key: cord-0041905-pnxqu50b
authors: Moore, B. D.; Cheng, S.‐H.; Rice, J.; Seemann, J. R.
title: Sucrose cycling, Rubisco expression, and prediction of photosynthetic acclimation to elevated atmospheric CO(2)
date: 2002-03-01
journal: Plant Cell Environ
DOI: 10.1046/j.1365-3040.1998.00324.x
sha: e05948739053176c90997d20e8c5f12095a80363
doc_id: 41905
cord_uid: pnxqu50b

Photosynthetic acclimation to elevated CO(2) cannot presently be predicted due to our limited understanding of the molecular mechanisms and metabolic signals that regulate photosynthetic gene expression. We have examined acclimation by comparing changes in the leaf content of RuBP carboxylase/oxygenase (Rubisco) with changes in the transcripts of Rubisco subunit genes and with leaf carbohydrate metabolism. When grown at 1000 mm(3) dm(–3) CO(2), 12 of 16 crop species at peak vegetative growth had a 15–44% decrease in leaf Rubisco protein, but with no specific association with changes in transcript levels measured at midday. Species with only modest reductions in Rubisco content (10–20%) often had a large reduction in Rubisco small subunit gene mRNAs (> 30%), with no reduction in large subunit gene mRNAs. However, species with a very large reduction in Rubisco content generally had only small reductions in transcript mRNAs. Photosynthetic acclimation also was not specifically associated with a change in the level of any particular carbohydrate measured at midday. However, a threshold relationship was found between the reduction in Rubisco content at high CO(2) and absolute levels of soluble acid invertase activity measured in plants grown at ambient or high CO(2). This relationship was valid for 15 of the 16 species examined. There also occurred a similar, albeit less robust, threshold relationship between the leaf hexose/sucrose ratio at high CO(2) and a reduced photosynthetic capacity ≥ 20%. These data indicate that carbohydrate repression of photosynthetic gene expression at elevated CO(2) may involve leaf sucrose cycling through acid invertase and hexokinase.

Growth of many herbaceous and woody plant species at elevated [CO 2 ] often results in a decrease in leaf photosynthetic capacity due to a decrease in leaf content of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco; Sage, Sharkey & Seemann 1989; Webber, Nie & Long 1994) . However, the magnitude of this acclimation response can vary not only between different species, but also among different varieties or cultivars of the same species grown under controlled conditions (e.g. Campbell 1997; Kalina & Ceulemans 1997) . Furthermore, the extent by which photosynthetic capacity may be reduced at elevated CO 2 in a given species can be modulated by a number of factors including light (Ghannoum et al. 1997; Sims, Seemann & Luo 1998) , nutrient availability (Tissue et al. 1993; Seneweera et al. 1994; Nakano et al. 1997) , water availability (Roden & Ball 1996; Scarascia-Mugnozza et al. 1996) , and tissue developmental stage (Van Oosten & Besford 1995; Xu, Gifford & Chow 1996) . However, there presently is no basis to predict photosynthetic acclimation to elevated CO 2 , in large part due to a lack of understanding of the regulatory mechanisms and metabolic signals involved.

Leaf carbohydrates commonly accumulate during longterm growth at high CO 2 (Long & Drake 1992 ) as result of a limited capacity for their utilization in growing sink tissues (Stitt 1991 ). An increase in leaf carbohydrates has long been associated with an inhibition of photosynthesis (e.g. Neals & Incoll 1968 ) and may account for the downregulation of photosynthetic capacity that can occur in plants exposed to elevated CO 2 (Van Oosten, Wilkins & Besford 1994) . However, end-product inhibition of photosynthesis is not due to increased carbohydrate content per se, but instead is associated with the metabolism of hexoses probably derived from sucrose hydrolysis (Goldschmidt & Huber 1992; Krapp et al. 1993) . Hexokinase has been proposed to function as a sugar sensor in the cytosol of plant mesophyll cells (Jang & Sheen 1994; Jang et al. 1997) . During phosphorylation of hexoses, hexokinase is hypothesized to initiate a signal cascade that results in the repressed expression of a number of photosynthetic genes (Sheen 1994; Smeekens & Rook 1997) , but there are many aspects of this process that are not known.

