Groucho/Tup1 family co-repressors in plant development Zhongchi Liu 1 and Vidyadhar Karmarkar 1, 2 1 Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD 20742, USA 2 Department of Plant Science and Landscaping Architecture, University of Maryland, College Park, MD 20742, USA Transcription repression is emerging as a key regulatory mechanism underlying cell fate specification and body patterning in both animals and plants. In animals and fungi, the Groucho (Gro)/Tup1 family co-repressors gen- erate the repressed chromatin state in genetic loci that control major developmental decisions ranging from dorsal–ventral patterning to eye development. In higher plants, information about the Gro/Tup1 co-repressors is beginning to emerge. Several recent publications have revealed both conserved and unique structural and mechanistic features of plant Gro/Tup1 co-repressors, including LEUNIG (LUG), TOPLESS (TPL) and WUSCHEL- INTERACTING PROTEINS (WSIPs). These co-repressors regulate key developmental processes in floral organ identity specification, embryo apical-basal fate determi- nation, and stem cell maintenance at the shoot apex. Introduction: Gro/Tup1 family co-repressors in animals and fungi Transcription repression is an important regulatory strategy that inhibits the expression of key regulatory genes, the inappropriate expression of which often leads to abnormal development. As with all co-repressors, Drosophila Groucho (Gro) and its mammalian homolog Transducin-like enhancer of split (TLE) lack intrinsic DNA-binding ability and are recruited by sequence-specific DNA-binding transcription factors to regulate target gene expression [1]. Gro/TLE proteins are characterized by an N-terminal glutamine (Q)-rich domain and C-terminal WD-repeats (Figure 1a). Each WD-repeat motif consists of 40 amino acid residues with characteristically located tryptophan (W) and aspartate (D) residues [2]. The ‘b- propeller’ structure of the WD-repeats mediates their interactions with transcription factors such as Engrailed and Dorsal [3,4]. By contrast, the Q-rich domain is prim- arily involved in homo-tetramerization of the co-repressors [1]. The Q-rich and WD-repeat domains are separated by a less-conserved region implicated in transcription repres- sion and nuclear localization [1]. In Drosophila, Gro can be recruited by several tran- scription factors, including the basic helix–loop–helix (bHLH) repressor Hairy, the Rel-family protein Dorsal, and the homeodomain protein Engrailed. In doing so it is able to regulate diverse developmental processes, in- cluding lateral inhibition, segmentation, sex determi- nation, dorsal–ventral and terminal pattern formation, and eye development [5]. Gro interacts directly with short peptide motifs (e.g. Trp-Arg-Pro-Trp) present in these DNA-binding transcription factors. The mammalian TLE proteins interact with the transcription factors AML1 and AML2 (acute myeloid leukemia 1 and 2) to regulate hema- topoiesis and osteoblast differentiation. Similarly, TLE interacts with the transcription factor LEF1 (lymphocyte enhancer factor 1) to regulate Wnt signaling during cell fate determination [6,7]. Other than the C-terminal WD-repeats, there is no significant sequence homology between Saccharomyces cerevisiae Tup1 and Drosophila Gro. However, Tup1 and Gro are regarded as functional analog, owing to their similar domain organization and mechanisms of repres- sion [1]. Unique to Tup1 is the presence of an N-terminal domain (Figure 1a) required for its tight association with Ssn6 (suppressor of snf1), a Q-rich protein with 10 tetra- tricopeptide repeats that function to mediate protein– protein interactions [8]. When tethered to reporter genes by LexA, Tup1 is capable of repressing genes in the absence of Ssn6, but Ssn6 is not capable of repressing genes in the absence of Tup1, suggesting that Ssn6 mainly functions as an adaptor protein between the Tup1 repressor and particular DNA-binding transcription factors [9–11]. In a similar fashion to Drosophila Gro, yeast Tup1–Ssn6 co- repressor complex represses genes involved in diverse pathways, including glucose metabolism, DNA damage repair, mating-type switch, and anaerobic respiration [12]. As with other co-repressors, the specificity of repres- sion is determined by its interaction with sequence-specific DNA-binding proteins, including Mig1 (multicopy inhibi- tor of GAL), Crt1 (constitutive RNR transcription) and a2 [13]. Tup1–Ssn6 mediates repression by multiple mechan- isms [12]. First, both Tup1 and Ssn6 were found to phy- sically interact in vivo with multiple histone deacetylases (HDACs; Box 1), including Rpd3 (reduced potassium dependency), Hos1 (HAD one similar) and Hos2 [14]. By recruiting HDACs, Tup1–Ssn6 is able to deacetylate histones at target promoters, which in turn stabilizes the interaction between the histones and the Tup1 com- plex, creating a self-reinforcing repressive state and an effective long-range repression [5,12,15] (Box 1). Second, nucleosome repositioning is observed in genes that are subject to Tup1 repression, limiting the accessibility of promoter and/or enhancer elements to the transcription activation machinery [12]. Third, Tup1–Ssn6 interacts with several components of the Mediator complex [12] Review Corresponding author: Liu, Z. (zhongchil@gmail.com). 1360-1385/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tplants.2007.12.005 137