INTRODUCTION

Other than acting as an energy source for photosynthesis, light is an important signal for many growth and developmental processes in plants. Plants thus possess multiple photoreceptors, which enable them to perceive and adapt to light intensity, direction, and quality in their environment (1). Phytochrome, the most well characterized of these light receptors, mediates a diverse array of photomorphogenetic responses in various organisms from green algae to higher plants (2-4). Molecular analyses have established that most plants contain multiple phytochrome genes for which expression appears to be responsible for this diversity of phytochrome-mediated phenomena (3, 5-7). The ability of phytochrome to sense the light environment is due to a covalently attached linear tetrapyrrole chromophore, which allows the photoreceptor to photointerconvert between red light-absorbing Pr¹ form and far red light-absorbing Pfr form. Light-grown plants deficient in phytochrome chromophore biosynthesis exhibit aberrant growth and development (reviewed in Refs. 8 and 9).

It is well established that (3E)-phytochromobilin (PB) is the immediate precursor of the chromophore of phytochrome A from higher plants, and it has implicitly been assumed that other phytochromes also contain a PB chromophore (reviewed in Ref. 10). Two types of observations suggest that some phytochromes could contain an alternative bilin chromophore. First, the ability to reconstitute photoactive phytochromes with phycobilin pigments supports this hypothesis. One of these is the phycobiliprotein chromophore precursor, (3E)-phycocyanobilin (PCB), assembly of which with apophytochrome A yields a photoactive species with blue-shifted absorption maxima for both Pr and Pfr forms (11). Second, phytochromes with blue-shifted absorption spectra have been reported for the fern Adiantum capillus-veneris (12), the mosses Ceratodon purpureus (13) and Atrichum undulatum (14), light-grown higher plants (15, 16), and the green alga Mesotaenium caldariorum (17, 18). With the notable exception of green algal phytochrome, these blue shifts were restricted to the Pfr forms, a result that can readily be ascribed to limited proteolysis or denaturation of phytochrome during its purification (19). The observations that full-length algal phytochrome displays blue-shifted Pr and Pfr absorption maxima (18), and that the action spectra for chloroplast reorientation in Mesotaenium is similarly blue-shifted (20), suggest that this atypical phytochrome spectrum is an inherent property of the native algal photoreceptor.

The present study was undertaken to establish the molecular basis for the blue-shifted absorption maxima of the green algal phytochrome. In one avenue of investigation, we examined algal plastid protein extracts for the presence of enzymatic activities capable of converting biliverdin IX (BV), a known intermediate in the biosynthesis of the phytochrome chromophore in higher plants, to PB and other bilin products. These studies take advantage of an improved HPLC assay system for PB synthase (21), the plastid-localized enzyme responsible for the reductive conversion of BV to PB in higher plants (22, 23). With this assay, which can readily distinguish between (3Z)- and (3E)-isomers of PB and PCB, the production of PCB by a non-phycobiliprotein containing green algal species has been documented for the first time.

EXPERIMENTAL PROCEDURES

M. caldariorum strain 41 was obtained from the University of Texas at Austin Culture Collection of Algae. Liquid suspension cultures were grown at 16-18 °C in Sager and Granick medium II (24) supplemented with 2% yeast extract as recommended (17) under cool white fluorescent continuous light (50-100 µmol m² s¹). Cyanidium caldarium strain CPD was a generous gift from Dr. S. I. Beale (Brown University, Providence, RI) and was grown at 37 °C in pH 2.0 glucose-based heterotrophic medium (25).

Mesotaenium suspension cultures (300 ml/flask) were allowed to grow to late log phase in continuous light. For isolation of intact chloroplasts, algal cultures were placed in darkness for 7 days to consume the starch storage granules. Dark-adapted cells were harvested by centrifugation at 160 × g for 1 min and washed with deionized H₂O. Cell pellets were resuspended in 50 ml of protoplast buffer (0.3 M potassium succinate, pH 5.6, 0.2 M sorbitol, 2 mM CaCl₂) containing 0.5% Onozuka RS cellulase (Karlan) and incubated with shaking (60 rpm) under green safe light for 4-6 h at room temperature until >90% of the cells had become protoplasts. Protoplasts were harvested by centrifugation in a swinging-bucket rotor at 425 × g for 3 min, washed with 50 ml of protoplast buffer, and re-centrifuged. Protoplast pellets were resuspended in 11 ml of cold protoplast lysis buffer (25 mM TES-KOH, pH 7.3, 2.4 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1% polyvinylpyrrolidone 40, 2 mg/ml bovine serum albumin, 1 µg/ml leupeptin) by pipetting and inverting for 2 min, whereupon 1 ml of 4 M sorbitol was immediately added for osmotic stabilization of chloroplasts. Intact chloroplasts were collected by centrifugation in a swinging-bucket rotor at 760 × g for 3 min. Chloroplasts were then osmotically lysed by adding 25 ml of chloroplast lysis buffer (100 mM potassium phosphate, pH 7.3, 1 mM EDTA, 1 mM Na₂EDTA, 1 mM MgCl₂, 1.5 mg/ml sodium ascorbate, 1 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride) and incubated on ice for 10-20 min. The supernatant obtained, after centrifugation at 100,000 × g for 1 h, represented the chloroplast soluble protein fraction. This solution was further concentrated by using an Amicon ultrafiltration cell with a YM 10 membrane. Phosphate buffer was chosen for chloroplast lysis because of its necessity for solubilizing bilin reductase activity in higher plant.²

