Inhibitor binding electron density maps

Type I (uninhibited) crystals. In a rather marked contrast, Fig. 6b shows the same view of the 4-A electron density map of the Type II (inhibited) crystals. A mass of electron density now appears in the pocket, and there also appears to be some added density at the left side of the perimeter of the pocket itself. It should perhaps be reemphasized here that these structures of the Type I and Type II crystals were determined completely independently using one set of heavy atoms and unsubstituted data for the uninhibited nuclease and another entirely different combination for the nuclease-inhibitor complex. However, the fact that heavy-atom positions appear very distinctly when intensities from Type I crystals are combined with Type II phases (and vice versa) indicates that the nuclease undergoes no gross conformational changes with pdTp and Ca2+ binding (31) thus, the kind of comparison presented... 

The problem is a general one of placing a rigid molecular structure in an electron density map in the most objective fashion. There are a number of different types of such problems in protein crystallography. One particular instance is the placement of small substrate (or equally well an inhibitor) structures in poor electron density maps of protein binding sites. Another case is when homologue structures are used in the initial location of new protein structures within the unit cell of poor electron density maps. The location of a known structure within a poor electron density map from a different space group is yet another variant. 

The cofactors linked to the amino-acid residues on the L- and M-subunits might be labeled with subscripts L and M. However, this designation becomes a little complicated when it comes to the quinones, as the quinone that is huried in a pocket formed from amino-acid residues ofthe M-subunit is actually located in a region dominated by a branch of the L-subunit. We will therefore adopt a nomenclature which uses the subscript A for an L-dominated branch and B for an M-dominated branch, i.e., Ba and Bq for BChls b, BOa and BOg for BOs b, and Qa and Qb for the two quinones. Thus Qa, menaquinone-9, although hound to the M-subunit is located at the end of an L-branch, while Qb, ubiquinone-9, is bound to the L-suhunit but located near the end of an M-branch. It is to be noted that -70% of the Qs-binding sites in the Rp. viridis crystals appear as empty cavities on the electron-density map, as most of the Qb was lost during isolation ofthe Rp. viridis reaction center. This cavity, however, can be filled with some exogenous quinone or a molecule of the inhibitors o-phenanthroline or terbutryn. 

Difference electron density maps are often calculated to show changes between the native protein and some altered state where this includes deletion from or addition to the molecule such as binding an inhibitor to it.13 Such maps are calculated by use of the phases for the native protein and differences in the amplitudes between the native and altered state as coefficients in the Fourier series. Such difference density maps are interpreted in the same way as electron density maps and with the same qualifications. They are subject to rather more uncertainty, however, since the phases are for the native protein rather than for the difference vectors between the native and altered state reflections. 

Fig. 27. A difference electron-density map showing inhibitor binding (Courtesy of H. L. Carrell).

In this paper we discuss the results obtained from an examination of the 2.5 A resolution electron density map, and present some of the data obtained from binding pepstatin (an acid-protease-specific inhibitor) to the enzyme in the crystals. 

When a substrate or inhibitor binds to a protein, it displaces water. As a result, electron density for ordered water is replaced by electron density for part of the ligand molecule. This means that there may be no appreciable peak in the difference map. In addition, because of somewhat incorrect phases, the substrate or inhibitor will appear in a difference map with reduced electron density, usually about half that of a well-phased map (half-weight AF map). Consequently, the practice has sometimes been to multiply coefficients. Often a map that combines the features of a difference map Fph Fp, enhanced by a factor of two, and of the native protein map Fp, is used. The coefficients of the Fourier synthesis are then ... 

---