Blue Light Signaling through the Cryptochromes and Phototropins. So That's What the Blues Is All About

Sunlight is the ultimate energy source for nearly all life on Earth. Yet, its importance to life extends far beyond a source of energy because it is also a critical information carrier. Plants and animals sample the light environment to gain information about local surroundings, time of day, and season of the year. Although animals can use opsin-based visual systems to capture information from their light environments, plants don't have eyes in a metazoan sense. How, then, do plants sample their light environment? In short, plants have evolved a number of diverse, nonopsin photoreceptors capable of perceiving a broad range of light qualities and intensities. Just like children who sort M&Ms into different color groups before eating them, plants utilize various photoreceptors to sort the colors of incident light. However, unlike children who ultimately eat all the M&Ms independent of color and prior sorting, plants have preference for particular colors of light, making them more selective than the child in what they consume. For example, if one considers just the visible and near-infrared regions of the light spectrum (400–850 nm), plants have evolved three major classes of photoreceptors—the cryptochromes (cry), the phototropins (phot), and the phytochromes (phy)—capable of absorbing the blue (400–500 nm) and red/far-red (600–800 nm) portions in particular (Quail, 2002). Thus, like parents who have an apparent proclivity for green M&Ms, these plant photoreceptors have a consumption preference with blue and red being the colors of choice. On the surface, it appears that plants have restricted their information gathering capacity relative to the light environment. However, it makes adaptive sense to tune ones morphogenic and developmental program to the same light qualities, blue and red wavelengths, utilized by the photosynthetic apparatus. In fact, a majority of the photomorphogenesis (light-dependent changes in morphology or development) occurring in plants is induced by blue or red/far-red signals through the aforementioned three classes of photoreceptors. Gene duplication and selection events have led to the evolution of multiple cry, phot, and phy receptors in higher plants that allow efficient sampling of the blue and red/far-red portions of the spectrum over a wide range of environmental and developmental conditions. Recent developments in our understanding of red/far-red light-dependent photomorphogenic events induced by the phy class of receptors are the subject of another Update. Here, we wish to highlight some of the exciting advances that have occurred over the past couple of years in the area of blue light sensing and response associated with the cry and phot classes of receptors. Although this Update is focused on cry and phot signaling, it is important for the reader to keep in mind that other BL-absorbing pigments, such as xanthophylls (see Talbott et al., 2003 in this issue), although not predominant in their action, can influence cry- and phot-dependent and -independent processes. The cry family of blue light receptors regulates various aspects of plant development, most notably seedling de-etiolation—the transition from a pale nonautotrophic seedling to a green photosynthetically competent one, entrainment of the circadian clock, and day length-sensitive timing of flowering (Lin and Shalitin, 2003). Crys are members of a larger blue light-absorbing chromoprotein superfamily that also includes the DNA photolyases. All members of the cry/photolyase family share an amino-terminal photolyase-related (PHR) domain that is responsible for chromophore binding (both a primary/catalytic flavin and a second light-harvesting deazaflavin or pterin) and light-absorbing capacity (Fig. 1). What, then, separates a cry from a photolyase? In short, crys from both plants and animals lack DNA repair activity that is the hallmark of the photolyases (Sancar, 2003). A recent crystallographic study has provided structural information that may explain this catalytic difference between crys and photolyases (Brudler et al., 2003). Domain features of the plant crys and phots. Most crys can be divided into two major domains: the PHR and cry carboxylterminal (CCT) domains. The PHR domain is the sensor domain and binds two chromophores: (a) a light-harvesting deazaflavin (D) or pterin (P), and (b) a primary/catalytic FAD. The CCT is thought to represent the output domain (see Fig. 2). Within the CCT, there is a sub-domain of conserved sequence designated DAS for the presence of three amino clusters: (a) DQXVP near the amino end of the CCT, (b) a region containing varying numbers of acid residues (E or D), and (c) STAES and GGXVP at the carboxyl end separated by a short non-conserved spacer. The recently discovered cry DASH is the exception to this general domain organization because it lacks the CCT and is essentially just a PHR domain (not shown). The phots can also be divided into two major regions: the LOV domain region that functions as the sensor domain and a Ser/Thr protein kinase domain that functions as the signal output domain. All phots described to date have two LOV domains, each of which bind one FMN as a chromophore. Another difference between crys and photolyases is the presence of a carboxyl-terminal extension beyond the PHR domain in the crys that is absent in the photolyases (Cashmore et al., 1999; Fig. 1). Although little sequence similarity is observed in the CCT from different crys, there are three recognizable sequence motifs in this region from plant crys: (a) DQXVP near the amino end of the CCT, (b) a region containing varying numbers