DIFFERENCES IN THE DNA-BINDING PROPERTIES OF THE HMG-BOX DOMAINS OF HMG1 AND THE SEX-DETERMINING FACTOR SRY

High-mobility-group protein 1 (HMG1) is an abundant, non-sequence-spec ific, chromosomal protein with two homologous, HMG-box, DNA-binding do mains, A and B, and an acidic tail. The HMG-box motif also occurs, as a single copy, in some sequence-specific transcription factors, e.g. t he sex-determining factor, SRY. We have investigated whether or not th ere are differences in the DNA-binding properties of the isolated A. a nd B HMG-box domains of HMG1 and SRY and whether, in the case of A and B, there might also be differences due to different sequence contexts within the native protein. The basic regions that flank the HMG1 B bo x, giving B', enhance its DNA-binding, supercoiling and DNA-bending ac tivities, and promote the self-association of the DNA-bound B-box. All the HMG-box domains bind with structure specificity to four-way junct ions, but the structure selectivity is significantly greater for A and the SRY box than for the HMG1 B or B' domains, as judged by competiti on with excess plasmid DNA. The domains self-associate to different ex tents on supercoiled DNA and this may explain differences in the abili ty to discriminate between four-way junctions and supercoiled DNA. The HMG1 A, B and B' domains constrain negative superhelical turns in DNA , but the SRY HMG box does not. Only the full B domain (B') bends DNA in a ligase-mediated circularisation assay; the minimal B box, the A d omain and the SRY box do not. Thus, despite a common global fold, the HMG box appears to have been adapted to Various functions in different protein contexts.