Functional domain structure of human centromere protein B. Implication of the internal and C-terminal self-association domains in centromeric heterochromatin condensation.

Abstract

Centromere protein B (CENP-B) is a common centromere DNA-binding protein among mammalian centromeres. CENP-B possesses the specific DNA binding activity to the 17-base pair sequence dispersed in centromeric repetitive DNA sequences. In the previous study, we have shown that its DNA-binding domain exists within the N-terminal 134 amino acid residues. Here, to clarify the whole domain structure, another functional unit required for CENP-B self-association was examined. Recombinant CENP-B was expressed in Escherichia coli. First, a chemical cross-linking reagent, disuccinimidyl suberate, was used to fix the physical association without losing the DNA-binding activity. The complexes with the same molecular weight as homodimer and trimer were identified after a separation by SDS-polyacrylamide gel electrophoresis. With a series of CENP-B deletion constructs, the area responsible for this oligomer formation was located at the internal region. Second, an electrophoretic mobility shift assay was also used to survey the minimum regions required for the CENP-B self-association. Three separate elements were identified by assaying the capacity to form the additional slow-migrating complex, two in the internal region and one in the C terminus. These results suggest that CENP-B molecules interact with each other at the multiple sites to fold the centromeric DNA repeats into a heterochromatin structure.