On the structure and spectroscopic properties of two 13-hydroxysparteine epimers

On the structure and spectroscopic properties of two 13-hydroxysparteine epimers

Journal of Molecular Structure 832 (2007) 90–95 www.elsevier.com/locate/molstruc On the structure and spectroscopic properties of two 13-hydroxyspart...

140KB Sizes 0 Downloads 45 Views

Journal of Molecular Structure 832 (2007) 90–95 www.elsevier.com/locate/molstruc

On the structure and spectroscopic properties of two 13-hydroxysparteine epimers Tadeusz Brukwicki *, Jacek Włodarczak, Waleria Wysocka Faculty of Chemistry, Adam Mickiewicz University, Grunwaldzka 6, 60-780 Poznan, Poland Received 14 July 2006; received in revised form 8 August 2006; accepted 8 August 2006 Available online 18 September 2006

Abstract The 13C and 1H NMR spectra of 13a-hydroxysparteine and 13b-hydroxysparteine in CDCl3 have permitted identification of their conformation with ring C a boat, the same as that of sparteine. Very weak hydrogen bond involving hydroxy group does not shift a possible conformation equilibrium as it occurs in 13a-hydroxy-2-oxosparteine and 13b-hydroxy-2-oxosparteine.  2006 Elsevier B.V. All rights reserved. Keywords: Quinolizidine alkaloids; Sparteine; Lupanine; NMR; IR; Conformation; Hydroxy group

1. Introduction Sparteine (1), the main tetracyclic bis-quinolizidine alkaloid, is built of two quinolizidine systems: a relatively resistant to configurational – conformational changes double-chair trans-quinolizidine system A/B and a much more flexible cis-quinolizidine system C/D, susceptible to inversion at the N16 atom connected with a conformational– configurational change. The DFT calculations derived by Galasso et al. [1] have conclusively proved that 1 has a strong preference (3.4 kcal/mol, 99.95%) to conformation 1a (with ring C a boat) over 1b (with ring C a chair). Most other bis-quinolizidine alkaloids with the sparteine skeleton adopt the same configuration and conformation of the C/D part of the molecule in the solid but many of them exist in a conformational equilibrium in solution [2]. Some chemical modifications of the sparteine skeleton or the presence of some groups attached to it can influence the proportion of conformers in this equilibrium. For instance, 2-oxosparteine (lupanine, 2) which has the conformation with ring C a boat in the solid [3] occurs in the equilibrium with 90% dominance of the boat form [2,4] in chloroform *

Corresponding author. Tel.: +48 61 829 1005; fax: +48 61 8658008. E-mail address: [email protected] (T. Brukwicki).

0022-2860/$ - see front matter  2006 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2006.08.005

solution. We have recently published our X-ray, NMR and IR investigation on the conformation of two epimers of 13-hydroxy-2-oxosparteine (13-hydroxylupanine, 3 and 4) [5]. Surprisingly, in the solid state, 3 and 4 adopt the conformation with a chair ring C but they prefer the conformation with ring C a boat in solution. However, the fraction of the chair conformer is a little larger in 3 and 4 than in lupanine (2) [5]. In the present paper, we report our investigation on the conformation of epimeric 13-hydroxysparteines (5 and 6) in solution. We are particularly interested in finding out the influence of the axial and equatorial hydroxy group attached to the carbon atom in position 13 of sparteine (1) on the conformational equilibrium of the compounds. 2. Experimental 2.1. General techniques Melting points were determined on a Boetius apparatus (PHMK 05 VEB Wagetechnik Rapido, Radebeul). IR spectra were recorded on a FT-IR Bruker IFS 113v spectrometer (KBr pellets technique). 13C NMR spectra were obtained on a Varian 300 Mercury spectrometer at 75.462 MHz number of transients 10,000, acquisition time