of acid residues (E or D), and (c) STAES and GGXVP at the carboxyl end separated by a short non-conserved spacer. Collectively, these motifs are referred to as the DAS domain (Lin and Shalitin, 2003; Fig. 1). Molecular genetic studies of Arabidopsis cry1 and cry2 indicate that the DAS domain is crucial for function (Lin and Shalitin, 2003). As will be discussed below, considerable evidence now suggests that the CCTs are important as protein-protein interaction domains to transduce the activated cry state to downstream signaling partners. Although it is generally accepted, based on comparisons of sequences and domain architectures, that the crys arose from a photolyase progenitor, it has remained an open question whether all crys have a common progenitor and whether the split between animal and plant CRY evolution occurred before or after the appearance of eukaryotes (Kanai et al., 1997; Cashmore et al., 1999). However, as will be discussed in the following subsection, recent identification of a new class of crys provides new clarity on this issue. Genome sequencing projects provided the impetus for the identification of a new class of crys, dubbed cry DASH because orthologs are present in fruitfly (Drosophila melanogaster), Arabidopsis, Synechocystis sp., and humans (Homo sapiens; Brudler et al., 2003). Synechocystis sp. PCC6803 cry DASH has been shown to be structurally most similar to class I CPD photolyases. Interestingly, both Synechocystis sp. cry DASH and Arabidopsis cry DASH (also called cry3) bind DNA in a sequence-independent fashion, although neither exhibit DNA repair activity (Brudler et al., 2003; Kleine et al., 2003). Although Arabidopsis cry2 has been shown to associate with chromatin (Cutler et al., 2000), cry3 is the only plant cry with clear DNA-binding activity. Arabidopsis cry3 is also distinct from other plant crys examined to date in that it lacks the CCT and is targeted to chloroplasts and mitochondria (Kleine et al., 2003). Relative to intracellular targeting, Arabidopsis cry2 appears to be constitutively nuclear in residence, whereas cry1 can either be nuclear (in darkness) or cytoplasmic (in light; Lin and Shalitin, 2003). But what do cry DASH holoproteins do? Comparison of microarray expression profiles between Synechocystis sp. wild-type and cry DASH knockout mutant strains suggests that cry DASH functions as a transcriptional repressor (Brudler et al., 2003). Nothing is known at present about the function of Arabidopsis cry3; however, it is intriguing to speculate that, given its localization and ability to bind DNA, it functions as a transcriptional regulator of chloroplast and mitochondrial-encoded genes. The regulation of transcription by crys appears not to be unique to the cry DASH subfamily. For example, mouse CRY1 and CRY2 appear to repress CLOCK:BMAL1-activated transcription in the circadian clock (Kume et al., 1999). More relevant to plants, Arabidopsis cry1 and cry2 appear to derepress gene transcription by interacting with and repressing COP1 activity (Wang et al., 2001; Yang et al., 2001; see below). Do the cry DASH proteins provide any clarity with respect to the evolutionary history of the cry family? Some. The discovery of cry DASH in the Synechocystis sp., a cyanobacterium, supports the idea that crys evolved from photolyases before (Kanai et al., 1997), rather than after (Cashmore et al., 1999), the divergence of prokaryotes and eukaryotes. However, the possibility of convergent evolution within the cry family still remains. For example, Kleine et al. (2003) have speculated that plant crys evolved from two independent horizontal transfer events, with CRY1 and CRY2 having an endosymbiotic α-proteobacteria-like ancestry and CRY3 having an endosymbiotic cyanobacteria-like ancestor. How does light alter cry activity? This is one of the fundamental questions one faces when studying any photoreceptor system. Fortunately for cry investigators, a vast photolyase literature exists that can be used as a springboard to ask very direct questions about cry photochemistry. For example, comparisons of the catalytic FAD-binding pocket of Synechocystis sp. cry DASH and class I CPD photolyase from Escherichia coli demonstrate the structural conservation of three Trp residues required for electron transfer from an external reducing agent to the photoexcited flavin in photolyases (Aubert et al., 2000; Brudler et al., 2003). Results from a recent time-resolved difference spectroscopy study suggest that the reduction of photoexcited FAD in Arabidopsis cry1 also requires intramolecular electron transfer from a Trp residue (Giovani et al., 2003). So, what consequences do these photochemical events have on the function of crys? Two obvious possibilities exist. First, the photoreduction of the excited flavin in photolyase provides the electron that is used to isomerize the pyrimidine dimer in damaged DNA (Sancar, 2003). Although crys do not catalyze repair of damaged DNA, and, with the exception of cry DASH, plant crys do not bind DNA, a similar mechanism of electron transfer to a non-DNA substrate molecule could be essential for the induction of cry-dependent photomorphogenesis. Alternatively, an intramolecular redox reaction—a transfer of electrons to or from the flavin— could result in structural changes within cry itself that allow for signal transmission. Although no direct evidence exists in support of either of these possibilities, results from studies utilizing Arabidopsis cry1 and cry2 CCTs have been interpreted as supporting an intramolecular redox model (Yang et al., 2001). It has been shown that transgenic Arabidopsis plants expressing cry1 and cry2 CCTs fused to β-glucuronidase (GUS) exhibit a constitutive