LysR-type transcriptional regulators (LTTRs) constitute the largest family of regulators in prokaryotes. To date, more than 3800 proteins have been found to be members of this family and the number is rising steadily (Henikoff et al., 1988; Pareja et al., 2006, http://www.era7.com/ExtraTrain). They regulate a diversity of functions, including the synthesis of virulence factors, CO2- and N2-fixation, antibiotic resistance, catabolism of aromatic compounds and the biosynthesis of amino acids, among others (Schell, 1993; Gerischer, 2002; Tropel and van der Meer, 2004; Maddocks and Oyston, 2008). Their common features comprise sequence lengths of around 300 residues, high sequence similarity at the N-terminal winged-Helix–turn–helix (wHTH) motif for DNA binding (first 60 residues), an inducer binding C-terminal domain and, with few known exceptions (Sainsbury et al., 2009), a homotetrameric quaternary structure of the active species. Numerous studies have described individual LysR-type regulators and established the defining characteristics but a detailed molecular characterization is only available for a few model systems, e.g. BenM and CatM (Collier et al., 1998; Clark et al., 2004; Ezezika et al., 2006; 2007a; Craven et al., 2009), AtzR (Porrúa et al., 2007), CatR and ClcR (Parsek et al., 1995; McFall et al., 1998), CbbR (Van keulen et al., 2003; Dangel et al., 2005), CysB (Lochowska et al., 2001; Lochowska et al., 2004), CbnR (Ogawa et al., 1999), OxyR (Kullik et al., 1995; Choi et al., 2001) and OccR (Wang et al., 1992; Wang & Winans, 1995a,b). LTTRs may act as activators or repressors of transcription. After binding an inducer (often a metabolite of the regulated pathway) ‘classical’ LTTRs positively regulate transcription from their target promoters, while repressing their own expression. However, this is not an exclusive mechanism, and subgroups of LTTRs are now also known that catalyse positive autoregulation, or act as transcriptional repressors of other target genes (Maddocks and Oyston, 2008). Footprints upon DNase I digestion studies and circular permutation assays (Wang et al., 1992; Leveau et al., 1994; Wang & Winans, 1995a,b; McFall et al., 1998; Ogawa et al., 1999; Porrúa et al., 2007) suggest a common regulatory mechanism for many LTTR members in which in the absence of inducer, a high affinity Recognition Binding Site (RBS), upstream from the transcription start site, characterized by the presence of an imperfect dyad symmetry within the consensus sequence [T-N11-A] (Schell, 1993), is necessary and sufficient for anchoring the LTTR onto the DNA. An inducer-responsive, Activation Binding Site (ABS), closer to the transcription start site binds the LTTR with less affinity. In addition, multiple binding sites, even on different promoters (Tropel & van der Meer, 2004) have been described for a number of LTTRs along with an induced bending of the DNA bound and the need for conformational flexibility of the regulator multimer (Maddocks and Oyston, 2008). In particular, a conformational change upon ligand binding causes some LTTRs like AtzR to move a variable number of base pairs from a more proximal ABS subsite (ABS′) to a more distal ABS subsite (ABS″), usually accompanied by a relaxation of the DNA bending angle. This is the basis of the so-called ‘sliding dimer’ hypothesis of LTTR activation (Fig. 1A) (Porrúa et al., 2007). Cases are known as well, where very slight differences in binding–site positions or bend angle are observed, but still, differences in the positioning of the activator relative to RNA polymerase can alter the mechanism of activation by influencing interactions with alpha-CTD or sigma 70 (Fritsch et al., 2000).
