Changes in Growth CO₂ Result in Rapid Adjustments of Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase Small Subunit Gene Expression in Expanding and Mature Leaves of Rice

Abstract

The accumulation of soluble carbohydrates resulting from growth under elevated CO₂ may potentially signal the repression of gene activity for the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase (rbcS). To test this hypothesis we grew rice (Oryza sativa L.) under ambient (350 μL L⁻¹) and high (700 μL L⁻¹) CO₂ in outdoor, sunlit, environment-controlled chambers and performed a cross-switching of growth CO₂ concentration at the late-vegetative phase. Within 24 h, plants switched to high CO₂ showed a 15% and 23% decrease in rbcS mRNA, whereas plants switched to ambient CO₂ increased 27% and 11% in expanding and mature leaves, respectively. Ribulose-1,5-bisphosphate carboxylase/oxygenase total activity and protein content 8 d after the switch increased up to 27% and 20%, respectively, in plants switched to ambient CO₂, but changed very little in plants switched to high CO₂. Plants maintained at high CO₂ showed greater carbohydrate pool sizes and lower rbcS transcript levels than plants kept at ambient CO₂. However, after switching growth CO₂ concentration, there was not a simple correlation between carbohydrate and rbcS transcript levels. We conclude that although carbohydrates may be important in the regulation of rbcS expression, changes in total pool size alone could not predict the rapid changes in expression that we observed.

The inevitability of the increase in atmospheric CO₂ concentration makes understanding its potential effects on plant growth and development of great importance. Increased growth and yield responses for many plant species grown under elevated CO₂ are well documented (Kimball, 1983; Cure and Acock, 1986). However, many C₃ species when grown for long periods in elevated CO₂ exhibit decreased photosynthetic capacity (Delucia et al., 1985; Sage et al., 1989; Besford et al., 1990). This acclimation response to elevated CO₂ is often accompanied by an increase in soluble carbohydrate pools and a decrease in Rubisco protein content, activity, and activation state (Bowes, 1993; Drake et al., 1997).

Carbohydrate source-sink balance under growth at elevated CO₂ is believed to play a major role in the regulation of photosynthesis through feedback inhibition (Arp, 1991; Stitt, 1991). Source-sink imbalances may occur during exposure to elevated CO₂ when photosynthetic rate exceeds the export capacity or the capacity of sinks to use the photosynthate for growth, resulting in the accumulation of carbohydrates in photosynthetically active source leaves (Farrar and Williams, 1991; Stitt, 1991). It is hypothesized that accumulation of soluble carbohydrates in the cell may signal the repression of Rubisco small and large subunit gene expression (Sheen, 1994). Regulation of the expression of several genes coding for key photosynthetic enzymes has been shown to be influenced by soluble carbohydrate levels in plant cells (Koch, 1996). Transcription of the small subunit of Rubisco (rbcS) and to a lesser extent the large subunit (rbcL) has been shown to be strongly repressed by Glc and Suc (Sheen, 1990; Van Oosten and Besford, 1994). Repression of rbcS and rbcL occurs in field-grown wheat (Nie et al., 1995) and in growth-chamber-grown Arabidopsis (Cheng et al., 1998) exposed to elevated CO₂. The signal transduction pathway for the regulation of sugar-sensing genes may involve the phosphorylation of hexose sugars via hexokinase (Jang and Sheen, 1997). However, the contributions that carbohydrate pool sizes or fluctuation of pools via metabolism make to the regulation of gene activity are still uncertain.

We reasoned that by nonintrusively modifying the source-sink balance via a cross-switching of growth [CO₂], it should be possible to rapidly reverse the source-sink balance of plants grown at either ambient or elevated [CO₂]. If photosynthetic acclimation to elevated CO₂ results from carbohydrate feedback inhibition, then switching growth [CO₂] from high to ambient should cause a shift from sink-limited to source-limited photosynthesis, and thus feedback-inhibited effects would disappear. Conversely, the opposite should be observed for plants switched from ambient to high CO₂.

Rice (Oryza sativa L.) shows marked acclimation to growth under elevated atmospheric CO₂ (Rowland-Bamford et al., 1991) and thus provides a good model system for studying the processes involved. In the present study rice was grown under field-like conditions in outdoor, sunlit, environment-controlled chambers under either ambient (350 μL L⁻¹) or high (700 μL L⁻¹) atmospheric [CO₂] until the late-vegetative stage, at which time the cross-switching of growth [CO₂] was imposed. It was hypothesized that altering the source-sink relationship would result in a change in the amount of rbcS mRNA, followed by corresponding changes in Rubisco total activity and protein content. Furthermore, if leaf carbohydrate pool size was a direct component of the signal transduction pathway, then changes in rbcS mRNA should correlate with changes in soluble carbohydrate pools.

