Acclimation of photosynthesis to high irradiance in rice: gene expression and interactions with leaf development

Abstract

Rice (Oryza sativa L.) has been used to study the long-term responses of photosynthesis to high irradiance focusing on the composition of the photosynthetic apparatus and leaf morphology. Typical sun/shade differences in chloroplast composition are seen in the fifth leaf following growth in high irradiance compared with low irradiance (1000 and 200 μmol m⁻² s⁻¹, respectively): higher light-saturated rates of photosynthesis (P_max), higher amounts of Rubisco protein, and a lower chlorophyll a:b ratio. In addition, leaves were thicker under high light compared with low light. However, responses appear more complex when leaf developmental stage is considered. Using a system of transferring plants from low to high light in the laboratory responses that occur before and after full leaf extension have been studied. Acclimation of photosynthesis is limited by leaf age: the transfer to high light, post-leaf extension, is characterized by alterations in chlorophyll a:b but not in Rubisco protein, which may be limited by leaf morphology. Microarray analysis of gene expression was carried out on plants that were transferred to high light post-leaf extension. A down-regulation of light-harvesting genes was seen. No change in the expression level of Rubisco genes was observed. Up-regulation of genes involved in photoprotection was observed. It was also shown that high-light leaf morphology is established prior to formation of the zone of cellular elongation and division. The endogenous and environmental factors which establish the characteristics of high light acclimation may be important for attaining high rates of assimilation in leaves and crop canopies, and the fifth leaf in rice provides a convenient model system for the determination of the mechanisms involved.

Acclimation, gene expression, high light, leaf development, photosynthesis, rice

Introduction

Light can be an unpredictable resource for plants. Sustained changes in irradiance induce alterations in biochemical composition and the morphology of whole plants and leaves. These have the effect of optimizing photosynthetic efficiency and the acclimation of photosynthesis to irradiance in leaves is well documented (Björkman, 1981; Anderson et al., 1995; Walters et al., 2003; Murchie and Horton, 1997; Bailey et al., 2001). In general, low-light responses occur to enhance the efficiency of photon capture, whilst high light responses occur to maximize light-saturated rates of photosynthesis. Under controlled growth conditions it is common to see acclimation maintaining ambient photosynthetic rates at a point below that of light-saturation.

Leaf thickness and architecture affect the content of photosynthetic components per unit leaf area. Thus, leaves grown in high irradiance often have higher rates of photosynthesis due to a higher content per unit leaf area of most photosynthetic components, including Rubisco and components of electron transport and ATP synthesis. However, changes also occur on the single chloroplast level: for example, the ratio of PSII to PSI has been shown to vary according to irradiance level (Murchie and Horton, 1998; Yamazaki et al., 1999). Plants grown under high-light conditions have fewer peripheral light-harvesting complexes per PSII reaction centre and a lower amount of Rubisco per unit chlorophyll, and cytochrome b/f complex per unit chlorophyll (Anderson et al., 1995; Murchie and Horton, 1998).

Low-light-grown leaves are generally thin compared with those grown under high light, with a wider overall area and require less investment in terms of N and C. High-light-grown leaves are thicker, have a higher capacity for N accumulation, and their construction demands a higher investment (Sims and Pearcy, 1994). In the context of maximizing biomass production, such leaf-level changes are often part of a set of integrated mechanisms which include, for example, whole plant biomass partitioning and night-time respiration (Sims and Pearcy, 1994). Morphological adjustments can actually maintain the rate of photosynthesis per unit dry mass at a constant value, despite a change in the rate of photosynthesis per unit leaf area (Sims and Pearcy, 1994; Evans and Poorter, 2001). However, it is unclear whether a low-light-type leaf would be as productive if subsequently transferred to a high-light environment. If the reduced leaf thickness is offset by an increase in area, then such leaves may have the same rate of photosynthesis per leaf when exposed to saturating light (Sims and Pearcy, 1992, 1994).

In terms of photosynthetic productivity it is therefore of importance to distinguish two fundamental types of acclimation: that which occurs before or whilst cellular morphology is established and that which occurs afterward. Features of low or high light leaf anatomy are established at an early point during leaf expansion (Leech et al., 1980; Mullet, 1988; Sims and Pearcy, 1992; Oguchi et al., 2003). Leaf cell size may also be a determinant of chloroplast number (Pyke and Leech, 1991). Although some expansion of mesophyll cells can be induced by light after full leaf extension (Bunce et al., 1977; Yano and Terashima, 2004), leaves need to be transferred from low to high light prior to, or in the process of, expansion and division in order to undergo large changes in acclimation of leaf anatomy. Parameters may vary in magnitude, rather than possessing a threshold where either a high light leaf or a low-light leaf is formed. (Jurik et al., 1979; Sims and Pearcy, 1992). There is also evidence for species-specific differences in the relative durations of cell division and cellular expansion during leaf formation (Van Volkenburgh, 1999; Stiles and Van Volkenburgh, 2002).

