Figure 11.29

of the purified products ("Product 2") shows two !H resonances in the "alcohol" region (CH-OH, 3-4.5 ppm), each with an integral of one proton. Because testosterone itself has one triplet resonance in this region (H-17), this indicates that one CH2 carbon was oxidized to CH-OH to give Product 2. There are eight positions that could have been oxidized: 1, 2, 6, 7, 11, 12, 15, and 16. The "new" CH-OH resonance at 4.15 ppm is a ddd (J = 7.5, 5.8, 2.4 Hz), so we can test whether this coupling pattern is consistent with oxidation at each of these positions. The ddd resonance is "new" because H-17, the existing CH-OH group in testosterone, cannot have more than two vicinal 1H coupling partners. C-16 is ruled out because oxidation would remove one of the couplings to H-17, which appears as a triplet. To be a ddd, the CHOH must be between a CH and a CH2 carbon. This rules out C-1, C-2, C-6, and C-12 because they are next to quaternary carbons. Only positions 7, 11, and 15 are possible: Each is between a CH and a CH2 group.

Figure 11.30 shows a portion of the HMBC spectrum of Product 2, including the "alcohol" region in both dimensions (3.2-4.4 in F2 = 1H and 67-85 ppm in F1 = 13C). The HSQC peaks are superimposed and shown in parentheses, and the 1H resonances are shown above or below these peaks with an expansion of the ddd at 4.15 ppm. One-bond artifacts are clearly visible in the HMBC spectrum as wide doublets centered on the position of the 13C-decoupled HSQC crosspeaks (compare to the diagram in Fig. 11.11). In addition a "fat" HMBC crosspeak is observed between the "new" 1H resonance at 4.15 ppm in F2 and the C-17 13C resonance at 82 ppm in F1 (Fig. 11.30, lower left side). This means that the CH-OH proton at the position of oxidation is two or three bonds away from C-17: Of the three possibilities (7, 11, and 15), it must be H-15. H-11 would be four bonds away and

4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3 Proton (ppm)

4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3 Proton (ppm)

H-7 would be five bonds away, so an HMBC crosspeak would be impossible. Note that the corresponding crosspeak between H-17 and C-15 (F2 = 3.48 ppm, F\ = 70 ppm, upper right side) is not observed; this is because the H-C17-C16-C15 dihedral angle is different from the C17-C16-C15-H dihedral angle, leading to a small 3 JCH for the former and a large 3JCH for the latter. HMBC does not exhibit the same symmetry as the COSY spectrum because the relationships are not equivalent.

Having established the position of hydroxylation (the regiochemistry), the stereochemistry is the next important question. The orientation of the activated oxygen species in the enzyme with respect to the plane of the steroid is expected to lead to hydroxylation on one side only, and determining which side could lead to an understanding of the geometry of substrate binding at the enzyme active site. Although we normally think of NOE experiments to determine stereochemistry, it is actually more useful to look at dihedral angles and J couplings. Two energy-minimized models of this relatively rigid molecule, one for oxidation at C-15 on the a face and one for oxidation on the j face, give the following dihedral angles:

0 0

Post a comment