractelylated derivative (2) in both enantiomeric forms ( and ) coherent with the relative stereochemistry of rhizopine [1].
difference in absorption is usually very small and often only 10−4 of the absorption in regular IR spectroscopy. VCD is a very configuration- and conformation-sensitive technique hence its use to probe stereochemistry [6–10]. VCD is therefore related to IR spectroscopy as electronic circular dichroism is related to UV–VIS spectroscopy albeit VCD has much higher information content due to the much larger number of vibrational transitions compared to electronic transitions. VCD spectra are, however, not easy to interpret and their interpretation is carried out using quantum chemical calculations. Briefly, an absolute configuration is assumed in a quantum chemical calculation of a VCD spectrum and if a match between the calculation and the experiment is found, this allows concluding that the absolute configuration of the experimental sample matches that used in the calculation. If the match is insufficiently good, calculations are performed for a different absolute configuration until the match is found. Of special importance here is that enantiomers have mirror image VCD spectra and that the calculated VCD spectra should be exactly the same for two enantiomers, but mirror images. Despite experimental and theoretical challenges, VCD has become one of the premier methods for the determination of the AC of chemical compounds. The number and complexity of structures analyzed more or less routinely con-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tinually increases [11–17], making VCD a method of choice for AC determination of complex natural products such a rhizopine.
2 Materials and methods 2.1 Materials and reagents A sample containing ca. 50 g of natural rhizopine (with unknown stereochemistry) was isolated in Isfahan, Iran using the original protocol [1] after inoculation of alfalfa roots by Sinorhizobium meliloti, strain L5–30. Racemic synthetic rhizopine was prepared as in ref. [ 5]. Both samples were derivatized into the peracetyl derivatives (2 in Scheme 1) using the following protocols. rac-Rhizopine hydrochloride (3-O-methyl-scyllo-inosamine hydrochloride (100 mg, 0.44 mmol) [5] and sodium acetate (150 mg, 1.8 mmol), dissolved in acetic acid anhydride (0.9 mL, 9.8 mmol) are heated at reflux for 1.5 h. After cooling to room temperature, water (25 mL) is added and the resulting solution is extracted with dichloromethane (50 mL, four times). The resulting solution is washed with brine, dried over sodium sulfate, filtered and then evaporated, under low pressure on a rotatory evaporator, leaving a crude product
www.jss-journal.com
J. Sep. Sci. 2015, 38, 2545–2550
Other Techniques
2547
Figure 1. SFC chromatograms of the peracetylated synthetic racemate (A) and the peracetylated natural extract (B).
(180 mg). The latter was purified by chromatography on a silica gel plate 7747, 60PF254 from Merck (Overijse, Belgium) using ethyl acetate/hexane mixture (2:1) leading to the peracetylated rac-rhizopine (80 mg). Spectral details are as follows: IR (IR, BIO-RAD FTS-165, Temse, Belgium) max cm−1 (KBr) 3286, 3091, 2993, 2943, 2851, 1758, 1661, 1560, 1436, 1371, 1314, 1221, 1159, 1083, and 1034; 1 H NMR ␦H (JNM EX400, from JEOL, Zaventem, Belgium, 400 MHz, CDCl3 as solvent, TMS as internal standard) ␦ ppm : 5.64 (d, 1 H, J : 10 Hz), RNHAc); 5.21 (t, 1 H, J : 10 Hz), 5.13 (t, 1 H, J: 10 Hz), 5.00 (t, 1 H, J: 10 Hz), 4.96 (t, 1 H, J: 10 Hz), 4.33 (dd, 1 H, J1 : 11 Hz et J2 : 11 Hz) five cyclic hydrogen atoms, 3.49 (m, 4 H, one cyclic hydrogen + ROMe); 2.12 (s, 3 H), 2,10 (s, 3 H), 2.04 (s, 3 H), 2,03 (s, 3 H) hydrogen atoms from four acetoxy groups (ROC(O)Me); 1,92 (s, 3 H, RNHC(O)Me). 13 C NMR (100,4 MHz, CDCl3 as solvent and internal standard) ␦ ppm: 169.71, 79.93, 71.97, 70.52, 60.79, 51.47, 22.99, 20.50. The same procedure was carried out on the crude extract of natural rhizopine (50 g) using an excess of reagents (acetic acid anhydride, 0.45 mL, 4.9 mmol) and sodium acetate (75 mg, 0.9 mmol) and the crude mixture obtained after extraction has been directly used for SFC separation. All used solvents and reagents were from Sigma–Aldrich, Bornem, Belgium. For analytical chromatographic analysis 5 mg of the synthetic sample was dissolved in 10 mL of acetone, for preparative separation the concentration was tenfold higher. The peracytelated natural rhizopine was dissolved in 100 L of acetone. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Mass spectra corresponding to (A) peak 1 in the peracetylated synthetic racemate, (B) peak 2 in the peracetylated synthetic racemate and (C) the peak of the peracetylated natural rhizopine that, in retention time, corresponds to peak 1 of the peracetylated synthetic racemate (see Fig.1).