In order to understand and therefore possibly predict photosynthetic acclimation at elevated CO 2 , we are interested in determining the process by which carbohydrates generate a mesophyll signal and the molecular targets and processes that are modulated at high CO 2 . If hexokinase can function as a flux sensor (Graham, Denby & Leaver 1994; Koch 1996) , an internal signal could be generated simply by increased provision of substrate in the mesophyll cytoplasm. However, there was no evidence for the presence of a significant level of daytime cytosolic hexoses in leaves of three species grown at elevated CO 2 (Moore, Palmquist & Seemann 1997) . Alternatively, sucrose hydrolysis by soluble acid invertase (Goldschmidt & Huber 1992) or hydrolytic starch degradation (Kruger 1990 ) could generate increased substrates for the hexokinase reaction. Thus, hexoses might not accumulate in the cytosol depending on their rate of transport from the vacuole or chloroplast and the catalytic activity of hexokinase.

Plant growth at elevated CO 2 can result in repression of a number of photosynthetic genes, but the effects are known to vary between species as well as within species. Nuclearencoded genes such as rbcS (Rubisco small subunit genes) and cab (chlorophyll a/b binding protein genes) are much more sensitive to repression than are chloroplast-encoded genes such as rbcL (Rubisco large subunit genes) or psaA-B (photogene; Van Oosten & Besford 1995) . Furthermore, growth at high CO 2 can result in a differential repression of rbcS gene-specific mRNAs within even a given species (Arabidopsis; Cheng, Moore & Seemann 1998) .

In this study, we have two main objectives in examining photosynthetic acclimation in a number of species grown at high CO 2 . First, we have evaluated acclimation (as a reduction in Rubisco content) in relation to leaf carbohydrate metabolism and in relation to rbcS and rbcL transcript levels to determine which, if any, parameters may be useful for predicting acclimation. Carbohydrate metabolism was examined by measuring the accumulation of specific leaf sugars and the activities of sucrose-phosphate synthase (SPS) and soluble acid invertase. Activities of these two enzymes are important to consider since they largely control the level of sucrose accumulation and hexose production within leaves (Huber 1989; Goldschmidt & Huber 1992) . Secondly, we have used the correlative information to help elucidate the mechanism for carbohydrate signalling at elevated CO 2 and to identify affected molecular mechanisms that control Rubisco content.

Bean (Phaseolus vulgaris L. cv. Little Linden), bugle (Ajuga reptans L. cv. Bronze Beauty), corn (Zea mays L. cv. Early Sunglow hybrid, cotton (Gossypium hirsutum L. cv. Delta Pine), cucumber (Cucumis sativus L. cv. Poinsett 76), parsley (Petroselinum crispum cv. Curled Leaf), plantain (Plantago lanceolata L.), soybean (Glycine max [L.] Merr. cv. Williams), spinach (Spinacea oleracea L. cv. Noble), sunflower (Helianthus annuus L. cv. Pioneer Brand #1650), tobacco (Nicotiana sylvestris L.) and tomato (Lycopersicon esculentum Mill. cv. Early Girl) were grown to a peak vegetative growth stage (5-8 weeks depending on species) in 5 dm 3 pots in greenhouses with natural irradiance, a 28°/15°C day/night thermoperiod, and either 400 or 1000 cm 3 m -3 CO 2 . Pea (Pisum sativum L. cv. Little Marvel), radish (Raphanus sativus L. cv. Cherry Belle), and wheat (Triticum aestivum L. cv. Twin Spring) were grown for 3-4 weeks in flats in the greenhouses. Plants were watered daily with Peter's nutrient solution (15, 16, 17% N, P, K; 5·7 mM N, 1·5 mM P, 2·4 mM K) . Arabidopsis (Arabidopsis thaliana [L.] Henyh. ecotype Columbia) were grown in growth chambers for 4-5 weeks as described previously (Cheng, Moore & Seemann 1998) . Youngest, full-sized leaves or rosettes (Arabidopsis) were collected into liquid N 2 at midday (12:00) and visible leaf veins were removed prior to biochemical analyses.

Leaf material (0·25 g FW) was extracted in boiling 80% ethanol as previously described (Moore, Palmquist & Seemann 1997) . Residual material from whole-leaf extracts was autoclaved and the starch from replicate aliquots was hydrolysed as described by Schulze et al. (1991) . Soluble carbohydrates were measured by HPLC using a Dionex DX 300 system with a pulsed amperometric detector and a CarboPac PA1 or occasionally an MA1 column (Dionex Corp., Sunnyvale, CA, USA) as previously described for parsley (Moore, Palmquist & Seemann 1997).

Rubisco content was measured as described elsewhere (Evans & Seemann 1984) . Chl content of whole-leaf material was determined after extraction in ethanol (Wintermans & De Mots 1965) .

