Elimination of POR expression correlates with red leaf formation in Amaranthus tricolor

Authors


For correspondence (fax: +81 35841 4462; e-mail: iwamoto@biol.s.u-tokyo.ac.jp).

Summary

Amaranthus tricolor L. tricolor cv. Earlysplendor, an ornamental amaranth, generates red leaves instead of green leaves in late summer to early autumn. Red leaf formation was promoted under short-day conditions and delayed by night-break treatments. Red leaves were characterized by lower levels of chlorophyll accumulation rather than higher levels of red pigment (betacyanin) accumulation. However, the metabolic activity toward the production of Mg-protoporphyrin, an intermediate in the biosynthesis pathway for chlorophyll, was detected in red leaves as well as in green leaves.

RNA gel blot analysis was performed to assess the expression of nine genes encoding eight enzymes involved in chlorophyll biosynthesis. Among these enzymes, red-leaf-specific reduction of gene expression was observed only for NADPH-protochlorophyllide oxidoreductase (POR), a key enzyme catalyzing a later step of chlorophyll biosynthesis. In addition, immunoblot analysis showed no accumulation of POR protein(s) in red leaves. These data indicate that the repression of POR gene expression and resultant loss of chlorophyll synthesis activity plays a role in red leaf formation of A. tricolor.

Introduction

Greenness due to chlorophyll characterizes the shoot of photoautotrophic plants. However, shoot organs can display a wide range of colors through accumulation of various pigments instead of chlorophyll (or through no accumulation of any pigment in white organs). With respect to the biosynthesis of these pigments involved in plant coloration, there have been a number of reports. For example, regulatory mechanisms for the synthesis of anthocyanin, one of the major plant pigments, have been well documented at the molecular level (Holton and Cornish, 1995). In contrast, suppression of chlorophyll accumulation, which is often important for the exhibition of bright colors except for green, has not been studied extensively to date. Since chlorophyll accumulation seems closely related to the formation of shoot organs, this default linkage is assumed to be cancelled in chlorophyll-less shoot organs. Conversely, chlorophyll-less shoot organs may provide clues to answer how chlorophyll accumulation is coupled with organogenesis in the shoot.

Typical examples of chlorophyll-less shoot organs are found in floral organs such as sepals, petals and bracts. Sepals and petals are different from foliage leaves in aspects other than color, including shape and internal structure. In these organs, chlorophyll accumulation seems to be controlled in tight linkage with their organ identities. This has been suggested by the fact that most green flower mutants are the result of homeotic conversion of petals to sepals or foliage leaves (Bowman et al., 1989; Wiering et al., 1979). Formation of bracts, which are relatively similar to foliage leaves, are also integrated into the program of reproductive development. It is thus very difficult to study the repression of chlorophyll accumulation in floral organs separately from extensive physiological changes associated with the transition of developmental programs.

During the late phase of development, ornamental amaranths (Amaranthus tricolor L. tricolor) produce red, yellow or multicolored leaves that are the same shape as green leaves. It was reported that non-green leaves of this plant accumulated strikingly less chlorophyll than the green leaves (McCormac et al., 1997). Here we show that the formation of red leaves in a cultivar Earlysplendor of amaranth can be controlled by photoperiod independently of switching from vegetative growth to reproductive growth, and propose this phenomenon as a simple model for studying the mechanisms of coloring realized by uncoupling chlorophyll accumulation and shoot organogenesis. As the first characterization of this model system, the present paper describes the relationship of red leaf formation with metabolic activity and the expression of genes involved in chlorophyll biosynthesis.

Results

A cultivar of ornamental amaranths, Earlysplendor, begins to form red leaves after a period of producing fully green leaves (Figure 1). The first several red leaves are only partially red. Each one consists of distal green and proximal red areas (occasionally separated by a yellow belt). Partially red leaves formed later have the boundaries at positions that are more distal and have greater proportions of red areas. Finally, fully red leaves emerge over the partially red leaves. Thus, in the shoot of a mature plant, nodes with a partially red leaf can be seen between the last (highest) of nodes with a fully green leaf and the first (lowest) of the nodes with a fully red leaf. The color pattern of each leaf appears to be determined and fixed at the very early stage of leaf development, because it remains unchanged after leaf emergence.

Figure 1.

Red leaf formation of Amaranthus tricolor L. tricolor cv. Earlysplendor.

A plant cultured for 60 days under 12-h light/12-h dark condition. Bar = 2 cm.

