354 Chromo-domain proteins: linking chromatin structure to epigenetic regulation Giacomo Cavalli and Renato Paro Chromo-domain proteins appear to be a central component in the epigenetic regulation of heterochromatin function and euchromatic gene expression. The recent discovery of a variety of interacting partners of chromo-domain proteins is yielding new molecular insights into epigenetic regulatory processes acting at the level of higher order chromatin structure. Addresses Zentrum for Molekulare Biologie, University of Heidelberg, Irn Neuenheirner Feld 282, 62120 Heidelberg, Germany Correspondence: Renato Paro; e-mail: paro@sun0.urz.uni- heidelberg.de Current Opinion in Cell Biology 1998, 10:354-360 http://biomednet.com/elecref/095506 7401000354 © Current Biology Ltd ISSN 0955-0674 Abbreviations HP1 heterochromatin associated protein 1 MoMOD1 mouse modifier protein 1 ORe origin recognition complex PC Polycomb PcG Polycomb group PEV position effect variegation PRE PcG response element TIF transcriptional intermediary factor TSA trichostatin A Introduction Stable and heritable inactivation of transcription is sus- tained at the level of higher order chromatin structures. Classic examples are heterochromatic sequences around centromere regions that act as graveyards of transpos- able elements. Due to their enormous size, achieved through multiple repeats of small building blocks, het- erochromatic regions are microscopically distinguishable as highly condensed material. This suggests that, at the molecular level, specific factors must exist which organise the chromatin fibre into a densely packaged and thus transcriptionally inert unit. Recent genetic analysis in model organisms has uncovered several players involved in regulating and structuring heterochromatin. Interestingly, the stable inactivation of euchromatic genes, needed for maintaining differential expression patterns of developmental regulators, appears to utilise similar factors and principles in order to generate locally silenced, heritable chromatin structures. Several proteins associated with silenced chromatin domains were found to carry a common motif: the chromo domain. This review will focus on new information gathered on the chromo domain during the past year that yielded a first glimpse of the three-dimensional structure, its involvement in protein complex formation, and some interesting new functional roles chromo-domain proteins are exerting. Conservation and structure of the chromo domain The chromo domain was originally identified as a con- served protein motif in the Drosophila heterochromatin associated protein 1 (HP1) and Polycomb (PC) protein [1]. HP1 is encoded by Su(var)2-5, a gene first identified as a modifier of heterochromatin-induced position effect variegation (PEV), a phenomenon whereby euchromatic genes, such as the eye pigment gene white, become silenced in a metastable (variegated) manner when placed close to large blocks of heterochromatin [21. PC is encoded by a member of a large class of genes, termed the Polycomb group (PcG), that is necessary for the stable repression of developmental regulators like the homeotic genes [3,4]. The conservation of the chromo domain allowed the isolation of a family of related proteins from Drosophila as well as from other species. This family contains the functional homologues of PC and HP1, as well as other chromo-domain-containing proteins which also share additional motifs often found in nuclear proteins having potential roles in chromatin organisation [5]. HPl-like proteins are characterised by the presence of a related motif in the carboxy-terminal part, termed the chromo shadow domain [6]. In most cases chromo-domain proteins are found in repressed chromatin domains and were functionally assigned as transcriptional repressors; however, recent analysis has shown that the chromo domain is also present in proteins with activating functions, such as the Drosophila male-specific lethal (MSL)-3 protein, involved in the hyperactivation of the male X chromosome, or the human retinoblastoma-bind- ing protein [5]. Interestingly, the chromo domain was found to be specifically targeted by the autoimmune response in scleroderma patients, although the functional relationship between chromo-domain proteins and the disease is unknown [7,8]. Laue and colleagues [9°°] have resolved the structure of the chromo domain from the mouse modifier pro- tein 1 (MoMOD1; a homologue of HP1) using nuclear magnetic resonance spectroscopy. The structure consists of an amino-terminal three-stranded anti-parallel 13 sheet folding against a carboxy-terminal c~ helix (see Figure 1). Remarkably, the most striking similarity to other solved structures of known proteins was found with two small histone-like DNA binding proteins from archaebacteria (Sac7d and Sso7d). These proteins are considered to be involved in the formation of chromatin structures; however, since the particular charge distribution necessary