Bacterial Photoreceptor with Similarity to Photoactive Yellow Protein and Plant Phytochromes

See allHide authors and affiliations

Science  16 Jul 1999:
Vol. 285, Issue 5426, pp. 406-409
DOI: 10.1126/science.285.5426.406


A phytochrome-like protein called Ppr was discovered in the purple photosynthetic bacterium Rhodospirillum centenum. Ppr has a photoactive yellow protein (PYP) amino-terminal domain, a central domain with similarity to phytochrome, and a carboxyl-terminal histidine kinase domain. Reconstitution experiments demonstrate that Ppr covalently attaches the blue light–absorbing chromophorep-hydroxycinnamic acid and that it has a photocycle that is spectrally similar to, but kinetically slower than, that of PYP. Ppr also regulates chalcone synthase gene expression in response to blue light with autophosphorylation inhibited in vitro by blue light. Phylogenetic analysis demonstrates that R. centenum Ppr may be ancestral to cyanobacterial and plant phytochromes.

All photosynthetic organisms respond in some manner to light quality and quantity. Multicellular plants control development, floral induction, and phototropism through photoreceptors that absorb specific wavelengths of light. Algae, cyanobacteria, and anoxygenic photosynthetic bacteria control motility and gene expression in response to light.

Until recently, phytochrome was thought to be a plant- and algal-specific red and far-red light photoreceptor (1). However, the cyanobacteria Synechocystis and Fremyella diplosiphon have proteins with similarity to plant phytochromes (2). Plant and cyanobacterial phytochromes contain similar NH2-terminal chromophore (bilin) binding domains as well as one or two COOH-terminal kinase domains (Fig. 1). Cyanobacterial phytochromes exhibit similarity to histidine sensor kinases (2). Plant phytochromes contain limited sequence similarity to histidine kinases, lacking some critical sequence motifs such as the highly conserved histidine residue of autophosphorylation (3). Nevertheless, plant phytochromes do undergo autophosphorylation, suggesting that they function as a sensor kinase in a signal transduction cascade (4).

Figure 1

Domains conserved among various phytochrome-like sequences.

Plant blue-light receptors cryptochrome CRY1, CRY2 (5), and NPH1 (6) contain flavin as chromophores. Cryptochromes exhibit similarity to photolyases, whereas NPH1 has similarity to serine kinases. Prokaryotic homologs of cryptochrome and NPH1 have not been described; however, many photosynthetic prokaryotes do contain a blue light–absorbing photoreceptor termed photoactive yellow protein (PYP) (7). PYP has the chromophorep-hydroxycinnamic acid attached by a covalent thioester linkage to a conserved cysteine (8). Blue-light excitation of PYP results in a reversible dark-light photocycle (9). High-resolution structural analyses of PYP in its ground state, as well as several photocycle intermediates, have been determined (10).

Despite detailed structural and biophysical knowledge of PYP, little is known of its biological role (11). To define a function for PYP, we cloned and disrupted a PYP homolog from the purple photosynthetic bacterium Rhodospirillum centenum, an organism that exhibits phototactic motility in response to blue light (12). Amplification of an internal domain of PYP by polymerase chain reaction (PCR) was accomplished with degenerate primers (13). Sequence analysis indicated that the amplified PYP domain was part of a much larger open reading frame encoding a ∼96-kD polypeptide. A similarity (BLAST) search with the deduced polypeptide indicated the presence of a 135–amino acid PYP domain at the NH2-terminus followed by a ∼500–amino acid central domain with similarity to the chromophore (bilin) attachment domain of phytochromes (Fig. 1). The central domain is followed by a ∼250–amino acid COOH-terminal domain with features typical of prokaryotic histidine kinases (14). Because of these structural features, we have named this gene ppr for PYP-phytochrome-related.

Alignment of the PYP domain of Ppr with PYPs from other bacteria showed high similarity, including the cysteine used for attachment of the chromophore p-hydroxycinnamic acid (Fig. 2A). Alignment of the middle domain with the bilin attachment domains from phytochromes shows that the conserved cysteine used for covalent attachment of bilin is absent (Fig. 2B). Indeed, the lack of the conserved bilin attachment cysteine in Ppr is congruent with the observation that anoxygenic photosynthetic bacteria do not synthesize linear tetrapyrroles. Thus, Ppr appears to have substituted the blue light–absorbing chromophorep-hydroxycinnamic acid in the PYP domain for red and far-red light–absorbing bilin that is attached to phytochromes.

Figure 2

DNA sequence alignments of R. centenum Ppr with selected PYP (A) and the bilin domains from phytochromes (B). Abbreviations in (A) are Eha_PYP,Ectothiorhodospira halophila PYP; Rsp_PYP, Rhodobacter sphaeroides PYP; Rca_PYP, Rhodobacter capsulatus PYP. Abbreviations in (B) are PhyC, phytochrome C fromArabidopsis; PhyE, phytochrome E fromArabidopsis; Cph1, phytochrome Cph1 from SynechocystisPCC6803. Conserved Cys residues used for chromophore attachment are indicated with a plus sign in the PYP domain and an asterisk in the phytochrome chromophore domain. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

We purified Ppr apoprotein and covalently attachedp-hydroxycinnamic acid using a procedure developed for PYP from Ectothiorhodospira halophila (E-PYP) (15). Reconstituted Ppr had a yellow color with an absorbance maximum at 434 nm (Fig. 3A). When irradiated with white light, steady-state photobleaching occurred followed by dark recovery with a time constant of 46 s (Fig. 3A). This recovery is 330-fold slower than that observed withE-PYP (16). The light minus dark difference spectrum indicated two spectral species, a red-shifted form at ∼470 nm and a blue-shifted form at ∼330 nm that have isosbestic points at 465 and 365 nm, respectively (Fig. 3B). These species are comparable to the I1 and I2 intermediates observed with E-PYP (16). A microsecond photolysis experiment by laser flash excitation at 445 nm resulted in a rapid decrease of the 434-nm absorbance followed by a fast partial recovery with a time constant of 0.21 ms (Fig. 3C). This was followed by a slow recovery component. The photocycle kinetics obtained with Ppr are different from those observed with E-PYP in which I1 is formed with a time constant of 3 ns, I2with a time constant of 200 μs, and a return to the dark-adapted state in 140 ms (16).

