A Cyanobacterial Phytochrome Two-Component Light Sensory System

See allHide authors and affiliations

Science  05 Sep 1997:
Vol. 277, Issue 5331, pp. 1505-1508
DOI: 10.1126/science.277.5331.1505


The biliprotein phytochrome regulates plant growth and developmental responses to the ambient light environment through an unknown mechanism. Biochemical analyses demonstrate that phytochrome is an ancient molecule that evolved from a more compact light sensor in cyanobacteria. The cyanobacterial phytochrome Cph1 is a light-regulated histidine kinase that mediates red, far-red reversible phosphorylation of a small response regulator, Rcp1 (response regulator for cyanobacterial phytochrome), encoded by the adjacent gene, thus implicating protein phosphorylation-dephosphorylation in the initial step of light signal transduction by phytochrome.

The ability to cope with a continuously changing light environment is essential to the survival of all organisms that rely on sunlight for energy. Photosynthetic organisms, from bacteria to higher plants, possess numerous light-sensing molecules for perception and adaptation to fluctuations of intensity, direction, duration, polarization, and spectral quality of light (1). Most well known of these photoreceptors are the phytochromes, which sense ambient light conditions by their ability to photointerconvert between red (Pr) and far-red (Pfr) light-absorbing forms (2). The hypothesis that phytochrome is a light-regulated enzyme was proposed nearly 40 years ago (3). Despite evidence that purified plant phytochromes exhibit protein kinase activity (4) and possess a COOH-terminal domain similar to that of bacterial histidine kinases (5), the enzyme hypothesis remains controversial.

Identification of the rcaE gene from the cyanobacterium Fremyella diplosiphon, which encodes a protein that is structurally related to higher plant phytochromes and bacterial histidine kinases, has renewed interest in the possibility that phytochrome is a protein kinase (6). Other phytochrome-like open reading frames (ORFs) have been noted in the cyanobacterium Synechocystis sp. PCC6803 genome (6,7). One of these ORFs, locus slr0473, encodes a 748-residue polypeptide whose expression in Escherichia coli and incubation with phycocyanobilin (PCB), yielded an adduct with a red, far-red photoreversible phytochrome signature (8). Closer inspection of this phytochrome locus, which we have namedcph1 for cyanobacterial phytochrome 1, reveals another ORF only 10 base pairs (bp) downstream, locus slr0474, which we have namedrcp1 for response regulator for Cph1 based on this study (Fig. 1A). Because the COOH-terminal domain of Cph1 contains all conserved features of histidine kinase transmitter modules (Fig. 1B) and rcp1 encodes a 147–amino acid protein related to the CheY superfamily of bacterial response regulators (Fig. 1C), which contain aspartate kinase receiver modules, we investigated whether these proteins represent a functional light-regulated transmitter-receiver pair (9).

Figure 1

Phytochrome operon of Synechocystis sp. PCC6803. (A) Genomic organization of the phytochrome-related gene cph1 (locus slr0473) (GB:D64001, locus 1001165) and the adjacent small response regulator genercp1 (locus slr0474) (GB:D64001, locus 1001166). (B) Deduced amino acid sequence of Cph1 (26). Highlighted residues are 100% conserved between Cph1 and 21 full-length eukaryotic phytochrome sequences in the nonredundant GenBank and EMBL databases (13). The conserved cysteine for bilin attachment for eukaryotic phytochromes is shown in a black box. Underlined protein sequences represent the five conserved motifs of transmitter modules (9). Outlined H represents the conserved histidine autophosphorylation site. (C) Multiple sequence alignment of Rcp1 and response regulators RcaF (17), CheY (27), and SpoOF (18). Invariant aspartate, threonine, and lysine residues of the CheY superfamily are boxed, and conserved residues are shaded.

Affinity-tagged versions of both proteins were cloned by polymerase chain reaction (PCR) and expressed in E. coli(10). That Cph1 is a functional phytochrome homolog was demonstrated by its ability to catalyze its own chromophore attachment to yield photoreversible adducts with the higher plant chromophore precursor phytochromobilin (PΦB) and its phycobilin analog PCB (Fig.2A). Assembly with phycoerythrobilin (PEB), a phycobilin analog that lacks the C15 double bond found in PCB and PΦB, also produced a covalent adduct as visualized by zinc-blot analysis (Fig. 2B). The PEB adduct of Cph1 was photochemically inactive, thus demonstrating that photoisomerization of the C15 double bond is required for Cph1 photoactivity, as is the case for higher plant phytochromes (11). The Cph1 deletion mutant N514, which lacks the transmitter domain (12), also bound all three bilins covalently (Fig. 2B), yielding PΦB and PCB adducts with absorption difference spectra indistinguishable from the full-length photoreceptor (13). These data indicate that the NH2-terminal region of Cph1 delimits a functional photosensory domain (Fig. 2C) consistent with the structure and photochemistry of eukaryotic phytochromes (2, 14).