Influence of daylength upon red leaf formation and growth

Horticulturists manipulate the timing of red leaf formation in amaranth plants by shading them from sunlight or illuminating them in the night. This suggests that red leaf formation in amaranth is a kind of photoperiodic response. Therefore, the first experiment was conducted to evaluate the effects of daylength on red leaf formation. Plants were grown under various photoperiods throughout culture and the number of plants that formed red leaves were scored every day (Figure 2). Under 8-h light/16-h dark conditions, the first plant with a partially red leaf appeared after approximately 30 days of growth, and the proportion of plants with red leaves increased up to 100% within the subsequent 2 weeks (Figure 2a). The timing of red leaf formation was later as the daylength was longer. When plants were grown under 16-h light/8-h dark conditions, red leaves were not formed during 60 days of culture (but all plants formed red leaves before dying). Vegetative growth was accelerated under long-day conditions, as clearly indicated by the number of leaves produced on each plant (Figure 2b).

Figure 2.

Effects of daylength upon red leaf formation and vegetative growth.

Between 46 and 48 plants were cultured under the 16-h light/8-h dark (squares, 16 L), 14-h light/10-h dark (circles, 14 L), 12-h light/12-h dark (diamonds, 12 L) or 8-h light/16-h dark (triangles, 8 L) conditions.

(a) The percentage of plants that formed at least one partially red leaf.

(b) The mean number of leaves per plant.

Standard errors for the data of the leaf number are less than 0.62. 

In the light of our knowledge that most photoperiodic responses depend on the night length rather than the daylength, the effect on red leaf formation of interrupting the night period by a pulse of light (night-break treatment) was examined in the next experiment. To this end, plants were grown under the 16-h light/8-h dark (long-day), 8-h light/16-h dark (short-day) or 7-h light/8-h dark/1-h light/8-h dark (night-break) condition. Results indicated that the night-break treatment delayed red leaf formation by about 10 days compared with red leaf formation under the short-day conditions (Figure 3a). Accordingly, the promotive effect of the short day on red leaf formation was at least partially attributable to the long period of night. As plants grew at almost the same rate under the short-day and night-break conditions (Figure 3b), red leaf formation could be manipulated without affecting vegetative growth by the night-break treatment. Under the 16-h light/8-h dark conditions, inflorescences were induced at the leaf axiles 40–50 days after germination (data not shown). Such floral induction was delayed by shortening the daylength, and never observed during 80 days of culture under the 8-h light/16-h dark and night-break conditions.

Figure 3.

Effects of night-break treatment upon red leaf formation and vegetative growth.

Between 34 and 36 plants were cultured under the 16-h light/8-h dark (squares, LD), 8-h light/16-h dark (triangles, SD) or 7-h light/8-h dark/1-h light/8-h dark (circles, NB) conditions.

(a) The percentage of plants that formed at least one partially red leaf.

(b) The mean number of leaves per plant.

Standard errors for the data of the leaf number are less than 0.19. 

Quantitation of chlorophyll and betacyanin content

A 60-day-old plant cultured under short-day (8-h light/16-h dark) conditions was used for quantitation of chlorophyll and betacyanin. This plant had nine expanded leaves: the 1st and the 2nd fully green, the 3rd through the 5th partially red and the 6th through the 9th fully red (Figure 4a). The 3rd leaf consisted of a fully green blade and a partially red petiole. The 4th and the 5th leaves that had partially red blades were cut into green parts and red parts before processing. As summarized in Figure 4(b,c), red areas contained a relatively large amount of betacyanin and an extremely small amount of chlorophyll irrespective of the leaf position. Similar results were obtained from several independent experiments using plants cultured under different conditions.

Figure 4.

Chlorophyll and betacyanin contents of amaranth leaves.

(a) Leaves excised from a 60-day-old plant cultured under the 8-h light/16-h dark condition. Numbers indicate leaf positions from basipetal to acropetal on the stem.

(b) Chlorophyll content in green and red areas of the excised leaves.

(c) Betacyanin content in green and red areas of the excised leaves.

Bar in (a) = 1 cm.