2.2 Instrumentation and operating conditions A 1260 Infinity Analytical SFC System combined with a 6130 Quadrupole mass spectrometer (MS) from Agilent Technologies (Brussels, Belgium) was used in this study. The chiral separation was performed on a Chiral-Pak AD-H column (250 L × 4.6 mm ID, 5 m particles) from Chiral Technologies, Illkirch, France. Carbon dioxide was of Alphagaz quality from Air Liquide, Liege, Belgium. The modifier was methanol with 0.05% TFA and the gradient applied was 5– 40% in 0–10 min. The flow rate was 2.0 mL/min, the outlet pressure 120 bar and the temperature 40⬚C. For analytical work, the injection volume was 5 L while 100 L was injected for semi-preparative work. For analytical SFC an APCI source was used and the MS settings were: make-up MeOH at 0.2 mL/min, scan range 100 to 550 amu, +3000 V, drying gas www.jss-journal.com
2548
A. Krief et al.
Figure 3. Stable conformations of 4-acetamido-6-methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 calculated at the B3LYP/ cc-pVTZ SCRF (solvent = chloroform) level. The Boltzmann populations of the conformations, obtained using standard enthalpies in solution, are 95.3% (left) and 4.7% (right).
at 8 L/min and 350⬚C, nebulizer at 35 psig and Vap at 325⬚C. The MS was disconnected for semi-preparative application and fractions were directly collected after the backpressure regulator. IR and VCD spectra were obtained using a ChiralIR-2X dual PEM spectrometer (BioTools, Jupiter, FL, USA). For the spectroscopic characterization of the enantiomer of the synthetic sample considered (see Section 3), a solution containing 2.0 mg of the compound was dissolved in deuterated chloroform in a 100 micron liquid cell equipped with BaF2 windows. For the solution 80 000 scans were recorded at 4 cm−1 resolution and averaged. In addition, to allow background subtraction, a spectrum of the pure solvent was recorded. As explained above, proper assignment of the AC of the chiral molecules in the fractions requires quantum chemical calculations for relevant absolute configurations. Conformational analyses were performed using molecular mechanics force fields instead of direct ab initio potential energy hypersurface exploration. A combination of MMFF, MMFF94S, and SYBYL force fields was used, applying the search algorithms from the commercially available software packages Conflex (‘reservoir filling’), Spartan (Monte Carlo stochastic search) and ComputeVOA (Monte Carlo stochastic search). Combination of the resulting conformer libraries led to 328 unique geometries. For all geometries obtained, full geometry optimizations were then initiated at the B3LYP/ccpVTZ and B3PW91/cc-pVTZ levels, using a SCRF = IEFPCM [18] model to account for solvent polarization (chloroform, g = 4.809). All ab initio calculations were performed using Gaussian09 [19]. Boltzmann-weighted IR and VCD spectra were obtained assuming Lorentzian band profiles with a full width at half height (FWHH) of 15 cm−1 . The Boltzmann populations used were based on the standard enthalpies obtained.
3 Results and discussion 3.1 Separation by chiral SFC The analytical SFC analysis of the peracetylated racemic synthetic sample 2, showed two well separated compounds in the total ion chromatogram (TIC) eluting at 3.4 min (peak 1) C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2545–2550
and at 6.7 min (peak 2), respectively (Fig. 1A). The resolution (Rs ) is higher than 10. Note that this resolution is only achieved for the acyl derivative on the chiral selector amylose tris(3,5-dimethylphenylcarbamate). No chiral separation was obtained for the underivatized racemate analyzed on the same column or on a cellulose based stationary phase [20]. The mass spectral data for peak 1 and peak 2 are shown in Fig. 2 A and B, and both spectra are identical. The main ion is m/z 404.2 corresponding to the [M+H]+ of 2. The chromatogram of the peracetylated natural product sample (rhizopine, Iran) analyzed under the same conditions exhibited several peaks with a major peak eluting at the same retention time as peak 1 of the synthetic sample (Fig. 1B). The mass spectrum of this peak agrees with the mass spectra of the two enantiomers of the synthetic sample (Fig. 2C) meaning first of all that the composition of the natural rhizopine and the synthetic products is the same. Second, as the retention time of the main peak of natural rhizopine is the same, the stereochemistry of the structure that corresponds to this peak equals that of the structure that corresponds to peak 1 in the synthetic sample. Fraction collection was then performed on the racemic synthetic sample using the same SFC configuration but the injection volume was increased from 5 to 100 L of a tenfold more concentrated solution and the MS was disconnected to allow collection of two fractions corresponding to the enantiomer of peak 1 and that of peak 2. The windows for collection were set at 2.25–4.50 min and 5.50–8.00 min. After 25 collections, the weight of fraction 1 was 4.1 mg but the purity as ascertained by analytical SFC was only 79%. On the other hand, fraction 2 was 5.8 mg in weight with a purity > 99% (enantiomeric excess >98%). Due to the higher purity of this fraction, it was used for the determination of its AC using VCD spectroscopy. Note also that performing VCD directly on natural peracetylated rhizopine was not an option due to the presence of several unknown compounds that may make interpretation of the VCD spectrum very difficult. As natural peracetylated rhizopine corresponds with the AC in fraction 1 of the synthetic racemic mixture, but VCD analysis is performed on fraction 2 of the synthetic racemic mixture, the AC of natural peracetylated rhizopine corresponds to the enantiomer of the AC established with VCD for fraction 2.
3.2 Conformational analysis Calculations of 4-acetamido-6-methoxycyclohexane-1,2,3,5tetrayl tetraacetate 2 led to two unique conformations that differ in the relative orientation of the methoxy group. The B3LYP/cc-pVTZ and B3PW91/cc-pVTZ Boltzmann populations of the conformations, shown in Fig. 3, were based on the standard enthalpies and are estimated to be 95.3 and 4.7% and 93.6 and 6.4%, respectively. The differences in relative stability of both conformations are due to changes in steric hindrance between the methoxy group and the nearby carbonyl bonds upon internal rotation. www.jss-journal.com
J. Sep. Sci. 2015, 38, 2545–2550
Other Techniques
2549
Figure 4. IR and VCD Spectra of the peracetylated rhizopine enantiomer in peak 2 of Figure 1. The left panels show the calculated IR and VCD spectra for (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 obtained at the B3LYP/cc-pVTZ level. The right panels show the experimental IR and VCD spectra obtained for a solution in CDCl3 . The theoretical spectra were obtained using a uniform scale factor of 0.990.