Oguchi et al. (2003) examined the role of leaf anatomy in acclimation to high light in fully expanded leaves of Chenopodium album and concluded that while acclimation of Rubisco and photosynthetic rate was seen, it was indeed limited. The most significant post-expansion acclimation feature was an enlarged chloroplast volume and a larger chloroplast surface area exposed to intercellular spaces. Other light acclimation studies have noted, on transfer of mature leaves from low to high irradiance, an increase in light-saturated photosynthesis (Sims and Pearcy, 1992) and measurable Rubisco activity in leaves in pea (Chow and Anderson, 1987).

The extent to which these features of acclimation limit photosynthesis depend on other specific mechanisms of adaptation. For example, in fast-growing species the partitioning of resources into newly synthesized leaves may be of more benefit than investment in already existing leaves. By contrast, the stress-tolerant epiphyte Guzmania monostachia has long-lived leaves which have a high capacity for reversible acclimation characterized by the loss and gain of entire photosynthetic units according to seasonally dependent variations in light level (Maxwell et al., 1999).

There is little known concerning the interaction between acclimation of photosynthesis to irradiance and processes of leaf ageing following leaf extension. It is likely to be greatly dependent on the species under study. For example, in rice leaves, a decline in Rubisco content and photosynthesis can be observed shortly following full leaf extension (Makino et al., 1985). On the other hand it is clear that certain features of chloroplast acclimation can be retained even during advanced leaf senescence (Humbeck and Krupinska, 2003). The above has mostly considered dicotyledons. In this paper, the interaction of leaf development with irradiance using a monocotyledonous plant, rice, is examined. Rice has recently been shown to show acclimation according to irradiance level (Murchie et al., 2002). Grasses such as rice have certain developmental features that differ from Arabidopsis such as development of leaves within a leaf sheath. How these features affect the dependency of acclimation on leaf development is investigated. The nature of acclimation of photosynthesis post-leaf-extension is also examined by measuring changes in gene expression following a transfer of rice plants from low light to high light.

Materials and methods

Growth of plant material

All growth took place in a custom-built growth room (Stiells, Glasgow) maintained at 28 °C, 50–60% RH, and a 12/12 h dark/light cycle. Light was provided by a fluorescent PL-L tubes (Phillips, 55 W,/840/4P, colour 84), supplemented by low intensity tungsten filament bulbs (GE, 60 W, S15). Irradiance was varied by positioning the plants at different distances from the light sources. Air circulation and heat shields constructed from transparent, neutral density material ensured that conditions were the same regardless of position. Leaf temperature was found to be the same for both positions. Rice plants (Indica variety IR72 supplied by the International Rice Research Institute, IRRI) were grown hydroponically in containers of 8.0 l capacity using the following medium, adapted from Makino et al. (1985): 1.4 mM NH₄NO₃, 0.6 mM NaH₂PO₄.2H₂O, 0.5 mM K₂SO₄, 0.8 mM MgSO₄, 0.2 mM CaCl₂, 0.009 mM MnCl₂.4H₂O, 0.001 mM (NH₄)₆Mo₇O₂₄.4H₂O, 0.037 mM H₃BO₃, 0.003 mM CuSO₄.5H₂O, 0.000138 mM NH₄VO₃, 0.00075 mM ZnSO₄.7H₂O, 0.07mM Fe-EDTA, 10 ml per 60 l potassium silicate (cat. 2965463 stock purchased from BDH, Leicester UK). pH was adjusted to 5.5 with 3 M HCl. The solution was replaced every week and replenished daily with water as necessary. Holes 1.5 cm in diameter were drilled into a light-proof support which rested above the nutrient solution and plants were inserted through these and held in place with sponge. Seedlings were planted out 10 d after germination.

Leaf 5 was used for all measurements. This was the fifth leaf to emerge on the primary tiller, ignoring the seed leaf. During leaf extension, following emergence from the leaf sheath, leaf length was measured daily in the morning on every plant. When no overnight change in length could be observed, the day previous to this measurement was taken as the point of full leaf extension, or day 0. All time-points were calculated relative to this. The differences in growth rate between individual plants were generally small. However, plants were transferred individually when they had reached the desired time-point which meant that transfer according to developmental stage was extremely accurate. Leaf temperature was measured immediately after transfer to monitor any adverse stress effects and no significant change was noted. Except for the microarray experiment, all experiments were repeated two to three times independently.

Measurements of photosynthesis and assays of photosynthetic components

Photosynthetic gas exchange was measured using a Li-Cor (Lincoln Nebraska) 6400 portable photosynthesis system with a fluorometer attachment (6400-02) which provided irradiance by means of an array of red and blue LEDs. Measurements were made in the growth room, using ambient humidity (50–60% RH). Sample chamber CO₂ concentration was maintained at 360 μl l⁻¹and light of 2000 μmol m⁻² s⁻¹. Flow rate was 500 μmol s⁻¹. Chamber air temperature was maintained at 28 °C. Photosynthesis and sampling took place on a point of the leaf that was a distance from the tip equivalent to one-quarter of the length of the entire leaf. Given the upright but slightly curved posture of the rice leaves, this point was well exposed to prevailing light conditions. Following insertion of the leaves into the chamber, photosynthetic rate was monitored until a steady-state was reached; this was a minimum of 3 min and a maximum of 5 min. No diurnal alteration in P_max was noted in these plants: measurements were taken between 1 h and 5 h after the lights in the chamber were switched on.