MATERIALS AND METHODS

Plant Material and Growth Conditions

Rice (Oryza sativa L. cv Lemont) was grown in four sunlit environment-controlled chambers, known as Soil-Plant-Atmosphere-Research chambers, located in Gainesville, Florida. The chambers were constructed of an aluminum frame covered with transparent polyethylene terephthalate (Sixlight, Taiyo Kogyo Co., Tokyo, Japan), allowing approximately 90% of incident PAR to penetrate. Aboveground chamber dimensions were 2.0 × 1.0 m in the cross-section and 1.5 m in height. Chamber tops were attached to aluminum vats of 2.0 × 1.0 m in the cross-section and 0.6 m deep, providing a watertight, flooded rooting environment for rice culture. The chamber vats were filled with soil to a depth of 0.5 m. Pregerminated rice seed was sown in 1.0-m rows that were 0.17 m apart on October 3, 1996. Nine days after planting the rice was thinned to approximately 250 plants m⁻² and a 0.05-m flood was applied and maintained for the duration of the experiment. Before planting, the soil was fertilized with 8.4 and 13.5 g m⁻² phosphorous and potassium, respectively. Nitrogen as urea was applied to the soil at 12.6, 6.3, and 6.3 g m⁻² at 8, 18, and 25 d after planting, respectively.

Two chambers of rice were grown at a daytime atmospheric [CO₂] of 350 mL L⁻¹ (ambient CO₂) and the other two were grown at 700 mL L⁻¹ (high CO₂). At the late-vegetative phase of growth, 34 d after planting, a cross-switching of growth CO₂ concentration was performed. One chamber of rice at ambient CO₂ was switched to high CO_2, whereas another chamber at high CO₂ was switched to ambient CO₂. Dry bulb temperatures were maintained at 28°C/21°C (day/night) and the dew point temperatures were maintained at 18°C/14°C (day/night). The data for microclimate conditions, canopy carbon-exchange rate, and evapotranspiration were collected every 2 s, and 300-s averages were computed and recorded. Each chamber [CO₂] was monitored with a dedicated IR gas analyzer (Ultramat 21P, Siemens, Haguenan, France). Gas analyzers were calibrated and checked for linearity before and after the experiment with a range of standard [CO₂] in nitrogen. Calibration was checked daily with a span gas. Daily printouts of environment-control conditions were monitored to ensure identical performance of the chambers throughout the study. Variability around the desired set points was ±0.25°C for dry bulb temperature, ±0.5°C for dew point temperature, and ±5 μL L⁻¹ for [CO₂]. Details of the methods used for controlling chamber environmental set points (hardware and control algorithms) are given by Pickering et al. (1994).

Photosynthesis Measurements

Photosynthesis was measured at the treatment growth [CO₂] on the attached uppermost fully expanded leaf (mature leaf) and the emerging leaf just above it (expanding leaf) with a portable photosynthesis system (model LI-6200, Li-Cor, Lincoln, NE) equipped with a 0.25-L cuvette. The calibration of the instrument was checked daily with CO₂ standards of 200, 400, 700, and 1000 μL L⁻¹ CO₂ in air. At 32 d after planting, 1 d before the start of the experiment, for each chamber the main culms of approximately 80 to 100 plants at the same developmental stage were tagged. The mature leaf was number 7, and the expanding leaf was number 8 on the main rice culm. Emerging leaves were approximately 25% expanded at the start of the measurements and fully expanded by the time the last measurement was taken. Photosynthesis was measured between 11:30 am and 12:30 pm eastern standard time, when the solar PPFD was saturating at 1200 to 1500 mmol m⁻² s⁻¹. Three mature and three expanding leaves were measured per chamber on the day before the CO₂ switch (d 1), the day of the CO₂ switch (d 2), and for several days thereafter.

Leaf Sampling

The mature and expanding leaves were detached at the ligule from 10 to 20 plants per treatment, at d 1, 3, and 10 of the CO₂-switching experiment. Leaves were sampled between 1 pm and 3 pm eastern standard time. After detachment, the leaves were immediately immersed in liquid nitrogen. Sampled leaves were pooled by treatment and developmental stage (mature or expanding), ground to a fine powder in liquid nitrogen with a mortar and pestle, and stored in liquid nitrogen until analysis for Rubisco activity, Rubisco content, rbcS transcripts, and soluble carbohydrates. Ratios of fresh weight to leaf area were determined for a subset of plants at the same time that samples were taken for biochemical and molecular analysis. Leaf area was measured with a leaf area meter (model LI-3100, Li-Cor).