T. Brukwicki et al. / Journal of Molecular Structure 832 (2007) 90–95

1.5 s, spectral width 13,718 Hz, number of points 27,372, digital resolution 0.50 Hz. 1H NMR spectra: number of transients 64, acquisition time 3.0 s, the 90 pulse width 8 ls, the 45 pulse width 4 ls, spectral width 9000 Hz, number of points 54,016, digital resolution 0.167 Hz per point The 1H and all 2D correlation spectra were recorded on a Bruker AVANCE 600 (600.31 MHz for 1H and 150.052 MHz for 13C) spectrometer, with a 5 mm triple – resonance inverse probe head (1H/31P/BB) with actively shielded z gradient coil (90 1H pulse width 90 ls, 13C pulse width 13.3 ls). All 2D spectra were acquired and processed using standard Bruker software. Spectral width of 6313.13 and 25,000 Hz were used for 1H and 13C, respectively. Relaxation delays of 2.0 s were used for all 2D experiments and mixing time 0.8 s for 1H–1H NOESY spectrum was applied. All 2D spectra were collected with 2K points in F2 and 256 increments (F1) with 4 (g-COSY) and 64 (NOESY and g-HSQC) transients each and zero filling in F2 to 2048 · 1024 data matrix.

91

2.2. Substances 13a-Hydroxylupanine was extracted from seeds of Lupinus angustifolius according to the method described previously [6] and characterized like in Ref. [5]. 2.2.1. 13a-Hydroxysparteine 0.100 g PtO2 were suspended in 25 ml of 2 N HCl in a two-necked 100 ml round bottom flask equipped with a hydrogen supplying pipe dipped into the liquid, with another pipe used to carry away the air, replaced by a stopper after some minutes and a magnetic stirrer. Gas hydrogen was bubbled into the suspension. After 20 min, brown PtO2 was converted into black Pt catalyst. 13a-Hydroxylupanine (0.800 g) suspended in 10 ml of 2 N HCl was added to the suspension of the catalyst and hydrogen was bubbled in. The progress of the reaction was controlled by TLC (silca gel, acetone–methanol–aqueous 25% NH3 10:1:1 mixed immediately before use). The reduction was

8

B A

C

N

7

D

17

6

N

N

4 2

12

11

5

9

10

N16

13

1

14

15

3

1 8 17

N

12

R2 R1

R2 N

N

N 16 1

R1

14

O

O 2 R1 = H, R2 = H 3 R1 = OH, R2 = H 4 R1 = H, R2 = OH

8 17

N

12

R2

R2 N

N

R1

N 16

1

R1 14

5 R1 = OH, R2 = H 6 R1 = H, R2 = OH 8 17

N O

O N

OH

N

12

N 16

OH

1 14

7

92

T. Brukwicki et al. / Journal of Molecular Structure 832 (2007) 90–95

completed after 155 h. The solution was filtrated through a column with a small amount of asbestos and the column was washed by three portions of water (2 ml). The solution was concentrated to about 15 ml on the rotary evaporator under a reduced pressure and the rest was transferred into a mortar. Pellets of potassium hydroxide (8 g) were cautiously added and all was mixed with 12 g of diatomaceous earth. The mixture was placed in a column with basic diatomaceous earth (prepared of 5 g of diatomaceous earth and 4 ml of 20% aqueous solution of potassium hydroxide). The alkaloids were eluted with 300 ml of petroleum ether (40–60), 100 ml of diethyl ether and 50 ml of methylene chloride. From the petroleum ether fraction 0.696 g of chromatographically pure colourless crystals of 13ahydroxysparteine were obtained (92%). Mp. 145. EI-MS m/z (%): 250 (30) [M+Æ], 251 (7.5), 252 (0.87), 233 (11), 209 (21), 153 (11), 152 (22), 150 (16), 138 (15), 137 (100), 136 (19), 134 (11), 126 (12), 122 (16), 114 (13), 113 (22), 108 (13), 98 (84). IR (KBr pellets, cm1) 3364 (OH), 2828, 2798, 2777, 2753, 2681 (Bohlmann band); 13C and 1 H NMR – Tables 1 and 2, respectively. 13b-Hydroxylupanine was obtained from 13-oxolupanine [7] by NaBH4 reduction according to the method reported previously [8] and characterized in Ref. [5]. 2.2.2. 13b-Hydroxysparteine PtO2 (0.050 g) was suspended in 10 ml of 1 N HCl and the suspension was stirred by a magnetic stirrer. Gas hydrogen was bubbled into the suspension. After 30 min, 13b-hydroxylupanine (0.090 g) (dissolved in 5 ml of 1 N HCl) was added to the suspension. The progress of the reaction was controlled by TLC (silica gel, acetone–methanol–methanol/NH3, 5–0.5–0.5). The reaction was completed in 150 h. The reaction mixture was filtrated through a column with 5 g Al2O3 and alkalized by KOH to pH 14. The solution was extracted by petroleum ether (100 ml), diethyl ether (100 ml) and methylene chloride (100 ml).