photomorphogenic response in darkness (Yang et al., 2000). This phenotype appears to result from the suppression of COP1 activity through the direct interaction with the CCTs (Wang et al., 2001; Yang et al., 2001). COP1 is an E3 ubiquitin ligase that tags positive regulators of photomorphogenesis, such as the transcription factors HY5 and LAF1, for proteolysis in darkness (Wang and Deng, 2002; Seo et al., 2003; Fig. 2, top). In the light, COP1 activity is repressed, resulting in the accumulation of the positive regulators and transcriptional of proteins for seedling (Wang and Deng, 2002; Fig. 2, cry and CCT bind to COP1 with similar independent of light that interaction itself is not for of COP1 (Wang et al., 2001; Yang et al., 2001). It then, that in wild-type plants, the COP1 requires a light-dependent in the CCT to COP1 (Fig. 2, Yang et al. have that an intramolecular redox between the PHR and the CCT results in a that cry to What, then, is in the transgenic the constitutive photomorphogenic of the plants, one that the by the proteins that of the CCT in a cry that has been activated by light (Fig. 2, that is with all the is that the PHR domain functions as a repressor of the CCT and that a light-dependent in whether induced by an intramolecular or this Although these are results discussed suggest that the protein is more similar to the CCT it is to the CCT in a wild-type for signaling through of gene In wild-type plants crys protein with PHR and are to with COP1 their In COP1 with transcription such as and the The the of the transcription them for by the Thus, in are not because the positive regulators are not In blue light, the PHR domain of crys is activated resulting in a structural that of the This structural to of its interaction with transcriptional proteins and the which in the transcriptional to bind to and transcription of for photomorphogenesis. Plants expressing a protein exhibit a constitutive photomorphogenic phenotype that results from a of the plants to of the COP1 of these and the that the is constitutively it is that the CCT of protein a similar to that of the CCT from wild-type cry to blue light. other cry function in et al. have recently that the CCT of Arabidopsis cry2 in response to blue light whereas in it exists in an The and of suggest that this response is important to the of In to being a signal for may also a in as cry2 is targeted for proteolysis et al., an of cry2 regulation that been to be light et al., 1999). It is to that in to protein is in both and plants et al., 2002). results suggest that the constitutive photomorphogenic of the discussed may result from the constitutive of the CCT What, if is there between these events and the redox changes that in crys? can speculate that in wild-type plants, a blue intramolecular electron transfer between the PHR and the CCT results in a that, although not to on the CCT to a constitutively protein kinase and that of the cry COP1 (Fig. 2, In the the are light is not required for COP1 (Fig. 2, the studies discussed it be obvious that a short can between crys and their ultimate in gene the cry1 and cry2 signaling seedling the receptors the and of COP1 such as HY5 (Fig. 2). this it is not that very have been Yet, recent studies have three new and (also called and a transcription that appears to function as a positive regulator of photomorphogenesis downstream of both cry1 and although functions in both is and 2003). a binding protein that appears to function as a repressor of light-dependent HY5 accumulation et al., 2001). functions both in cry and Although the mechanism by which functions is the that crys can with and repress which the of suggests that as a to accumulation of HY5 and other a Ser/Thr protein with sequence to the fruitfly that appears to function as a positive and regulator of cry signaling et al., 2003). The function of an obvious to cry2 and This appears because exhibit rather than blue light as be because cry2 is the et al. (2003) have that may the function of a nuclear signaling to the of signaling downstream of The phot blue light receptors a number of blue in plants chloroplast and and 2002). do all these have in requires the of some of the the of an to either the or the of blue light et al., the of two in response to blue light et al., and the of or from blue light 2002). However, as will be discussed these are not the only phot The phots are the members of a larger superfamily known as the LOV domain family that includes such proteins as of and of et al., 2003). The hallmark of the superfamily is the LOV and domain acid that is responsible for chromophore and capacity in LOV proteins (Fig. 1). What, then, separates the phots from other members of the domain First, the phots are the only members of the family to two LOV domains and all other members a LOV domain et al., 2003). The of having two LOV domains in the phots is not However, as will be discussed below, some important have been relative to the functions of and although a Ser/Thr protein kinase domain is to the LOV domain region in the phots (Fig. other members of the superfamily couple and output domains to the LOV sensor domain et al., 2003). Thus, although all members of the domain superfamily are capable of light sensing through the LOV output signals to downstream events from one to Although no downstream have been to it is clear that activity of the protein kinase domain of is for of signals et al., 2002). How is the activated state of a phot recent developments in the identification of downstream signaling events will be discussed The couple of years have a number of blue light relative to what them, to the of First, it has been that the of by blue light that the and response requires