Rubisco Activity and Content Assays

Rubisco total activity was determined by the method described by Vu et al. (1997). Approximately 150 mg of liquid-nitrogen-frozen leaf powder was ground in a prechilled Ten Broek homogenizer at 4°C containing 3 mL of extraction medium consisting of 50 mm CO₂-free Tricine-NaOH, pH 8.0, 10 mm MgCl₂, 5 mm DTT, 10 mm isoascorbate, 0.1 mm EDTA, and 2% (w/v) PVP-40. The homogenate was microcentrifuged for 45 s at 4°C, an aliquot of the supernatant was immediately assayed for Rubisco activity, and a 0.2-mL aliquot of the supernatant was frozen in liquid nitrogen and later analyzed for Rubisco content. Assay reactions were performed at 30°C in a volume of 0.5 mL. The reaction mixture consisted of 50 mm CO₂-free Tricine-NaOH, pH 8.0, 10 mm MgCl₂, 5 mm DTT, 0.1 mm EDTA, 0.5 mm RuBP, and 10 mm NaH¹⁴CO₃ (2 GBq mmol⁻¹). A 0.1-mL aliquot of the supernatant was incubated with the reaction mixture with RuBP omitted for 5 min and the reaction initiated by the addition of RuBP. Reactions were stopped after 45 s by adding 0.1 mL of 6 n HCl. The assays were dried at 60°C and the acid-stable ¹⁴C radioactivity determined by scintillation spectrometry. Blank reactions consisting of the reaction mixture without the leaf extract were used to subtract background radioactivity from the assays.

Rubisco content was determined by a radioimmunoprecipitation procedure described by Vu et al. (1997). To a 0.2-mL aliquot of the leaf extract obtained from the Rubisco activity assays, NaHCO₃ was added to 10 mm and the mixture allowed to incubate on ice for 20 min to activate Rubisco. A 0.025-mL aliquot of this mixture was added to 0.05 mL of buffer (100 mm Bicine, pH 7.8, 20 mm MgCl₂, and 1 mm EDTA) containing 4 nmol of [2-¹⁴C] carboxyarabinitol bisphosphate and 0.05 mL of antiserum to purified tobacco Rubisco raised from rabbits. After incubation at 37°C for 2 h, the precipitate was collected on 0.45-mm pore-size Millipore cellulose acetate/nitrate filters and washed with 5 mL of a solution containing 145 mm NaCl and 10 mm MgCl₂. The amount of bound ¹⁴C was determined by liquid-scintillation spectrometry. Blanks containing everything except the leaf extract were used to subtract background radioactivity. Assays were performed in triplicate on two to three extractions per treatment.

Rubisco Transcript Analysis

Total RNA was isolated from approximately 100 mg of liquid-nitrogen-frozen leaf tissue using a kit (RNeasy, Qiagen, Chatsworth, CA) following the manufacturer's instructions, but with the addition of two phenol/chloroform extraction steps. Absorbance of individual RNA samples was scanned between 320 and 220 nm and quantified by its A₂₆₀. Total RNA (10 μg) was denatured in 50% (v/v) formamide, 18% (v/v) formaldehyde, and separated by electrophoresis on a 1.5% agarose gel containing 18% (v/v) formaldehyde. RNA was transferred by northern blotting to positively charged nylon membranes (Boehringer Mannheim). Electrophoresis and northern blotting were carried out using the procedures of Sambrook et al. (1989). Immediately after blotting, membranes were baked for 30 min at 120°C to achieve cross-linking. A duplicate set of samples was run on the same gel and stained with ethidium bromide (1 mg mL⁻¹ stock solution). In addition, after hybridizations, the membranes were stained with ethidium bromide to verify integrity and equal loading of RNA.

Detection of mRNA for rbcS was performed by hybridization with a DIG-labeled DNA probe. A 176-bp probe was synthesized using PCR with a cDNA of the rice rbcS gene as the template (Xie and Wu, 1988). DIG labeling was performed using a probe-synthesis kit (PCR DIG, Boehringer Mannheim). Hybridization and detection of the probe were accomplished using the protocols described by the manufacturer. Hybridizations were carried out in 5× SSC (1× SSC = 150 mm NaCl, 15 mm sodium citrate), 50 mm sodium phosphate, pH 7.0, 50% (v/v) formamide, 7.0% (w/v) SDS, 0.1% (w/v) N-lauroylsarcosine, and 2% (w/v) blocking reagent (Boehringer Mannheim) for 16 h at 50°C. After hybridization, membranes were washed twice for 15 min at room temperature in 2× SSC, 0.1% (w/v) SDS, and twice for 15 min at 70°C in 0.1× SSC, 0.1% (w/v) SDS. The DIG label was detected by chemiluminescence (CSPD, Tropix, Bedford, MA) and the membranes were exposed to x-ray film (Fugi, Tokyo, Japan) at room temperature. Quantification of probe hybridized to rbcS mRNA of rice was achieved by image-density scanning of blots (IS-1000 version 2.0, Alpha Innotech Corp., San Leandro, CA). Northern-blot analysis was performed for two separate RNA extractions with similar results.

