The relationships between CO2 assimilation rate, RuP2 carboxylase activity and sizes of the pools of ribulose 1,5-bisphosphate (RuP2) and 3-phosphoglyceric acid (PGA) were examined using a freeze clamp device to rapidly freeze sections of attached leaves of R. sativus which previously had gas-exchange measurements made on them. At high irradiance and ambient partial pressures of CO2 and O2, RuP2 carboxylase was fully active in vivo. Activity was less at very low CO2 pressures and at high CO2 pressures, particularly when combined with low O2 pressures. In vivo RuP2 carboxylase activity and both RuP2 and PGA pool sizes increased with increasing irradiance. RuP2 pool sizes were high at low CO2 pressures and decreased at high CO2 pressures. PGA pool sizes, on the other hand, were low at low CO2 and high at high CO2 pressures. A model of RuP2 carboxylase-oxygenase (Rubisco) kinetics is used to examine the quantitative relationship between in vivo RuP2 carboxylase activity, CO2 assimilation rate and RuP2 and PGA pools. The model predictions fit in vivo data, except at high CO2 pressures, if it is assumed that RuP2 does not bind tightly to the inactive enzyme form in vivo. It is shown that a large fraction of the RuP2 and PGA pools may be chelated by magnesium in the stroma and that the high RuP2 pools (e.g. at low irradiance) may represent an optimal concentration rather than be truly saturating. We conclude that RuP2 pool sizes above Rubisco site concentration do not necessarily indicate a Rubisco limitation of photosynthetic rate.

The catalytic activity of ribulosebisphosphate carboxylase (Rubisco) declined as soon as catalysis was initiated by exposure to its substrate, D-ribulose-1,5-bisphosphate (ribulose-P2). The decline continued exponentially, with a half-time of approximately 7 minutes until, eventually, a steady state level of activity was reached which could be as low as 15% of the initial activity. The ratio of the steady state activity to the initial activity was lower at low C02 concentration and at low pH. The inhibitors 6-phosphogluconate and H202 alleviated the inactivation, increasing the final/initial rate ratio and the half-time. Varying ribulose-P2 concentration in the range above that required to saturate catalysis did not affect the kinetics of inactivation. The affinities for C02 and ribulose-P2 were unaffected by the inactivation. The decline in activity occurred with preparations of ribulose-P2 which contained no detectable D-xylulose-1,5-bisphosphate and also with ribulose-P2 which had been generated enzymatically immediately before use. Inclusion of an aldolase system for removing Dxylulose-1,5-bisphosphate also did not alter the inactivation process. The inactivated Rubisco did not recover after complete exhaustion of ribulose-P2. We conclude that the inactivation is not caused by readily-reversible binding of ribulose-P2 at a site different from the active site and that it is unlikely to be attributable to inhibitory contaminants in ribulose-P2 preparations.Rubisco2 catalyzes the carboxylation and oxygenation of ribulose-P2, thus initiating both of the mutually opposing plant processes of photosynthetic CO2 fixation and photorespiration (3). When assayed in vitro, the activity of purified Rubisco from higher plants decreases with a half-time of several minutes after contact with its substrate, ribulose-P2 (2,5,12,15,17,22,24,26). This loss in activity, to which we have applied the term 'fallover,' is not due to substrate exhaustion or product accumulation (2) and no explanation for this unusual behavior has been established. The phenomenon is displayed to a much lesser degree, ifat all, by cyanobacterial (1) and algal (26) Rubiscos.Proposed explanations for fallover have fallen into three categories: (a) Rubisco becomes catalytically competent only ' Present address: Centre for Molecular Biology and Biotechnology, University of Queensland, St Lucia QLD 4067, Australia. 2Abbreviations: Rubisco, ribulose-P2 carboxylase-oxygenase (EC 4.1.1.39); ribulose-P2, o-ribulose-1 ,5-bisphosphate; P-glycerate, 3-phospho-D-glycerate; xylulose-P2, D-xylulose-1,5-bisphosphate; ribose-P, D-ribose-5-phosphate; glycerol-P, glycerol-3-phosphate. after carbamylation, by C02, of the e-amino group of lys-20 1 at the catalytic site, which enables the catalytically essential divalent metal ion to bind (reviewed by Andrews and Lorimer [3]). There have been many proposals that ribulose-P2 binds more tightly to the uncarbamylated enzyme (E) than to the carbamylated, metal-complexed form (ECM), thus promoting decarbamylation with ...