For a standard 1-ml assay, chloroplast soluble protein extracts (730 µl) containing roughly 1 mg of total protein as determined by Bradford assay (26) were used as enzyme source. Reaction mixtures consisted of a NADPH-regenerating system (6.5 mM glucose 6-phosphate, 0.82 mM NADP⁺, and 1.1 units/ml glucose-6-phosphate dehydrogenase), 10 µM bovine serum albumin, 4.6 µM spinach ferredoxin, and 0.025 units/ml ferredoxin-NADP⁺ reductase (all final concentrations). Assays were initiated by addition of BV (or other bilin) in Me₂SO to give a final bilin concentration of 5-10 µM and Me₂SO concentration of 1% (v/v). Reaction mixtures were incubated at 28 °C under green safe light for the appropriate time indicated and stopped by placing on ice. Crude bilins were isolated with a C18 Sep-Pak Light cartridge (Waters-Millipore Corp., Milford, MA) as described (23). Crude bilins were analyzed by C18 reversed phase HPLC using a Varian 5000 liquid chromatograph equipped with a 4.6 mm × 250-mm Phenomenex Ultracarb 5µ ODS20 and a 4.6 mm × 30-mm guard column of the same material. The mobile phase consisted of a 50:50 mixture of acetone and 20 mM formic acid in water. Column eluates were monitored at 380 nm with Varian UV100 absorbance detector.

Two different recombinant phytochrome apoproteins were used for in vitro assembly experiments. Oat apophytochrome A was prepared from Saccharomyces cerevisiae strain 29A (MAT leu2-3 leu2-112 his3-1 ade1-101 trp1-289) expressing pMphyA3 (27). For Mesotaenium apophytochrome, an XhoI-NotI fragment containing MCphy1b cDNA was subcloned into an engineered pMAC105 vector³ digested with SalI and NotI to generate pMMCphy1b. This plasmid was then used to transform S. cerevisiae strain 29A, and Mesotaenium apophytochrome was expressed as described (27).

For in vitro spectrophotometric assays, ammonium sulfate was added to the yeast soluble protein fraction to produce a concentration of 0.23 g/ml. The mixture was incubated on ice for at least 1 h, and precipitates were collected by centrifugation for 30 min at 17,000 × g at 4 °C. Protein pellets were dissolved in TEGE buffer (25 mM Tris-HCl, pH 8.0, 25% ethylene glycol, 1 mM EDTA) containing 1 mM dithiothreitol and 1 mM phenylmethylsulfonyl fluoride. Holophytochrome concentrations were determined with a HP8450A UV/visible spectrophotometer using the absorbance difference assay described previously (28). For in vitro assembly, (3E)-PCB, (3E)-PB, or bilin pigment as indicated was added to reaction mixture to produce a final concentration of 4 µM. Assay mixtures were incubated at room temperature under green safe light for 30 min prior to spectrophotometric measurements.