Figure 1. Transcription regulation by TsaR, a LysR-type transcription regulator. A. Schematic representation of the sliding dimer hypothesis of LTTR activation. Left: non-activated LTTR tetramer, bound to the high affinity RBS site and to the proximal ABS′ subsite of the promoter, with a high angle DNA bending. Right: the ligand-induced conformational change leaves the regulator bound to the RBS site and to the more distal ABS″ subsite, while relaxing the bending angle of the underlying DNA. B. Degradation of TSA as sole source of carbon and energy by C. testosteroni T-2. TSA is transformed to para-sulfobenzoate (PSB). The latter is desulfonated to form protocatechuate (PCA), which is subsequently mineralized via the tricarboxylic acid cycle, following meta-ring cleavage. C. Sketch of the tsa-operon, including the genes for the LTTR TsaR and the TSA transporter TsaST. The tsaR/tsaMBCD promoter region is shown enlarged. Putative transcription elements are highlighted above (TsaR) and below (TsaM) the DNA sequence. DNA fragments described to bind TsaR in the presence and in the absence of TSA are represented as grey bars (Tralau et al., 2003b). Region displaying two overlapping pseudopalindromic [T-N11-A] consensus motifs required for LTTR-boxes (Schell, 1993) are shown framed in black background. Enzymes: p-toluenesulfonate methylmonooxygenase (TsaMB; oxygenase M, reductase B), p-sulfobenzylalcohol dehydrogenase (TsaC), p-sulfobenzaldehyde dehydrogenase (TsaD), p-sulfobenzoate-3,4-dioxygenase (PsbA(C)).
Download figure to PowerPoint
Despite the abundance of LTTRs, only two full-length structures have been reported, those of CbnR (Muraoka et al., 2003a,b) and CrgA (Sainsbury et al., 2009). Additionally, the full-length structure of a putative LTTR from Pseudomonas aeruginosa is deposited in the PDB under code 3FVZ as a result of a structural genomics study. This scarcity is presumably due to the generally observed low solubility and inherent structural flexibility, as multi-meric proteins with flexible linkers (Verschueren et al., 2001; Clark et al., 2004; Stec et al., 2004; Ezezika et al., 2007b). However, as the co-inducer binding C-terminal domain is of particular interest, several studies have reported partial LTTR structures in which the DNA-binding N-terminal domain is not resolved, including structures of BenM and CatM (Ezezika et al., 2006; Craven et al., 2009), DntR (Smirnova et al., 2004), OxyR (Choi et al., 2001) and CysB (Verschueren et al., 1998). While these studies have yielded valuable insight into details of substrate binding and how it affects the molecular biology of the corresponding regulator, the approach of studying partial domains is known to miss the context of domain-domain interactions (Bjerrum and Biggin, 2008; Kobe et al., 2008).
TsaR is a member of the LTTR-family, regulating the degradation of para-toluenesulfonate (TSA) as sole source of carbon and energy by the soil bacterium Comamonas testosteroni T-2 (Cook et al., 1999; Tralau et al., 2003a). TSA is a commonly found pollutant as it is widely used as a catalyst in the chemical industry, in the production of foundry moulds and in laundry detergents. It is degraded by C. testosteroni T-2 via p-sulfobenzoate and protocatechuate (Fig. 1B). A TSA-transport system and the first three metabolic steps in the degradation of TSA are encoded on the 72 kb plasmid pTSA by the genes tsaST and the operon tsaMBCD respectively (Junker et al., 1997; Tralau et al., 2001; Mampel et al., 2004) (Fig. 1C). The TsaR transcriptional regulator is encoded divergently with respect to tsaMBCD. It consists of 299 amino acids and has a molecular weight of 32.7 kDa. Transcription of TsaR is induced by TSA but not with terephthalate, protocatechuate or succinate (Tralau et al., 2003a). Knockout studies have shown that TsaR is essential for the transcription of tsaMBCD and specific binding of TsaR occurs at three different fragments within the tsaR/tsaMBCD promotor region (Tralau et al., 2003a,b). A consensus [T-N11-A] LTTR-box was found on three of the corresponding DNA fragments, and the same study shows that binding to three of them requires the presence of TSA as co-inducer. TsaR also binds specifically to the promoter region of the gene encoding the inducible transporter component TsaT, where it probably regulates transcription together with a second regulator, TsaQ (Tralau et al., 2003b).
In this paper we present a structural study of the full-length TsaR, ligand-free in two crystal forms, and in complex with the natural inducer TSA. Our study reveals a novel conformation of a tetrameric LysR-type regulator, which in accordance with previous insights from structural and biochemical studies on other members of the family, should broaden the structural interpretation of LTTR activation.