All solutions were concentrated and controlled by TLC. The petroleum ether fraction containing chromatographically pure fraction of alkaloids was dissolved in a small amount of diethyl ether and filtrated through a column with 10 g of silica gel. The alkaloids were eluted with diethyl ether (50 ml) and methylene chloride (50 ml). From diethyl ether fraction, colourless crystals of 13b-hydroxysparteine were obtained (0.076 g, 85%). Mp. 124 C. EI-MS m/z (%): 250 [M+Æ] (45), 251 (10), 209 (32), 153 (34), 137 (100), 134 (21), 133 (19), 126 (15), 121 (15), 114 (13). IR (KBr pellets, cm1) 3342 (OH), 2819, 2795, 2772, 2679, 2664 (Bohlmann band); 13C and 1H NMR – Tables 1 and 2, respectively. 3. Results and discussion 3.1.

13

C NMR and 1H NMR spectra

The 13C and 1H NMR signals in the spectra of 13ahydroxysparteine (5) and 13b-hydroxysparteine (6) were assigned using 2D techniques. Tentative results of the 13C NMR spectra were starting-points for HSQC spectra analysis and the assignment of 1H signals was verified by means of 1H–1H COSY. The results are shown in Tables 1 and 2, respectively. The data for 13a- (3) and 13b-hydroxylupanine (4) [5] were included into the same tables. The 13C NMR spectra of 13a-hydroxysparteine (5) are in accordance with the results of Bohlmann (Bohlmann interchanged signals for C7 and C9; the only difference, ca 0.6 ppm, is for carbon atom C13) [9], the 13C and 1H NMR spectra for 13b-hydroxysparteine (6) and 1H NMR for 5 have been recorded, as far as we know, for the first time. A comparison of the chemical shifts of carbon atoms of compounds 5 and 6 with those of the appropriate carbon atoms of sparteine (1) indicates the differences involving the appropriate atoms forming ring A can be neglected.

Table 1 13 C NMR chemical shifts of 13a-hydroxysparteine (5), 13b-hydroxysparteine (6), 13a- hydroxylupanine (3) and 13b-hydroxylupanine (4), (CDCl3, ppm from TMS) C atom

13a-Hydroxysparteine (5)

D = d5  d1

13a-Hydroxylupanine (3)

13b-Hydroxysparteine (6)

D 0 = d6  d1

13b-Hydroxylupanine (4)

Sparteine (1)

2 3 4 5 6 7 8 9 10 11 12 13 14 15 17

56.12 25.77 24.69 29.28 66.41 32.98 27.43 35.48 61.63 57.25 41.54 65.01 32.76 49.13 53.18

0.16 0.19 0.11 0.13 0.14 0.22 0.31 0.75 0.38 7.21 6.76 40.06 6.70 6.31 0.47

171.19 33.09 19.78 27.46 60.86 32.40 26.62 34.30 46.75 57.11 40.17 64.46 31.73 49.16 52.43

56.27 25.92 24.75 29.40 66.45 33.12 27.48 35.81 61.66 61.86 43.57 69.34 35.24 52.83 52.61

0.01 0.04 0.05 0.01 0.10 0.08 0.26 0.42 0.35 2.60 8.79 44.39 9.18 2.66 1.04

171.25 32.42 19.50 27.31 60.57 32.24 26.62 34.35 46.71 61.17 41.56 68.97 33.84 52.78 51.49