Amounts of Rubisco were measured by SDS-polyacrylamide gel electrophoresis followed by densitometric analysis of Coomassie-stained gels according to Murchie et al. (2002). Chlorophyll content was analysed by extraction and analysis in 80% acetone according to Murchie and Horton (1997).

Microscopy

Leaf segments that were approximately 1 mm wide and from a freshly excised leaf were cut in water with a brand-new razor blade and fixed in 3% glutaraldehyde in 0.1 M phosphate buffer for a minimum of 24 h. These were washed in two changes of 0.1 M phosphate buffer for 1 h and then dehydrated through graded alcohol solutions (70%, 90%, 100%) for a minimum of 1 h per solution. Tissue was then infilitrated in JB-4® solution A plus catalyst (as per the manufacturer's instructions) overnight at 4 °C (fridge) and then embedded in fresh JB-4® solution A with catalyst (100 ml) and 0.8 ml solution B. Polymerization was overnight at 4 °C. Unpolymerized resin was removed from the block by rinsing briefly in 70% alcohol and air-drying. Sections were cut using an LKB Historange microtome and a glass knife. Sections of 4 μm thickness were collected over water and stained in 0.05% Toluidine Blue in acetate buffer pH 4.4 for 2 min and washed in distilled water for 2 min, dried on a hotplate, and mounted in DPX. Mounted sections were used for measurements of leaf thickness and analysis of leaf structure. Hand-cut sections of fresh leaves gave similar results (data not shown).

Microarray analysis

Whole plants were transferred from low light to high light at two time-points during the leaf development process: firstly at the point of full leaf extension (day 0); secondly at exactly 7 d after full leaf extension. Transferred plants were sampled at the new irradiance level at two points: 24 h and 72 h after transfer. At each point, samples were also taken from plants maintained in low light. In this way, comparisons could be made from the equivalent time points at 24 h and 72 h after transfer. This took into account any developmental changes in the leaves which would have occurred in the leaf within this time.

Plants were both transferred and sampled at the mid-point of the diurnal cycle ±20 min. Leaves were approximately 300 mm in length: the section of leaf removed for RNA extraction was 30–150 mm from the leaf tip, i.e. the portion of the leaf most exposed to the prevailing light. Leaf 5 from the primary tiller was used throughout. Leaves were cut from the plant, two leaf discs punched out of the leaf for confirmation of protein and Rubisco content, and the leaves immediately frozen in liquid N₂. The entire process took less than 30 s. Leaves from five individual plants were used per sample.

Leaves were kept frozen at −80 °C for less than 1 month prior to extraction. Leaf material was ground in liquid nitrogen using a mortar and pestle and added to RNAwiz solution (Ambion, Texas) at an approximate ratio of 1 ml:100 mg tissue. Chloroform was added (0.2× starting volume), shaken vigorously and incubated at room temperature for 20 min. The upper aqueous phase was removed and a second chloroform extraction carried out. RNase-free water and isopropanol were added (0.5× and 1× starting volume, respectively) and mixed. This was incubated at room temperature for 10 min and centrifuged at 13 000 g for 15 min at 4 °C. The pellet was washed by vortexing with 1× starting volume of 75% ethanol and then centrifugation at 13 000 g for 5 min at 4 °C. The pellet was dried and resuspended in Rnase-free water. Leaves were extracted individually and the appropriate RNA samples were pooled at the end. For shipping, RNA was dissolved in a storage solution consisting of 70% ethanol and 0.08 M sodium acetate (pH 5.2) at a concentration of 1 mg ml⁻¹.

For microarray analysis, custom-made 21K Affymetrix (Santa Clara CA) Rice genechip™ arrays were used. These annotated arrays contained 23 212 rice genes, 32 Affymetrix control probe sets, and 15 positive control probe sets. More information on these arrays is available at http://www.tmri.org/en/partnership/array_information.aspx. Quality control, hybridization, and imaging were carried out exactly as described in Zhu et al. (2003). Following staining and scanning, the average intensity of all probe sets of each array was scaled to 100 so that the hybridization intensity of all arrays was equivalent. Analysis of the data was performed in Sheffield using Genespring (Silicon Genetics, CA), Vizard (UC Berkeley), and Microsoft Excel. For this dataset, a simple approach was taken: differences in gene expression between samples was made by filtering out any genes which showed a less than 2-fold change and only included genes which had at least one gene flagged as ‘present’. Comparisons were made between plants at 24 h and 72 h after transfer with those maintained at low light and sampled at the same timepoints after full leaf extension. This gave three lists of genes, one which showed-light responsiveness after 24 h one which showed light-responsiveness after 72 h and a third in which a response was seen at both 24 h and 72 h, after transfer. A double replication, combined with the fact that five leaves were pooled per sample provided sufficient confidence that these changes were not due to chance variation.

Results and discussion

Acclimation to light before and after full leaf extension

Figure 1 shows P_max and Rubisco content of the fifth leaf in rice plants grown under low light (200 μmol m⁻² s⁻¹) or high light (1000 μmol m⁻² s⁻¹). Higher values for both parameters were seen in high-light plants compared with low-light plants until the sixth day after full leaf extension. After this, a decline in Rubisco amount and P_max was observed. P_max fell to less than half of its initial value by day 8. Rubisco is the most abundant protein in these leaves, and represents a major store of leaf N. Such changes in Rubisco content represent a well-documented redistribution of amino acids to rapidly growing sinks of rice plants (Mae and Ohira, 1981; Mae, 1997) and the dynamics of N movement within rice plants are thought to play a major role in determining the leaf Rubisco content at any given time. This is important in assessing the role of irradiance acclimation in leaf photosynthesis.