The inhibition of purified spinach ribulo~bisphosphate carboxylase-oxygenase which occurs progressively during catalysis in vitro is caused by accumulation of at least two tight-binding inhibitors at the catalytic site. Reduction of these inhibitors with NaB3H,, followed by dephosphorylation, produced a mixture of xylitol and arabinitol, thus identifying one of them as D-xylulose 1,Sbisphosphate. It was formed during carboxylation, presumably by a stereochemically incorrect reprotonation of the 2,3-enediolate intermediate bound at the catalytic site. Under the conditions used, this epimerization occurred approximately once for every 400 carboxylation turnovers. Another inhibitor may be 3-keto-D-arabinitoi l$bisphosphate which would also be formed by misprotonation of the enediolate intermediate, but at C-2 rather than at C-3.

The slow inactivation which accompanies catalysis by higherplant ribulose-P2 carboxylase-oxygenase (Rubisco) (2), such reactions could occur at the catalytic site as well as in solution.In this study, we show that a slow, tight-binding inhibitor is, indeed, responsible for fallover. The data are most consistent with the inhibitor being formed as a by-product during catalysis. MATERIALS AND METHODSRubisco2 becomes slowly inactivated during catalysis after exposure to ribulose-P2. This inactivation, or "fallover," continues until eventually a steady rate of catalysis is reached which is substantially less than the initial rate (6). The phenomenon is not a result of substrate depletion or product accumulation and it is worsened by lowering the pH or the CO2 concentration and alleviated by inhibitors which bind at the active site (6). Earlier papers of this series indicated that it is unlikely that fallover is caused by readily reversible, inhibitory binding of ribulose-P2 at a non-catalytic, regulatory site (6) and showed that fallover is not due to decarbamylation of the active-site E-carbamyl-lysyl residue that is required for catalysis (7).One possibility which remains is that fallover is caused by ' Present address:

An investigation was made of the proposal that the slow inactivation of ribulosebisphosphate carboxylase (Rubisco) activity, which occurs during in vitro assays, is due to decarbamylation of the enzyme. The level of carbamylation was compared with catalytic activity during assay conditions in which activity was both increasing and decreasing. Carbamylation level was measured using the reaction-intermediate analogue 2'-carboxy-D-arabinitol-1,5-bisphosphate (carboxyarabinitol-P2). A dual isotope procedure was used in which [3H]carboxyarabinitol-P2 measured total active sites and 14CO2 reported the level of carbamylation.The efficacy of the procedure was verified both in the presence and in the absence of the substrate D-ribulose-1,5-bisphosphate (ribulose-P2). These measurements showed that changes in activity during assays were not correlated with carbamylation status. Inactivation during assays initiated with both fully and partially carbamylated enzyme was not associated with any change in carbamylation level. This implies that the loss of activity during assays is not due to ribulose-P2 binding and sequestering the E form of the enzyme. Ribulose-P2 did not appear to alter the equilibrium between carbamylated and uncarbamylated enzyme, but it did slow the rate at which enzyme was both decarbamylated and carbamylated. The most likely explanation for the loss of activity during assays appears to be the sequestration of carbamylated, Mg2 -bound active sites by an inhibitor. (ECM). This theory has been based on experimental findings which show that phosphorylated sugars, including ribulose-P2, may bind to both the E and ECM forms of the enzyme at the catalytic site (2, 14). When bound to the ECM form, the activating CO2 and Mg2' are stabilized on the enzyme. When bound to the E form, carbamylation is retarded. Depending on the relative affinity ofthese compounds for E and ECM, they may shift the equilibrium between these forms of the enzyme.There have been several indications that ribulose-P2 binds tightly to the E form of Rubisco. For example, Jordan and Chollet (9) measured a dissociation constant for the E-ribulose-P2 complex of 21 nM at 2°C, and Laing and Christeller ( 11) showed that activation of E was very slow in the presence of ribulose-P2. With such tight binding of ribulose-P2 to E, it could readily be expected that there should be a decline in the level of ECM during carboxylation, and conversely, that there should be very little increase in the level of ECM when E is added to ribulose-P2 in the presence of CO2 and Mg2'. In fact, it is thought that the enzyme, Rubisco activase, is required in vivo to promote activation in the presence of ribulose-P2 (20). Contrary to this, there are a number of reports of ribulose-P2 promoting carbamylation of Rubisco. Vater et al. (22) showed that ribulose-P2 induced a decrease in the dissociation constant for the EC complex similar to that seen with sugar phosphate activators such as 6-phosphogluconate, and NADPH (2,5,14). Additionally, Laing and Christelle...