BV was prepared from bilirubin and purified as described (29). Crude (3E)-PCB and (3E)-PB were obtained by methanolysis of Spirulina platensis (10) and Porphyridium cruentum (30), respectively. Crude pigments were further purified by C-18 reversed phase HPLC as described above, except that a 10 mm × 250-mm semipreparative column with a 10 mm × 50-mm guard column was used. (3Z)-PCB was obtained by co-incubation of BV with Cyanidium extracts (31) and purified by HPLC as described above. (3Z)-PB was similarly biosynthesized from BV using a partially purified oat etioplast fraction containing PB synthase.⁴ HPLC-purified bilins were diluted with 2-3 volumes of 0.1% trifluoroacetic acid, applied to C-18 Sep-Pak cartridge, and washed with 0.1% trifluoroacetic acid, and bilins were eluted with 60:40 acetonitrile:0.1% trifluoroacetic acid. Eluted pigments were concentrated in vacuo with a Speed Vac concentrator (Savant), wrapped with foil, and stored at 20 °C. For biosynthetic studies, bilins were dissolved in Me₂SO to give a final concentration of 1 mM. Bilins were quantitated in 2% (v/v) HCl in methanol using molar absorption coefficients of 46,773 M¹ cm¹ at 368 nm for (3Z)-PCB (32), 47,900 M¹ cm¹ at 374 nm for (3E)-PCB (33), 66,200 M¹ cm¹ at 377 nm for BV (29), 38,019 M¹ cm¹ at 382 nm for (3Z)-PB (32), and 64,565 M¹ cm¹ at 386 nm for (3E)-PB (32).

RESULTS

Mesotaenium Chloroplasts Synthesize a Phytochrome Chromophore Precursor Other than PB

Previous studies have shown that the biosynthetic pathway for the chromophore of phytochrome resides in the plastid compartment of higher plants (22). To test the hypothesis that the green alga Mesotaenium is capable of synthesizing phytochrome chromophore precursor(s) other than PB, we prepared chloroplast soluble protein extracts to study the metabolism of BV. An NADPH-regenerating system, ferredoxin, and ferredoxin-NADP⁺ reductase were included in our assay mixture, since reduced ferredoxin has been shown to be required for the synthesis of PCB in red algae (34) and (3Z)-PB in higher plants.⁴ After prolonged incubation of chloroplast extracts with BV, the total pigment mixture was partially purified and analyzed for its ability to assemble with recombinant oat apophytochrome A. Fig. 1A shows the difference spectrum of the oat phytochrome A-bilin adduct(s) produced. Comparisons with the spectra of PCB and PB adducts of recombinant oat apophytochrome A show that this spectrum is indistinguishable with that of the PCB adduct (Fig. 1B). This result strongly suggests that the Mesotaenium chloroplast soluble fraction contains enzymatic activities that can metabolize BV either to PCB or a novel bilin, for which assembly with apophytochrome A yields a phytochrome species with a difference spectrum very similar to the PCB adduct.

To identify the pigment(s) produced by incubation of the Mesotaenium chloroplast soluble protein extracts with BV, a more detailed time-course study was conducted (Fig. 2). At time zero, the predominant pigment detected by HPLC was the reaction substrate BV. A new pigment, labeled A, emerged after 5 min of incubation and increased during the initial 15-min period. Pigment A was accompanied by another minor pigment, labeled C, with an earlier retention time. A third pigment, labeled B, appeared after most of the BV had been metabolized. Pigment B became the predominant species along with the production of a fourth pigment, labeled D. At later time points, pigments B and D appeared to be the end products of the BV conversion. The kinetics of BV turnover and the production of pigments A and B are illustrated in Fig. 3. These data strongly suggest that the metabolism of BV proceeds via pigment A, which is subsequently converted to pigment B. That these conversions were enzyme-mediated was established by the inability of heat-denatured chloroplast soluble protein extracts to convert BV to pigments A-D (data not shown). Spectrophotometric analysis of the pigments eluting prior to 10 min (Fig. 2) revealed these to be rubinoid pigments, possibly thiol adducts, which were not further characterized.

PCB and PB Are Both Phytochrome Chromophore Biosynthetic Intermediates in Mesotaenium

Several lines of research were undertaken to address the identities of the pigments derived from BV in this green alga. First, pigments A and B were HPLC-purified and tested for their ability to assemble with recombinant oat apophytochrome A. Fig. 4A shows that oat apophytochrome A adducts of pigments A and B exhibited difference spectra typical for PB and PCB adducts, respectively (see Fig. 1A for comparison). Second, the absorption spectra of pigments A and B were determined and proved indistinguishable from those of (3Z)-PB and (3Z)-PCB (Fig. 4B and data not shown). Third, HPLC co-injection experiments with the products of BV metabolism by soluble chloroplast extracts were performed using authentic (3Z)- and (3E)-isomers of PCB and PB (Fig. 5). The retention times of pigments A and B were shown to be identical with those for (3Z)-PB and (3Z)-PCB (Fig. 5, c and d). These experiments also demonstrate that pigments C and D represent the (3E)-isomers PB and PCB, respectively (Fig. 5, e and f). That (3Z)-PB is a bona fide precursor of (3Z)-PCB in Mesotaenium was established by the time-dependent conversion of purified (3Z)-PB to (3Z)-PCB (Fig. 6). Taken together, it is clear that Mesotaenium chloroplasts possess the enzyme activities for conversion of BV to (3Z)-PB and subsequently to (3Z)-PCB.

