Index References

Light plays an important role in the photoperiodic control of flowering in many plant species (Imaizumi and Kay, 2006). Expression of the Arabidopsis thaliana gene CONSTANS (CO) is regulated by the circadian clock and subsequently induces FLOWERING LOCUS T (FT) expression when exposed to light under long day conditions. This molecular regulation of FT expression in the long day plant (LDP) Arabidopsis is at the core of the day-length measurement mechanism (reviewed in Imaizumi and Kay, 2006; Kobayashi and Weigel, 2007). FT protein produced in leaves was recently shown to act as the long distance fluorogenic signal and thought to be conserved in many plant species as a key element in the control of flower induction (Tsuji et al., 2008; Corbesier et al., 2007; Jaeger and Wigge, 2007; Lin et al., 2007; Mathieu et al., 2007; Tamaki et al., 2007).

In rice, a short day plant (SDP), the rice homologs of Arabidopsis CO and FT, Heading date 1 (Hd1) and Heading date 3a (Hd3a) also play important roles in the regulation of flowering (Yano et al., 2000; Kojima et al., 2002). Hd1 activates Hd3a expression under inductive short day (SD) conditions, whereas Hd1 suppresses Hd3a expression under non-inductive long day (LD) conditions. This finding suggests that the reversal of Hd3a regulation by Hd1 under LD conditions is the molecular basis of the difference between light-regulated flowering in rice, an SDP, and in Arabidopsis, an LDP (Izawa et al., 2002; Hayama et al., 2003; Hayama and Coupland, 2004). In addition to the conserved component of photoperiodic control of flowering between Arabidopsis and rice, the original component regulating flowering in rice has been identified so far. Early heading date1 (Ehd1), a gene encoding a B-type response regulator and that might not have an ortholog in the Arabidopsis genome, regulates SD promotion of flowering independently of Hd1 (Doi et al., 2004). Recently, analysis of natural variation in rice cultivars identified Ghd7, which encodes a protein with CCT domain and delays flowering under LD condition by downregulating Ehd1 and Hd3a (Xue et al., 2008). Furthermore, OsMADS51, a type I MADS box transcription factor, and RID1/Ehd2/Osld1, an ortholog of the maize INDETERMINATE1 (ID1), were identified, and both were demonstrated to upregulate Ehd1, hence promoting flowering (Wu et al., 2008; Matsubara et al., 2008, Park et al., 2008, Kim et al., 2007). Other member of FT-like gene, RICE FLOWERING LOCUS T 1 (RFT1/FT-L3) was also studied and shown to have a role in the floral activation under SD conditions in concert with Hd3a (Komiya et al., 2008).

In many SDPs, night break (NB), a short exposure to light in the middle of night, was extensively used to understand the role of the circadian clock and light on the regulation of flowering (Hamner and Bonner, 1938; Thomas and Vince-Prue, 1997). Identification of action spectra required for the NB effect in many SDPs showed that red light is the most effective (Thomas and Vince-Prue, 1997), and phytochrome has become established as an important photoreceptor for the NB response ever since the red and far-red light photoreversible effect on flowering was discovered in the early 1950’s (Borthwick et al., 1952). We have previously reported that NB causes a delay in rice flowering by suppressing mRNA expression of Hd3a. Moreover, plants with a phyB mutation lost the ability to suppress Hd3a with a NB, indicating that phyB is required for NB (Ishikawa et al., 2005). In Pharbitis, an SDP, NB experiment was also conducted. NB suppressed PnFT1 and PnFT2, orthlogs of Arabidopsis FT and rice Hd3a, and this correlates with a strong reduction in flowering (Hayama et al., 2007). These findings that NB reduce expression of FT homologs in rice and Pharbitis, the two SDPs tested so far, suggest that a light-sensitive mechanism regulates transcription of these genes in both species.