3.3 Spectroscopic methods The experimentally measured IR and VCD spectra for fraction 2 and the computationally simulated spectra for the (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane-1, 2,3,5-tetrayl tetraacetate (2 ) obtained at the B3LYP/cc-pVTZ level are depicted in Fig. 4. The theoretical spectra correspond to a Boltzmann average over both conformations found. Quantum chemically computed spectra that rely on the harmonic approximation with a finite basis set generally need to have their wavenumbers scaled and here a factor of 0.990 was used based on a best match between computation and experiment. Inspection of the data presented shows that good agreement is found between the theoretically calculated spectrum for 2 and the experimental spectrum for fraction 2. Nearly all bands observed in the IR and/or VCD spectra are neatly reproduced. The agreement observed for (1R,2R,3S,4R,5R,6S)-4-acetamido-6methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 , and the lack of agreement observed for its mirror image belonging to (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1,2,3,5-tetrayl tetraacetate 2 unambiguously allows the AC of fraction 2 (or peak 2 in Fig. 1A) to be determined as (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 . Combination of this result for fraction 2 (peak 2) and the data obtained by chiral chromatography showing that peak 1 and peak 2 form an enantiomeric pair, thus allows to conclude that peracetylated natural rhizopine corresponds to (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 and therefore that natural rhizopine corresponds to the (1R,2S,3R,4R,5S,6R)-4-amino-6methoxycyclohexane-1,2,3,5-tetraol structure 1 in Scheme 1. The above assignment is confirmed by the CompareVOA algorithm [21] describing the level of agreement between the experimental VCD spectra and the calculated data derived for the assigned AC, and the lack of agreement observed the enantiomer, respectively. Based on the B3LYP/cc-pVTZ data, the agreement for the above suggested assignment for (1R,2R,3S,4R,5R,6S)-4-acetamido-6methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 results in a degree of similarity of 67.0%, whereas only 10.4% is found
www.jss-journal.com
2550
A. Krief et al.
for (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 . Using these similarities as input in the spectral database in the CompareVOA program [21] yields a confidence level of 99%, meaning that the current assignment lies statistically among previous assignments of dozens of compounds for which the AC was independently confirmed using other techniques. This thereby strengthens the assignment reported above using an unbiased numerical criterion. Calculations supporting the above assignment were also performed at the B3PW91/cc-pVTZ level. The results obtained for these calculations are in excellent agreement with the data reported using the B3LYP/cc-PVTZ level and need no further comment.
4 Conclusions The structure and absolute configuration of natural rhizopine has been established as natural rhizopine is (1R,2S,3R,4R,5S,6R)-4-amino-6-methoxycyclohexane-1,2, 3,5-tetraol using state-of-the-art chromatography combined with vibrational circular dichroism spectroscopy applied to the separated fractions. This combination of techniques forms a strong symbiotic duo allowing individual enantiomers or diastereomers to be obtained from a racemic mixture with VCD establishing per fraction what the absolute configuration is. A.K. thanks Prof. Davide Bonifazi, for carrying out the peracetylation reactions in his laboratory (COMS, U Namur). P.B. and W.H. acknowledge the Flemish Fund for Scientific Research (FWO Vlaanderen) for continuous support. This work was carried out using the CalcUA Supercomputer Infrastructures at the University of Antwerp. Funding through Ghent University, the CalcUA core facility, the Hercules Foundation and the Flemish Government department EWI is acknowledged. The authors have declared no conflict of interest.
5 References [1] Heinrich, K., Ryder, M. H., Murphy, P. J., Can. J. Microbiol. 2001, 47, 165–171. [2] Murphy, P. J., Heycke, N., Trenz, S. P., Ratet, P., De Bruijn, F. J., Schell, Proc. Natl. Acad. Sci. USA 1988, 85, 9133– 9137. [3] Murphy, P. J., Saint, C. P., “Rhizopines in the legumeRhizobium symbiosis”. In Molecular Signals in PlantMicrobe Communication; Varma, D. P. S., CRC Press: Boca Raton, 1992, pp 377–390. [4] Murphy, P. J., Heycke, N., Banfalvi, Z., Tate, M., De Bruijn, F., Kondorosi, A., Tempe, J., Schell, J., Natl. Acad. Sci. U.S.A. 1987, 84, 493–497. [5] Krief, A., Dumont, W., Billen, D., Letesson, J.-J., Murphy, P. J., Lacroix, D., Tetrahedron Lett. 2004, 45,1461–1463.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2545–2550