Carbohydrate Analysis

Soluble carbohydrates were extracted from approximately 150 mg of liquid-nitrogen-frozen leaf tissue using 80% (v/v) ethanol at 85°C. Glc, Fru, and Suc were quantified using the microtiter method described by Hendrix (1993). Carbohydrate assays were performed in triplicate on three separate extractions.

Statistical Methods

A repeated-measures analysis of variance was used to model Rubisco and soluble carbohydrate data according to a split-plot-in-time approach using the Mixed Repeated Measures Analysis procedure (SAS Institute, Cary, NC) to obtain the estimates for the model. The factors used for the model were [CO₂] treatment, day of experiment, and [CO₂] treatment × day of experiment.

RESULTS

Leaf Photosynthesis

Light-saturated rates of photosynthetic CO₂ assimilation by both mature and expanding leaves of rice, measured at their respective growth [CO₂], were greatest for plants growing at high CO₂ (700 μL L⁻¹ CO₂). Photosynthetic rates of mature leaves gradually declined during the course of the experiment, whereas those of expanding leaves tended to increase (Fig. 1). Values for rice grown continuously under ambient CO₂ ranged from 19 to 28 and 19 to 25 μmol CO₂ m⁻² leaf area s⁻¹ for mature and expanding leaves, respectively, whereas values for rice grown continuously under high CO₂ were as high as 33 to 37 μmol m⁻² s⁻¹.

Figure 1.

Photosynthetic CO₂-assimilation rate from mature (A) and expanding (B) leaves of rice. The switch in [CO₂] was made on d 2. Values are the means ± se, n = 3.

Upon switching growth [CO₂] from ambient to high and vice-versa there was a rapid adjustment in photosynthesis rate measured at the new [CO₂]. Mature leaves of rice switched to ambient CO₂ initially showed a 10% reduction in photosynthesis compared with ambient controls on d 2, the day of the CO₂ switch, when both were measured at ambient [CO₂]. By d 3, 1 d after the switch, the photosynthesis rate had increased to that of the ambient controls (Fig. 1A). Mature leaves of rice switched to high CO₂ had an 8% higher photosynthetic rate on d 2 and a 15% higher rate on d 3, compared with the high-CO₂ controls. By d 5, the rate had declined to that of the high-CO₂ controls. The photosynthetic rates of expanding leaves of rice switched to either high or ambient CO₂ follow the same pattern as their respective unswitched controls and were statistically indistinguishable from them (Fig. 1B).

Stomatal conductances of leaves for plants grown at high CO₂ were 35% to 40% lower than those for plants grown at ambient CO₂ (data not shown). For each of the CO₂ treatments the C_i/C_a ratios were determined (data not shown). The C_i/C_a ratio of high-CO₂ controls was 1% higher than that of the ambient controls throughout the study. In plants switched to either high or ambient CO₂, there was a slight decline in C_i/C_a ratio of 3% compared with ambient controls on d 3, but by d 5 it was again within 1% of that of the ambient controls.

Rubisco Total Activity and Content

Rubisco total activity and content declined with age in mature leaves, with the greatest reduction being in plants grown continuously under high CO₂ (Table I). During the course of the experiment, mature leaves of plants kept at ambient CO₂ showed 28% and 33% decreases in Rubisco activity and content, respectively, whereas plants kept at high CO₂ showed 43% and 41% decreases, respectively (Table I).

Table I.

Rubisco total activity and protein content for mature and expanding leaves of rice on d 1, 3, and 10 of the experiment

Leaf Type and Day of Experiment	Rubisco Total Activity				Rubisco Protein Content
	350	350 → 700	700	700 → 350	350	350 → 700	700	700 → 350
	μmol CO2 m−2 leaf area s−1				g m−2 leaf area
Mature
1	88.3a  ± 1.4		83.7a  ± 0.8		2.9a  ± 0.05		2.7b  ± 0.05
3	82.2a  ± 0.7	81.0ab  ± 2.0	65.8c  ± 1.1	73.9b  ± 2.2	2.8a  ± 0.05	2.8a  ± 0.03	2.4c  ± 0.05	2.6b  ± 0.03
10	63.8ab  ± 3.2	57.5b  ± 4.8	47.7c  ± 3.6	65.5a  ± 4.2	2.0a  ± 0.03	2.1a  ± 0.03	1.6b  ± 0.03	2.0a  ± 0.05
Expanding
1	77.8a  ± 2.0		75.4a  ± 1.9		2.7a  ± 0.05		2.5a  ± 0.05
3	74.6a  ± 3.6	73.4a  ± 5.1	67.8a  ± 4.4	69.9a  ± 4.4	2.6a  ± 0.08	2.8a  ± 0.07	2.5a  ± 0.10	2.6a  ± 0.07
10	75.7a  ± 5.0	69.4ab  ± 2.6	66.4b  ± 4.1	76.8a  ± 5.7	2.7ab  ± 0.10	2.5b  ± 0.07	2.3c  ± 0.08	2.9a  ± 0.10

The switch in [CO₂] was made on d 2. Values are the means ± se, n = 6 to 9; values followed by a different letter across treatments on the same day are significantly different at the P ≤ 0.05 level.

