DNA sequence context and protein composition modulate HMG-domain protein recognition of cisplatin-modified DNA.

Proteins containing the high mobility group (HMG) DNA-binding domain form specific complexes with cisplatin-modified DNA which shield the major intrastrand d(GpG) and d(ApG) cross-links from excision repair. The molecular basis for the specificity of binding was investigated for the two isolated domains of HMG1 with a series of 15-bp oligonucleotides, d(CCTCTCN1GGN2TCTTC). (GAAGAN3CCN4GAGAGG), where asterisks denote N7-modification of guanosine with cisplatin. Alteration of the nucleotides flanking the platinum lesion modulated HMG1domA recognition in this series by over 2 orders of magnitude and revealed an unprecedented preference for N2 = dA > T > dC. The flanking nucleotide preference for HMG1domB interaction with this oligonucleotide series was less pronounced and had only a 20-fold range of binding affinities. For the N1 = N2 = dA 15-bp probe, 100-fold stronger binding occurred with HMG1domA (Kd = 1.6 +/- 0.2 nM) compared to HMG1domB (Kd = 134 +/- 18 nM). The platinum-dependent recognition of the N1 = N2 = dA 15-bp probe saturates at 1 equiv of HMG1domA and is highly specific, as evidenced by the 1000-fold decrease in HMG1domA binding affinity for the corresponding unplatinated oligonucleotide. HMG domains were unable to bind specifically to cisplatin-modified DNA-RNA hybrids, revealing the need for a deoxyribose sugar backbone for specific complex formation with HMG-domain proteins. Protein-DNA contacts which may account for these observed binding preferences are proposed, and potential implications for the biological processing of cisplatin-DNA adducts are discussed.