Figure 3.

Predicted structures of NBS domains. Structural models for the NBS domain of TNL RPS4 and CNL RPS5 of Arabidopsis were generated using a self-consistent mean-field homology modeling technique [95], in the absence of ADP. ADP was added to the two NBS models by inference from the APAF-1-ADP complex without further refinement of the models to illustrate the position of the nucleotide relative to the conserved motifs. (a)The structures of the NBS domains of RPS4 and RPS5, showing the positions of the conserved motifs. The protein structures are shown as ribbon diagrams and ADP is shown as a stick model. TIR-type and CC-type NBS domains are made up of motifs, in order from the amino terminus [3]: P-loop (or Walker A site, blue); RNBS-A (green); kinase-2 (or Walker B site, magenta); RNBS-B (green); RNBS-C (green); GLPL (yellow); RNBS-D (green); MHDV (orange). (b) The binding sites of human APAF-1 (PDB code 1z6tA), Arabidopsis RPS4, and RPS5, showing the residues interacting with ADP and ATP. The coordination of ADP in the three proteins involves three different conserved motifs. A small anchor region at the amino terminus of the NBS domain coordinates the adenine of ADP or ATP, the P-loop coordinates the α- and β-phosphates, and the MHDV motif (in the winged-helix subdomain in APAF-1) coordinates either the sugar or the β-phosphate of ADP. The two terminal aspartic acids from the kinase-2 motif are located in the pocket in which the γ-phosphate of ATP would sit. Images were generated using PyMol [96].