Kinetic plot of (3Z)-PB metabolism by plastid extracts.

Although Mesotaenium chloroplasts synthesize PCB from BV, it was still unclear whether PCB is the chromophore precursor for algal phytochrome. To address this issue, recombinant algal apophytochrome was therefore expressed in yeast and used for in vitro assembly with either PB or PCB. Fig. 7A indicates that, when assembled with PB, the spectrum was quite similar to the native spectrum for the higher plant phytochrome A (compare with Fig. 1B). When assembled with PCB, the difference spectrum of recombinant Mesotaenium phytochrome was indistinguishable from that of native algal phytochrome (Fig. 7B). Together with the metabolism of BV to PCB by algal plastid extracts, these results demonstrate that PCB is the immediate chromophore precursor for the algal photoreceptor phytochrome.

DISCUSSION

This work was undertaken to determine the nature of the observed blue-shifted difference spectrum of phytochrome from the green alga Mesotaenium. We have demonstrated that this blue shift arises from the use of PCB as the precursor of the algal phytochrome chromophore rather than an altered chromophore environment. To our knowledge, this is the first report of the existence of PCB in an organism lacking phycobiliproteins and the first phytochrome identified to use a chromophore other than PB. Our studies also establish that Mesotaenium possesses a novel biosynthetic pathway for the synthesis of PCB. Instead of the major route employed by the red algae which synthesizes PCB via the intermediates of 15,16-dihydrobiliverdin and phycoerythrobilin (34-36), Mesotaenium converts BV to (3Z)-PB, which is subsequently converted to (3Z)-PCB (Fig. 8). While similar to that proposed for phytochrome chromophore biosynthesis in higher plants, an additional step for the reductive conversion of (3Z)-PB to (3Z)-PCB catalyzed by a hypothetical PB reductase has been included in the green algae phytochrome chromophore biosynthetic pathway. In this scheme, we have also included 3Z to 3E bilin isomerization steps based on the observed production of 3E isomers of PB and PCB. Whether one or both isomerization steps are enzyme-mediated and/or required for assembly with Mesotaenium apophytochrome is unknown at present.

The ability of Mesotaenium to synthesize both PB and PCB raises the possibility that PB could also be used as a precursor of the algal phytochrome chromophore. This might occur should PB escape from the chloroplast compartment or should further reduction of PB to PCB be inhibited and/or regulated under various physiological conditions. Since the observed blue-shifted difference spectrum was observed for Mesotaenium holophytochrome isolated from the dark-adapted cells, and the enzyme used for BV metabolism was also obtained from dark-adapted cells, it is conceivable that PB can be used as a natural phytochrome chromophore precursor for cells grown under different conditions. These possibilities remain to be addressed in future study.

The identification of an enzyme activity responsible for the reduction of PB to PCB raises a number of interesting questions. Does this represent a second bilin reduction enzyme or reflect a dual substrate specificity for PB synthase in green algae? What is the ecological advantage for this green alga to possess an enzyme that yields a blue-shifted phytochrome species? Could introduction of the green algal PB synthase gene into higher plants lead to the synthesis of phytochrome molecules with PCB as chromophore? A 10-16-nm blue shift in both red- and far red-absorbing forms by using PCB as chromophore could provide a powerful tool for studying the impact of light quality in plant physiology and development. Whether higher plants possess phytochrome molecules with different chromophores other than PB remains to be addressed.

The presence of PCB in the non-phycobiliprotein containing green alga Mesotaenium raises the interesting question of whether green algae can synthesize PCB via same pathway as that reported for the red alga Cyanidium. Preliminary experiments in our laboratory failed to detect either 15,16-dihydrobiliverdin or phycoerythrobilin as BV metabolites by plastid extracts from Mesotaenium (data not shown). These results suggest that phycobilin-containing organisms that do not produce phycoerythrobilin might synthesize PCB via the intermediacy of PB. Furthermore, a closer examination of the PCB biosynthetic pathway in Cyanidium has revealed the evidence for the production of (3Z)-PB as the biosynthetic intermediate.² Although it is not known at present, cyanobacteria may be capable of synthesizing PB based on the hypothesis that the cyanobacterial family of eubacteria are the progenitors of the rhodophyte plastids. Together with the presence of PB synthase activity in yeast, Pichia pastoris (21), it is tempting to speculate that PB synthase, the enzyme catalyzing BV reduction to PB, was present in the primitive ancestor for all extant organisms.