In this work we characterized the molecular events connected with NB in more detail to understand the relationship between light and phytochrome-mediated control of Hd3a expression. Monochromatic blue, red and far-red light sources were prepared and their relative spectral irradiance was measured (Fig. 1). The effect of NB with red light was shown to be the most effective (Thomas and Vince-Prue, 1997). One of the most basic steps in studies of the physiology of phytochrome action is an analysis of the quantitative relationships between the effects of light and phytochrome-mediated responses. Therefore, we analyzed the quantitative relationship between light intensities and Hd3a expression. NB effects with different intensities of red light were tested, and fluence–response curves of Hd3a expression and flowering time were plotted for the wild type and each of three phytochrome mutants, phyA-4, phyB-1 and phyC-1 (Fig. 2). Plants were grown in climate chambers at 70% humidity under SD conditions with daily cycles of 10 h of light at 30°C and 14 h of dark at 25°C. Transcript levels of Hd3a were measured by real time PCR according to a previous report (Ishikawa et al., 2005).

View Details

Fig. 1 Spectra of monochromatic light. The spectra of experimental light sources [cool white, blue (B), red (R) and far-red (FR)] are shown as relative spectral irradiance in the wavelength range of 380–780 nm. R was obtained from FL-20S.Re-66 (Toshiba Ltd., Tokyo) filtered through 3-mm red acrylic, Acrylight K5-102 (Mitsubishi Rayon Ltd., Tokyo) and FR was obtained from FL-20S.FR-74 (Toshiba Ltd., Tokyo) filtered through 3-mm far-red acrylic, Deraglass A-900 (Asahi Kasei Ltd., Tokyo). For B, FL-20S.B (Toshiba Ltd., Tokyo) was filtered through 3-mm blue acrylic, Acrylight K5-302 (Mitsubishi Rayon Ltd., Tokyo). Light spectra and fluence rates of monochromatic lights were measured using a CS-1000 (Konica Minolta Sensing, Osaka, Japan). The wavelengths of transmission peaks for filtered B, R, and FR light were 436, 659, and 763 nm.

Suppression of Hd3a by red light in wild type was dependent on the fluence dosage and the delay of flowering correlated with the level of Hd3a expression (Fig. 2A and 2E). A slightly elevated response was detected in phyA-4 mutant (Fig. 2B and 2F). In the phyB-1 mutant, Hd3a expression was not completely suppressed at all intensities tested (Fig. 2C). Consequently, flowering time was not affected (Fig. 2G). This result supports a previous finding that the phyB-1 mutant is insensitive to NB (Ishikawa et al., 2005). We found that the phyC-1 mutant was not sensitive to red light as measured by Hd3a expression at the lower intensities tested but downregulation of Hd3a expression depends on photon fluence (Fig. 2D). The flowering time of phyC mutant was almost similar to that of wild type (Fig. 2H). Previous finding suggests that phyC protein is destabilized in each of the phyB mutants, but phyB protein is not affected in phyC mutant (Takano et al., 2005). Flowering time of the phyC mutant was similar to the phyB mutant under LD conditions, and changes in the flowering times of both were due to the upregulation of Hd3a (Takano et al., 2005; Ishikawa et al., in preparation). Based on these results, the effect of NB on Hd3a expression in the phyB-1 mutant may be mediated by phyC, whose activity, however, is reduced by loss of phyB function, whereas that in the phyC-1 mutant is mediated by phyB. Although the slight difference was detected in the regulation of Hd3a expression under NB conditions in phyC mutant, it still remained to be elucidated whether phyC is partially involved in the regulation of NB effect on Hd3a transcription.

View Details

Fig. 2 The red light NB effect is dependent on fluence dosage. (A) to (D) Suppression of Hd3a by red light is dependent on the dosage of photon fluence. NB with red light with several intensities (25.7, 217.2, 1974.2, 19741.7 μmolm–2) was given as indicated during the middle of a normal cycle dark phase. Hd3a expression levels were measured at the beginning of light (ZT = 0) in wild type (A), phyA-4 (B), phyB-1 (C) and phyC-1 (D). Data are means ± S.D. of 3–4 independent RNA extractions on a semi-log plot. (E) to (H) Delay in flowering due to red light NB is dependent on photon fluence dosage. Wild type (E), phyA-4 (F), phyB-1 (G) and phyC-1 (H) plants were treated with NB for 3 weeks and their flowering times were recorded. Data are means ± S.D. of 8–12 plants. One flowering time data for phyA-4 mutants (F) is missing, so we do not show one data at > 10000 μmolm–2.

