Shedding (blue) light on algal gene expression
Light in the blue region of the spectrum [blue light (BL), 400–480 nm] is an ubiquitous environmental signal. BL can penetrate marine water to a depth greater than all other wavelengths, up to the limits of the photic zone (≈1,500 m depth) and may be linked to the evolution of photosynthesis (1). BL is also potentially dangerous because it is readily absorbed by intracellular photosensitizers (e.g., porphyrin derivatives and flavins) (2). Therefore, living organisms detect and respond to BL either by photoprotection mechanisms or by maximally exploiting this environmental input, e.g., to entrain circadian rhythms and optimize photosynthetic efficiency. Given its high penetrability, BL is of utmost importance for marine species, but little is known about the BL detection mechanisms of sea plants. The AUREOCHROMES (AUREOs) described by Takahashi and colleagues (3) in a recent issue of PNAS are the first BL receptors identified in stramenopile algae and show a clear link to the photomorphogenesis of these organisms.
The existence of BL photoreceptors in plants has long been proposed, but only recently have the flavin-binding cryptochromes and phototropins (phot) been characterized at a molecular level (4). Phot are conserved in higher plants and in several lower plant species, where they mediate a variety of BL responses (e.g., phototropism, gametogenesis) (5). What is making this research field increasingly exciting is the awareness that BL photoreceptors are widespread among distant taxa and are well represented in both eukaryotes and prokaryotes (4, 6, 7). The common feature conserved among phot-related proteins is the light-sensing, flavin-binding light–oxygen–voltage (LOV) domain, a small protein unit of ≈110 aa belonging to the PerArntSim (PAS) superfamily (8). Takahashi and colleagues (3) now show the presence of LOV proteins in the photosynthetic stramenopiles Vaucheria frigida and Fucus distichus and in a diatomean species, Thalassiosira pseudonana. These novel phot-related proteins have been named AUREO in reference to the typical golden-yellow color of stramenopiles. The authors identified the sequence of AUREOs, their photoinduced reactions, and their light-driven regulation of gene expression and photomorphogenesis in V. frigida. This represents an important step in the understanding of photoreceptor systems and BL-driven responses in these marine plants, which originated by means of secondary endosymbiosis from red algal symbionts and nonphotosynthetic eukaryotic hosts. The finding of AUREOs could provide new information on the phylogenetic link between these plants and other eukaryotes.
V. frigida hosts two AUREOs (AUREO1 and AUREO2) composed of an N-terminal, DNA-binding basic leucine zipper (bZIP) motif and a C-terminal LOV domain (Fig. 1 a), but it does not possess phot-encoding genes. Conversely, a search through various plant genomes rules out AUREO-like proteins in green plants. AUREO1 shows the spectral features and light-induced reactions typical of the well known LOV paradigm. LOV domains noncovalently bind a fully oxidized flavin mononucleotide (FMN) molecule in the dark, absorbing maximally at 447 nm. BL triggers a photocycle that involves the transient formation of an FMN–cysteine C(4a) thiol adduct, slowly reverting to the dark state on a seconds-to-hours timescale (7).
Architecture of selected LOV proteins and structure of a PAS-HtH protein. (a) Examples of LOV proteins with DNA-binding domains. From the top: AUREO from V. frigida; white-collar 1a (WC-1a) from Phycomyces blakesleeanus; helix–turn–helix (HtH) proteins from Erythrobacter litoralis (ELI_04755) and Thiomicrospira denitrificans (Tmden_2087); and opsin activator from the archaeon Natronomonas pharaonis. ZnF, zinc finger; RR, response regulator; GAF, domain present in phytochrome and cGMP-specific phosphodiesterases. (b) A possible conformation for a DNA-binding LOV protein. Shown is the crystal structure of the protein TraR from Agrobacterium tumefaciens, complexed with DNA (Protein Data Bank ID code 1H0M) (16). The N-terminal PAS-like domain (in blue) binds N-(3-oxooctanoyl)-l-homoserine lactone (in black) in a similar position as FMN within LOV domains. Dimerization is mediated both by the HtH (in red) and by the PAS-like domain, via the helical regions flanking the PAS core.