Leaf samples (0·25 g FW) were extracted with a mortar and pestle in 2 cm 3 of 50 mM Mops-KOH (pH 7·5), 5 mM MgCl 2 , 1 mM EDTA, 2·5 mM dithiothreitol, 0·05% Triton X-100, and 10% (w/w) polyvinylpolypyrrolidone. Extracts were filtered through Miracloth and centrifuged at 10 000 g for 1 min. The supernatants were immediately desalted by centrifugal filtration using 5 cm 3 Sephadex G-25 columns equilibrated in extraction buffer minus the Triton X-100.

Soluble acid invertase (β-fructosidase, EC 3·2.1·16) was assayed at 25°C using components described by Huber (1989) . Reactions were initiated by addition of extract and aliquots were removed for boiling at 0 and 30 or 60 min. Amounts of glucose plus fructose were determined by endpoint enzyme assays coupled through glucose 6-phosphate dehydrogenase (from Leuconostoc) as described by Doehlert (1989) , but with all coupling enzymes used at 2 units cm 3 .

Sucrose-phosphate synthase (EC 2.4.1.14) was assayed at 30°C by measuring under V max conditions the fructose-6-P-dependent formation of sucrose-6-phosphate + sucrose from UDP-glucose . Sucrose products were measured using an anthrone reagent (Van Handel 1968).

Total leaf RNA was isolated as in Cheng & Seemann (1997) . The purity and quantity of the RNA samples was monitored by measuring UV absorbance at 230, 260 and 280 nm . The ratios of A 260 /A 230 and A 260 /A 280 were each about 2, indicating little or no contamination by carbohydrates or proteins. The average RNA yield was about 500 µg (g FW) -1 of leaf tissue.

Northern analyses were carried out using 1 µg (for tobacco and Arabidopsis) or 5 µg (for all other species) of total RNA as described by Cheng, Moore, & Seemann (1998) . cDNAs used as probes were a 520 bp HindIII fragment of tobacco rbcS (pTSSU3) kindly provided by Dr Graham Hudson, a 1·2 kb EcoR1/Xba1 fragment of corn rbcS (SS1, a generous gift from Dr Jen Sheen), a 750 bp Sal1/Not1 fragment of Arabidopsis rbcS (ID no. 11C1T7P) obtained from the Arabidopsis Biological Research Center at Ohio State University, and a 1·2 kb plastid BamH1/EcoR1 fragment of tobacco rbcL (Shinozaki & Sugiura 1982) . These DNA fragments were labelled with [α-32 P]dCTP using the Prime-a-Gene Labeling system (Promega). For hybridization of rbcS mRNA, 32 P-labelled Arabidopsis rbcS cDNA was used for Arabidopsis, bean, bugle, cotton, parsley, pea, plantain, radish, spinach, sunflower and wheat; labelled tobacco rbcS cDNA was used for cucumber, soybean, tobacco, and tomato, and labelled corn rbcS cDNA was used for corn. 32 P-labelled tobacco rbcL cDNA was used for all species. In each blot, a dilution series of an RNA sample was included to ensure that 32 Plabelled probes were in excess. The hybridization conditions were as described by Cheng et al. (1998) . Following hybridization, the blots were washed twice for 5 min at room temperature in 2X SSC/0·1% SDS, and then twice for 10 min either in 0·2X SSC/0·1% SDS at 60°C (for rbcS mRNA of Arabidopsis, corn and tobacco), in 0·5X SSC/0·1% SDS at 56°C (for rbcS mRNA of all other species), or in 0·5X SSC/0·1% SDS at 60°C (for rbcL mRNA). Hybridization signals were quantified with a Phosphor Imager (Bio-Rad Laboratories, Inc.) to determine the relative amount of RNA present in each lane.

In this study, we have examined possible correlations between photosynthetic acclimation and specific biochemical and molecular characteristics of mature leaves of 16 species grown at ambient or elevated CO 2 . Rubisco protein content was measured as a biochemical indicator of leaf photosynthetic capacity. The species were grouped according to the relative reduction in leaf Rubisco protein content that occurred in plants grown at high CO 2 ( Table 1) . Some of the species in the different groups include corn and spinach with no reduction (group 1), cotton and wheat with a moderate reduction (10-20%, group 2), cucumber and soybean with a substantial reduction (20-30%, group 3), and Arabidopsis and tobacco with a very large reduction (30-45%, group 4).