[6] Nafie, L. A., Vibrational Optical Activity: Principles and Applications, 1st ed.; John Wiley & Sons, Chichester, 2011. [7] Stephens, P. J., Devlin, F. J., Cheeseman, J. R., VCD Spectroscopy for Organic Chemists, CRC Press: Boca Raton, 2012. [8] Barron, L. D., Buckingham, A. D., Chem. Phys. Lett. 2010, 492, 199–213. [9] He, Y., Wang, B., Dukor, R. K., Nafie, L. A., Appl. Spectrosc. 2011, 65, 699–723. [10] Magyarfalvi, G., Tarczay, G., Vass, E., Comput. Mol. Sci. 2011, 1, 403–425. [11] Batista, J. M., Batista, A. N. L., Rinaldo, D., Vilegas, W., Cass, Q. B., Bolzani, V. S., Kato, M. J., Lopez, S. N., Furlan, M., Nafie, L. A., Tetrahedron: Asymmetry 2010, 21, 2402– 2407. [12] Batista, J. M., Batista, A. N. L., Mota, J. S., Cass, Q. B., Kato, M. J., Bolzani, V. S., Freedman, T. B., Lopez, S. N., Furlan, M., Nafie, L. A., J. Org. Chem. 2011, 76, 2603– 2612. [13] Batista, J. M., Batista, A. N. L., Bolzani, V. S., Kato, M. J., Nafie, L. A., Lopez, S. N., Furlan, M., Pharm. Biol. 2012, 50, 587–589. [14] Batista, Jr, J. M., Batista, A. N. L., Kato, M. J., Bolzani, ´ V. S., Lopez, S. N., Nafie, L. A., Furlan, M., Tetrahedron Lett. 2012, 53, 6051–6054. ˜ [15] Munoz, M. A., Mart´ınez, M., Joseph-Nathan, P., Phytochem. Lett. 2012, 5, 450–454. ´ [16] Gordillo-Roman, B., Camacho-Ruiz, J., Bucio, M. A., Joseph-Nathan, P., Chirality 2012, 24, 147–154. [17] Gordillo-Roman, B., Reina, M., Ruiz-Mesia, L., RuizMesia, W., Joseph-Nathan, P., Tetrahedron Lett. 2013, 54, 1693–1696. [18] Tomasi, J., Mennucci, B., Cammi, R., Chem. Rev. 2005, 105, 2999–3093. [19] Frisch, M. J. T., Schlegel, G. W., Scuseria, H. B., Robb, G. E., Cheeseman, M. A., Scalmani, J. R., Barone, G., Mennucci, V., Petersson, B., Nakatsuji, G. A., Caricato, H., Li, M., Hratchian, X., Izmaylov, H. P., Bloino, A. F., Zheng, J., Sonnenberg, G., Hada, J. L., Ehara, M., Toyota, M., Fukuda, K., Hasegawa, R., Ishida, J., Nakajima, M., Honda, T., Kitao, Y., Nakai, O., Vreven, H., Montgomery, Jr., T., Peralta, J. A., Ogliaro, J. E., Bearpark, F., Heyd, M., Brothers, J. J., Kudin, E., Staroverov, K. N., Kobayashi, V. N., Normand, R., Raghavachari, J., Rendell, K., Burant, A., Iyengar, J. C., Tomasi, S. S., Cossi, J., Rega, M., Millam, N., Klene, N. J., Knox, M., Cross, J. E., Bakken, J. B., Adamo, V., Jaramillo, C., Gomperts, J., Stratmann, R., Yazyev, R. E., Austin, O., Cammi, A. J., Pomelli, R., Ochterski, C., Martin, J. W., Morokuma, R. L., Zakrzewski, K., Voth, V. G., Salvador, G. A., Dannenberg, P., Dapprich, ¨ Ortiz, J. B., J. J., Daniels, S., Farkas, A. D., Foresman, O., Cioslowski, J. V., Fox, J., Gaussian, D. J., Inc., Rev. D.01, Wallingford CT, 2009. [20] Layton, C., Ma, S., Wu, L., Lee, H., Fandrick, K., Fandrick, D., Senanyake, C., Yee, N., Grinberg, N., Clark, J., J. Sep. Sci. 2013, 36, 3004–3010. [21] Debie, E., De Gussem, E., Dukor, R. K., Herrebout, W., Nafie, L. A., Bultinck, P., ChemPhysChem. 2011, 12, 1542– 1549.
www.jss-journal.com
J S S
ISSN 1615-9306 · JSSCCJ 38 (14) 2371–2558 (2015) · Vol. 38 · No. 14 · August 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
14 15
Vol. 38 (2015) · No. 14 · Pages 2371–2558
Methods Chromatography · Electroseparation
www.jss-journal.com
Applications Biomedicine · Foods · Environment
JSSC_38_14_cover.indd 1
10/07/15 4:54 PM
2545
J. Sep. Sci. 2015, 38, 2545–2550
Alain Krief1 Melissa Dunkle2 Masoud Bahar3 Patrick Bultinck4,5 Wouter Herrebout5,6 Pat Sandra2 1 Department
of Chemistry, University of Namur, Namur, Belgium 2 Research Institute for Chromatography (R.I.C.), Kortrijk, Belgium 3 Department of Plant Protection & Agricultural Biotechnology, College of Agriculture, Isfahan University of Technology, Isfahan, Iran 4 Department of Inorganic and Physical Chemistry, Ghent University, Krijgslaan 281 (S3), Ghent, Belgium 5 Scaldis Spectroscopy www.scaldis-spectroscopy.com 6 Department of Chemistry, University of Antwerp, Antwerp, Belgium
Research Article
Elucidation of the absolute configuration of rhizopine by chiral supercritical fluid chromatography and vibrational circular dichroism The absolute configuration of rhizopine, an opine-like natural product present in nitrogenfixing nodules of alfalfa infected by rhizobia, is elucidated using a combination of state-ofthe-art analytical and semi-preparative supercritical fluid chromatography and vibrational circular dichroism spectroscopy. A synthetic peracetylated racemate was fractionated into its enantiomers and subjected to absolute configuration analysis revealing that natural rhizopine exists as a single enantiomer. The stereochemistry of non-derivatized natural rhizopine corresponds to (1R,2S,3R,4R,5S,6R)-4-amino-6-methoxycyclohexane-1,2,3,5-tetraol. Keywords: Absolute configuration/Infrared spectroscopy/Rhizopine/Supercritical fluid chromatography/Vibrational circular dichroism DOI 10.1002/jssc.201500138
Received February 2, 2015 Revised April 13, 2015 Accepted April 15, 2015
1 Introduction Rhizopine is a natural compound produced in nitrogen-fixing nodules [1] resulting from infection of the root hairs of alfalfa (Medicago sativa) by rhizobia bacteria (Sinorhizobium meliloti and Rhizobium leguminosarum bv. Viciae) [2]. This infection occurs in soil with low bioactive nitrogen content and is known to favor the growth of alfalfa as well as the freeliving rhizobia, which are not only able to catabolize rhizopine but also use it as a source of carbon and nitrogen [3, 4]. The structure of rhizopine has been postulated as 4amino-6-methoxycyclohexane-1,2,3,5-tetraol (Scheme 1) more than 25 years ago but has not yet been firmly established. Moreover, although its relative stereochemistry has been proposed as the one shown in Scheme 1 [1], there remain some doubts whether rhizopine in nature exists as a racemate (1 + 1 ) or as a single enantiomer (1 or 1 ), hence no absolute configuration (AC) information is available nor shown in Scheme 1. Correspondence: Prof. Alain Krief, University of Namur, Rue de Bruxelles 61, 5000 Namur, Belgium E-mail: [email protected]