At d 3, Rubisco activity and content in mature leaves of rice switched to ambient CO₂ were significantly greater than those of high-CO₂ controls but less than the ambient-CO₂ controls. By d 10, the plants switched to ambient CO₂ did not differ from those kept at ambient CO₂, and had significantly higher activity and content than those kept at high CO₂. Rubisco activity and content in mature leaves of rice switched to high CO₂ differed little from those of the ambient-CO₂ control, and remained greater than those of the high-CO₂-grown plants, indicating a lack of adjustment despite the change in growth [CO₂].

Expanding leaves of rice kept at ambient or high CO₂ showed no significant differences in Rubisco total activity and content at d 1 and 3 (Table I). However, at d 10, both total activity and content were significantly less in the high-CO₂ plants. One day after switching to high or ambient CO₂ there were no changes in either total activity or content. By d 10, total activity and content in plants switched to ambient CO₂ had increased to levels similar to those of plants kept at ambient CO₂. Conversely, plants switched to high CO₂ appeared to have less Rubisco activity and content than the ambient-CO₂ plants, although the differences were not statistically significant.

Rubisco Small Subunit mRNA

Differences in steady-state rbcS transcript levels between plants grown at ambient and high CO₂ were detected in both mature and expanding leaves (Fig. 2). Levels of rbcS transcripts for mature leaves were similar in both ambient- and high-CO₂-grown rice on d 1. However, by d 3 the amount of rbcS transcripts in plants kept at ambient CO₂ was 13% greater than that of plants kept at high CO₂, and by d 10 it was 43% greater (Table II).

Figure 2.

Northern blots of rbcS mRNA for mature and expanding leaves of rice on d 1, 3, and 10. The switch in [CO₂] was made on d 2.

Table II.

Abundance of rbcS transcripts for mature and expanding leaves of rice on d 1, 3, and 10

Leaf Type and Day of Experiment	rbcS mRNA
	350	350 → 700	700	700 → 350
	% of 350 control
Mature
1	100		94  ± 4
3	100	77  ± 4	87  ± 3	98  ± 6
10	100	87  ± 4	57  ± 3	111  ± 5
Expanding
1	100		85  ± 8
3	100	85  ± 6	70  ± 7	97  ± 4
10	100	107  ± 7	70  ± 12	86  ± 2

The switch in [CO₂] was made on d 2. Values are the means ± sd of two experiments.

After switching the growth [CO₂], changes in rbcS transcript levels occurred within 24 h. In mature leaves of rice switched to ambient CO₂, the transcript level increased to that of ambient-CO₂ controls by d 3, whereas for plants switched to high CO₂ it decreased to 23% less than that of the ambient-CO₂ controls. At d 10 the level of rbcS transcripts for rice switched to ambient CO₂ was 11% greater than that of plants kept at ambient CO₂, whereas plants switched to high CO₂ showed 13% less.

In expanding leaves, rbcS transcript levels remained about 15% to 30% greater for plants kept at ambient CO₂ compared with plants kept at high CO₂ (Table II). After switching growth [CO₂], expanding leaves showed a similar response to that of mature rice leaves. At d 3 plants switched to ambient CO₂ showed an increase in transcripts. Conversely, plants switched to high CO₂ showed a 15% decrease compared with plants kept at ambient CO₂. At d 10 plants switched to high CO₂ showed the same transcript level as plants kept at ambient CO_2, whereas those switched to ambient CO₂ showed less. Expanding leaves of rice were approximately 20% and 50% expanded on d 1 and 3, respectively, and were fully mature by d 10.

Soluble Carbohydrates

Mature leaves of plants kept in high CO₂ had considerably greater Glc and Fru pool sizes than those of plants kept in ambient CO₂ (Table III). However, Suc pool size did not differ until d 10, when it was significantly greater in plants kept in high CO₂. Plants kept at ambient CO₂ showed little change in Glc and Fru from d 1 to 10, whereas Glc increased in plants kept at high CO₂. Suc pool sizes for both ambient- and high-CO₂ plants were greatest at d 1 and declined thereafter.

Table III.