Figure 3

Absorption characteristics of Ppr. (A) Absorption spectra of a dark-adapted Ppr (solid line) and the photobleached product (dashed line) obtained after 2 min of irradiation with a 40-W fiber optic lamp. (Inset) The kinetic trace observed during dark recovery of the 434-nm peak with 10-s data points. The line through the data points represents a single exponential fit with k = 2.17 × 10−2 s−1(time constant = 46 s). (B) Light minus dark difference spectrum shows two spectral bands at 470 and 330 nm with isobestic points at 465 and 365 nm. (C) Microsecond laser flash photolysis transient observed at 434 nm after 445-nm excitation. Irradiation of Ppr resulted in a rapid loss of absorption with a biphasic recovery. The smooth line through the data is a single exponential fit to the partial fast recovery phase with k = 4.7 × 103 s−1(time constant = 1.5 × 10−4 s).

In vitro kinase assays of both apoprotein and chromophore-reconstituted Ppr were performed under dark and light (400 to 900 nm) conditions to see if excitation affected kinase activity (17). Analysis of the apoprotein indicated that dark and illuminated kinase preparations had similar high rates of autophosphorylation (Fig. 4A). In contrast, the photochemically active reconstituted Ppr demonstrated two- to threefold higher phosphorylation under dark conditions than when illuminated as assayed after a 20-min incubation (Fig. 4, A and B). The inhibitory effect of light was reproducible and evident in four independent experiments (18). Because the apoprotein has kinase activity, the presence of significant amounts of apoprotein in these preparations (∼40%) (19) tends to mask the inhibitory effect of light on the reconstituted protein.

Figure 4

Kinetics of autophosphorylation of Ppr when incubated with [γ -32P]ATP (17). (A) Autoradiography of phosphorylated holoenzyme (lanes 1 and 2) and apoprotein (lanes 3 and 4) after a 20-min incubation in the dark (lanes 1 and 3) or in blue light (lanes 2 and 4). (B) Ppr autophosphorylation in reactions that were incubated in the dark (•) or that were illuminated with >400-nm light (○).

We also assayed for a role of Ppr in phototaxis by the construction of a ppr deletion through allelic exchange with a spectinomycin cassette (20). The resulting strain exhibited no defects in photosensory behavior when compared with the wild-type parent strain (18). This indicates that a Ppr photosensory pathway does not interact with the chemotaxis sensory transduction cascade that governs photosensory behavior in this organism (21).

We next tested whether Ppr affects gene expression by assaying expression of the chalcone synthase gene (chs). Chalcone synthase is an early enzyme in the flavonoid biosynthetic pathway, the products of which serve protective roles in plants (22). The function of chalcone synthase in bacteria is unknown but it is present in several species, including R. centenum (23). In plants, chs expression is regulated by both blue and red light through the phytochrome and cryptochrome signal transduction pathways (22). Consequently, we tested whetherchs expression is affected by Ppr. For these assays, we constructed a transcriptional fusion of the R. centenum chs promoter region to a lacZ reporter gene (24) and then assayed β-galactosidase (β-Gal) activity after growth under different illumination conditions (25). Maximal expression of chs was observed when wild-type cells were illuminated with only infrared light (Fig. 5). When these cells were illuminated with infrared light and white light, or infrared light and blue light (400 to 450 nm), chs expression was reduced twofold. Expression ofchs is clearly dependent on Ppr as evidenced by low-level unregulated expression in the ppr-disrupted strain, C145. The finding that the ppr-deleted strain has much lower expression than the blue light–irradiated wild-type cells suggests that Ppr may be providing a certain basal level of phosphorylation to a signal transduction cascade under illuminated conditions that is ramped up to a higher level in the dark.

Figure 5

Light-regulated chalcone synthase expression of a chs::lacZ fusion in wild-type andppr-disrupted cells (strain C145). Cells were illuminated with infrared light (IR), IR + white light, or IR + blue light (400 to 450 nm) and measured for β-Gal activity (25). Activity units represent micromoles ofo-nitrophenyl-β-d-galactopyranoside hydrolyzed min−1 mg−1 protein.

Given that R. centenum and plants both regulate chalcone synthase with related photoreceptors, it raises the intriguing question of whether Ppr is ancestral to phytochromes. For analysis we constructed a phylogenetic tree which indicated that bacterial phytochromes appear to be ancestral to plant phytochromes, withR. centenum phytochrome being the most distant (Web Fig. 6) (18, 26). The phytochrome tree is similar to rooted trees depicting comparison of photosynthesis genes, which indicate that the photosystem synthesized by purple bacteria is ancestral to cyanobacterial and plant photosystem (27). Thus, plant phytochromes may be derived from an ancient bacterial signal transduction system that has been retained among many diverse photosynthetic organisms.

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


View Abstract

Navigate This Article