Abbreviations: AC, absolute configuration; LCPL, left circularly polarized light; RCPL, right circularly polarized light; VCD, vibrational circular dichroism
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ten years ago one of us described the synthesis of racemic 4-amino-6-methoxycyclohexane-1,2,3,5-tetraol [5] and some preliminary experiments indicated that natural rhizopine and its synthetic racemate exhibited similar biological properties. This, however, is no guarantee that natural rhizopine does indeed have exactly the same structure as the synthetic product and moreover that it would also occur as a racemate. We here report on the structure, relative and absolute configuration of natural rhizopine by means of analytical and semi-preparative supercritical fluid chromatography (SFC) and spectroscopic methods including vibrational circular dichroism (VCD) spectroscopy. State-of-the-art chromatography, allowing to effectively separate enantiomers, is followed by VCD that is used to establish the AC of the compound in the fractions of interest. The use of VCD as a technique to establish the stereochemistry of chirally separated molecules is not yet very widespread albeit very promising. VCD is a technique based on absorption of IR radiation although using circularly polarized light. For achiral molecules, the absorption of left circularly polarized light (LCPL) is the same as for right circularly polarized light (RCPL) and hence also for unpolarized light. IR spectra therefore consist of bands that all have the same sign. In the case of chiral molecules, the handedness of the molecule (e.g., the question whether a single chiral center has R or S configuration) causes a difference in interaction with the IR light of different handedness. For instance, the R enantiomer may interact slightly more strongly (or weakly) with RCPL than with LCPL. The VCD spectrum is a plot of the difference in absorption between RCPL and LCPL by the molecule of interest. For some bands, RCPL is slightly more absorbed than LCPL leading to a negative difference RCPL– LCPL whereas for a different band it may be opposite. VCD spectra therefore can have both signs for every band. The
www.jss-journal.com
2546
J. Sep. Sci. 2015, 38, 2545–2550
A. Krief et al.
Scheme 1. Structure of rhizopine (1) and its peractelylated derivative (2) in both enantiomeric forms ( and ) coherent with the relative stereochemistry of rhizopine [1].
difference in absorption is usually very small and often only 10−4 of the absorption in regular IR spectroscopy. VCD is a very configuration- and conformation-sensitive technique hence its use to probe stereochemistry [6–10]. VCD is therefore related to IR spectroscopy as electronic circular dichroism is related to UV–VIS spectroscopy albeit VCD has much higher information content due to the much larger number of vibrational transitions compared to electronic transitions. VCD spectra are, however, not easy to interpret and their interpretation is carried out using quantum chemical calculations. Briefly, an absolute configuration is assumed in a quantum chemical calculation of a VCD spectrum and if a match between the calculation and the experiment is found, this allows concluding that the absolute configuration of the experimental sample matches that used in the calculation. If the match is insufficiently good, calculations are performed for a different absolute configuration until the match is found. Of special importance here is that enantiomers have mirror image VCD spectra and that the calculated VCD spectra should be exactly the same for two enantiomers, but mirror images. Despite experimental and theoretical challenges, VCD has become one of the premier methods for the determination of the AC of chemical compounds. The number and complexity of structures analyzed more or less routinely con-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tinually increases [11–17], making VCD a method of choice for AC determination of complex natural products such a rhizopine.
2 Materials and methods 2.1 Materials and reagents A sample containing ca. 50 g of natural rhizopine (with unknown stereochemistry) was isolated in Isfahan, Iran using the original protocol [1] after inoculation of alfalfa roots by Sinorhizobium meliloti, strain L5–30. Racemic synthetic rhizopine was prepared as in ref. [ 5]. Both samples were derivatized into the peracetyl derivatives (2 in Scheme 1) using the following protocols. rac-Rhizopine hydrochloride (3-O-methyl-scyllo-inosamine hydrochloride (100 mg, 0.44 mmol) [5] and sodium acetate (150 mg, 1.8 mmol), dissolved in acetic acid anhydride (0.9 mL, 9.8 mmol) are heated at reflux for 1.5 h. After cooling to room temperature, water (25 mL) is added and the resulting solution is extracted with dichloromethane (50 mL, four times). The resulting solution is washed with brine, dried over sodium sulfate, filtered and then evaporated, under low pressure on a rotatory evaporator, leaving a crude product
www.jss-journal.com
J. Sep. Sci. 2015, 38, 2545–2550
Other Techniques
2547
Figure 1. SFC chromatograms of the peracetylated synthetic racemate (A) and the peracetylated natural extract (B).