The leaf amounts of rbcS and rbcL mRNAs were measured from the same samples used to measure Rubisco contents ( Fig. 1 , Table 1 ). Relative leaf Rubisco protein content at high CO 2 generally did not correlate with the expression level of Rubisco transcripts (both measured at midday). In many species, rbcS and rbcL mRNAs were the same or higher at elevated CO 2 (e.g. corn and spinach), including some species which had a reduced Rubisco content (e.g. plantain and tobacco; Table 1 ). In other species, though, rbcS mRNA amounts were substantially reduced, but rbcL mRNA levels were not reduced much, if any (e.g. parsley, cotton and wheat). Only in Arabidopsis were transcript amounts of both gene products reduced at high CO 2 to about the same extent as Rubisco content. Species of group 2 had highly reduced rbcS mRNA levels (> 30%), with only a moderate reduction in Rubisco protein content (e.g. cotton and sunflower). Species in group 4 had the most reduction in Rubisco content, yet both rbcS and rbcL transcript levels were frequently not much affected (e.g. radish and tobacco). These responses indicate that photosynthetic acclimation cannot be predicted by changes in message level (Fig. 2) .

A decline in relative Rubisco content of species at high CO 2 was generally associated with some decrease in leaf Chl content, but the latter to a lesser extent (Fig. 3 , e.g. radish, 23% decrease in Chl, 34% decrease in Rubisco). However, there was substantial variation in the relative influence of high CO 2 on Rubisco protein and Chl contents. In plantain and sunflower, leaf Chl contents did not decline although Rubisco contents did. Such variation indicates that one cannot accurately predict photosynthetic acclimation based on the response of leaf Chl content. Whether photosynthetic components other than leaf Chl and/or Rubisco content were affected in this study by growth at elevated CO 2 is not known.

Next, we examined to what extent particular aspects of carbohydrate metabolism were correlated with photosynthetic acclimation. The accumulation of total non-structural carbohydrates (TNC) in leaves varied substantially between species grown at either ambient or elevated CO 2 (Fig. 4) . Species with the greatest amounts of TNC (cotton, cucumber, soybean, tomato and parsley) had relatively high levels of starch. The present carbohydrate data notably are closer to actual mesophyll levels than those reported in previous studies due to our removal of as much vein material as reasonably possible prior to analyses. None the less, the absolute amount of TNC at high CO 2 was not closely correlated with photosynthetic acclimation. For example, at high CO 2 Arabidopsis had relatively low carbohydrate levels and greatly reduced Rubisco protein content, while parsley had relatively high TNC levels but no reduction in Rubisco content. Furthermore, cotton, cucumber and soybean accumulated the greatest amounts of TNC, yet none of these species had a very large reduction in Rubisco protein content (all < 25%). Photosynthetic acclimation also was not associated with the accumulation of any particular sugar. For example, a number of species such as cotton, soybean and tomato had high levels of Photosynthetic acclimation to elevated CO 2 907 starch at high CO 2 and had a reduced Rubisco content (Fig. 5) . However, this relationship was not consistent. Parsley had a substantial amount of starch and no reduction in Rubisco content, while several species had a reduced Rubisco level but a relatively low amount of starch (e.g. Arabidopsis, plantain and radish).

Although total TNC levels were not associated with photosynthetic acclimation, we examined whether changes in the relative leaf Rubisco content in plants grown at high CO 2 may be a function of the relative change in TNC (Fig. 6 ) or in total hexoses (Fig. 7) . A decline in leaf Rubisco content was not necessarily associated with increased TNC. TNC changed very little in several species which had reduced Rubisco levels (e.g. sunflower, wheat and radish). Also, TNC increased 3-fold in spinach, but Rubisco content did not decline. Similarly, changes in leaf hexose content were not specifically associated with adjustments in Rubisco content at high CO 2 . For example, leaf hexoses increased 2·5-fold in spinach, but Rubisco content did not change. Also, there were minimal changes in leaf hexoses at high CO 2 in a number of species which had substantial declines in Rubisco content (e.g. bean, radish and sunflower). Next, we measured activities of SPS (under V max conditions) and soluble acid invertase (Table 2) in the same leaf material used for the previously described measurements. SPS activities declined by ≥ 35% in a few species grown at high CO 2 (e.g. bean, cotton, cucumber and plantago), with smaller differences in other species. Acid invertase activities also declined substantially in only a few species at high CO 2 (e.g. radish, soybean and wheat; about a 25% decline). For all species grown at high CO 2 , the relative mean SPS activity was 87% and the relative mean soluble acid invertase activity was 82%. The SPS activities of species did not correlate with measured leaf sucrose levels, nor were acid invertase activities closely associated with total hexose or total sucrose levels (analyses not shown). The general decline in leaf SPS activities at high CO 2 was not correlated with specifiec decreases in Rubisco content (Fig. 8) . Since a number of species had similar relative leaf SPS activities, but very different relative Rubisco contents, one cannot predict photosynthetic acclimation based on measured SPS activities.