Fig. 1.

(A) Light-saturated rates of photosynthesis, P_max, measured at 360 μl l⁻¹ CO₂ and 2000 μmol m⁻² s⁻¹ and (B) leaf Rubisco content for the first 12 d following full extension of leaf 5. Plants were entirely grown under 200 or 1000 μmol m⁻² s⁻¹. Values represent means ±standard error of means, n=6.

Leaf sections shown in Fig. 2 show differences in leaf morphology between low-light and high-light-grown leaves. Two features are noticeable: high-light leaves are thicker than low-light leaves and generally have a larger cell size than low-light-grown leaves. This is consistent with the increased Rubisco content per unit leaf area in high-light leaves. There was not a noticeable difference in cell number, however (number of cells measured from abaxial to adaxial surface), either measured at the position of bulliform cells or between bulliform cells and vascular bundles. These features contrast with those from dicotyledonous species which can show both periclinal division of palisade mesophyll cells and also elongation of mesophyll cells perpendicular to the plane of the leaf (Weston et al., 2000; Yano and Terashima, 2004). These anatomical features of high-light and low-light-grown rice leaves are similar to, although more pronounced than those noted by Makino et al. (1997), who used older plants and observed smaller differences in Rubisco content and P_max.

Fig. 2.

Cross-sections of rice leaves grown under low light and high light, stained with toluidine blue, showing the differences in leaf thickness. Bar = 0.1 mm.

The increase in leaf thickness in high light was accompanied by a small reduction in area per leaf blade of approximately 20% (data not shown). Depending on the distribution of Rubisco along the length of the leaf, it suggests this would not be sufficient to account for the 2-fold increase in Rubisco protein content seen in Fig. 1. In other words, the decline in Rubisco content in the low-light leaves was not due to a ‘dilution’ effect whereby the production of thinner leaves could be offset by an increase in leaf area.

Rice plants grown under low light were transferred to high light at the point of full leaf extension. Figure 3 shows that this was not sufficient to induce an increase in the content of Rubisco protein. Therefore the high-light condition is established prior to this point. There are in fact few examples in the literature that specify whether alterations in Rubisco content occur as a result of an increase in irradiance imposed after full leaf expansion. Recently, Oguchi et al. (2003) observed a small increase in leaf Rubisco content in C. album following transfer to high light in fully expanded leaves. In the case of rice Suzuki et al. (2001) showed that following full leaf emergence it is the rate of degradation of Rubisco that determines levels within the leaf, rather than the rate of synthesis, although light-dependency of the relative rates of degradation and synthesis were not measured. An important point to note is that the decline in Rubisco and P_max following full leaf emergence of the plants used in the current study was more rapid than that seen in older plants in which leaves have longer life spans (Zhang et al., 2003). Therefore, the turnover of N in leaf 5 may be particularly rapid and could have restricted any alteration in relative rates of protein synthesis and degradation.

Fig. 3.

Transfer of whole plants from low light to high light at the point of full leaf extension of leaf 5 does not result in an increase in the amount of Rubisco protein (A), but does result in an increase in Chl a:b ratio (B). Data for plants maintained in low light or high light throughout as in Fig. 1. Values represent means ±standard error of means, n=6.

Two reports have examined the changes in Rubisco content in high-light-grown rice plants transferred to low light. Hidema et al. (1991) observed a small delay in the age-related decline of Rubisco content, whereas Murchie et al. (2002) who imposed shade upon field-grown plants observed a decline in Rubisco content in shade-grown plants over a period of around 18 d compared with plants maintained at high light. It was previously uncertain what the response of low-light-grown leaves would be when introduced to a high-light environment. These data suggest that Rubisco content per unit leaf area is light-regulated, but is strongly linked to morphology and N turnover.

Figure 3 shows that chloroplast-level acclimation can occur independently of leaf age. Chl a:b is a reliable indicator of the proportion of Chl which is associated with light-harvesting complexes compared with other Chl-containing complexes.

Figure 4 shows the responses of light-saturated photosynthetic rate (P_max), stomatal conductance (g), and substomatal CO₂ concentration (C_i) to the same transfer to high light given in Fig. 3. By day 4, the rates of P_max were equivalent to those seen in the high-light-grown plants (Fig. 1) indicating that the content of Rubisco in low-light-grown plants was sufficient to attain high photosynthetic rates. This followed a defined pattern: on day 2 there was a decline in C_i and no change in g or P_max. By day 4 a higher g and P_max was observed. This may involve the regulation of existing proteins, for example, by Rubisco activation state. Rubisco in rice is regulated by the inhibitor CA1P and removal of this molecule is likely to be involved in attaining a higher in vivo assimilation rate (Yamauchi et al., 2001). Interestingly, the decline in P_max of transferred plants followed a similar pattern to that seen for plants maintained at high light (ie it declined between days 4 and 8). This suggests photosynthetic gas exchange was affected by an as-yet unidentified feature of the ageing process. It is concluded that low-light-grown rice leaves accumulate sufficient Rubisco to support high rates of photosynthesis, but at least 2 d of sustained high light are required to attain these higher rates.

