The nitrogen-15 (¹⁵N) NMR technique has become one of the most powerful tools in the structural elucidation of nitrogenous organic compounds ^[1]. Several reports that covered the ¹⁵N NMR studies of many nitrogenous organic compounds including alkaloids are available ^{[2, 3]}. Due to the low natural abundance of ¹⁵N isotope,the ¹H-¹⁵N correlation spectroscopy,which showed an about 310-fold sensitivity increase over the direct detection method,was employed in the practical applications ^[1]. By comparison,the ¹H-¹⁵N HMBC spectra showed ³J long-range correlations from H to N atoμsof the detected molecules,whereas ³J signals from H to C atoμswere observed in the traditional ¹H-¹³C HMBC techniques. In other words,the former may act as a valuable supplement to the conventional NMR experiments in the structural elucidation of nitrogen-containing organic compounds.

Following our previous investigations ^{[4, 5]},we studied the ¹⁵N NMR of C₁₈-diterpenoid alkaloids,a group of complex and polycyclic diterpenoid alkaloids bearing eighteen carbons and one nitrogen on the skeletons ^[6]. The present paper describes the long-range 2D ¹H-¹⁵N NMR spectra data of thirteen C₁₈-diterpenoid alkaloids (Fig. 1),including lappaconitine (1) ^[7],ranaconitine (2) ^[7], sinomontanines A (3) and B (4) ^[8],excelsine (5) ^[7],8-O-acetylexcelsine (6) ^[7],sinomontanine H (7) ^[9],tuguaconitine (8) ^[10],lappaconidine (9) ^[7],lappaconine (10) ^[7],anthriscifolcines E (11),A (12),and C (13) ^[11]. In addition,the effects of the substituents of nitrogen atoms,the conformations of A ring,as well as protonation,on the nitrogen-15 chemical shifts are observed.

Fig. 1

	Download: JPG larger image
Fig. 1.The structures of compounds 1-13.

All NMR spectra were recorded in CDC_l3,CD₃OD,or acetone-d₆ as solvents,using a Bruker Avance III 800 spectrometer operating at frequencies of 800.25 MHz (¹H) and 81.1 MHz (¹⁵N). The temperature of measurement was 303 K. Sample concentrations ranged from 400 μL to 450 μL in 500 μL deuterium solvents. The ¹⁵N chemical shifts were referenced to the signal of 1 mol/L urea in DMSO-d₆ (77.0 ppm). The chemical shifts were measured by the HMBC experiments and under identical conditions to those for the described alkaloids.

The pulse conditions were as follows: 90° pulse,duration 8.69 μs for ¹H and 35 μs for ¹⁵N (5 mm CPQCI ¹H/³¹P/¹³C/¹⁵P/D equipped with a self-shielded z-gradient coil). Computer processing was performed with TOPSPIN Bruker software.

For ¹H-¹⁵N chemical shift correlations,the HMBC experiments were used. Spectra were acquired as 2048×128 data matrices with 80 transients accumulated per t₁ increment with final digital resolution better than 1 and 5 Hz in the ¹H and ¹⁵N dimensions, respectively. The delay used for evolution of ¹H-¹⁵N long-range couplings was set to 100 ms in both types of experiments. A gradient pulse duration of 1 ms and a post-gradient recovery delay of 20 μs were used. Gradient ratios for HMBC of 70: 30: 50.1 G cm^-1 were applied. The gradient pulses used for HMBC experiments were shaped to a SMSQ10.100 sine envelope and in HMBC experiments square gradients were applied.

All the alkaloids in this study,including lappaconitine (1), ranaconitine (2),sinomontanines A (3) and B (4),excelsine (5),8-O-acetylexcelsine (6),sinomontanine H (7),tuguaconitine (8), lappaconidine (9),lappaconine (10),anthriscifolcines E (11),A (12),and C (13),were isolated by our laboratory from the plants of the genera Aconitum and Delphinium.

The ¹⁵N chemical shifts of compounds 1-13 and their trifluoroacetates were listed in Table 1,from which crucial information was observed as below:

Table 1

Table 1 15N chemical shifts for compounds 1-13 (15N: 81.1 Hz).