Accumulation of chlorophyll biosynthesis intermediates in 2,2′-dipyridyl-treated leaves

Segments of a fully green or fully red leaf excised from a plant of the same age grown under the night-break or short-day conditions were incubated with 2,2′-dipyridyl and then assessed for the biosynthesis activity of Mg-protoporphyrins (Mg-protoporphyrin and its monomethyl ester) (Figure 5 and Supplementary material, Figure S-1). In all cases, a fluorescence peak representing Mg-protoporphyrins was observed in the extracts from 2,2′-dipyridy-treated leaf segments (Figure 5, solid line) but not in the extracts from mock-treated leaf segments (Figure 5, broken line). This result indicates accumulation of Mg-protoporphyrins by treatment with 2,2′-dipyridyl. 2,2′-Dipyridyl induced accumulation of greater amounts of Mg-protoporphyrins in red leaf segments than in green leaf segments. This may indicate that red leaves have a relatively higher activity of biosynthesis pathway from ALA to Mg-protoporphyrins.

Figure 5.

Accumulation of chlorophyll biosynthesis intermediates in 2,2′-dipyridyl-treated leaves.

Plants were cultured for 60 days under the 8-h light/16-h dark (short day) or 7-h light/8-h dark/1-h light/8-h dark (night break) condition. Fully green leaves at the 2nd node of the short-day plant and at the 2nd and the 5th leaves of the night-break plant and a fully red leaf at the 5th node of the short-day plant were excised and treated with 2,2′-dipyridyl. Panels show fluorescence emission spectra (excitation at 419 nm) of acetone/NH4OH extracts of 2,2′-dipyridyl-treated (solid line) and mock-treated (broken line) leaf segments. Under the present experimental conditions, protoporphyrin has a fluorescence peak at 633 nm, and both Mg-protoporphyrin and its monomethyl ester (Mg-protoporphyrins) have a fluorescence peak at 595 nm. Data are typical of the results of four independent experiments.

Figure S‐1.

  Chlorophyll biosynthesis pathway and action sites of 2,2′-dipyridyl.

Activity of porphyrinogen synthesis in vitro

Activity of porphyrin synthesis was also assessed in vitro with extracts prepared from fully green and fully red leaves. During incubation of the extracts with porphobilinogen, significant amounts of uroporphyrinogen and coproporphyrinogen were produced (Table 1). The activity of producing these porphyrinogens was higher in the extracts from leaves at the upper position. When green and red leaves at the 5th position were compared, the activity of porphyrinogen synthesis was found to be much higher in the red leaf extract than in the green leaf extract.

Table 1.  Accumulation of porphyrinogens in cell-free extracts of green and red leaves
Temp.
Culture conditiona
Leaf positionLeaf colorPorphyrinogens formed (pmol mg protein-1 h-1) b
UrogenCoprogenProtogen
  • a Plants were cultured under night-break (7-h light/8-h dark/1 -h light/8-h dark) or short-day (8-h light/16-h dark) condition as described in the legend of Figure 6.

  • b Crude enzyme extracts were added to reaction mixtures containing 0.2 mm porphobilinogen as substrate. After incubation in darkness for 16 h, uroporphyrinogen (Urogen), coproporphyrinogen (Coprogen), and protoporphyrinogen (Protogen) formed in the reaction mixtures were extracted with perchloric acid and methanol, and quantified fluorometrically. Data are typical of the results of three independent experiments.

  • c

    ND: not detected.

Night break5thGreen40.44.3NDc
Short day5thRed89.422.3NDc
Night break2ndGreen16.51.7NDc
Short day2ndGreen17.32.0NDc

Isolation and sequence analysis of genes involved in chlorophyll biosynthesis pathway and in photosynthesis

Thirteen cDNA fragments for amaranth genes involved in photosynthesis or chlorophyll biosynthesis were isolated by the polymerase chain reaction procedure (Supplementary material, Table S-1): Attcab1a (DDBJ/GenBank/EMBL accession number AB050123) and Attcab1b (AB050125) for chlorophyll a/b-binding protein (cab) type I, Attcab3 (AB050126) for cab type III, AttrbcL (AB050127) for large subunit of ribulose 1,5-bisphosphate carboxygenase/oxigenase (RuBisCO), AttGluTR (AB050119) for glutamul-tRNA reductase (GluTR), AttGSA-AT (AB050120) for glutamate-1-semialdehyde aminotransferase (GSA-AT), AttALAD (AB050115) for aminolevulinic acid dehydratase (ALAD), AttPBGD (AB050116) for porphobilinogen deaminase (PBGD), AttUROD (AB050117) for uroporphyrinogen decarboxylase (UROD), AttCOPROX (AB050118) for coproporphyrinogen oxidase (COPROX), AttMg-CH (AB050121) for H subunit of Mg-chelatase (Mg-CH) and AttPOR1 (AB050122) and AttPOR2 (AB050124) for NADPH-protochlorophyllide oxidoreductase (POR). Sequence identities at the nucleotide level between Attcab1a and Attcab1b and between AttPOR1 and AttPOR2 were 80.5% (371/461) and 76.3% (514/674), respectively. Southern-hybridization analyzes with AttPOR1 and AttPOR2 cRNA probes suggested that no additional POR genes are present on the amaranth genome (Supplementary material, Figure S-2).