Fig. 4.

(A) Transfer of whole plants from low light to high light at the point of full leaf extension of leaf 5 resulted in an increase in light-saturated photosynthetic rate (P_max, measured at 360 μl l⁻¹ CO₂ and 2000 μmol m⁻² s⁻¹) (A). Also shown are stomatal conductance (B) and intercellular CO₂ concentration, C_i (C). Values are means ±standard error of means, n=6.

The role of basal growth in photoacclimation of rice leaves

Leaf thickness and cell size may have restricted acclimation in low-light-grown leaves transferred to high light. Figure 5 investigates this: whole plants were transferred from low light to high light at progressively earlier points from the point of full extension. It is clearly shown that in order to induce a high-light condition it was necessary to transfer plants at a time which corresponded to a preliminary stage of development of leaf 5 (dissection of plants at this point of transfer (Fig. 5, F) revealed that leaf 5 was small, a maximum of 10 mm in length). The level of Rubisco protein in fully expanded leaves closely matched leaf thickness. Even when plants were transferred whilst leaf 5 was still encased by the leaf sheath (Fig. 5, E), no increase in leaf thickness was seen. It is important to note that the part of the leaf blade on which measurements were made would have been fully expanded within the leaf sheath by this point (see below). Importantly, when the thickness of leaf 5 was measured whilst in the leaf sheath just prior to emergence, it had a thickness (at minor vein position) almost equivalent to that of a high-light leaf (0.1 mm: data not shown). In other words, exposure to high light on emergence from the leaf sheath was not sufficient to induce further cell expansion to attain high-light leaf morphology because this had already been attained within the leaf sheath. This may be related to a greater photosynthate supply in high-light conditions.

Fig. 5.

Rubisco content (A) and leaf thickness in leaf 5 of plants transferred from low light to high light at points in time which incrementally precede the point of full leaf extension. Thickness was measured approximately half way between the midrib and the leaf margin. Distance was measured across a minor vein (not adjacent to a major vein) (B) and at a point where bulliform cells occurred (C). Plants were maintained at high light (A) and low light (B) throughout or grown under low light and then transferred to high light at the point of full leaf extension (C), 4 d before the point of full leaf extension (D), 8 d before the point of full leaf extension (just prior to emergence from the leaf sheath (E), and 14 d before full leaf extension (at this point leaf 3 is approaching full leaf extension) (F). The final point (F) occurred prior to the formation of the zone of rapid division and elongation of leaf 5. Bar length=0.1 mm. Values represent means ±standard error of means, n=6.

This observation is consistent with what is known about the relationship between division and expansion of chloroplasts and cells. Grasses, such as rice, grow from the base of the plant. Leaves are formed from an intercalary meristem which is positioned within the leaf sheaths of subtending leaves. Following cell division, cells enlarge creating a zone of cell elongation, typically no more than a few centimetres long above the meristem, and also within the leaf sheath. Cells cease elongation before emergence from the leaf sheath (Murayama, 1995). Therefore the phase of cellular expansion in the rice leaf is encased and protected within the leaf sheath and is not directly exposed to external conditions of humidity and CO₂. The amount of light is also restricted by the leaf sheath. Chloroplasts divide during the phase of cell elongation, but remain small and undeveloped. Immediately on emergence from the leaf sheath and on exposure to light, chloroplasts enlarge significantly, transcriptional activity increases, and carotenoids and chlorophyll accumulate (Ishii, 1995). A significant increase in leaf N content also occurs at this time (Murayama, 1995). Within the leaf sheath a certain amount of development can be expected: in contrast to dicotyledonous plants, chloroplast division and development can occur in the absence of light in barley (Mullet, 1988). In dark-grown barley leaves, accumulation of most chloroplast soluble proteins can be observed, and plastid volume was 70% of that of light-grown leaves (Mullet, 1988).

In the experiments described in this paper, greening was observed in the upper portions of the leaf prior to emergence and it was concluded that light was penetrating the leaf sheath. In older plants with greater numbers of subtending sheaths light may be more restricted. In a field situation the meristematic tissue may even be positioned underground or underwater. The amount of light available to the zone of division and elongation, therefore, can be severely limited and it is questionable whether or not it plays a direct role in light acclimation.