(180 mg). The latter was purified by chromatography on a silica gel plate 7747, 60PF254 from Merck (Overijse, Belgium) using ethyl acetate/hexane mixture (2:1) leading to the peracetylated rac-rhizopine (80 mg). Spectral details are as follows: IR (IR, BIO-RAD FTS-165, Temse, Belgium) max cm−1 (KBr) 3286, 3091, 2993, 2943, 2851, 1758, 1661, 1560, 1436, 1371, 1314, 1221, 1159, 1083, and 1034; 1 H NMR ␦H (JNM EX400, from JEOL, Zaventem, Belgium, 400 MHz, CDCl3 as solvent, TMS as internal standard) ␦ ppm : 5.64 (d, 1 H, J : 10 Hz), RNHAc); 5.21 (t, 1 H, J : 10 Hz), 5.13 (t, 1 H, J: 10 Hz), 5.00 (t, 1 H, J: 10 Hz), 4.96 (t, 1 H, J: 10 Hz), 4.33 (dd, 1 H, J1 : 11 Hz et J2 : 11 Hz) five cyclic hydrogen atoms, 3.49 (m, 4 H, one cyclic hydrogen + ROMe); 2.12 (s, 3 H), 2,10 (s, 3 H), 2.04 (s, 3 H), 2,03 (s, 3 H) hydrogen atoms from four acetoxy groups (ROC(O)Me); 1,92 (s, 3 H, RNHC(O)Me). 13 C NMR (100,4 MHz, CDCl3 as solvent and internal standard) ␦ ppm: 169.71, 79.93, 71.97, 70.52, 60.79, 51.47, 22.99, 20.50. The same procedure was carried out on the crude extract of natural rhizopine (50 g) using an excess of reagents (acetic acid anhydride, 0.45 mL, 4.9 mmol) and sodium acetate (75 mg, 0.9 mmol) and the crude mixture obtained after extraction has been directly used for SFC separation. All used solvents and reagents were from Sigma–Aldrich, Bornem, Belgium. For analytical chromatographic analysis 5 mg of the synthetic sample was dissolved in 10 mL of acetone, for preparative separation the concentration was tenfold higher. The peracytelated natural rhizopine was dissolved in 100 L of acetone. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Mass spectra corresponding to (A) peak 1 in the peracetylated synthetic racemate, (B) peak 2 in the peracetylated synthetic racemate and (C) the peak of the peracetylated natural rhizopine that, in retention time, corresponds to peak 1 of the peracetylated synthetic racemate (see Fig.1).
2.2 Instrumentation and operating conditions A 1260 Infinity Analytical SFC System combined with a 6130 Quadrupole mass spectrometer (MS) from Agilent Technologies (Brussels, Belgium) was used in this study. The chiral separation was performed on a Chiral-Pak AD-H column (250 L × 4.6 mm ID, 5 m particles) from Chiral Technologies, Illkirch, France. Carbon dioxide was of Alphagaz quality from Air Liquide, Liege, Belgium. The modifier was methanol with 0.05% TFA and the gradient applied was 5– 40% in 0–10 min. The flow rate was 2.0 mL/min, the outlet pressure 120 bar and the temperature 40⬚C. For analytical work, the injection volume was 5 L while 100 L was injected for semi-preparative work. For analytical SFC an APCI source was used and the MS settings were: make-up MeOH at 0.2 mL/min, scan range 100 to 550 amu, +3000 V, drying gas www.jss-journal.com
2548
A. Krief et al.
Figure 3. Stable conformations of 4-acetamido-6-methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 calculated at the B3LYP/ cc-pVTZ SCRF (solvent = chloroform) level. The Boltzmann populations of the conformations, obtained using standard enthalpies in solution, are 95.3% (left) and 4.7% (right).
at 8 L/min and 350⬚C, nebulizer at 35 psig and Vap at 325⬚C. The MS was disconnected for semi-preparative application and fractions were directly collected after the backpressure regulator. IR and VCD spectra were obtained using a ChiralIR-2X dual PEM spectrometer (BioTools, Jupiter, FL, USA). For the spectroscopic characterization of the enantiomer of the synthetic sample considered (see Section 3), a solution containing 2.0 mg of the compound was dissolved in deuterated chloroform in a 100 micron liquid cell equipped with BaF2 windows. For the solution 80 000 scans were recorded at 4 cm−1 resolution and averaged. In addition, to allow background subtraction, a spectrum of the pure solvent was recorded. As explained above, proper assignment of the AC of the chiral molecules in the fractions requires quantum chemical calculations for relevant absolute configurations. Conformational analyses were performed using molecular mechanics force fields instead of direct ab initio potential energy hypersurface exploration. A combination of MMFF, MMFF94S, and SYBYL force fields was used, applying the search algorithms from the commercially available software packages Conflex (‘reservoir filling’), Spartan (Monte Carlo stochastic search) and ComputeVOA (Monte Carlo stochastic search). Combination of the resulting conformer libraries led to 328 unique geometries. For all geometries obtained, full geometry optimizations were then initiated at the B3LYP/ccpVTZ and B3PW91/cc-pVTZ levels, using a SCRF = IEFPCM [18] model to account for solvent polarization (chloroform, g = 4.809). All ab initio calculations were performed using Gaussian09 [19]. Boltzmann-weighted IR and VCD spectra were obtained assuming Lorentzian band profiles with a full width at half height (FWHH) of 15 cm−1 . The Boltzmann populations used were based on the standard enthalpies obtained.