56.28 25.96 24.80 29.41 66.55 33.20 27.74 36.23 62.01 64.46 34.78 24.95 26.06 55.44 53.65

T. Brukwicki et al. / Journal of Molecular Structure 832 (2007) 90–95

93

Table 2 1 H NMR chemical shifts of 13a-hydroxysparteine (5), 13b-hydroxysparteine (6), 13a-hydroxylupanine (3) and 13b-hydroxylupanine (4), (CDCl3, ppm from TMS) H atoms

13a-Hydroxysparteine (5)

2a eq 2b ax 3a ax 3b eq 4a eq 4b ax 5a ax 5b eq 6 7 8a eq 8b ax 9 10a eq 10b ax 11 12a eq 12b ax 13a ax 13b eq 14a eq 14b ax 15a ax 15b eq 17a ax 17b eq

2.67 1.93 1.50 1.54 1.70 1.20 1.38 1.24 1.74 1.85 2.06 1.08 1.43 2.53 2.00 2.42 1.61 1.61 1.7 (OH) 4.08 1.68 1.86 2.41 2.59 2.44 2.71

13a-Hydroxylupanine (3)

13b-Hydroxysparteine (6)

2.32 2.46 1.84 1.60 1.58 1.76 3.30 2.06 2.17 1.27 1.60 4.48 2.53 2.11 1.66 1.66 2.35 (OH) 4.08 1.65 1.78 2.37 2.54 2.02 2.85

2.69 1.95 1.55 1.52 1.71 1.21 1.37 1.25 1.75 1.85 2.05 1.09 1.54 2.54 2.03 2.09 1.84 1.37 3.63 1.65 (OH) 1.92 1.55 2.11 2.82 2.35 2.76

The minor differences are seen for the atoms of ring B, especially for C9. The most distinct differences are visible for the carbon atoms of ring D due to the hydroxy group effect. They are different for the two epimers: for C11 and C15 it is a c-gauche (c-synclinal) effect amounting to 7.21 and 6.31 ppm, respectively, in 13a-hydroxysparteine (5), whereas it is a c-anti effect amounting to 2.60 and 2.66 ppm, respectively in 13b-hydroxysparteine (6). Also the a and b-effects are different for the two epimers: they are distinctly greater for the b-isomer. This is in accordance with similar results published for the epimers of 13hydroxylupanine by Bohlmann [9] and us [5]. Chemical shifts for the carbon atoms of ring D are similar in the appropriate pairs: 13a-hydroxysparteine (5) and 13ahydroxylupanine (3) as well as 13b-hydroxysparteine (6) and 13b-hydroxylupanine (4), especially those of C15 are almost the same in the two pairs and of C11 in 5 and 3. Chemical shifts of C12, C14 and C17 are known to be very sensitive to conformational changes [2,10]. The differences in the chemical shifts of these atoms in the appropriate pairs of alkaloids, 5 and 3 as well as 4 and 6, can be attributed to different conformational equilibria for the derivatives of lupanine and those of sparteine. For the hydroxy derivatives of lupanine, the conformational equilibrium with about 10–15% fraction of the minor conformer with ring C a chair [5] has been estimated assuming that the hydroxy derivatives of sparteine have probably the conformational equilibrium shifted to the almost 100% dominance of the conformer with ring C a boat. Now, we try

13b-Hydroxylupanine (4)

Sparteine (1) [12]

2.30 2.46 1.82 1.62 1.56 1.77 3.32 2.02 2.15 1.30 1.68 4.53 2.55 1.82 1.82 1.43 3.59 2.21 (OH) 1.82 1.52 2.05 2.77 2.00 2.87