The architecture of AUREO underscores the modularity of LOV proteins. In phot, two LOV domains (LOV1 and LOV2) are coupled to a kinase effector module whose activity is enhanced upon light activation. LOV2 is necessary and sufficient for such light activation, which leads us to question the role of LOV1. In all other LOV proteins, only one LOV domain is present, coupled to a broad variety of effector functions, such as kinases and transcriptional regulators, constituting modular systems presumably switchable by light. Other than phot, fungal BL sensors of the LOV family are the best understood systems (9), but recently, bacterial LOV proteins have also begun to be examined. YtvA from Bacillus subtilis is the first bacterial protein for which the LOV paradigm has been demonstrated, followed by LOV proteins from proteobacteria and cyanobacteria (7, 10). At the functional level, light excitation increases the level of phosphorylation in bacterial LOV kinases, showing that a typical bacterial two-component system can be BL-activated (10–12). In Brucella abortus, this light-regulated kinase activity was importantly linked to the infectivity of the bacterium for mammalian cells (12). In Caulobacter crescentus, a LOV kinase is the BL receptor that regulates cell–cell attachment (10). Takahashi and colleagues al. (3) show that BL treatment of AUREO1 strongly enhances binding to its target DNA sequence, implying that AUREO1 functions as a BL-regulated transcriptional factor. The link to a photomorphogenetic function for AUREO 1—in this case BL-induced branching during plant development—is elegantly established by RNAi experiments. AUREO2 does not bind a flavin chromophore in vitro and appears to function as a secondary switch during BL-induced branching.
AUREO: Not Just Another Leucine Zipper
To our knowledge, AUREOs are the first example of a PAS domain coupled to a bZIP DNA-binding motif. More commonly associated with PAS domains are basic helix–loop–helix (bHlH), helix–turn–helix (HtH), and Zn-finger motifs. All of these proteins contain helical dimerization regions, which are essential for DNA recognition and binding. bZIPs dimerize through the stereotypical leucine zipper sequence, forming a coiled-coil structure, whereas at the N terminus, the basic region interacts with DNA via a classical scissor-like motif. Through hetero- and/or homodimerization networks, bZIP proteins increase their sequence specificity (13). It is also becoming clear that high specificity is acquired with the cooperation of additional proteins or domains, forming the so-called “enhanceosome” (14). A similar concept applies to the bHlH superfamily of transcription factors, where associated PAS domains control secondary dimerization and play a role in substrate binding by influencing the conformation of target DNA (15).
HtH transcriptional regulators also bind to DNA as dimers or multimers and additionally contain a sensing/acceptor domain, typically a PAS (e.g., in TraR) (16) or a response regulator domain (e.g., in NarL) (17). Dimerization allows target site recognition, and activation of the sensor domain is thought to unmask binding sites for DNA through alternative conformation of the interdomain linker region. The conformational flexibility of the linker region is underscored by the structure of the TraR dimer bound to DNA, where the two PAS-like domains are arranged asymmetrically to each other because of a different orientation of the linker (Fig. 1 b). Finally, we note that fungal white-collar 1 (WC-1; see Fig. 1), a Zn-finger photoreceptor LOV protein, is active as a dimer, where dimerization is mediated by the nonphotosensing PAS domains (18).
BL photoreceptors are widespread among distant taxa and are well represented in both eukaryotes and prokaryotes.
The structural/functional aspects described above for DNA-binding proteins address central questions in LOV-protein research, namely the role that dimerization, competitive surfaces, protein regions flanking the LOV core, and intraprotein interactions play during signal transduction (7). The ability to homo- or heterodimerize is a feature typical of PAS domains and has also been observed, in some cases, for LOV domains, probably with functional significance (7). It has been suggested that the β-scaffold mediates dimerization and also is involved in direct interactions with partner domains and/or with the N- and C-terminal regions flanking the LOV core, thus representing a competitive surface for multiple possible partners. In phot-LOV2, the helical linker region folds beneath the β-scaffold and becomes unfolded upon light activation (19). In the crystal structure of YtvA–LOV, the linker instead protrudes outside of the LOV core, and the β-scaffold is engaged in the tight LOV–LOV dimer (20). In full-length YtvA, the β-scaffold is involved in intraprotein/interdomain interactions, and there is no light-induced unfolding of the linker, whose exact orientation remains to be determined (21).
Light-Switching of Genes by the Versatile LOV Module
The photomorphogenesis process described for V. frigida by Takahashi and colleagues (3) relies on BL-regulated direct activation of genes by AUREO, eventually resulting in a specific growth pattern. Up to now, no DNA-binding bacterial LOV-transcriptional regulator has been studied, although such regulators are well represented within the group (7) and, together with AUREOs, are good candidates for structural and functional studies in a similar way as the TraR protein (16). These studies could open the way for photocontrolled gene expression through the design and engineering of DNA-binding proteins that are readily photoswitchable by the versatile LOV module.
Footnotes
- †To whom correspondence should be addressed. E-mail: aba.losi{at}fis.unipr.it
-
Author contributions: A.L. and W.G. wrote the paper.
-
The authors declare no conflict of interest.
-
See companion article on page 19625 in issue 49 of volume 104.
- © 2008 by The National Academy of Sciences of the USA