3 Results and discussion 3.1 Separation by chiral SFC The analytical SFC analysis of the peracetylated racemic synthetic sample 2, showed two well separated compounds in the total ion chromatogram (TIC) eluting at 3.4 min (peak 1) C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2545–2550
and at 6.7 min (peak 2), respectively (Fig. 1A). The resolution (Rs ) is higher than 10. Note that this resolution is only achieved for the acyl derivative on the chiral selector amylose tris(3,5-dimethylphenylcarbamate). No chiral separation was obtained for the underivatized racemate analyzed on the same column or on a cellulose based stationary phase [20]. The mass spectral data for peak 1 and peak 2 are shown in Fig. 2 A and B, and both spectra are identical. The main ion is m/z 404.2 corresponding to the [M+H]+ of 2. The chromatogram of the peracetylated natural product sample (rhizopine, Iran) analyzed under the same conditions exhibited several peaks with a major peak eluting at the same retention time as peak 1 of the synthetic sample (Fig. 1B). The mass spectrum of this peak agrees with the mass spectra of the two enantiomers of the synthetic sample (Fig. 2C) meaning first of all that the composition of the natural rhizopine and the synthetic products is the same. Second, as the retention time of the main peak of natural rhizopine is the same, the stereochemistry of the structure that corresponds to this peak equals that of the structure that corresponds to peak 1 in the synthetic sample. Fraction collection was then performed on the racemic synthetic sample using the same SFC configuration but the injection volume was increased from 5 to 100 L of a tenfold more concentrated solution and the MS was disconnected to allow collection of two fractions corresponding to the enantiomer of peak 1 and that of peak 2. The windows for collection were set at 2.25–4.50 min and 5.50–8.00 min. After 25 collections, the weight of fraction 1 was 4.1 mg but the purity as ascertained by analytical SFC was only 79%. On the other hand, fraction 2 was 5.8 mg in weight with a purity > 99% (enantiomeric excess >98%). Due to the higher purity of this fraction, it was used for the determination of its AC using VCD spectroscopy. Note also that performing VCD directly on natural peracetylated rhizopine was not an option due to the presence of several unknown compounds that may make interpretation of the VCD spectrum very difficult. As natural peracetylated rhizopine corresponds with the AC in fraction 1 of the synthetic racemic mixture, but VCD analysis is performed on fraction 2 of the synthetic racemic mixture, the AC of natural peracetylated rhizopine corresponds to the enantiomer of the AC established with VCD for fraction 2.
3.2 Conformational analysis Calculations of 4-acetamido-6-methoxycyclohexane-1,2,3,5tetrayl tetraacetate 2 led to two unique conformations that differ in the relative orientation of the methoxy group. The B3LYP/cc-pVTZ and B3PW91/cc-pVTZ Boltzmann populations of the conformations, shown in Fig. 3, were based on the standard enthalpies and are estimated to be 95.3 and 4.7% and 93.6 and 6.4%, respectively. The differences in relative stability of both conformations are due to changes in steric hindrance between the methoxy group and the nearby carbonyl bonds upon internal rotation. www.jss-journal.com
J. Sep. Sci. 2015, 38, 2545–2550
Other Techniques
2549
Figure 4. IR and VCD Spectra of the peracetylated rhizopine enantiomer in peak 2 of Figure 1. The left panels show the calculated IR and VCD spectra for (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 obtained at the B3LYP/cc-pVTZ level. The right panels show the experimental IR and VCD spectra obtained for a solution in CDCl3 . The theoretical spectra were obtained using a uniform scale factor of 0.990.
3.3 Spectroscopic methods The experimentally measured IR and VCD spectra for fraction 2 and the computationally simulated spectra for the (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane-1, 2,3,5-tetrayl tetraacetate (2 ) obtained at the B3LYP/cc-pVTZ level are depicted in Fig. 4. The theoretical spectra correspond to a Boltzmann average over both conformations found. Quantum chemically computed spectra that rely on the harmonic approximation with a finite basis set generally need to have their wavenumbers scaled and here a factor of 0.990 was used based on a best match between computation and experiment. Inspection of the data presented shows that good agreement is found between the theoretically calculated spectrum for 2 and the experimental spectrum for fraction 2. Nearly all bands observed in the IR and/or VCD spectra are neatly reproduced. The agreement observed for (1R,2R,3S,4R,5R,6S)-4-acetamido-6methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 , and the lack of agreement observed for its mirror image belonging to (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1,2,3,5-tetrayl tetraacetate 2 unambiguously allows the AC of fraction 2 (or peak 2 in Fig. 1A) to be determined as (1R,2R,3S,4R,5R,6S)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 . Combination of this result for fraction 2 (peak 2) and the data obtained by chiral chromatography showing that peak 1 and peak 2 form an enantiomeric pair, thus allows to conclude that peracetylated natural rhizopine corresponds to (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 and therefore that natural rhizopine corresponds to the (1R,2S,3R,4R,5S,6R)-4-amino-6methoxycyclohexane-1,2,3,5-tetraol structure 1 in Scheme 1. The above assignment is confirmed by the CompareVOA algorithm [21] describing the level of agreement between the experimental VCD spectra and the calculated data derived for the assigned AC, and the lack of agreement observed the enantiomer, respectively. Based on the B3LYP/cc-pVTZ data, the agreement for the above suggested assignment for (1R,2R,3S,4R,5R,6S)-4-acetamido-6methoxycyclohexane-1,2,3,5-tetrayl tetraacetate 2 results in a degree of similarity of 67.0%, whereas only 10.4% is found
www.jss-journal.com
2550
A. Krief et al.
for (1S,2S,3R,4S,5S,6R)-4-acetamido-6-methoxycyclohexane1,2,3,5-tetrayl tetraacetate 2 . Using these similarities as input in the spectral database in the CompareVOA program [21] yields a confidence level of 99%, meaning that the current assignment lies statistically among previous assignments of dozens of compounds for which the AC was independently confirmed using other techniques. This thereby strengthens the assignment reported above using an unbiased numerical criterion. Calculations supporting the above assignment were also performed at the B3PW91/cc-pVTZ level. The results obtained for these calculations are in excellent agreement with the data reported using the B3LYP/cc-PVTZ level and need no further comment.
4 Conclusions The structure and absolute configuration of natural rhizopine has been established as natural rhizopine is (1R,2S,3R,4R,5S,6R)-4-amino-6-methoxycyclohexane-1,2, 3,5-tetraol using state-of-the-art chromatography combined with vibrational circular dichroism spectroscopy applied to the separated fractions. This combination of techniques forms a strong symbiotic duo allowing individual enantiomers or diastereomers to be obtained from a racemic mixture with VCD establishing per fraction what the absolute configuration is. A.K. thanks Prof. Davide Bonifazi, for carrying out the peracetylation reactions in his laboratory (COMS, U Namur). P.B. and W.H. acknowledge the Flemish Fund for Scientific Research (FWO Vlaanderen) for continuous support. This work was carried out using the CalcUA Supercomputer Infrastructures at the University of Antwerp. Funding through Ghent University, the CalcUA core facility, the Hercules Foundation and the Flemish Government department EWI is acknowledged. The authors have declared no conflict of interest.