Figure S‐2.

  Southern-hybridization analysis of attPOR1 and attPOR2.

Expression pattern of genes involved in chlorophyll biosynthesis pathway

To clarify the relationships between the formation of red leaves and the expression levels of chlorophyll biosynthesis genes, RNA gel blot analysis was carried out using cRNA probes prepared from the above-described cDNA fragments (Figure 6). Messenger RNAs of all genes examined were accumulated at detectable levels in deetiolated seedlings. The size of mRNA was 2.0 kb, 2.3 kb, 1.8 kb, 1.7 kb, 1.5 kb, 2.0 kb, 5.0 kb, 2.2 kb and 2.2 kb for AttGluTR, AttGSA-AT, AttALAD, AttPBGD, AttUROD, AttCOPROX, AttMg-CH, AttPOR1 and AttPOR2, respectively. The expression of most of these genes appeared to be up-regulated in association with light-induced deetiolation. Such pattern of gene expression was most obvious for AttMg-CH.

Figure 6.

RNA gel blot analysis of genes involved in chlorophyll biosynthesis pathway.

Total RNA was isolated from seedlings grown in darkness for 7 days followed by 1-day-light exposure, seedlings grown in darkness for 8 days, and green and red leaves of different ages. These leaves were prepared from short-day (SD) plants and night-break (NB) plants as described in the legend of Figure 5. Two micrograms of RNA was electrophoresed in a 1.2% agarose gel and blotted onto a nylon membrane. Hybridization was performed with digoxigenin-labeled cRNA for AttGluTR, AttGSA-AT, AttALAD, AttPBGD, AttUROD, AttCOPRO, AttMg-CH, AttPOR1 and AttPOR2 as probes. The ethidium bromide-stained gel is shown below.

Red leaves (5th SD) as well as green leaves (5th NB, 2nd NB and 2nd SD) accumulated considerable amounts of mRNAs for all genes except for AttPOR1 and AttPOR2. The mRNA levels of AttGluTR, AttGSA-AT and AttPBGD were rather higher in red leaves than in green leaves. In contrast to these genes, the expression of AttPOR genes was suppressed specifically in red leaves.

Expression pattern of genes involved in photosynthesis

Figure 7 shows the expression pattern of nuclear-encoded cab genes (Attcab1a, Attcab1b and Attcab3) and a plastid-encoded rbcL gene (AttrbcL). While AttrbcL mRNA was equally abundant in red leaves as in green leaves, the mRNA levels of Attcab genes in red leaves were extremely low compared with those in the corresponding green leaves.

Figure 7.

RNA gel blot analysis of genes involved in photosynthesis.

Total RNA was isolated from seedlings grown in darkness for 7 days followed by 1-day-light exposure, seedlings grown in darkness for 8 days and green and red leaves of different ages. These leaves were prepared from short-day (SD) plants and night-break (NB) plants as described in the legend of Figure 5. Two micrograms of RNA was electrophoresed in a 1.2% agarose gel and blotted onto a nylon membrane. Hybridization was performed with digoxigenin-labeled cRNA for Attcab1a, Attcab1b, Attcab3 and AttrbcL as probes. The ethidium bromide-stained gel is shown below.

Accumulation of NADPH-protochlorophyllide oxidoreductase (POR) protein(s)

Accumulation patterns of POR protein(s) were examined by immunoblot analysis using an anti-arabidopsis POR A antibody (Kusnetsov et al., 1998). In total proteins prepared from seedlings and various organs of amaranth, one band of POR polypeptide(s) with an apparent molecular mass of about 37 kDa was detected under a stringent condition (Figure 8). This band may represent a mixture of products of AttPOR1 and AttPOR2 genes, which can not be separated by standard SDS-PAGE because their molecular masses are too close. In amaranth seedlings, POR protein(s) decreased markedly after 24-h light irradiation. The accumulation pattern of POR protein(s) in green and red leaves was quite similar to the expression pattern of AttPOR1 and AttPOR2 genes: green leaves accumulated POR protein(s) at a substantial level while red leaves accumulated only a trace amount of POR protein(s).

Figure 8.

Immunoblot analysis of POR.