Glc, Fru, and Suc content for mature and expanding leaves of rice on d 1, 3, and 10

Leaf Type and Sugar	Day of Experiment	Sugar Content
		350	350 → 700	700	700 → 350
		mg m−2 leaf area
Mature
Glc	1	111b  ± 10		190a  ± 14
	3	111b  ± 12	95b  ± 25	221a  ± 11	79b  ± 14
	10	79c  ± 14	111bc  ± 11	269a  ± 33	158b  ± 24
Fru	1	221b  ± 18		348a  ± 25
	3	253b  ± 29	205b  ± 8	411a  ± 37	253b  ± 24
	10	252b  ± 24	284ab  ± 32	363a  ± 45	363a  ± 27
Suc	1	4377a  ± 143		4550a  ± 99
	3	3887a  ± 175	3271b  ± 132	4266a  ± 95	2481c  ± 146
	10	3113c  ± 82	3539bc  ± 80	4076a  ± 139	3681b  ± 214
Expanding
Glc	1	340b  ± 13		448a  ± 23
	3	394a  ± 23	304b  ± 26	430a  ± 21	304b  ± 10
	10	286b  ± 26	537a  ± 13	537a  ± 31	251b  ± 6
Fru	1	412b  ± 16		501a  ± 37
	3	412ab  ± 41	358b  ± 59	465a  ± 42	376b  ± 28
	10	555b  ± 45	806a  ± 44	770a  ± 35	394c  ± 17
Suc	1	2541a  ± 119		1915b  ± 84
	3	2775a  ± 116	2775a  ± 180	2523ab  ± 147	2220b  ± 113
	10	4618b  ± 260	5871a  ± 239	5334a  ± 221	3437c  ± 71

The switch in [CO₂] was made on d 2. Values are the means ± se, n = 9; values followed by a different letter across treatments on the same day are statistically different at the P ≤ 0.05 level.

Switching the growth [CO₂] of rice from high to ambient resulted in a dramatic decline in soluble carbohydrates the day after the switch. Glc and Fru pools decreased to the same level as in the plants kept at ambient CO₂, whereas Suc decreased to significantly less. Compared with plants kept at high CO₂, Glc, Fru, and Suc pool sizes for plants switched to ambient CO₂ were 64%, 38%, and 42% less, respectively. At d 10 soluble sugar pool sizes for plants switched to ambient CO₂ increased to levels greater than those of plants kept at ambient CO₂, but Glc and Suc pools were significantly less than those of high-CO₂ controls.

Mature leaves of rice switched to high CO₂ showed no significant difference in Glc and Fru pool sizes by d 3 compared with ambient-CO₂ controls (Table III). However, Suc decreased dramatically by d 3, similar to that of plants switched to ambient CO₂. By d 10 soluble sugar pool sizes in leaves of plants switched to high CO₂ were not significantly different from those of plants kept at ambient CO₂.

Glc and Fru pool sizes of expanding leaves generally exceeded those of mature leaves (Table III). Suc content increased continuously in all treatments during the course of the experiment, and by d 10 the high-CO₂ plants and plants switched to high CO₂ showed significantly greater amounts than plants kept in ambient CO₂. By d 10 plants kept at high CO₂ showed 47% and 23% greater Glc and Fru pool sizes, respectively, than plants kept at ambient CO₂. Expanding leaves of plants switched to ambient CO₂ showed a decrease in both Glc and Fru on d 3, with Glc attaining a level significantly less than that of plants kept at ambient CO₂. From d 3 to 10 there was little change in the level of Glc and Fru in ambient-CO₂-switched rice, and although Suc increased, it remained significantly less than in the other three treatments. Plants switched to high CO₂ showed a small decline in Glc and Fru at d 3. However, by d 10 the amount of Glc, Fru, and Suc had increased dramatically to levels similar to that in plants kept at high CO₂.

DISCUSSION

A cross-switching of growth [CO₂] for rice previously grown at either ambient (350 μL L⁻¹) or high (700 μL L⁻¹) CO₂ resulted in a rapid (24 h) change in rbcS gene expression. Both expanding and mature leaves of rice switched to high CO₂ showed a decrease in transcripts with respect to plants maintained at ambient CO₂. In contrast, high-CO₂ plants switched to ambient CO₂ showed an increase in rbcS transcripts. Whatever signaled the response in the expression of rbcS with the switching of growth [CO₂] apparently was sensed by both expanding and mature leaves. This change in transcript levels persisted in mature leaves at d 10. Rapid responses in rbcS mRNA have been detected as a consequence of exogenously added sugars in detached tomato leaves (Van Oosten and Besford, 1994), maize mesophyll protoplasts (Sheen, 1990), and autotrophic cell-suspension cultures (Krapp et al., 1993). However, to our knowledge, this is the first report of a rapid response in rbcS expression for plants growing under field-like conditions and exposed to different [CO₂].

