2.2.2. Vincristine (**5**)

The alkaloid vincristine (**5**) from the pink-colored catharanthe (*Catharanthus roseus*) was chosen as very complex natural product. Vincristine (**5**) is an approved drug in cancer therapy. It has nine stereogenic centers, six of which are arranged consecutively in a six-membered ring and three are located in a remote ring fragment connected to the former segmen<sup>t</sup> by a single rotatable bond only, and therefore **5** is a challenging goal for a configurational analysis by NMR spectroscopy. For the configurational assignment of vincristine (**5**) altogether 23 NOEs and 24 RDCs in up to three AM, respectively, were used (all restraints were again used without assignments of diastereotopic methylene protons as described for **4**). The RDC data for **5** were taken from Ref. [22] for three independent alignment scenarios, respectively (see also Supplementary Tables S6 and S7). It also must be noted that due to the absence of NOE and RDC associated with the substituents of C-42 (a quaternary carbon carrying a hydroxyl group and a COOMe ester moiety), the configuration of this stereogenic center is not assignable based on the data used here, and thus C-42 was excluded from any further analysis.

Results for 1000 structures of vincristine (**5**) are shown in Figure 9a ("best 100") as a graphical representation of the total error for each structure, ordered according to ascending total errors. Again, a single stereogenic center of **5** was set as reference and fixed (C-9). Using NOE data only, the first wrong structure is already the structure No. #1 ranked best (black curve/circle in Figure 9a), which is a clear indication that NOEs alone are insufficient to accomplish the configurational analysis of vincristine (**5**). In analogy to the methodology employed in the case of **4**, the results can be improved by further adding RDC data to the restraints.

**Figure 9.** (**a**) Plot of the total "*pseudo energy*" of ranked rDG structures of vincristine (**5**), showing the first 100 out of 1000 structures generated (*K*NOE = 10.0 Å−2, *K*RDC = 1.5/*M* Hz−2) using only NOE restraints (black symbols) and an increasing number of additional RDC data sets (blue, green, and dark red with *M* = 1–3 alignment media). The dashed lines and Δ*E* values on the right indicate the energy levels of the first wrong configuration identified, respectively, and thus the increasing differentiability of the correct configuration when using an increasing number of RDC restraints. The inset plot shows the corresponding data obtained using only RDCs (*M* = 1–3 data sets) without any NOE restraints. (**b**) Superposition of the first 86 DG structures (top plot, *M* = 3) of correct configuration, and backbone representation of the best-fit geometry (lowest pseudo energy, bottom plot, only hydrogen atoms connected to stereogenic centers are shown) of vincristine (**5**).

Successively adding multiple AM, RDC data sets slowly increases the certainty with which the correct configuration of **5** can be assigned: with *M* = 1 (blue curve), 2 (green), and 3 (dark red) AMs (see Figure 9a). However, the level of differentiability of the correct configuration from alternate wrong diastereomers is at first very low (*M* = 1: Δ*E* = 0.04, and *M* = 2: Δ*E* = 0.12), but raises constantly to Δ*E* = 1.63 when using RDCs from three alignment media (*M* = 3). In the latter case, the first wrong structure is No. #87 (dark red triangle in Figure 9a), and this diastereomer is already separated from the best-fit (pseudo energy minimum) correct diastereomer of **5** by a now significant step in the error function, which is due to a configurational change of the quaternary carbon C-17 (at structure #87). Another step (not shown) follows at structure #245 (Δ*E* = 6.04) due to a misassignment of C-41. In conclusion, the most problematic stereogenic centers to be determined for vincristine (**5**) are–as discussed above–C-42, C-17, and C-41 (in this order), whereas the reliability with which any of the other six stereogenic centers is differentiated from alternate configurations is high when NOE and RDC data is used in combination (see Figure 10 for a traffic-light type encoded pictorial description of these assignment probabilities). The problems associated with C-17 arise from limited RDC data available for the rotating ethyl side chain, and for C-41, only a single CH RDC and two NOEs indicate some preference for the correct configuration over a wrong assignment. However, Figure 9a and the inset plot therein clearly indicate the importance of using combinations of both NOE and RDC data sets, as neither NOEs nor RDCs alone provide conclusive evidence for the correct configuration of **5**, and in particular calculations relying on RDCs only gave much less conclusive results as compared to the combined approach.

**Figure 10.** Structure of vincristine (**5**). The color-coded atom markers in the formula of **5** correspond to di fficulties in the configurational assignment (for details, see text).