(a) Total protein was extracted from 7-day-old etiolated seedlings of arabidopsis, seedlings grown in darkness for 7 days followed by 1-day-light exposure, seedlings grown in darkness for 8 days, green and red leaves of different ages (prepared as described in the legend of Figure 5) of amaranth. Twenty micrograms of total protein was loaded in each lane of a polyacrylamyde gel, electrophoresed and blotted onto a nitrocellulose membrane. Proteins immunoreactive to anti-arabidopsis POR A antibody on the blot were detected by chemiluminescence. The CBB-stained gel is shown below.

(b) Two micrograms of total protein from etiolated amaranth seedlings and 20 µg of total protein from green and red leaves of different ages were subjected to immunoblot analysis as described above. Time of exposure to chemiluminescence was prolonged to enhance immunoreactive signals. Molecular weights are indicated on the right of each panel.

Discussion

Daylength responses of amaranth

Cultures of amaranth plants under various photoperiods demonstrated the promotive effect of short-day conditions on red leaf formation (Figure 2). In addition, red leaf formation was delayed significantly by the night-break treatment, which did not affect the rate of vegetative growth (Figure 3). These results indicate that the red leaf formation of amaranth is a photoperiodically sensitive reaction preferring long night length. The promotion of red leaf formation by short-day conditions was rather quantitative, and critical night length could not be detected. Thus, the red leaf formation is not absolutely but facultatively dependent on photoperiods.

Under the photoperiods of 7-h light/8-h dark/1-h light/8-h dark and 16-h light/8-h dark, which gave the same continuous night length, amaranth plants exhibited different responses with respect to red leaf formation (Figure 3). This difference might be attributed to the incompleteness in cancellation by the 1-h light irradiation of the influence by the preceding dark period. Such a decrease in sensitivity to night-break treatments is often observed with photoperiodic control of flowering in long-day plants (Bernier, 1988).

Kohli and Sawhney (1979) examined the flowering responses of three species of Amaranthaceae (A. caudatus f. albiflorus, A. caudatus f. caudatus and A. tricolor var. tristis) to various photoperiods, and concluded that all were absolute short-day plants. However, amaranth plants employed in the present study (A. tricolor L. tricolor cv. Earlysplendor) formed floral buds earlier under the longer daylength (data not shown). As night-break treatment had no effect on the timing of the floral bud formation, flowering in this plant seems not to be photoperiodically controlled. Long-day conditions might promote flowering in this case indirectly through the acceleration of vegetative growth. Accordingly, this cultivar may be regarded as a day-neutral plant of which flowering had been released from photoperiodic control. As mentioned above, effects of daylength on red leaf formation and on flowering were quite different. These properties enable us to analyze the red leaf formation independently of growth phase transition from vegetative to reproductive states.

Biosynthesis activities toward chlorophyll production in red leaves

Fully red leaves and red areas of partially red leaves of amaranth were characterized by little accumulation of chlorophyll and higher content of betacyanin (Figure 4). The difference in chlorophyll content between green and red leaves was qualitative, while the difference in betacyanin content was quantitative. It follows from this that most critical for the formation of red leaves in amaranth is the deficiency of chlorophyll, the green color of which masks the red color of betacyanin in green leaves.

Since red leaves of amaranth appeared to emerge congenitally as chlorophyll-less leaves, the low level of chlorophyll accumulation in red leaves was expected to result from the lack of activation of chlorophyll synthesis. This view was supported by immunoblot analysis of POR, which plays an essential role in the late stage of chlorophyll biosynthesis pathway. The result demonstrating only negligible accumulation of POR protein(s) in red leaves (Figure 8) suggested strongly that red leaves do indeed have a poor activity of chlorophyll synthesis. However, the pathway toward the formation of Mg-protoporphyrins, intermediates of chlorophyll synthesis, was active in red leaves as well as in green leaves (Table 1, Figure 5).

In general, mutants or transgenic plants that are defective at a specific step in the chlorophyll synthesis pathway accumulate intermediates upstream of that step (Gough, 1972; Jensen et al., 1996; Kruse et al., 1995; Mock and Grimm, 1997). Mutational blockage at the POR reaction brings about protochlorophyllide accumulation, which is reported for xan4 mutants of arabidopsis (Runge et al., 1995). The excess intermediates generate highly reactive oxygens in the light, causing photooxidative damages (Reinbothe et al., 1996). Unlike such cases, red leaves of amaranth deficient in POR suffered neither unusual accumulation of protochlorophyllide nor photooxidative damage. It remains an unsettled question how amaranth plants can avoid protochlorophyllide accumulation despite eliminating POR.