Figure 2

Spectroscopic and biochemical properties of bilin adducts of recombinant Cph1. (A) Phytochrome difference spectra of 40% ammonium sulfate–fractionated, Cph1-containing protein extracts (11) after incubation with PΦB (short dashes), PCB (solid line), or PEB (long dashes). (B) Visualization of PΦB, PCB, and PEB adducts of Cph1 and N514 mutant on polyvinylidene difluoride (PVDF) membranes treated with zinc acetate (upper) (11) or alkaline phosphatase conjugated to streptavidin (lower) (23). Molecular mass markers at 119, 83, and 47 kD (top to bottom) are indicated with dots; apo, apoprotein. (C) Structural model for prototypical eukaryotic andSynechocystis apoprotein phytochromes. The phytochromes share a similarly sized photosensory domain (open rectangle) containing a conserved cysteine chromophore binding site (*) and a COOH-terminal transmitter-related module (dark-shaded rectangle). Prototypical phytochromes also contain a small NH2-terminal extension and a second transmitter-related module (light-shaded rectangle) that contains the PAS A and B repeats (13, 28). aa, amino acids. (D) Dark reversion of PCB and PΦB adducts of full-length Cph1 (• and ○, respectively) and N514 mutant (▴ and ▵, respectively) (29).

Cph1 is smaller than eukaryotic phytochromes; it lacks a 60– to 100–amino acid NH2-terminal fragment found on the photosensory domains of prototypical phytochromes and approximately one-half of the COOH-terminal region (Fig. 2C). Removal of the NH2-terminal portion of higher plant phytochromes blue-shifts its Pfr absorption maximum and attenuates its biological activity (14). Consistent with these observations, Pfr absorption maxima of Cph1-bilin adducts are blue-shifted relative to higher plant phytochrome bilin adducts whereas Pr absorption maxima are similar (Fig. 2A). The dark reversion properties of the two bilin adducts of Cph1 are particularly interesting (Fig. 2D). PCB adducts of Cph1 and the N514 deletion mutant display little dark reversion, whereas PΦB adducts show considerable dark reversion, with respective half-lives of 10 and 24 hours. In addition to demonstrating that Pfr stability depends on chromophore structure, these results indicate that the transmitter domain influences the conformational stability of the chromophore domain. In view of the potential role of dark reversion in the perception of photoperiod (3) and light direction (15), the identity of the natural Cph1 chromophore is of great interest.

To test whether Cph1 and Rcp1 represent functional transmitter and receiver molecules, we purified affinity-tagged versions of Cph1 and Rcp1 fusion protein (16) and analyzed the PCB adduct of Cph1 for protein kinase activity (Fig. 3A). Surprisingly, the Pr form of Cph1 exhibited adenosine triphosphate (ATP)–dependent autophosphorylation activity, whereas phosphorylation of the Pfr form was greatly reduced. Consistent with a histidine residue as the phosphorylation site, Cph1 autophosphorylation was base stable and acid labile. Similar experiments with the N514 mutant demonstrated that the transmitter domain was required for Cph1 autophosphorylation (Fig. 3A). That Rcp1 is a functional receiver substrate for Cph1 was established by phosphotransfer from Cph1 to Rcp1 (Fig. 3B). No phosphotransfer occurred with the Asp68 to Ala (D68A) mutant of Rcp1 (16), which lacks the conserved phosphate-accepting aspartate residue of receiver domains (9). The inability of the His538 to Lys (H538K) mutant of Cph1 to autophosphorylate and to support phosphotransfer to Rcp1 demonstrated that the conserved histidine at amino acid residue 538 in the Cph1 transmitter module is required for both activities (13). These data, taken together, demonstrate that Cph1 is a histidine kinase that mediates light-dependent phosphotransfer to Rcp1. Cph1 and Rcp1 thus represent a two-component regulatory system in cyanobacteria that is modulated by red and far-red light.