2.53 1.79 1.38 1.38 1.55 1.08 1.24 1.12 1.58 1.69 1.91 0.90 1.32 2.38 1.84 1.83 1.21 1.35 1.15 1.55 1.43 1.43 1.86 2.63 2.20 2.54

to verify this assumption on the basis of the 1H NMR spectra. Before we do that, we must point out the fact that the c-gauche effect is different for C11 and C15 in 13a-hydroxysparteine (5). It can be caused by a little torsion of the skeleton of 5 in the ring C/D connection region. The most conspicuous feature of the 1H NMR spectra of compounds 3–6 is the chemical shift of the hydroxyl proton. For the hydroxy derivatives of lupanine 3 and 4, it amounts to about 2.3 ppm, whereas for the hydroxy derivatives of sparteine 5 and 6 – only to about 1.7 ppm. It means that the intermolecular hydrogen bond involving the hydroxyl group is very weak in 5 and 6 [11], much weaker than those in 3 and 4 [5]. Chemical shifts of the corresponding protons attached to the carbon atoms of rings A and B are almost the same in 5 and 6 and differ from those of sparteine (1) [12] by 0.12  0.19 ppm. In Ref. [12], chemical shifts were referred to the signal of CDCl3 and it seems to be reasonable to assume that they are shifted by about 0.14 ppm relative to those for 5 and 6 referred to TMS. The only small deviation is for H9 (0.11 ppm for 5 and 0.22 ppm for 6), probably due to a different effect of the hydroxy group in the two epimers. The differences in the chemical shifts of the protons attached to ring D carbon atoms in 13a-hydroxysparteine (5) or 13b-hydroxysparteine (6) when compared with those of the appropriate protons in sparteine (1) (taking into account the correction of ca. 0.14 ppm – see above) are very similar to the respective differences observed for

94

T. Brukwicki et al. / Journal of Molecular Structure 832 (2007) 90–95

Table 3 The effect of the hydroxy groupa in 13a- and 13b-hydroxyspareteines and lupanines H atoms 11 12a 12b 13a 13b 14a 14b 15a 15b

eq ax ax eq eq ax ax eq

c-ax b-eq b-ax a a beq bax c-ax c-eq

13a-Hydroxysparteineb

13a-Hydroxylupanine

13a-Hydroxymultiflorinec

Steroid modeld

13b-Hydroxysparteineb

13b-Hydroxylupanine

0.45 0.12 0.26

0.49 0.12 0.31

0.57 0.09 0.28

0.47 0.09 0.27

0.12 0.35 0.02 2.34

0.20 0.28 0.08 2.33

2.39 0.11 0.29 0.41 0.18

2.39 0.09 0. 25 0.47 0.21

2.40 0.06 0.32 0.51 0.21

2.48 0.09 0.27 0.47 0.28

0.35 0.06 0.11 0.05

0.26 0.01 0.15 0.02

a

A difference in the chemical shift at the proton in the hydroxy derivatives and that of the respective proton in sparteine or lupanine (ppm). With correction for changing the solvent reference to the TMS reference (see text). c Ref. [5]. d A difference in the chemical shift at the proton in 3a,6a-dihydroxy-5a,17a-pregnan-2-one and that of the respective proton in 5a,17a-pregnan-2-one (ppm) [15]. b

13a-hydroxylupanine (3) or 13b-hydroxylupanine (4) when compared with lupanine (2) [13]. The only exception is the difference in chemical shifts for protons attached to the carbon atom C12 in 5 or 6 compared with 1. We think that Duddeck et al., have mistakenly interchanged the values of chemical shifts of the protons H12ax and H12eq in sparteine (1) (we are sure of our values for the same protons in the spectra of lupanine (2) because the signal at 1.35 ppm has three large coupling constants and must be attributed to the axial proton). So, when we interchange the values for the chemical shifts of H12ax and H12eq in the spectra of sparteine reported by Duddeck, we will obtain more reasonable results for the b-effect of hydroxy group for 5 and 6 (Table 3). The effects for 13a-hydroxysparteine (5) and 13ahydroxylupanine (3) are similar to those obtained for 13a-hydroxymultiflorine (7) [10]. The regularity of the effects of the hydroxy group and the similarity of the chemical shifts of the ring A and B protons in 13a-hydroxysparteine (5) and 13b-hydroxysparteine (6) with those of the appropriate protons in sparteine (1) can be the evidence for the conformations of 5 and 6 similar to that of 1 in solution. A more direct proof can be the value of the coupling constant JH7-H17b [2,10]. For sparteine (1) it amounts to 10.8 Hz [2,14]. Unfortunately, for 5 and 6 it is difficult to determine JH7–H17b because it is almost the same as the geminal coupling constant JH17a–H17b. As a result, the signals of H17b both in 5 and 6 are triplets with the coupling constants amounting to 10.9 Hz. The geminal coupling constant determined from the signal of H17a is 11.1 Hz both in 5 and 6. As the sum of the coupling constants is equal to the difference in the frequencies of the first and the last line of the signal, we can assume that JH7–H17b is ca. 10.7 Hz. This is almost the same as the value determined for H17b proton in sparteine (different from the values determined for 3 and 4, 9.9 and 8.9, respectively [5]). So, we can conclude that 13a-hydroxysparteine (5) and 13b-hydroxysparteine (6) have the same or very similar conformation as that of sparteine (1). It

