I have looked over a number of high resolution models with NAD+ and NADH in the PDB as well as small molecule structures. I also have some familiarity with similar chemistry in the decorations on the edge of bacteriochlorophyll-a molecules. The CONH2 group does flip over when the hydrogen bonding environment calls for it. It is very hard to tell the difference between the oxygen atom and the nitrogen atom from the appearance of the electron density so you always have to check the hydrogen bonding environment when building an NAD? model.
I have seen one case where a Ser -> Ala mutation in the protein caused the group to flip with interesting consequences on the far side of the co-factor. My go-to QM person tells me that flipping this group will change the energies of the molecular orbitals and therefor the redox potential of the NAD? molecule so this conformational change may be important to the action of your catalysis. I have also seen a number of NAD? models in the PDB where this group is clearly misorientated. As you note, the torsion angle should be close to zero or 180. However it is unlikely to have exactly those values because there are non-bonded clashes when everything is in one plane. Some restraint libraries inappropriately restrain this group to be co-planar with the six-membered ring. As always, check you CIF! Dale Tronrud On 5/17/2017 12:46 PM, Jorge Iulek wrote: > Dear all, > > I came across some difficulty to refine a NAD molecule in a > structure, specially its amide of the nicotinamide moiety. > A (very) brief search in deposited structures seems to point that > not so ever the C2N-C3N-C7N-N7N dihedral is close to either 0 or 180 > degrees, but in most cases it is to one of these, with a preference > towards 0 degrees. Another search in the literature, and I could not > find any study on either NAD or even the nicotinamide alone to calculate > the energy barrier to rotate around this bond (in vacuum, eg). > My data quality and resolution do not put much confidence on > B-factor differences, but they seem to indicate that the cited dihedral > angle should be close to 180 degrees, id est, O7N is "closer" to C2N > (and, consequently, to N1N) than N7N is. In fact, I have a glutamine > nearby whose terminal amide is interacting with the nicotinamide amide, > so my idea is to make one's nitrogen to interact with other's oxygen. > Concerning b-factor differences for this glutamine, they favor its NE2 > to point to nicotinamide amide, what would imply that the > C2N-C3N-C7N-N7N dihedral to would be close to 180 degrees rather than 0 > degree. > Is there any wide study on NAD nicotinamide amide conformation? > Specially, bound to protein structures? > Thanks, > > Jorge >
