Abstract
XPA, an essential protein in nucleotide excision repair (NER), interacts with damaged DNA and other proteins (RPA, ERCC1 and TFIIH) to remove a wide variety of chemically and structurally distinct DNA lesions from the eukaryotic genome. To understand the structural basis for the role of XPA in the repair process, the structure of the minimal DNA binding domain of human XPA [XPA-MBD (M98-F219)] was studied by NMR spectroscopy. A three-dimensional structure for XPA-MBD was generated using distance geometry and simulated annealing methods from NOE-based distance restraints, hydrogen bond and Zn-S distance restraints, and dihedral restraints. The structure calculations indicate that XPA-MBD contains elements of well-defined secondary structure interspaced with disordered loops organized into two non-interactive sub-domains: a zinc-binding core (D101-K137) and a loop-rich domain (L138-F219). The zinc-associated core contains an antiparallel beta-sheet (Y102-C105 and K110-M113) and an alpha-helix (C126-K137) separated by a poorly defined turn, reminiscent of the structure of the zinc-binding domain of the chicken erythroid transcription factor GATA-1 when bound to its cognate DNA sequence. The loop-rich domain contains a triple-strand antiparallel beta-sheet (L138-T140, L182-M178 and K163-K167), three loops (K151-L162, N169-D177 and Q208-F219) and three alpha-helices (K141-L150, K183-W194 and Q197-R207). The XPA-MBD structure is discussed in terms of known functions: binding single- and double-stranded DNA and binding RPA.