Expression patterns of genes involved in chlorophyll biosynthesis pathway

Expression levels of nine genes for eight enzymes involved in different steps of chlorophyll biosynthesis pathway were examined through RNA gel blot analysis for comparing expression patterns of these genes in green and red leaves of amaranth (Figure 6). The obtained results showed that red leaves as well as green leaves accumulated considerable amounts of mRNAs for all the genes except for AttPORs, of which expression was suppressed specifically in red leaves. This is consistent with the high catalytic activities toward the Mg-protoporphyrin formation and with the loss of POR proteins in red leaves. From these findings, we can reasonably speculate that the lack of expression of POR genes and the resultant loss of the accumulation of POR protein may interrupt chlorophyll production, leading to the formation chlorophyll-less, red leaves in amaranth.

In recent years, many POR genes have been isolated from various plant species. Expression patterns of these POR genes can categorize most of them into two groups referred to as A-type and B-type POR genes (Armstrong et al., 1995; Holtorf et al., 1995). A-type POR mRNAs accumulate only in dark-grown seedlings and rapidly become undetectable upon illumination of them. By contrast, B-type POR mRNAs persist throughout vegetative development irrespective of light conditions. With such difference in the expression patterns as a criterion, both AttPOR1 and AttPOR2, of which mRNAs accumulated in deetiolated seedlings and green leaves of adult plants, can be judged to belong to the B type.

Expression pattern of genes involved in photosynthesis

RNA gel blot analysis also showed that rbcL and cab genes, which are involved in photosynthesis, were differently controlled during red leaf formation (Figure 7). AttrbcL mRNA was accumulated equally both in red leaves and green leaves, whereas expression levels of all three Attcab genes were much lower in red leaves than in green leaves. This expression pattern of rbcL gene is consistent with a report by McCormac et al. (1997), which demonstrated that rbcL mRNA and protein were present at similar levels in green and non-green regions of three-color leaves of amaranth. With respect to chlorophyll-binding proteins of amaranth, there has been only one research article reporting protein amounts of LHCII polypeptides and PsaD (Wang et al., 1999). This report indicated that these chlorophyll-binding proteins did not accumulate in non-green regions of three-color leaves, while other photosynthesis-related proteins such as cytochrome f and plastocyanin accumulated in non-green regions as well as in green regions. Taking these findings together with our results, shows that proteins that bind directly to chlorophyll appear to be eliminated selectively from photosynthesis apparatus in chlorophyll-less leaves of amaranth.

In the present study, we established an experimental system in which red leaf formation in amaranth can be controlled independently of floral bud formation and of plant growth by manipulating photoperiodic conditions. This system may be very useful for studying the mechanisms of plant coloration through uncoupling of chlorophyll accumulation and shoot organogenesis. Taking advantage of this system, we have revealed that the selective shutoff of the expression of POR genes is a key event for red leaf formation.

Color pattern alteration unique to the amaranth leaves, from total green to total red via partial red consisting of an increasing red area at the proximal region and a decreasing green area at the distal region, is also notable. This interesting phenomenon might bring clues for understanding how cell fate with respect to chlorophyll accumulation is determined during leaf organogenesis. From a different angle, the red leaf formation of amaranth can be regarded as a good example of photoperiodic response. To date, most pieces of information about photoperiodism in plants have come from researches of flowering. Red leaf formation of amaranth, a relatively simple phenomenon compared with flowering, will be an advantageous subject in this research field as well.

Experimental procedures

Plant material

Seeds of Amaranthus tricolor L. tricolor (cv. Earlysplendor, purchased from Sakata Seed, Yokohama, Japan) were germinated on vermiculite in plastic pots, which were soaked in culture solution containing 16 mm KNO3, 2.5 mm KH2PO4-K2HPO4 (pH 5.5), 2 mm MgSO4, 2 mm Ca(NO3)2, 70 µm H3BO3, 50 µm Fe-EDTA, 14 µm MnSO4, 10 µm NaCl, 1 µm ZnSO4, 0.5 µm CuSO4, 0.2 µm Na2MoO4 and 10 nm CoCl2. The culture solution was aerated to supply oxygen. Plants were cultured under the following conditions: temperature, 25 ± 1°C; light intensity at the surface of soil, 56 µmol m−2 sec−1; light source, fluorescent lamps Palook (Matsushita Electric Industrial, Osaka, Japan) and Fishlux (Toshiba Lighting & Technology, Tokyo, Japan).