activity and 2001; et al., 2001). Although has no on the phenotype of in of blue light and do exhibit a in only through and 2001). and (2003) have shown that is essential for the blue light-dependent of the and in with this to and proteins when to whereas wild-type and cry1 the accumulation of these proteins et al., 2000). phot has been shown to be required for multiple aspects of in that are blue light, to of gene expression of and and 2003). is about these in is that these are the of developmental that have been associated with cry function in higher How does light by the LOV domains alter phot activity? the that both phots and crys utilize a flavin as the we these two classes of blue light receptors to exhibit conserved photochemical a more of however, and to a different For example, although crys utilize FAD as their chromophore (Lin and Shalitin, phots bind FMN et al., 2003; Fig. 1). only is the of flavin chromophore different between the crys and the amino acid sequences of their domains are also to the of two very different that different photochemical to the two classes of et al., 2003; 2003). the light of a phot LOV In phot LOV domains a in which a is between the within the conserved in the LOV domain and the of the FMN chromophore (Fig. This described by et al. and by crystallographic studies and 2002; et al., 2003). the of this in is essential for the function of in Arabidopsis et al., 2002). function does not appear to in of LOV domains from phots. In each phot LOV domain one molecule of FMN on of blue light results in the of the excited state FMN results in the of the and this excited state is by of the of the through the of a from the conserved in the LOV domain This of the state FMN the of the and of the at this resulting in the of the This is in The and are on the state and their excited are in blue on the and state This model is from that by et al. The LOV domain a different for the domain family with other For example, of most requires electron transfer to an excited flavin such as in the cry/photolyase family (Sancar, 2003). In electron transfer appear for the of phots et al., although an excited flavin can be if the critical is to et al., 2003). The in the phots appears to be a FMN state that is by transfer from the conserved to the of the excited FMN et al., 2003; et al., 2003). This at resulting in the (Fig. How does this signal comparisons of and of a domain from suggests that only changes the and 2002). results have been interpreted as that phot signal changes in the protein rather than changes et al., 2003). However, and spectroscopy studies of a domain from suggest that structural changes can et al., 2001; et al., 2002). studies from the crystallographic studies in that spectroscopy a LOV domain fused to the domain from light It then, that the protein in which the LOV domain in the structural changes that in response to light the structural ultimately to the of the protein kinase domain of the phots and of signal et al., 2003). there is only one known substrate for the phot phots As be from the of the phot is light and 2002). A recent study has that the of from is et al., 2003). In blue light of two amino-terminal and two carboxyl-terminal to whereas higher of all carboxyl-terminal to the still amino-terminal to there be and in of the signaling by protein-protein with downstream signaling such as and 1999). In at higher may be for 2002; et al., 2003). It to be whether these have and whether or not phots have other The regulation of activity by light has been to a in photomorphogenesis, blue light now known to be phot 2000). this information and the localization of both the phots et al., 2002; and and a majority of the to blue light it is not to that phots activity. However, does any evidence for such an three independent have recently that there are in First, et al. have used transgenic Arabidopsis expressing the to that a blue light-dependent in cytoplasmic in an seedling is in a In a second study utilizing transgenic plants, et al. (2003) that both and a blue light-dependent in in Arabidopsis the to use to that both and from the whereas only is capable of an in through from a in and of wild-type and phot used in a study to that a in in response to blue light et al., 2002). This like that of et al. also that of from whereas the are on et al. (2003) have used to that blue light of a is in from and is essentially in Although each of these studies a for phots in the blue light regulation of provides direct evidence for a of in any of the known phot-dependent However, the of the observed are with the of each of the known phot A more between and response has been provided by a study of et al., 2001). In this it shown that from a which essentially to exhibit blue to in response to blue light. This is known to be by an and to be essential for et al., 2001). It is intriguing to that blue light of this and 1999). the a substrate for phot kinase activity? This one of the open questions phot The of is only by the of its the new of and of will not is also an important in the of cry and phot having for proteins of one to at very As discussed in this information from structural and photochemical studies is important into cry and phot these with we will to important into understanding cry and phot we comparisons of cry/photolyase and domain family members different are also in the studies not just sequence also and function we will provide new not This is an exciting time to study time for with the like to and for and critical that because of and the to from broad in plant of the recent studies on blue light signaling, with respect to of the not discussed to all of our not discussed we all who have this Update to into the signaling