Recent work has demonstrated the role of long-distance signalling of acclimation in dicotyledonous plants. In Arabidopsis the amount of light given to mature leaves has influenced the stomatal index of newly developing leaves (Lake et al., 2001). Thomas and Quick (2004) recently demonstrated that stomatal density could be determined by long-distance irradiance signalling in tobacco. In Chenopodium album it was shown clearly that leaf anatomy (palisade differentiation) was dependent on signals from mature leaves whilst chloroplast ultrastructure was derived from local signals (Yano and Terashima, 2001). Systemic signalling has been observed for other plant stimuli, for example, light-induced stress (Karpinski et al., 1997) and biotic stress (Bowles, 1998). The mechanism of systemic signalling for developmental responses to abiotic stimuli and its interaction with exogenous signalling during expansion is the subject of much current research. It may involve photosynthate levels as well as hormonal signalling. A review for acclimation of this type to CO₂ concentration is provided by Lake et al. (2002). Such observations may have a clear physiological origin: dicotyledonous leaves in the early phases of growth are rolled and the adaxial surface may be somewhat isolated from external stimuli. For this reason, separate mechanisms of acclimation for regulation of stomatal density on adaxial and abaxial surfaces have been proposed (Lake et al., 2002). Given these features of early leaf development and expansion, signals from mature leaves may then provide the most accurate representation of a plant's direct external environment. Development within the leaf sheath may mean that grasses represent an extreme example of the case for dicotyledonous plants. The data shown in Fig. 4 are certainly consistent with such a view. Long-distance signalling of external abiotic stimuli has not been investigated in grasses and monocotyledonous species.

Gene expression following a transfer to high irradiance in rice leaves

Despite the lack of change in Rubisco content, acclimation of photosynthetic rate to high irradiance after full leaf emergence was seen to occur (Fig. 4). This took between 2 d and 4 d to occur which would be consistent with an alteration of leaf protein profile. It is important to ascertain the cellular mechanisms which give rise to this increase in photosynthetic capacity. For example, this could include an increase in the capacity for electron transport and ATP synthesis within the thylakoid membrane: amounts of the cytochrome b/f complex are known to be light-sensitive in mature rice leaves (Hidema et al., 1991). There may be alterations to relieve any end-product limitation of photosynthesis (Winder et al., 1998). One mechanism for the regulation of protein levels is by gene expression, which was examined here using microarray analysis.

Plants were transferred from low light to high light at two points, firstly at the point of full leaf extension (day 0) and secondly at 7 d after full leaf emergence. Sampling was made at 24 h and 72 h following transfer with control samples taken at the same time-points from plants maintained in low light. The intention was to characterize differences in gene expression according to whether acclimation occurs before or after the decline in P_max shown in Fig. 4. However, a lack of replication meant this was not possible. Nevertheless, there was a distinct subset of genes that demonstrated a change in expression in the same direction whether transferred on day 0 or day 7 and this was used as a means of biological replication. These light-responsive genes are shown in Tables 1, 2 and 3. The criteria for this list was strict: the change must be at least 2-fold and occurring in the same direction (up or down) on both transfer days.

Table 1.

Summary of gene expression following a transfer to high light after full leaf extension has been attained

Category	Up					Down
	24 h only	72 h only	24+72 h	24 h only	72 h only	24+72 h
Development and hormone-related (ho)			1	1	3	2
Kinase (kin)			1	1
Metabolism (met)	25	12	7	2	6	9
Membrane protein (mp)	3				1	1
Protease inhibitor (pi)	1		1		2
Protease (pase)		1
Photosynthesis (ps)			1	2	1	5
Stress (sp)	5	7	3	1	3	1
Ribosome (rib)		2
Signal transduction and transcription factor (st)		1		2	3
Other	9	1		2		1
Total unknown, putative etc	25	16	8	12	28	10
Total known+unknown	68	40	22	23	47	29
Grand total	130			99

Shown are categories of genes which showed a 2-fold change in expression on transfer to high light 0 d after full leaf extension and also 7 d after full leaf extension.

Table 2.

Genes down-regulated following a transfer to high light

Reference	Description	Category
24 h only
BAA83570	Similar to wak1	kin
AAF64042	Fatty acid alpha-oxidase	sp
AAB18404	Chlorophyll a/b binding protein	ps
CAA66373	10 kDa PSII protein	ps
AAA80595	Chlorophyll a/b binding protein	ps
P56660	Maize homeobox protein knotted-1 like	st
72 h only
AAC62209	Putative ethylene receptor; ERS2	ho
AF149816	IAA28	ho
AAA18534	Glucose transporter	met
BAB02065	Similar to nitrate-induced NOI protein	met
AAC98060	Putative protein translocase	mp
S51653	Glutathione synthase (fragment)	sp
AAC39465	Catalase 3	sp
AAB31685	Trypsin inhibitor, WTI	pi
BAA88536	Similar to wheat serpin	pi
P14276	Chlorophyll a/b binding protein 3A precursor	ps
24 h and 72 h
CAA72717	FPF1 protein	ho
AAC26203	Dormancy-associated protein	ho
AAD18145	Putative vetispiradiene synthase	met
P00843	ATP synthase C chain (lipid binding protein) (subunit III)	ps
AAB18404	Chlorophyll a/b binding protein	ps
AAA62700	Photosystem I hydrophobic protein	ps
AAC14566	Chlorophyll a/b-binding protein	ps
AAA64416	Chlorophyll a/b-binding apoprotein CP24 precursor	ps
AAF68391	Hypersensitive-induced response protein	sp
BAA99940	Transcription initiation factor IIB (TFIIB)-like	tf

Selected genes which showed a minimum 2-fold down-regulation of expression levels (compared with low-light controls at the same number of days after full leaf extension) at 24 h, 72 h or both 24 h and 72 h after transfer to high light. Plants were transferred at 0 d and 7 d after full leaf extension and genes were only included which showed a change in the same direction in both sets of data. Catogories are as in Table 1.