Extraction and quantitation of pigments

Leaves were homogenized in 100% acetone and centrifuged at 20 000 g for 5 min. After the pellet was re-extracted with 100% acetone, the resultant supernatants were combined, diluted with 2.5 mm sodium phosphate buffer (pH 7.8) to adjust the concentration of acetone to 80%, and then used for spectrophotometric determination of chlorophyll content. The remaining precipitation after chlorophyll extraction was dried and resuspended in 3.33 mm acetic acid. After centrifugation, the pellet was re-extracted with acetic acid. The supernatants were combined and used for spectrophotometric determination of betacyanin content. The concentrations of chlorophyll and betacyanin were calculated from absorbance using formulas by Porra et al. (1989) and by Elliott (1979), respectively.

Extraction and quantitation of chlorophyll biosynthesis intermediates from 2,2′-dipyridyl-treated leaves

Chlorophyll biosynthesis intermediates were extracted from 2,2′-dipyridyl-treated leaves and quantified spectrofluorometrically according to the procedure of Masuda et al. (1996). Leaves were excised and cut into pieces approximately 5 mm square. The leaf segments were placed on a filter paper moistened with 1 mm 2,2′-dipyridyl or pure water in Petri dishes, and incubated at 25°C for 16 h in darkness with gentle shaking. After homogenization of the leaf segments in a mixture of acetone and 0.1 mm NH4OH (9 : 1, v/v), the homogenates were centrifuged at 20 000 g for 30 min at 0°C. Chlorophyll was removed from the aqueous acetone solution by three extractions with hexane. The hexane-washed acetone layer was used for measuring the fluorescence emission spectrum at excitation wavelength of 419 nm.

Quantitation of the rate of porphyrin synthesis

The rate of porphyrin synthesis was measured by the method of Masuda et al. (1996). Leaves were homogenized in a buffer [20 mm Tris–HCl (pH 8.0), 0.5 mm MgCl2 and 1 mmβ-mercaptoethanol], and centrifuged at 10 000 g for 15 min. The resultant pellet was re-extracted with the same buffer supplemented with 1% (v/v) Triton X-100, and centrifuged at 10 000 g for 15 min. The extract was added to a reaction mixture to give final concentrations of 0.5 µg protein ml-1, 20 mm Tris–HCl (pH 8.0), 0.5 mm MgCl2, 1 mmβ-mercaptomethanol and 0.2 mm PBG. The reaction mixture was incubated at 25°C in darkness for 16 h. Reaction was stopped by adding 3 volumes of a 2 : 1 (v/v) mixture of 2 m perchloric acid and methanol. Thereafter reaction tubes were stood at room temperature for 3 h. During this period, porphyrinogens in the reaction mixture were converted to the corresponding porphyrins. After centrifugation, the concentrations of coproporphyrin, uroporphyrin and protoporphyrin in the supernatant were quantitated spectrofluorometrically by the method of Grandchamp et al. (1980).

Isolation and analysis of genes involved in chlorophyll biosynthesis and photosynthesis

With reference to the published amino acid sequences, forward and reverse primers specific to conserved regions of glutamyl-tRNA reductase (GluTR), glutamate-1-semialdehyde aminotransferase (GSA-AT), aminolevulinic acid dehydratase (ALAD), porphobilinogen deaminase (PBGD), uroporphyrinogen decarb oxylase (UROD), coproporphyrinogen oxidase (COPROX), H subunit of Mg-chelatase (Mg-CH), NADPH-protochlorophyllide oxidoreductase (POR), rbcL and cab genes were designed to amplify 0.7–1.8 kbp cDNA fragments by polymerase chain reaction (Supplementary material, Table S-1). The first-strand of cDNA of each gene except for rbcL was reverse-transcribed from poly (A)+ RNA of deetiolated seedlings using the corresponding reverse primer. Polymerase chain reaction was performed with the cDNAs as templates by repeating the following reaction 40 times: denatuation at 94°C for 1 min, annealing at 55°C for Glu-TR and at 50°C otherwise for 2 min, and polymerization at 72°C for 2 min. A DNA fragment for rbcL was amplified by PCR from total amaranth DNA with rbcL-specific primers under the same condition as described above. Amplified cDNAs with expected length were cloned into AspEI-digested BSII TSK vector (Ichihara and Kurosawa, 1993) or pGEM-T Easy vector (Promega, Madison, WI, USA).