Rubisco total activity and content measured at midday responded to the switching of rice to ambient CO₂, but this response was not as rapid nor did it parallel the response in transcript levels. For expanding leaves, Rubisco total activity and content among all treatments were not significantly different on d 3, but by d 10 increased significantly for rice switched to ambient CO₂ and decreased for plants switched to high CO₂, although the change was not significantly different from that in plants kept at ambient CO₂. A lag time between rbcS expression and changes in Rubisco total activity and content has been shown by others. Krapp et al. (1993) found that when Chenopodium rubrum cell-suspension cultures were supplied with Glc, the expression of rbcS transcripts decreased within 6 h, which was long before significant changes in Rubisco total activity and content occurred. In all CO₂ treatments, mature leaves showed an overall trend of decreasing Rubisco total activity and content over time. The decrease was considerably greater in rice maintained at high CO₂. Switching to ambient CO₂ slowed this decline so that it paralleled that of plants maintained at ambient CO₂. However, switching to high CO₂ had no apparent effect on Rubisco total activity and content in mature leaves.

Synthesis of photosynthetically active Rubisco requires the coordination and assemblage of small and large subunits in the chloroplast. Therefore, changes in Rubisco synthesis may be regulated by changes in transcription of rbcS and rbcL, transcript stability, translation, and protein turnover, all of which may potentially be affected by growth [CO₂] (Webber et al., 1994). The level and change in regulation through each of these steps may also vary with developmental stage (Deng and Gruissem, 1987). Besford et al. (1990) have shown for tomato leaves that Rubisco activity and protein content peak and then decline earlier during leaf expansion in high CO₂ than in ambient CO₂. Although transcriptional regulation of Rubisco small and large subunits has been shown to be affected by [CO₂] in some instances (Winder et al., 1992), the level of regulation by translation and posttranslational turnover is complicated by the fact that photosynthetically competent Rubisco has a relatively slow turnover rate (Peterson et al., 1973).

Our study has shown that upon switching from high to ambient CO₂, expanding leaves up-regulated rbcS mRNA, which apparently resulted in increased Rubisco synthesis, whereas switching to high CO₂ resulted in down-regulation of rbcS mRNA and a decrease in Rubisco synthesis. Cheng et al. (1998) also showed that rbcS mRNA levels were substantially repressed in Arabidopsis 6 d after transferring them from a growth [CO₂] of 360 to 1000 μL L⁻¹, followed later by a significant decrease in Rubisco protein content at 9 and 12 d. The response of rice leaves to changes in growth [CO₂] in our study is much more rapid than the acclimation to elevated CO₂ reported in some other field studies (Körner and Diemer, 1994; Jacob et al., 1995).

Regulation of rbcS transcription was also apparent in mature leaves. Rubisco turnover appeared to be greater than synthesis, because content declined in all treatments from d 1 to 10. However, plants switched to ambient CO₂ showed an increase in rbcS transcripts and by d 10 significantly greater Rubisco content than plants kept at high CO₂, suggesting that synthesis had increased in this treatment. The processes of synthesis and turnover may have been affected differently by the switching of growth [CO₂] depending on the developmental stage of the leaf. However, it is not evident from these data whether the turnover rate was affected by CO₂ switching.

Increased amounts of nonstructural carbohydrates are often associated with photosynthetic acclimation (Bowes, 1993). The down-regulation of photosynthesis associated with acclimation has been attributed to an accumulation of carbohydrates in photosynthetically active source leaves caused when photosynthetic rate exceeds the capacity of sinks to use assimilate for growth (Arp, 1991; Farrar and Williams, 1991). In our study leaf photosynthesis was generally 25% to 30% greater in mature leaves and 20% to 24% greater in expanding leaves of plants grown under high CO₂. Photosynthetic acclimation was evident in mature leaves of rice grown at high CO₂. This can be inferred by comparing photosynthetic rates of plants switched to high CO₂ with those of plants maintained at high CO₂ on the day of the switch and 1 d after, when both were measured at high CO₂. Photosynthesis was 15% greater in plants switched to high CO₂. Also, plants switched to ambient CO₂ showed a 10% lower photosynthetic rate than plants maintained at ambient CO₂ on the day of the switch.

Growth at elevated CO₂ often results in decreased stomatal conductance (Drake et al., 1997). To determine if stomatal conductance was a major limitation to leaf photosynthesis, in the present study the C_i/C_a ratios for various treatments were compared. Even though stomatal conductivity did decrease approximately 35% to 40% at high CO₂, the C_i/C_a ratio of high-CO₂ controls was nearly identical to that of ambient controls (1% higher) throughout the study. In plants switched to either high or ambient CO₂, there was a slight decline in C_i/C_a ratio on d 3 (3% decrease), but by d 5 it was again within 1% of the ambient controls. These data indicate that stomatal conductance was not a major limiting factor to photosynthesis in the high-CO₂-grown plants.

