Mutations in the polypeptide and the DNA

In order to mutate Gln50 to Lys, the O $\varepsilon_1$ and N $\varepsilon_2$ atoms, as well as the hydrogens at the nitrogen were removed in Gln50. Subsequently, two heavy atoms: C $\varepsilon$ and N$\zeta$ were added with the COFIMA program [14]. The following step in this operation was to calculate the missing hydrogens. Finally, possible van der Waals or steric conflicts were manually minimized by rotating the side chain 50 as well as neighbouring side chains, e.g. Met54 around some of their bonds (see Figure 2.1). The latter two operations were performed using the MOLMOL program [15].

  

Figure 2.1: The Gln50$\to$Lys mutation in the protein. The bonds between heavy atoms in the four DNA-contacting side chains of the recognition helix are shown: Ile47 (yellow), Gln50 in the wild-type (blue) respectively Lys50 in the mutant (magenta), Asn51 (red) and Met54 (white in the wild type, black in the mutant).

The mutation in the DNA was performed for the four bases by superimposing the three atoms of the sugar (C₁', C₂' and O₄') of the standard DNA base provided by MOLMOL onto the corresponding atoms of the original molecule. Next, the superimposed standard sugar and the original base were removed from the file. Then, each base was rotated about the glycosidic bond to C₁' of the sugar, so that steric conflicts were minimized as much as possible. The superposition of the mutant DNA onto the original is shown in Figure 2.2.

  

Figure 2.2: The C6C7$\to$GG mutation in the DNA. Bases 8 to 10 of the $\alpha$-strand are shown in red and of the $\beta$-strand in gray. Cytosines 6 and 7 of the $\alpha$-strand in the wild type molecule are dark orange, the mutation to guanines is in cyan. The corresponding guanines of the $\beta$-strand in the wild type complex are shown in black, and the mutation to cytosines in blue.