Table 3.

Genes up-regulated following a transfer to high light

Reference	Description	Category
24 h only
AAC72335	Starch branching enzyme IIa	met
BAA94257	Endo-1,4-β-glucanase	met
BAA09167	Starch debranching enzyme precursor	met
S58123	Thioredoxin	met
AAC39336	Cellulose synthase catalytic subunit	met/other
Q44262	β-carotene hydroxylase	ps
BAB01998	Aldehyde dehydrogenase	sp
JT0902	Chaperonin 60 β	sp
P08823	Rubisco sub-unit binding protein α subunit precursor	sp
AAD38281	Putative low molecular early light-inducible protein	sp
CAA61018	WIR1	sp
72 h only
CAA09867	Amino acid selective channel protein	met
CAA58823	NADH dehydrogenase	met
P07519	Serine carboxypeptidase I precursor	pase
P46526	Cold shock protein CS66	sp
AAB63591	10 kDa chaperonin	sp
AAF27638	Heat shock protein 70	sp
P31673	Class I heat shock protein 17.4 kDa	sp
24 h and 72 h
CAB96661	HY5	ho
CAA45776	Nitrate reductase	met
AAA66057	ADP-glucose pyrophosphorylase small subunit	met
BAA77282	Monodehydroascorbate reductase	sp
AAB17095	Proteinase inhibitor	pi
CAA74320	Chloroplast NAD-MDH	ps
CAA17528	Adrenodoxin-like protein	ps
BAA12337	Precursor of rice 22 kDa protein of photosystem II (psbs)	ps
AAA21819	Wcs66	sp

Selected genes which showed a minimum 2-fold up-regulation of expression levels (compared with low-light controls at the same number of days after full leaf extension) at 24 h, 72 h or both 24 h and 72 h after transfer to high light. Plants were transferred at 0 d and 7 d after full leaf extension and genes were only included which showed a change in the same direction in both sets of data. Categories are as in Table 1.

Table 1 shows that, of these light-responsive genes, 99 were down-regulated and 130 were up-regulated. Some distinct patterns could be seen: when considering the metabolism and photosynthesis-related genes in Table 1, the majority showed reduced levels of expression in response to high light whereas stress-related genes were more highly represented in the subset showing increased levels of expression. The number of genes involved in hormone responses and developmental processes that showed a decrease in transcript levels was greater than that showing an increase. It is unclear from these data of the existence of specific short-term and long-term responses to changes in light levels; however the majority of the responses in Table 1 were specific to 24 h or 72 h.

Down-regulation of genes encoding light-harvesting proteins may be interpreted as being a response to a large change in the range of absorbed versus utilizable irradiance (Table 2). The transfer from 200 to 1000 is likely over this short period to increase the proportion of irradiance that constitutes excess for photosynthetic requirements. In order to avoid over-excitation of chlorophyll protein complexes and photo-oxidation, a regulated degradation of light-harvesting complexes is frequently observed, and this is borne out by the changes in Chl a:b. Also shown is a decline in expression of CP24, a photosystem I gene, and a 10 kDa PSII gene. The latter corresponds to PSBR, which encodes a protein in the oxygen evolving complex of PSII.

Interestingly, no change in levels of Rubisco RNA was demonstrated by these analyses. Suzuki et al. (2001) showed that the amount of leaf Rubisco RNA (LSU and SSU) is determined before full emergence and declines throughout the lifetime of the leaf in parallel with the amount of Rubisco protein. Indeed these data showed, in low-light leaves only, a decline in RNA content between day 0 and day 12 (data not shown). Rossel et al. (2002) did not observe an increase in the level of Rubisco RNA on the transfer of Arabidopsis plants to high irradiance.

An increase in the expression of genes encoding ADP glucose pyrophosphorylase, starch debranching enzyme, and starch branching enzyme may represent an increase in the use of starch as a storage carbohydrate. This was possibly caused by the added photosynthate availability. Rice leaves are typically sucrose formers and synthesize starch under periods of stress or low sink size (Murchie et al., 2002). This may therefore indicate light stress or that the growing regions formed under high light were an insufficient sink for the extra photosynthate when transferred to high light. An up-regulation of the expression of a nitrate reductase gene was observed: this is interesting because no change in total leaf protein was observed, so any extra assimilated nitrogen may have been exported to growing tissue in shoots and roots.

The increase in P_max of plants transferred to high light showed a lag of 2–4 d. Without a photochemical sink for added excitation energy, the light during this period may have been excessive. This suggestion is supported by the observation that dark-adapted (30 min) F_v/F_m values declined from approximately 0.8 in high-light plants to approximately 0.7 within 24 h of transfer and remained at this level for 3 d (data not shown). This is borne out by the data which show an up-regulation of genes involved in photoprotection and photo-oxidative stress. An increase in the level of expression of monodehydroascorbate reductase was seen, which is an enzyme involved in the processes which remove active oxygen formed in the chloroplast via reduction of O₂ by PSI (Noctor and Foyer, 1998). However, the expression of glutathione synthase was found to decline on transfer to high light. No ascorbate peroxidase genes were up-regulated, by contrast with Rossel et al. (2002) and Karpinski et al. (1997). The induction of genes involved in protection against photo-oxidative stress suggests the presence of activated oxygen species and/or hydrogen peroxide, which are known to be involved in cellular signalling processes for cross–tolerance (Karpinski et al., 1999). This provides a mechanism for the induction of heat and cold shock proteins through an alteration in light intensity alone. This is consistent with the conclusions for Rossel et al. (2002) for Arabidopsis.