Pool sizes of Glc and Fru for mature leaves of plants kept in high CO₂ were up to 2-fold greater than those of plants kept in ambient CO₂, whereas Suc was only significantly higher on d 10. Transcript levels for rbcS did not differ on d 1, but were greater for the ambient controls on d 3 and 10. Differences in soluble carbohydrate pool sizes between ambient- and high-CO₂-grown plants were not as great in expanding leaves as in mature leaves, at least until d 10. This might be expected, because at d 1 and 3 these leaves were growing and thus acting as carbohydrate sinks, whereas by d 10 they were fully expanded. However, at all three sampling times, plants kept at ambient CO₂ contained a higher level of rbcS mRNA. Switching from high to ambient CO₂ led to a significant decline in Glc and Fru pool sizes for both mature and expanding leaves and a dramatic decrease in Suc in mature leaves on d 3. This was accompanied by an increase in rbcS transcripts for both mature and expanding leaves, which suggests that decreased carbohydrate pool sizes may have signaled the activation of rbcS transcription. However, for plants switched to high CO₂, even though photosynthesis increased and rbcS transcripts decreased, the pool sizes of Glc, Fru, and Suc either stayed the same or decreased in comparison with plants kept at ambient CO₂. These data are at odds with the hypothesis that rbcS expression is regulated by soluble carbohydrate pool sizes. This does not rule out the possibility that soluble carbohydrates are involved in the regulation of rbcS expression, but our data indicate that the overall leaf pool sizes during leaf development cannot be used to predict the degree of rbcS expression.

Rice is known to be a Suc accumulator when photosynthetic rates are relatively high (Rowland-Bamford et al., 1990, 1996). During vegetative growth, elongating nodes and culms act as a major sink for photosynthetic assimilate (Yoshida and Aln, 1968). Rowland-Bamford et al. (1990) have shown during vegetative growth in rice that as growth [CO₂] increases, so does the tendency for recently produced photosynthetic assimilate in leaves to be exported rather than stored. Because photosynthesis increased substantially in plants switched to high CO₂, the most likely reason for the observed decrease in soluble carbohydrates, or the lack of accumulation, was an increase in export of recent photosynthetic assimilate.

The lack of correlation between soluble carbohydrate content and photosynthetic acclimation during growth under elevated CO₂ seen here and in other studies (Wang and Nobel, 1996; Moore et al., 1997) suggests that fluctuations in soluble carbohydrate pools and metabolic cycling of carbon produced through photosynthesis, rather than total pool sizes themselves, may have a more vital role in signaling regulation of rbcS expression. The phosphorylation of hexose sugars via hexokinase has been hypothesized as a sugar-sensing and signal transduction mechanism for expression of sugar-sensitive genes in plants (Jang and Sheen, 1994). Cheng et al. (1998) showed that rbcS levels in Arabidopsis fluctuate diurnally, with a maximum occurring at the beginning and a minimum at the end of the light cycle; soluble carbohydrate levels generally followed an inverse pattern. Based on this and previous evidence (Moore et al., 1997), they suggest that hexokinase may act as a hexose flux sensor, primarily via nighttime activity, mediating the repression of rbcS transcription. Recently, Jang et al. (1997) have shown evidence for the involvement of hexokinase in the Glc-mediated repression of rbcS and cab1 (chlorophyll a/b-binding protein) and activation of nitrate reductase genes in transgenic Arabidopsis plants with altered hexokinase expression.

In conclusion, this study documents that a nonintrusive modification of atmospheric growth [CO₂] in vegetatively growing rice under field-like conditions leads to a rapid change in expression of the rbcS in mature and expanding leaves. Switching the growth [CO₂] eventually led to alterations of Rubisco protein content and total activity. Both mature and expanding leaves maintained the ability to up-regulate Rubisco upon being switched from high to ambient CO₂, whereas exposure to high CO₂ resulted in down-regulation of Rubisco. The signal for changing rbcS gene expression as a consequence of an alteration in [CO₂] evidently is detected soon after switching, certainly within 24 h. However, changes in total leaf soluble carbohydrate pool sizes alone could not explain the rapid change in rbcS expression that we detected. Further work is under way to determine the flux of carbon through soluble carbohydrate pools under growth at elevated CO₂ in an effort to elucidate the signal for rbcS expression as a function of [CO₂].

Abbreviations:

C_a

atmospheric CO₂ concentration

C_i

internal CO₂ concentration

DIG

digoxigenin

RuBP

ribulose-1,5-bisphosphate