RNA gel blot analysis

Digoxigenin-labeled cRNA probes were prepared from cDNA clones using DIG RNA Labeling Kit (Roche Diagnostics, Postfach, Switzerland). Two micrograms of total RNA derived from leaves or seedlings was electrophoresed in 1.2% (w/v) agarose/formamide gel, and transferred to a positively charged nylon membrane (Roche Diagnostics). The blot was irradiated with UV-light, and subjected to hybridization with the cRNA probes under the highly stringent condition. Hybridization signals were visualized by the immunological method with anti-digoxigenin-AP Fab fragments (Roche Diagnostics) according to the manufacturer's protocol.

Southern-hybridization analysis

Digoxigenin-labeled AttPOR1 and AttPOR2 cRNA probes were prepared from cDNA clones using DIG RNA Labeling Kit (Roche Diagnostics). Five micrograms of total DNA derived from mature green leaves was digested with restriction enzymes, electrophoresed in 0.8% (w/v) agarose gel, and transferred to a positively charged nylon membrane (Roche Diagnostics). After the blot was baked at 80°C for 15 min, hybridization with the cRNA probes and washing were done at 55°C (high-stringency condition) or at 30°C (low-stringency condition). Hybridization signals were visualized by the immunological method with anti-digoxigenin-AP Fab fragments (Roche Diagnostics) according to the manufacturer's protocol.

Immunoblot analysis with anti-arabidopsis POR A antibody

Plant materials were frozen in liquid nitrogen, pulverized with a mortar and pestle, and then homogenized in extraction buffer [100 mm Tris–HCl (pH 7.5), 10% SDS and 10 mm MgCl2] supplemented with 1/100 volume of ‘protease inhibitor cocktails for use with plant cell extracts’ (Sigma, St Louis, MO, USA). The homogenates were boiled for 10 min, and centrifuged at 10 000 g for 10 min. The resultant supernatants were loaded on a polyacrylamide gel, electrophoresed and electrically blotted onto a nitrocellulose membrane. The blot was incubated with 1: 25,000-diluted antiserum that had been raised against arabidopsis POR A (Kusnetsov et al., 1998), and subsequently with 1: 5,000-diluted peroxidase-conjugated rabbit anti-goat IgG (DAKO A/S, Glostrup, Denmark). Immunoreacting proteins were detected using ECL system (Amersham Pharmacia Biotech, Piscataway, NJ, USA), according to the manufacturer's instructions.

Acknowledgments

We thank Dr Ralf Oelmüller for providing the anti-arabidopsis POR A antiserum. We are grateful to Dr Eric Lam for careful reading of the manuscript. This work was supported in part by Grants-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan and from the Japan Society for the Promotion of Science.

Supplementary material

One additional table (Table S-1) and two additional figures (Figures S-1 and S-2) can be found in the on line version of this paper at http://www.blackwell-science.com/products/journals/suppmat/TPJ/TPJ1082/TPJ1082sm.htm

Figure S-1. Chlorophyll biosynthesis pathway and action sites of 2,2′-dipyridyl.

The C5 pathway producing aminolevulinic acid (ALA) from glutamic acid is a key regulatory step in the biosynthesis of protoporphyrin, a common precursor of chlorophyll and heme (Tanaka et al., 1996). This pathway is controlled through feedback inhibition by heme in vivo (Hoober and Stegemann, 1973). 2,2′-Dipyridyl, a chelator of Fe2+, reduces the concentration of heme by inhibiting its synthesis, which leads to the release of ALA synthesis from feedback inhibition by heme (Duggan and Gassman, 1974). Since 2,2′-dipyridyl also inhibits isocyclic ring formation, Mg-protoporphyrin and Mg-protoporphyrin monomethyl ester accumulates in 2,2′-dipyridyl-treated leaves in darkness (Kouji et al., 1989).

Figure S-2. Southern-hybridization analysis of attPOR1 and attPOR2.

Five micrograms of total DNA was digested with EcoRV (left in each panel), Kpn1 (middle) or PstI (right), electrophoresed in 0.8% (w/v) agarose gel, and transferred onto a nylon membrane. Hybridization with attPOR1 and attPOR2 probes and washing were performed at 55°C (high-stringency condition) or at 30°C (low-stringency condition).

Table S-1. Isolated genes encoding enzymes involved in chlorophyll biosynthesis and photosynthesis.

EMBL accession numbers: AB050115, AB050116, AB050117, AB050118, AB050119, AB050120, AB050121, AB050122, AB050123, AB050124, AB050125, AB050126 and AB050127

Ancillary