means that the conformational equilibrium for 5 and 6 is shifted towards ca. 100% domination of the conformation with ring C a boat. It is very tempting to calculate this value as ca. 99% and maintain that the tiny difference in relation to sparteine is caused by a very weak intermolecular hydrogen bond visible in IR and 1H NMR spectra, but we are not entitled to such a statement because JH7–H17b has not been determined to a good accuracy and must be considered as the same as that of sparteine within the limits of error. 4. Conclusion The conformation of 13a-hydroxysparteine (5) and 13bhydroxysparteine (6) in solution is very similar to that of sparteine (1) i.e. ca 100% of the conformation with ring C a boat. A possible very weak intermolecular hydrogen bond is not strong enough to be responsible for a shift of the conformational equilibrium towards a detectable amount of the conformation with ring C a chair, as it is in the case of epimers of 13-hydroxylupanine. The crucial factor shifting the equilibrium toward the almost entire domination of the conformation with ring C a boat in 5 and 6 compared with 3 and 4 is the repulsion of free electron pairs on N1 and N16 nitrogen atoms present in 5 and 6 and absent in 3 and 4. References [1] V. Galasso, F. Asaro, F. Berti, B. Kovacˇ, I. Habusˇ, A. Sacchetti, Chem. Phys. 294 (2003) 155. [2] W. Wysocka, T. Brukwicki, J. Mol. Struct. 385 (1996) 23. [3] H. Doucerain, A. Chiaroni, C. Riche, Acta Crystallogr. B32 (1976) 3213. [4] V. Galasso, F. Asaro, F. Berti, I. Habusˇ, B. Kovacˇ, C. de Risi, Chem. Phys. 301 (2004) 33. [5] T. Borowiak, I. Wolska, W. Wysocka, T. Brukwicki, J. Mol. Struct. 753 (2005) 27. [6] W. Wysocka, A. Przybył, Sci. Legumes 1 (1994) 37. [7] W. Wysocka, M. Wiewio´rowski, Bull. Acad. Polon. Sci. Ser. Sci. Chim. 21 (1973) 29.

T. Brukwicki et al. / Journal of Molecular Structure 832 (2007) 90–95 [8] W. Wysocka, M. Wiewio´rowski, Bull. Acad. Polon. Sci. Ser. Sci. Chim. 22 (1974) 831. [9] F. Bohlmann, R. Zeisberg, Chem. Ber. 108 (1975) 1043. [10] T. Brukwicki, W. Wysocka, B. Nowak-Wydra, Can. J. Chem. 72 (1994) 193. [11] R.M. Silverstein, G.C. Bassler, T.C. Morrill, Spectrometric Identification of Organic Compounds, fifth ed., John Wiley & Sons, New York, 1991, p. 183.

95

[12] H. Duddeck, J. Skolik, U. Majchrzak-Kuczyn´ska, Khim. Geterosikl. Soed. (1995) 1026. [13] R. Kolanos´, W. Wysocka, T. Brukwicki, Tetrahedron 39 (2003) 5531. [14] W.M. Gołe˛biewski, Magn. Res. Chem. 24 (1986) 105. [15] D.N. Kirk, H.C. Toms, Ch. Douglas, K.A. White, K.E. Smith, Sh. Latif, R.W.P. Hubbard, J. Chem. Soc. Perkin Trans. 2 (1990) 1567.