Although growth at high CO 2 did not substantially affect leaf acid invertase activity in most species (Table 2) Values are means of three or more extractions from mature leaves collected at midday, and are expressed as µmol of hexose equivalents per g FW. Values for bugle, cucumber, parsley and plantain include raffinose sugars, galactinol, mannitol and/or sorbitol. These sugars were not present at significant levels in the other species for whom values reflect only cumulative amounts of glucose, fructose, sucrose and starch. For a species, the average increase in soluble sugars was 29% (median increase 26%), and in starch 161% (median increase 135%). photosynthetic acclimation was related to the absolute level of acid invertase activity measured from plants grown at either ambient or high CO 2 (Fig. 9) . High acid invertase activities were always associated with a decrease in photosynthetic capacity at high CO 2 . Also, the absence of photosynthetic acclimation at high CO 2 (group 1 species in Table 1 ) occurred only in species with low activities of acid invertase. The data indicate that a threshold relationship exists between the level of acid invertase activity and photosynthetic acclimation to high CO 2 . Notably, in plants grown at ambient CO 2 , there was a 3-fold increase in invertase activity from the highest of the non-acclimating species to the lowest of the acclimating species. The threshold relationship between enzyme activity and acclimation was valid for 15 of the 16 species examined. The sole exception was bean which had only a low activity of acid invertase, but did have a substantial reduction in leaf Rubisco content when grown at high CO 2 . Since acid invertase catalyses the hydrolysis of sucrose to glucose plus fructose, we also examined to what extent one can predict photosynthetic acclimation at high CO 2 based on the amounts of leaf hexose-carbon that are present relative to sucrose-carbon (Fig. 10) . High hexose/sucrose ratios (> 2·0) were present only in Arabidopsis, radish, sunflower and tomato. These species all showed substantial photosynthetic acclimation (≥ 20% decline). A number of species that also showed photosynthetic acclimation had more moderate values of the leaf hexose/sucrose ratio, but these values increased substantially at high CO 2 . These species included cucumber, plantain, soybean and tobacco, with respective ratios at high CO 2 of 0·29, 1·4, 0·39 and 0·97. Thus, a hexose/sucrose ratio > 0·25 for leaves collected at high CO 2 was associated with a ≥ 20% decline in photosynthetic capacity in eight of Photosynthetic acclimation to elevated CO 2 911 © 1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 905-915 nine species. The sole exception again was bean. However, a moderate decline in Rubisco content (10-20%) at high CO 2 was not associated with either a high leaf hexose/sucrose ratio or an increased value in that ratio.

In this study, we found that species with relatively high leaf activities of soluble acid invertase showed photosynthetic acclimation (defined as a reduction in leaf Rubisco content) when grown at an elevated [CO 2 ] (Fig. 9) . The use of leaf Rubisco content as a biochemical indicator of the photosynthetic capacity of a species is generally reasonable, since within a given species adjustments in Rubisco content commonly result in corresponding adjustments in the photosynthetic rate (Evans & Seemann 1989; Stitt & Schulze 1994 ). However, the differential adjustments amongst species in leaf Chl and Rubisco contents that occurred at elevated CO 2 (Fig. 3) indicate that the responses of different photosynthetic components are not necessarily linked. None the less, the relationship between leaf acid invertase activity and Rubisco content is important for identifying the mechanism of photosynthetic acclimation, for use as a tool to predict acclimation, and also possibly as a target for plant transformation.

With regard to the mechanism, the data suggest that socalled 'futile' cycling of sucrose may in fact be a regulatory component involved in leaf sugar sensing. Goldschmidt & Huber (1992) provided evidence that sucrose cycling may occur in mesophyll cells of mature leaves under sink-limited conditions through sucrose hydrolysis by vacuolar acid invertase coupled with hexose phosphorylation by cytosolic hexokinase. If hexokinase can function as a flux sensor (Graham, Denby, & Leaver 1994; Koch 1996) , then the increased provision of hexoses from sucrose cycling could account for the generation of a primary carbohydrate signal. There is apparently no direct evidence from any metabolite labelling studies that sucrose cycling does occur in leaves. None the less, there are several lines of indirect evidence that such can occur. First, overexpression of a yeast invertase gene in the vacuole of mature tobacco leaves resulted in an accumulation of leaf hexoses and an inhibition of photosynthesis (Sonnewald et al. 1991) . Secondly, an analogous pathway of sucrose cycling that involves sucrose synthase rather than acid invertase has been shown to occur in potato tubers and certain other nonphotosynthetic tissues (Geigenberger & Stitt 1993) . Thirdly, sucrose cycling at high CO 2 through acid invertase might be expected to increase the leaf hexose/sucrose ratio as observed in many species in this study (Fig. 10 ), in part due to the relatively low activity of leaf hexokinase (e.g. Huber 1989) .