Phytochrome was first discovered as a key signal transduction component in the seed germination of lettuce in an experiment demonstrating that the irradiation of red light promotes seed germination but far-red light doesn’t (Thomas and Vince-Prue, 1997). Soon after, phytochromes were shown to participate in the NB reaction, mainly because the effect of NB on flowering can be reversed by subsequent irradiation of far-red light (Borthwick et al., 1952). Photobiological experiments led to the proposal that phytochrome exists in two spectral forms: the inactive Pr form (red light absorbing) phototransforms into the active Pfr form (far-red light absorbing) upon absorption of red light and vice versa. Pr and Pfr reversibility is the consequence of photo-conversion before the response system is activated.

We examined the effects of far-red light on Hd3a expression. NB with a single far-red (18000 μmolm^–2) light flash failed to suppress Hd3a expression (Fig. 3A), and had no effect on the flowering time when treated for three weeks (Fig. 3B). NB with far-red light alone thus has no effect on flowering. In the absence of phyB, far-red light increased Hd3a expression (Fig. 3A), however, flowering time was not accelerated as Hd3a expression may reach to the sufficient level to induce early flowering phenotype in phyB mutant (Fig. 3A). It may be possible that far-red light acting through phyA promotes Hd3a expression. Alternatively, phyB may antagonize phyA in the control of Hd3a expression, and phyB suppresses the phyA-dependent increase in Hd3a expression.

View Details

Fig. 3 The red light NB effect is reversible with subsequent far-red light irradiation. (A) NB with far-red light (FR-NB) alone did not cause downregulation of Hd3a. Far-red light was given in the middle of relative night. Hd3a expression levels were measured in wild type, phyA-4, phyB-1 and phyC-1 mutants at the beginning of the light period (ZT = 0). Data are means ± S.D. of 3–4 independent RNA extractions. (B) No effect on flowering time was detected due to NB with far-red light. Plants were treated with NB for 3 weeks and their flowering times were recorded. Data are means ± S.D. of 7–8 plants. (C) Effect on Hd3a expression by red light NB is reversible with subsequent far-red light irradiation. Wild type and the phyB-1 mutant were treated with red and far-red light as indicated. Light intensities of red and far-red light were 2500 μmolm–2 and 18000 μmolm–2, respectively. Hd3a expression levels were measured at the beginning of the light period (ZT = 0). Data are means ± S.D. of 3–4 independent RNA extractions. (D) Delay in flowering by red light NB is partially suppressed by the subsequent far-red light irradiation. Wild type and each phytochrome mutant were treated with the indicated conditions for 3 weeks. Data are means ± S.D. of 8–12 plants.

We next applied several combinations of red and far-red light as NB in the middle of 14 hours of darkness. The minimum photon fluence of red light (2500 μmolm^–2) that can induce the NB effect was deduced from the fluence–response curves (Fig. 2), and far-red light (9000 μmolm^–2), which is sufficient to reverse Pfr to Pr in the low fluence range, were used to determine the limits of Pfr/Pr reversal. We found that irradiation with far-red light immediately after NB with red light partially restored Hd3a expression in wild type (Fig. 3C). Further experiment demonstrated that Hd3a expression partially recovers depending on the presence or absence of the final far-red light irradiation signal, whereas Hd3a expression in phyB mutants remained relatively constant (Fig. 3C).

The effect of NB on flowering time was also examined. An NB-induced flowering delay with red light alone was suppressed by subsequent far-red light irradiation in both the phyA-4 mutant and wild type, but phyB-1 showed no detectable changes in flowering time (Fig. 3D). In the phyC-1 mutant, red light (2500 μmolm^–2) is apparently insufficient to suppress Hd3a expression completely as observed in wild type (Fig. 2D) and delay of flowering was slightly weaker than that in wild type (Fig. 2H). The hyposensitive response of flowering time by NB approximately at 2000 μmolm^–2 in phyC mutant may explain the no detectable effect of subsequent FR irradiation on flowering time (Fig. 3D).

