Quantitative analysis of HP1a binding to nucleosomal arrays Jun Y. Fan * , Jiansheng Zhou, David J. Tremethick The John Curtin School of Medical Research, The Australian National University, P.O. Box 334, Canberra, ACT 2601, Australia Accepted 14 November 2006 Abstract Elucidating how the metazoan genome is organised into distinct functional domains is fundamental to understanding all aspects of normal cellular growth and development. The ‘‘histone code’’ hypothesis predicts that post-translational modifications of specific histone residues regulate genomic function by selectively recruiting nuclear factors that modify chromatin structure. A paradigm supporting this hypothesis is the preferential binding of the silencing protein heterochromatin protein 1 (HP1) to histone H3 trimethylated at K9. How- ever, a caveat to several in vitro studies is that they employed histone N-terminal tail peptides to determine dissociation constants, thus ignoring any potential role of DNA and/or the underlying chromatin structure in the recruitment of HP1. Using a well-defined in vitro chromatin assembly system (employing a 12-208 DNA template), we describe here, the use of a fluorescence spectroscopic method that enabled us to measure and quantify the relative binding affinities of HP1a to unmodified and variant nucleosomal arrays. Using this approach, we previously demonstrated that mouse HP1a (i) binds with high affinity to naked DNA, (ii) has an intrinsic affinity for highly folded chromatin, (iii) has a 2-fold higher affinity for nucleosomal arrays when H2A is replaced with H2A.Z, and (iv) binds to DNA or chromatin in a non-cooperative manner. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Fluorescence spectroscopy; Dissociation constants; Nucleosomal arrays; HP1a; H2A.Z 1. Introduction It is clear that underpinning the generation of functional- ly distinct chromosomal domains is the specific recruitment of specialised chromatin binding proteins. Therefore, under- standing the formation of a particular chromatin functional state will depend upon the elucidation of the equilibrium dissociation constants for the required chromatin binding protein(s), and how this equilibrium is regulated. The in vitro method used to determine dissociation constants should obviously be accurate, reproducible, and highly sensitive; fluorescence spectroscopy provides such a method (if the chromatin binding protein of interest contains aromatic amino acid residues). HP1, a component of condensed chromatin, is a highly conserved protein found in organisms ranging from fission yeast to humans. In mammals, three isoforms exist (a, b, and c) [1]. HP1 has a tripartite structure, the chromodo- main (CD), a hinge region (H), and a chromoshadow domain (CSD) [2]. HP1 has been intensively studied because of the finding that the CD domain specifically interacts with the N-terminal tail of H3 methylated at K9, and this interaction is needed for the recruitment of HP1a/b to pericentric heterochromatin [2,3]. However, it is still unclear whether the CD domain is sufficient for heterochromatin binding. Examination of the literature shows that the hinge domain, which can bind to DNA, and the chromoshadow domain is also required for heterochromatin association [4]. Indeed, qualitative in vitro binding assays of HP1a to native chicken oligonu- cleosomes revealed that strong chromatin binding activity was dependent upon the hinge region and not the chrom- odomain [5]. Other factors may also contribute to HP1 binding; the conformation of heterochromatin [6] and the histone variant H2A.Z. Previously, we have shown that HP1a interacts with H2A.Z-containing nucleosomes [7] 1046-2023/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ymeth.2006.11.001 * Corresponding author. Fax: +61 2 6125 0415. E-mail addresses: jun.fan@anu.edu.au (J.Y. Fan), David.Tremethick@ anu.edu.au (D.J. Tremethick). www.elsevier.com/locate/ymeth Methods 41 (2007) 286–290