5 References [1] Heinrich, K., Ryder, M. H., Murphy, P. J., Can. J. Microbiol. 2001, 47, 165–171. [2] Murphy, P. J., Heycke, N., Trenz, S. P., Ratet, P., De Bruijn, F. J., Schell, Proc. Natl. Acad. Sci. USA 1988, 85, 9133– 9137. [3] Murphy, P. J., Saint, C. P., “Rhizopines in the legumeRhizobium symbiosis”. In Molecular Signals in PlantMicrobe Communication; Varma, D. P. S., CRC Press: Boca Raton, 1992, pp 377–390. [4] Murphy, P. J., Heycke, N., Banfalvi, Z., Tate, M., De Bruijn, F., Kondorosi, A., Tempe, J., Schell, J., Natl. Acad. Sci. U.S.A. 1987, 84, 493–497. [5] Krief, A., Dumont, W., Billen, D., Letesson, J.-J., Murphy, P. J., Lacroix, D., Tetrahedron Lett. 2004, 45,1461–1463.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2545–2550
[6] Nafie, L. A., Vibrational Optical Activity: Principles and Applications, 1st ed.; John Wiley & Sons, Chichester, 2011. [7] Stephens, P. J., Devlin, F. J., Cheeseman, J. R., VCD Spectroscopy for Organic Chemists, CRC Press: Boca Raton, 2012. [8] Barron, L. D., Buckingham, A. D., Chem. Phys. Lett. 2010, 492, 199–213. [9] He, Y., Wang, B., Dukor, R. K., Nafie, L. A., Appl. Spectrosc. 2011, 65, 699–723. [10] Magyarfalvi, G., Tarczay, G., Vass, E., Comput. Mol. Sci. 2011, 1, 403–425. [11] Batista, J. M., Batista, A. N. L., Rinaldo, D., Vilegas, W., Cass, Q. B., Bolzani, V. S., Kato, M. J., Lopez, S. N., Furlan, M., Nafie, L. A., Tetrahedron: Asymmetry 2010, 21, 2402– 2407. [12] Batista, J. M., Batista, A. N. L., Mota, J. S., Cass, Q. B., Kato, M. J., Bolzani, V. S., Freedman, T. B., Lopez, S. N., Furlan, M., Nafie, L. A., J. Org. Chem. 2011, 76, 2603– 2612. [13] Batista, J. M., Batista, A. N. L., Bolzani, V. S., Kato, M. J., Nafie, L. A., Lopez, S. N., Furlan, M., Pharm. Biol. 2012, 50, 587–589. [14] Batista, Jr, J. M., Batista, A. N. L., Kato, M. J., Bolzani, ´ V. S., Lopez, S. N., Nafie, L. A., Furlan, M., Tetrahedron Lett. 2012, 53, 6051–6054. ˜ [15] Munoz, M. A., Mart´ınez, M., Joseph-Nathan, P., Phytochem. Lett. 2012, 5, 450–454. ´ [16] Gordillo-Roman, B., Camacho-Ruiz, J., Bucio, M. A., Joseph-Nathan, P., Chirality 2012, 24, 147–154. [17] Gordillo-Roman, B., Reina, M., Ruiz-Mesia, L., RuizMesia, W., Joseph-Nathan, P., Tetrahedron Lett. 2013, 54, 1693–1696. [18] Tomasi, J., Mennucci, B., Cammi, R., Chem. Rev. 2005, 105, 2999–3093. [19] Frisch, M. J. T., Schlegel, G. W., Scuseria, H. B., Robb, G. E., Cheeseman, M. A., Scalmani, J. R., Barone, G., Mennucci, V., Petersson, B., Nakatsuji, G. A., Caricato, H., Li, M., Hratchian, X., Izmaylov, H. P., Bloino, A. F., Zheng, J., Sonnenberg, G., Hada, J. L., Ehara, M., Toyota, M., Fukuda, K., Hasegawa, R., Ishida, J., Nakajima, M., Honda, T., Kitao, Y., Nakai, O., Vreven, H., Montgomery, Jr., T., Peralta, J. A., Ogliaro, J. E., Bearpark, F., Heyd, M., Brothers, J. J., Kudin, E., Staroverov, K. N., Kobayashi, V. N., Normand, R., Raghavachari, J., Rendell, K., Burant, A., Iyengar, J. C., Tomasi, S. S., Cossi, J., Rega, M., Millam, N., Klene, N. J., Knox, M., Cross, J. E., Bakken, J. B., Adamo, V., Jaramillo, C., Gomperts, J., Stratmann, R., Yazyev, R. E., Austin, O., Cammi, A. J., Pomelli, R., Ochterski, C., Martin, J. W., Morokuma, R. L., Zakrzewski, K., Voth, V. G., Salvador, G. A., Dannenberg, P., Dapprich, ¨ Ortiz, J. B., J. J., Daniels, S., Farkas, A. D., Foresman, O., Cioslowski, J. V., Fox, J., Gaussian, D. J., Inc., Rev. D.01, Wallingford CT, 2009. [20] Layton, C., Ma, S., Wu, L., Lee, H., Fandrick, K., Fandrick, D., Senanyake, C., Yee, N., Grinberg, N., Clark, J., J. Sep. Sci. 2013, 36, 3004–3010. [21] Debie, E., De Gussem, E., Dukor, R. K., Herrebout, W., Nafie, L. A., Bultinck, P., ChemPhysChem. 2011, 12, 1542– 1549.
www.jss-journal.com