Table 1 15N chemical shifts for compounds 1-13 (15N: 81.1 Hz).

(1) The ¹⁵N₁ signal of lappaconitine (1) shifted upfield by 6.2 ppm compared to that of sinomontanine A (3). Apparently,the difference was attributed to the electron-donating effect of ethyl group on N1 atom,indicating that substituents (such as Et group) on nitrogen atomshad a remarkable influence to their chemical shifts. In contrast,the δ_N values were less affected by the substitution groups on the carbon frameworks,such as C-6 (between compounds 7 and 8),C-8 (between compounds 5 and 6),or C-10 (between compounds 12 and 13),even the presence of 3,4-epoxy units in this class of alkaloids.

(2) The ¹⁵N chemical shift data of lappaconitine (1) and ranaconitine (2) were close to each other (δ_N1～41,δ_N2～ 130/132), which suggested that the influence of different types of C₁₈- diterpenoid alkaloids (lappaconitine- and ranaconitine-type)was insignificant ^[6]. Related features were also observed between 8-O-acetylexcelsine (6) and sinomontanine H (7),as well between excelsine (5) and tanguacontine (8). Specifically, the ¹⁵N chemical shift values for all the four compounds (5-8) were in the range of δ_N 50-51 approximately as shownin Table 1.

(3) The δ_N peak of lappaconidine (9) shifted downfield by about 3.2 ppm by comparison with that of lappaconine (10),which may be ascribed to their different conformations of A ring: boat-form for the former and chair-form for the latter ^[12]. The same rule was applicable for the other alkaloids with boat conformation of A ring (e.g. sinomontanes A (3) and B (4), excelsine (5),8-O-acetylexcelsine (6),sinomontanine H (7), tuguaconitine (8),and lappaconidine (9) and those with chairform of A ring (e.g. lappaconitine (1),ranaconitine (2), lappaconine (10),and anthriscifolcine E (11). Therefore,the ¹⁵N NMR spectra could be employed to predict the conformation of A ring for the C₁₈-diterpenoid alkaloids.

(4) Similar to the ¹³C NMR,two isomeric δ_N2 signals for the alkaloids with an N-acetyl anthranilate at C-4 (such as 2 and 4) and their trifluoroacetates were observed in the ¹⁵N NMR spectra (Table 1).

(5) As reported in the literature ^[5],protonation by adding trifluoroacetic acid (TFA) to the free bases revealed a significant downfield shifting effect in the ¹⁵N NMR. It is noteworthy that protonation exerts the same influence on the chemical shift values of both nitrogen atoμs(Table 1).

(6) In the ¹H-¹⁵N HMBC spectra of the alkaloids bearing N-Et groups (compounds 1,2,and 5-13),the ³J_N,H long-range correlations between H-22 (δ_H～ 1.1) and ¹⁵N₁,and between H-2" (δ_H～2.2) and ¹⁵N₂,were the most obvious signals.

In conclusion,the ¹H-¹⁵N HMBC spectra of thirteen C₁₈- diterpenoid alkaloids reported in this paper could be summarized as follows: (1) The ¹⁵N chemical shifts were largely affected by the substituents of nitrogen atoμssuch as N-Et,rather than the substitution groups on C-6,C-8,C-10,or the presence of 3,4-epoxy units,or even the types of alkaloids. (2) The corresponding δ_N values were in the range of δ_N 44-51 for the alkaloids with boat conformation of A ring (with 1α-OH),while the values were between δ_N 40-42 for those with chair-form of A ring (with 1α- OCH₃). (3) After protonation,the ¹⁵N signals for all the compounds (1-13) shifted downfield significantly (δ_N 13-20 ppm). However, the reason for the unusually strong protonation effect observed for compound 10 (δ_N 43 ppm) is unclear at the moment. Overall,the application of long-range 2D ¹H-¹⁵N HMBC experiments serves as a strong complement to the normal NMR techniques in the structural elucidation of nitrogen-containing organic compounds.

We are grateful the National Natural Science Foundation of China (No. 81273387) for the financial support of this research.

Supplementary material related to this article can be found,in the online version,at http://dx.doi.org/10.1016/j.cclet.2015.04. 028.