Importantly, an up-regulation of a β-carotene hydroxylase gene, which encodes a step in xanthophyll cycle carotenoid biosynthesis, was observed. This has been shown to be involved in the stress tolerance of Arabidopsis thaliana: plants which overexpressed this gene had higher leaf contents of xanthophylls and a higher tolerance to high-light and heat stress (Davison et al., 2002). In addition, two genes associated with PSII and light-harvesting, PSBS and ELIP2 had two of the largest rises in expression levels on transfer to high light. These have been shown to have roles in the regulation of excitation energy dissipation and photoprotection, respectively (Li et al., 2002; Lindahl et al., 1997) strongly suggesting that these proteins are important for photosynthesis under high light in rice.

The experimental conditions used in this experiment were designed to change the light intensity available to the plant only. Leaf temperature and humidity remained comparable for all plants. Therefore any responses induced in these leaves are the result of an increased absorption of light by photosynthetic pigments or photoreceptors. The existence of light-stress responses has already been shown. However, there were indications of the induction of expression of a small number of other genes known to encode proteins involved in plant stress responses: heat and cold shock proteins and chaperonins. This does not imply a specific response to heat stress, as there is a certain amount of overlap for gene expression in response to different environmental stresses (Kreps et al., 2002).

Regulation by light of developmental processes during leaf development has been largely focused on the early stages of cell division and expansion. However, light has effects on features such as leaf longevity, ageing, and senescence. Work by Hidema et al. (1991) suggested that rice leaves developed in high light, but transferred to low light, showed a slight delay in the rate of senescence (decline in protein content). Significant changes in transcript levels of senescence related genes could not be found, so it was concluded that this study's high-light treatment did not hasten the progression of a senescent state, although changes in the expression of genes encoding proteases and protease inhibitors were noted. None of the transcription factors were those known to be senescence related (Chen et al., 2002). A number of genes encoding transcription factors and genes known to be involved in signal transduction were induced or repressed by a transfer to high light. The data also suggests the involvement of genes involved in hormonal signalling: IAA9 and an ethylene receptor gene. The repression of HY5 expression may be significant: it is thought to have a role in linking light and hormonal responses via interaction with auxin signalling pathways (Cluis et al., 2004).

Fewer genes involved in photoprotection and oxidative stress were induced by light in this study compared with that of Arabidopsis (Rossel et al., 2002) despite the use of largely similar treatments. This is not unexpected: rice has been bred for high performance under conditions of high light and has a very high rate of photosynthesis, especially in comparison with Arabidopsis. Photosynthesis in the latter will be saturated at much lower light intensities and will have to dissipate a larger fraction of excess excitation energy.

Concluding remarks

The endogenous and environmental factors which establish the characteristics of a high-light leaf may be important for attaining high rates of assimilation in leaves and crop canopies. This is important because the ability of mature leaves to acclimate to irradiance further is generally limited to existing chloroplasts and cells: no increase was shown in the amount of Rubisco or total protein on a transfer to high light following full leaf extension. A significant increase in P_max equivalent to that of high-light-grown plants did occur, however, indicating that Rubisco in low-light-grown plants was accumulated to excess proportions. This, combined with gene expression data, suggested that acclimation of P_max to high-light post-leaf extension mostly involves the modification of an existing protein profile. It is also clearly shown that the expression of genes involved in photoprotective processes can be induced simultaneously with a substantial increase in photosynthetic rate.

In high-light-grown leaves, the amount of Rubisco protein reached a peak just before the point of full leaf emergence and declined thereafter (Suzuki et al., 2001; Fig. 1 this paper). However, before emergence from the leaf sheath, Rubisco content is small (Suzuki et al., 2001). Given that chloroplast division ceases within the leaf sheath substantial chloroplast enlargement must occur post-emergence from the leaf sheath to accommodate the large amount of Rubisco protein (Muruyama, 1995). If this is accompanied by cellular enlargement, then these data suggest that this would have been pre-set within the leaf sheath and not determined by the light levels the leaf is exposed to following emergence from the leaf sheath. Rather, this may rely on signals from exposed plant parts or an as-yet unidentified environmental cue.

Leaf 5 in rice is a good system for further analysis of the signals and mechanisms that determine irradiance acclimation in grasses. Many questions remain unanswered, particularly with regard to the exact timing of events during development and the nature of the signalling mechanisms involved. Future work must analyse co-ordination of cell division, elongation, and events after emergence.

Abbreviations: Chl, chlorophyll; PSII, photosystem II; PSI, photosystem I; Rubisco, ribulose bisphosphate carboxylase/oxygenase.

We are indebted to collaborators at TMRI and Syngenta, NC, Tong Zhu, Sherman Chang and Steven Goff for microarray analysis and advice, and Chris Hill at Sheffield for microscopy of leaf sections.