Photosynthetic acclimation to elevated CO 2 was related to a threshold level of soluble acid invertase activity, resulting in an ability to predict in 15 of 16 species whether photosynthetic capacity would decrease (Fig. 9) . After petiole girdling, a comparable threshold relationship was previously observed between photosynthetic capacity or sucrose accumulation, and soluble acid invertase activity (Goldschmidt & Huber 1992) . Photosynthetic acclimation at high CO 2 similarly could be predicted from leaf hexose/sucrose ratios at high CO 2 , but only in cases of a ≥ 20% decrease in photosynthetic capacity (Fig. 10) . In coldstored tubers from a number of potato cultivars, the hexose/sucrose ratio was linearly dependent on the extractable activities of soluble acid invertase (Zrenner, Schuller & Sonnewald 1996) . In this study, species activities of soluble invertase were generally, but not rigorously, related to the leaf hexose/sucrose ratio (data not shown). Photosynthetic acclimation could not be predicted though from either the absolute activities or relative change in SPS activities (V max ; Tables 1 and 2; Fig. 8 ) or from the absolute amounts or relative changes in particular leaf carbohydrates (Figs 4, 5, 6 & 7) .

Leaf soluble acid invertase has two other characteristics that further strengthens its use in predicting photosynthetic acclimation. Leaf soluble invertases are not thought to be metabolically regulated in vivo other than by substrate availability (Kruger 1990) . Also, leaf soluble acid invertase activity is not much affected under sink-limited conditions that induce carbohydrate repression of gene expression (Table 2 ; see also Goldschmidt & Huber 1992 ). This response is in contrast to strong glucose-repression of yeast acid invertase expression (Gancedo 1992 ) and carbohydrate-dependent regulation of maize invertase genes in certain sink tissues (Koch 1996) .

Bean (Phaseolus vulgaris) had an anomalous response at elevated CO 2 , since photosynthetic acclimation occurred although soluble acid invertase activity was quite low (Fig. 9 ). There is some evidence that plant carbohydrate signalling can also occur by transport of external hexoses or even sucrose across the plasmalemma (Koch 1996; Smeekens & Rook 1997) . One such mechanism may involve leaf sucrose hydrolysis by cell wall invertase, thus establishing a futile sucrose cycle between the apoplast and mesophyll cell (Foyer 1987) . The transport of hexoses into the mesophyll cell and/or their subsequent phosphorylation may result in carbohydrate signalling, as apparently occurs in tobacco plants transformed to express yeast invertase in the leaf apoplasm (Stitt, von Schaewen & Willmitzer 1990) . Whether any such alternate mechanisms may account for bean photosynthetic acclimation remains to be determined (see also below). However, the low hexose/sucrose ratio in bean at elevated CO 2 is indirect evidence against any invertaserelated signalling mechanism.

Rubisco protein expression is normally co-ordinated with rbcS and rbcL gene expression during leaf development (Jiang & Rodermel 1995) . Growth at elevated CO 2 altered this coordination, and in different ways, in many of the species examined in this study (Table 1 ). The adjustments in plant development that commonly occur at elevated CO 2 (Bowes 1993) can complicate comparisons between species. In this study though, similar relative adjustments were repeatedly observed in leaf Rubisco protein and message levels in different, mature plants during vegetative growth and, when examined, in leaves just older than the youngest, full-sized (data not shown). The relative lack of coordination between Rubisco content and message levels at elevated CO 2 (Fig. 2) more probably reflects the complexity of molecular control of Rubisco content. The control of Rubisco content involves a number of processes, including transcription, post-transcription (e.g. message stability), translation and/or post-translation (e.g. protein turnover) events (Berry et al. 1986; Shirley & Meagher 1990; Wanner & Gruissem 1991) . In some species, there may have even occurred an altered diurnal rhythm in mRNA accumulation. Nonetheless, during growth of species at elevated CO 2 , there apparently did occur differential adjustments in the control of these processes, which are described as follows.

