SeeDev-binary@ldeleger:SeeDev-binary-17999645-1 / 3659-3700 JSONTXT

MADS-domain proteins comprise a large family of regulatory factors that have been identified in all eukaryotic kingdoms, and are involved in a diverse array of biological functions (for reviews, see Becker and Theissen, 2003; Messenguy and Dubois, 2003; Kaufmann et al., 2005). Plants uniquely possess a subclass of MADS factors referred to as MIKC, members of which contain a second weakly conserved coiled-coil motif (the K domain). Compared with other kingdoms plants have greatly expanded the number of MADS-domain proteins, with the Arabidopsis genome encoding 107 members, 39 members of which are in the MIKC subfamily (for a review, see Parenicováet al., 2003). AGAMOUS-like 15 (AGL15; At5g13790) is a member of the MIKC subfamily, which consist of four domains: the MADS (M), I, K, and C domains (for a review, see Kaufmann et al., 2005). The MADS domain itself is comprised of 55–60 highly conserved amino acids, and it is this domain that associates with the DNA (for a review, see Riechmann and Meyerowitz, 1997). The intervening (I) domain is less conserved, but forms part of the minimal DNA-binding domain (Riechmann et al., 1996). MADS-domain proteins bind as either homo- or heterodimers to an A/T-rich cis element named the CArG motif (C-A/T-rich-G, with a canonical sequence of CC[A/T]6GG; for a review, see Riechmann and Meyerowitz, 1997), and AGL15 has been shown to preferentially bind a CArG sequence with a longer A/T-rich core (C[A/T]8G) in vitro (Tang and Perry, 2003). The K domain is implicated in mediating protein–protein interactions (Riechmann et al., 1996). The carboxyl-terminal (C) domain is the most divergent, and in some cases functions as a transactivation domain (Cho et al., 1999; Honma and Goto, 2001; Lim et al., 2000; Moon et al., 1999; Ng and Yanofsky, 2001; Pelaz et al., 2001) or is involved in the formation of ternary complexes (Egea-Cortines et al., 1999). The literature contains a plethora of data demonstrating interactions between plant MADS-domain proteins (Causier et al., 2003; Cseke et al., 2007; Egea-Cortines et al., 1999; Fan et al., 1997; Favaro et al., 2002, 2003; de Folter et al., 2005; Honma and Goto, 2001; Immink et al., 2002; Jang et al., 2002; Lim et al., 2000; Mizukami et al., 1996; Moon et al., 1999; Shchennikova et al., 2004; Yang and Jack, 2004; Yang et al., 2003a,b), indicating the vast potential for modular-based regulation. There is also a rapidly growing body of knowledge where plant MADS are reported to interact with other factors, such as putative transcription factors (Causier et al., 2003; Masiero et al., 2002), or co-repressors (Sridhar et al., 2006), RNA-binding proteins (Pelaz et al., 2001), post-translational modifying factors (Fujita et al., 2003Gamboa et al., 2001; Yalovsky et al., 2000) and others (Cseke et al., 2007; Honma and Goto, 2001). MADS-domain transcription factors may function as both transcriptional activators and repressors, depending on these interactions. For instance, ectopic expression of SEPALLATA3 (SEP3) has been reported to induce AG expression outside of the floral context (Castillejo et al., 2005). However AP1 and SEP3 are also able to interact with the transcription co-repressors LEUNIG (LUG) and SEUSS (SEU) (Sridhar et al., 2006), which prevent ectopic AG transcription (Franks et al., 2002; Liu and Meyerowitz, 1995). Neither LUG nor SEU are predicted to encode a recognizable DNA-binding motif, but SEU has been shown to associate in vivo with an AG cis-regulatory region containing a putative CArG motif, perhaps through binding to DNA-bound SEP3 (Sridhar et al., 2006). AGL15 preferentially accumulates in a wide variety of tissues that develop in an embryonic mode (Heck et al., 1995; Perry et al., 1996, 1999; Rounsley et al., 1995), and constitutive expression promotes somatic embryogenesis (Harding et al., 2003). AGL15 is expressed at lower levels after completion of germination in restricted sets of cells (Fernandez et al., 2000). Research in our lab has identified a number of downstream targets of AGL15 (Tang and Perry, 2003; Wang et al., 2002; Wang et al., 2004; Zhu and Perry, 2005; unpublished data), and although some of these target genes are induced in response to AGL15, others are repressed. A number of direct target genes have been analyzed that exhibit strong association with AGL15 in vivo; yet in vitro, AGL15 binds only weakly. Taken together these data suggest that AGL15 may form heterodimers, or ternary complexes with other proteins, thus modulating the specificity and function of AGL15 in planta. AGL15 possesses the ability to directly interact with other MADS-domain proteins (de Folter et al., 2005; this study), some of which have overlapping expression patterns (de Folter et al., 2005; Lehti-Shiu et al., 2005). Although AGL15 has been reported as a protein co-purifying in the same complex as SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE1 (SERK1; Karlova et al., 2006), no direct interaction between AGL15 and a non-MADS domain protein has yet been reported. Here we report that AGL15 interacts with members of the SIN3 histone deacetylase (HDAC) complex, and that a conserved LxLxL motif present in the C-terminal domain of AGL15, which has been shown to be involved in transcriptional repression in other proteins (Ohta et al., 2001; Tiwari et al., 2004), is required for its association with one member of this complex, named SIN3-associated polypeptide of 18 kDa (SAP18), in yeast two-hybrid studies. We also show that AGL15 functions as a transcriptional repressor in vivo, and that the region where the LxLxL motif resides is essential to the repressive function. The interaction of AGL15 with members of the SIN3/HDAC1 complex suggests a mechanism that could explain its function as a transcriptional repressor in planta.

Annnotations TAB TSV DIC JSON TextAE

  • Denotations: 1
  • Blocks: 0
  • Relations: 0