Figure 3

Cph1 and Rcp1 represent a red, far-red light-regulated transmitter-receiver pair. (A) Autophosphorylation of purified PCB adducts of Cph1 and N514 mutant. Chemical stability of phospho-Cph1 on blots was assessed by treatment under neutral (50 mM tris HCl, pH 7.5), basic (3 M KOH), or acidic (1 M HCl) conditions at 25°C for 2.5 hours. (B) Phosphotransferase activities of purified Cph1 toward WT and D68A mutant of Rcp1. Normalized relative phosphorylation levels are indicated at the bottom of the figure. (C) Pfr* lacks Rcp1 phosphotransferase activity. Before addition of Rcp1, Pr* was prepared by Pr autophosphorylation for 30 min and either photoconverted to Pfr to produce Pfr* or kept as Pr*. Kinase assays were performed (30); proteins were resolved on SDS–10% polyacrylamide gels and transblotted to PVDF membranes for Coomassie blue staining (upper) or autoradiography (lower) (11). Molecular mass markers at 119, 83, and 47 kD (top to bottom) are indicated with dots.

The small amounts of Cph1 autophosphorylation and Rcp1 phosphotransferase activity exhibited by the Pfr sample (Fig. 3B) probably represent the presence of residual Pr and are consistent with the photoequilibrium mixture, containing 13% Pr, that results for higher plant phytochromes irradiated with saturating red light (2). To determine whether phosphorylated Pfr (Pfr*) was capable of phosphate transfer to Rcp1, we autophosphorylated Pr (Pr*), photoconverted it to Pfr*, and then incubated it with Rcp1 (Fig. 3C). By comparison with a control sample maintained in the Pr* form, phosphotransfer from Pfr* to Rcp1 was clearly prevented. Thus, the Pfr form of Cph1 lacks both autophosphorylation and Rcp1 phosphotransfer activities.

Organization of the cyanobacterial phytochrome operon is similar to the F. diplosiphon rcaEF operon, which encodes two elements of the complementary chromatic adaptation signal transduction pathway (6, 17). This and our biochemical data suggest that the molecular mechanism of Cph1 action involves light-regulated protein phosphorylation-dephosphorylation as depicted in Fig.4. In this model, Cph1 can exist as four species—Pr, Pr*, Pfr, and Pfr*—whose abundances are regulated both by light conditions and by Rcp1 phosphorylation status. By analogy to the multistep phospho-relay cascades proposed for complementary chromatic adaptation in F. diplosiphon(17), sporulation in Bacillus subtilis(18), and osmosensing in yeast (19), Rcp1 dephosphorylation could be mediated by phosphotransfer to another regulatory molecule. Alternatively, the two forms of the small receiver molecule—Rcp1 and phospho-Rcp1 (Rcp1*)—could have distinct regulatory activities like CheY (9).

Figure 4

Model for cyanobacterial phytochrome action. Cph1 can exist in four species—Pr, Pr*, Pfr, and Pfr*—whose relative levels are regulated by light, Pr autophosphorylation, and Rcp1 phosphotransferase activities. Species abundant in red (R) or far-red (FR) light are highlighted in white circles or black boxes, respectively. Phosphorylated Rcp1 (Rcp1*) is dephosphorylated by hypothetical molecule X by a multistep phospho-relay or protein phosphatase mechanism. See text for details. ADP, adenosine diphosphate; Pi, inorganic phosphate.

In higher plants, Pfr is thought to be the active form of phytochrome (2). Our studies suggest that the light signal transduced by Cph1 involves regulation of Pr abundance rather than that of Pfr. However, Pfr (or Pfr*) could perform an as yet unidentified role in the signal transduction process, such as allosterically regulating the activity of a Rcp1 phosphatase or influencing phosphotransfer to another regulatory molecule. In view of the evidence presented here, the presence of a transmitter-like domain in higher plant phytochromes (5) and the observed protein kinase activity of purified higher plant phytochromes (4,13), we expect that the molecular mechanism of phytochrome function in plants will involve phosphorylation-dephosphorylation of transmitter- and receiver-containing signaling proteins like those prevalent in eubacteria and archaebacteria. It is intriguing that two-component regulatory family members have been identified in plants, including the putative plant hormone receptors for ethylene (20) and cytokinin (21). Given the physiological interplay between light and hormone responses in plants (22), we speculate that these receptors may be targets for integrated transduction of multiple signals.

  • * To whom correspondence should be addressed. E-mail: jclagarias{at}


View Abstract

Navigate This Article