Blue light is also important for plant photomorphogenesis and development (Lin, 2000). Blue light promotes flowering in Arabidopsis and is also effective as a day-extension stimulus in crucifers (Thomas and Vince-Prue, 1997). We tested whether blue light causes the NB effect in rice. NB with monochromatic blue light suppressed Hd3a expression in the wild type and delayed flowering (Fig. 4A and 4B), whereas in the phyB-1 mutant, Hd3a expression was partially downregulated but not fully suppressed (Fig. 4A). No effect on flowering time was detected (Fig. 4B). This result indicates that monochromatic blue light is capable of inducing the NB effect in rice. Blue light NB inhibits flowering are also presented by the analysis of the spectral dependence of the NB effect on the photoperiodic control of flowering in Lemna paucicostata 441 (Saji et al., 1982). The absorption spectra of Pr and Pfr have the second absorbing peak around 380 nm, suggesting that NB with blue light may induce the Pfr-form of phyB. Blue light induced inhibition of rice coleoptiles can be reversed by far-red light, demonstrating that phytochrome is also a photoreceptor in this response pathway (Pjon and Furuya, 1967). It would be interesting to examine the involvement of phytochrome in the blue light effect of NB. However, we can not exclude the possibility that cryptochromes are blue light photoreceptors in the NB response pathway in rice.

View Details

Fig. 4 Blue light NB (B-NB) suppresses Hd3a expression and delays flowering. (A) Blue light suppresses Hd3a expression. Hd3a expression levels were measured in wild type and the phyB-1 mutant at the beginning of the light period (ZT = 0). Data are means ± S.D. of 2–4 independent RNA extractions. (B) Blue light NB alone delays flowering. Wild type and the phyB-1 mutant were treated with blue light NB for 3 weeks. Data are means ± S.D. of 7–8 plants.

Taken together, our results show that the NB effect on Hd3a expression in rice is under the control of a low-fluence response mediated by phyB, because the NB effect on Hd3a occurred in the low-fluence range of 1 to 1000 (μmolm^–2) of red light (Fig. 2) and it is far-red light reversible (Fig. 3C and 3D). Photoperiodic control of flowering operates at the transcription level of Hd3a through two changeable factors; light intensity and Hd1 protein, whose abundance seems to be diurnally controlled by the circadian clock. Under NB conditions, light signal transduction is the primary determinant of Hd3a transcription because circadian-regulated Hd1 expression is not affected (Ishikawa et al., 2005). Therefore, there is a measurable quantitative relationship between light intensity and Hd3a transcription. The role of light in photoperiodic flowering is two-fold. Firstly light acts as a signal to set the phase of the circadian rhythm which underlies photoperiodic timekeeping. Secondly, light acts in the Pfr-requiring reaction (Lumsden and Furuya, 1986; Thomas and Vince-Prue, 1997).

In Arabidopsis, CO protein stability appears to be directly regulated by light signals through phyA, phyB, cry1 and cry2 photoreceptors. phyA and cry signals protect CO protein from degradation whereas phyB signals promote degradation (Valverde et al., 2004). CO stability is controlled by SPA1 by direct interaction (Laubinger et al., 2006). More recently, CO was shown to interact with COP1, an RING motif containing ubiquitin ligase (Liu et al., 2008; Jang et al., 2008). Thus, our current understanding is that circadian controlled CO expression and light transduction meet at the level of CO stability. It would thus be of great interest to know whether Hd1 protein is degraded by NB treatment. An analysis of Hd1 protein stability would further provide the insight into how light signal mediated by phyB affects Hd3a transcription.

We thank Mr. Norio Ishikawa (Konica Minolta Sensing, Osaka, Japan) for the measurement of spectra irradiance, Hisayo Shimizu (Hitachi Central Research Laboratory) for excellent technical assistance and Drs. Hiroyuki Tsuji and Hann Ling Wong for critical reading of this manuscript. This research was supported by Grants-in-Aid for Scientific Research on Priority Areas (Grant 10182102 to K.S.) of the Ministry of Education, Culture, Sports, Science, and Technology of Japan and the Rice Genome Programs (Grant IP-1006 to T.S. and M.T.). R.I. was supported by fellowships from the Japanese Society for the Promotion of Science.