In a few species, transcript levels of both rbcS and rbcL genes were increased at high CO 2 even though Rubisco content was unchanged (e.g. corn) or even decreased (e.g. plantain). The reason for such stimulation of transcript levels is not clear, but the overall response suggests that translation of messages of both rbcS and rbcL genes can be directly or indirectly affected at high CO 2 . In a second response pattern to high CO 2 , there occurred a large decrease (> 30%) in rbcS mRNA, no decline in rbcL mRNA, and generally only a modest decrease in Rubisco content (e.g. parsley plus group 2 species). The simplest explanation for this response is that rbcS gene transcription may be reduced and either rbcL mRNA is not efficiently translated (as in tobacco rbcS antisense plants; Rodermel et al. 1996) or large subunit protein may accumulate in excess (as has been observed in tomato at high CO 2 ; Fig. 4 , Van Oosten & Besford 1995) . In these species, Rubisco protein content may be controlled in large part by the abundance of small subunit protein. In a third response pattern, there occurred a very large reduction in Rubisco content in the absence of appreciable adjustments in the levels of either transcript (e.g. bean, radish and tobacco). The control of Rubisco content in these species is probably rather complex. Possibly, either one or both of the subunit mRNAs are not efficiently translated, or Rubisco protein turnover may increase at high CO 2 . In these species, the primary target(s) for gene repression may involve specific translation factors for mRNA of either small or large subunit genes, specific proteases, and/or some dysfunction in the processing of small subunit protein or assembly of holoenzyme. Finally, in a fourth response pattern that occurred only in Arabidopsis, levels of rbcS and rbcL mRNAs and Rubisco content all substantially decreased at high CO 2 in an apparently co-ordinated manner. This response suggests that control of Rubisco content can be determined largely by message level, as discussed in more detail previously (Cheng, Moore & Seemann 1998) .

Carbohydrate modulation of photosynthetic gene expression probably also occurs in ambient CO 2 conditions since a decrease in [CO 2 ] also alters rbcS mRNA levels (Krapp et al. 1993; Majeau & Coleman 1996) . However, whether leaf sucrose cycling may occur under ambient CO 2 growth conditions is not known. Since a reduction of vacuolar invertase activity by co-suppression in tomato leaves did not affect leaf photosynthetic rates (Scholes et al. 1996) , sucrose cycling may be evident only under pronounced sink-limited conditions. This possibility is strengthened by the demonstration that in tobacco transformed to overexpress yeast invertase in the apoplast or vacuole, source but not sink leaves have bleached and/or necrotic sectors that are associated with reduced photosynthesis (Sonnewald et al. 1991) . Based on diurnal leaf responses, night-time starch metabolism was previously suggested as one component of carbohydrate signalling (Cheng, Moore & Seemann 1998) , and hydrolytic starch degradation also could generate substrate for the hexokinase reaction. Starch-accumulating species are often noted for their susceptibility to down-regulation of photosynthetic capacity under sink-limited conditions (e.g. Goldschmidt & Huber 1992; Farrar & Williams 1991) . In this study, though, we did not observe a close coupling between starch accumulation and photosynthetic acclimation (Fig. 5) . One extension of the idea that starch metabolism is in some aspects a buffer to sucrose metabolism (Stitt 1984) , is the possibility that starch metabolism in part may function to minimize leaf sucrose cycling. From this viewpoint, species that are more active in starch metabolism may be more likely to experience sink-limited growth conditions, as often occurs at high CO 2 .

1986) translational regulation of light-induced ribulose 1,5-bisphosphate carboxylase gene expression in amaranth

Facing the inevitable: plants and increasing atmospheric CO 2

Intraspecific variation of rubisco and rubisco activase protein levels in tomato leaves grown at elevated CO 2 concentration

Effects of short-and long-term treatments at elevated CO 2 on the expression of rubisco genes and leaf carbohydrate accumulation in Arabidopsis thaliana (L.) Henyh

Extraction and purification of RNA from plant tissue enriched in polysaccharides

Separation and characterization of four hexose kinases from developing maize kernels

Differences between wheat genotypes in specific activity of RuBP carboxylase and the relationship to photosynthesis

The allocation of protein nitrogen in the photosynthetic apparatus: costs, consequences, and control

The effects of increased atmospheric carbon dioxide and temperature on carbon partitioning, source-sink relations and respiration

The basis for source-sink interaction in leaves: an hypothesis

Carbon catabolite repression in yeast

Sucrose synthase catalyses a readily reversible reaction in vivo in developing potato tubers and other plant tissues

The effect of CO 2 enrichment and irradiance on the growth, morphology and gas exchange of a C 3 (Panicum laxum) and a C 4 (Panicum antidotale) grass

Regulation of photosynthesis by end-product accumulation in leaves of plants storing starch, sucrose, and hexose sugars

Carbon catabolite repression regulates glyoxylate cycle gene expression in cucumber

Biochemical mechanism for regulation of sucrose accumulation in leaves during photosynthesis

Variation among species in light activation of sucrose-phosphate synthase

Hexokinase as a sugar sensor in higher plants

Sugar sensing in higher plants

Regulation of photosynthesis during leaf development in rbcS antisense DNA mutants of tobacco

Clonal differences in the response of dark and light reactions of photosynthesis to elevated atmospheric CO 2 in poplar

Carbohydrate-modulated gene expression in plants

Regulation of the expression of rbcS and other photosynthetic genes by carbohydrates: a mechanism for the 'sink' regulation of photosynthesis

Carbohydrate synthesis and degradation

Photosynthetic CO 2 assimilation and rising atmospheric CO 2 concentrations

Effect of CO 2 concentration on carbonic anhydrase and ribulose-1,5-bisphosphate carboxylase/oxygenase expression in pea

Influence of plant growth at high CO 2 concentrations on leaf content of ribulose-1,5-bisphosphate carboxylase/ oxygenase and intracellular distribution of soluble carbohydrates in tobacco, snapdragon, and parsley

The effect of elevated partial pressures of CO 2 on the relationship between photosynthetic capacity and N content in rice leaves

The control of leaf photosynthesis rate by the level of assimilate concentration in the leaf: a review of the hypothesis

The effect of elevated [CO 2 ] on growth and photosynthesis of two eucalyptus species exposed to high temperatures and water deficits

A mechanism for intergenomic integration: abundance of ribulose-1,5-bisphosphate carboxylase small-subunit protein influences the translation of the large-subunit mRNA

Acclimation of photosynthesis to elevated CO 2 in five C 3 species

Long-term exposure to elevated [CO 2 ] in a natural Quercus ilex L. community: net photosynthesis and photochemical efficiency of PSII at different levels of water stress

The impact of reduced vacuolar invertase activity on the photosynthetic and carbohydrate metabolism of tomato

A quantification of the significance of assimilatory starch for growth of Arabidopsis thaliana L

Influence of elevated CO 2 and phosphorous nutrition on the growth and yield of a short-duration rice

Feedback control of gene expression

The nucleotide sequence of the tobacco chloroplast gene for the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase

A potential role for RNA turnover in the light regulation of plant gene expression: ribulose-1,5-bisphosphate carboxylase small subunit in soybean

The significance of differences in the mechanisms of photosynthetic acclimation to light, nitrogen and CO 2 for return on investment in leaves

Sugar sensing and sugar-mediated signal transduction in plants

Transgenic tobacco plants expressing yeast-derived invertase in either the cytosol, vacuole or apoplast: a powerful tool for studying sucrose metabolism and sink/source interactions

Degradation of starch in chloroplasts: a buffer to sucrose metabolism

Rising CO 2 levels and their potential significance for carbon flow in photosynthetic cells

Does rubisco control the rate of photosynthesis and plant growth? An exercise in molecular ecophysiology

Sink' regulation of photosynthetic metabolism in transgenic tobacco plants expressing yeast invertase in their cell wall involves a decrease of the Calvin-cycle enzymes and an increase of glycolytic enzymes

Long-term effects of elevated CO 2 and nutrients on photosynthesis and rubisco in loblolly pine seedlings

Direct microdetermination of sucrose

Some relationships between the gas exchange, biochemistry and molecular biology of photosynthesis during leaf development of tomato plants after transfer to different carbon dioxide concentrations

Regulation of the expression of photosynthetic nuclear genes by high CO 2 is mimicked by carbohydrates: a mechanism for the acclimation of photosynthesis to high CO 2

Expression dynamics of the tomato rbcS gene family during development

Acclimation of photosynthetic proteins to rising atmospheric CO 2

Spectrophotometric characteristics of chlorophylls a and b and their pheophytins in ethanol

Photosynthetic acclimation in pea and soybean to high atmospheric CO 2 partial pressure

Soluble acid invertase determines the hexose-to-sucrose ratio in cold-stored potato tubers

We wish to thank Shanti Rawat and Therese Charlet for expert technical assistance and Dianne Stortz-Lintz for setting up and maintaining the CO 2 growth facilities. This work was supported by NSF grant # IBN 1940424 to JRS and S-HC.