Oxygen Inhibition and Other Properties of Soybean Ribulose 1,5Diphosphate Carboxylase*

SUMMARY n-Ribulose-l , 5-di-P carboxylase, purified from soybeans, had K, values of 0.13 XnM for COS and 0.19 mM for ribulose-1,5-di-P under a Nz atmosphere. O2 inhibited 14C02 incorporation by the enzyme and this inhibition was rapidly reversed by Ns. Inhibition was competitive with respect to CO2 and uncompetitive with respect to ribulose-1,5-di-P. The Ki for O2 was 0.8 mM. This O2 inhibition, together with the ribulose-1,5-di-P carboxylase-catalyzed oxidation of ribu-lose-l ,5-di-P P-glycolate observed Biochem. Biophys. Res. Commun. 716-722), explains the “Warburg effect”: the rapidly reversible O2 inhibition of photosynthesis and stimulation of glycolate production seen in

45, 716-722), explains the "Warburg effect": the rapidly reversible O2 inhibition of photosynthesis and stimulation of glycolate production seen in plants and isolated chloroplasts.
These data may explain the different response to O2 by plants which utilize ribulose-1,5-di-P carboxylase for the initial photosynthetic carboxylation and those utilizing P-enolpyruvate carboxylase. The optimum temperature for purified ribulose-1,5-di-P carboxylase was 55" and activation energies, in kilocalories per mole, were 18.4 in Nt and 20.4 in OZ.
Phosphorylated compounds inhibited 14C02 incorporation by the enzyme.
Fructose-1,6-di-P was a competitive inhibitor with respect to ribulose-l,S-di-P, the Ki being 0.88 mM. Fructose-l, 6-di-P was a more effective inhibitor than fructose-6-P, fructose-l-P, and ribulose-5-P suggesting that both phosphate groups of ribulose-1,5-di-P are involved in binding to the enzyme.
HgCl, was a noncompetitive inhibitor with respect to COZ and a mixed inhibitor with respect to ribulose-1,5-di-P, suggesting that sulfhydryl groups are not involved in COZ binding but may be close to the site where ribulose-1,5-di-P binds to the enzyme. n-Ribulose-l , 5-di-P carboxylase (EC 4.1.1.39), the predominant component of Fraction I protein, is associated with the chloroplast and comprises up to 50% of the soluble protein in leaves (1). This enzyme, purified from spinach (2), catalyzes the reaction D-ribulose-1,5-di-P f CO, + 2 3.n-P-glycerate Evidence (3,4) indicates this enzyme is the rate-limiting step in light-saturated photosynthesis of Calvin cycle plants (C, species). Bn important factor in the photosynthesis of Ca species is the reversible O2 inhibition of photosynthetic COs fixation, termed the Warburg effect. Many schemes have been proposed to account for this inhibition (5,6), however, elucidation of the mechanism has been complicated by another phenomenon, photorespiration, which also occurs in Ca species. Photorespiration is characterized by a light-and OS-dependent efflux of CO2 this efflux being inhibited by increasing CO2 concentrations (3,7). These two phenomena are re-examined in view of the properties observed with purified soybean ribulose-1 ,5-di-P carboxvlase.
In certain plants another CO2 fixation pathway occurs (8). These plants, termed Cd dicarboxylic acid cycle or C4 species, utilize P-enolpyruvate carboxylase for the initial photosynthetic carboxylation.
However, ribulose-1 ,5-di-P carboxylase is also important in the photosynthesis of C4 species as carbon flows through this enzyme to carbohydrate (9). 02 levels below atmospheric do not inhibit the photosynthesis of Cd species and these species lack photorespiratory COa release (3,7). The different response to 02 by Cd and C& species was investigated with a Cd plant, corn, and a Cs plant, soybean.
Purified ribulose-1 ,5-di-P carboxylase has a reasonably high affinity for ribulose-1,5-di-P, the K, being 0.12 mM (10). Until recently the enzyme was considered to possess a low affinity for HCOe3 in vitro, with reports of K, values ranging from 2.5 to 30 mM (1). However COZ, and not HCO,, is the reactive species  (15,17), although whether they are an integral part of the catalytic site is still unresolved.
Part of the work reported here attempts to clarify some problems associated with the role of sulfhydryl groups and the mechanism of substrate binding to the enzyme.
A preliminary report of part of this work has appeared (18).

EXPERIMENTAL PROCEDURE
Purz&ztion-Ribulose-1 ,5-di-P carboxylase was purified from leaves of field-or greenhouse-grown soybeans (Glycine max (L.) Merrill var. Wayne) by a modification of a method described for spinach (10). Extraction difficulties with soybean necessitated two homogenization steps, with the filtrates being combined. Overnight storage at the ammonium sulfate I stage reduced activity, therefore purification was continued through to where the enzyme could be stored at -20".
The isolated enzyme was stored at 5" as a precipitate in 55% saturated ammonium sulfate, pH 6.5, with 0.1 mM EDTA and 5.0 mM 2-mercaptoethanol.
Prior to assay, an aliquot of the suspension was centrifuged at 10,000 X g for 10 min, the supernatant discarded, and the precipitated enzyme dissolved in 120 mM Tris buffer, pH 8.0, containing 0.25 mM EDTA, and 10.0 rnM MgC12.
After dissolving, enzyme activity gradually increased by up to 60%. Thus to achieve maximum activity the enzyme solution was kept at 2" for 4 hours before use.
Stored enzyme lost activity over 2 months, so control values in different experiments varied.
Although absolute values between experiments could not be compared, repetition of experiments with different preparations and enzyme ages showed that findings were comparable when t'he control values were taken into account.
The specific activity was lower than that reported for spinach (lo), being in the range of 10 to 100 nmoles of COZ fixed per min per mg of protein.
This can be partially attributed to inactivation during isolation and to the use of lower assay concentrations of ribulose-1 ,5-di-P and NaH14C03. Soluble protein in extracts was determined by the procedure of Lowry et al. (19) and in purified enzyme solutions by the method of Warburg and Christian (20). 1. Effect of ribulose-1,5-di-P concentration on the activity of soybean ribulose-1,5-di-P'carboxylase in atmospheres of Ni, Con-free air, and 02. The assay conditions were as described under "Experimental Procedure." Assay Procedures-Ribulose-l , 5-di-P carboxylase activity was assayed by 14C02 incorporation into acid stable products. At 5.0 mM MgC12, the pH optimum was 8.0. The reaction vessels contained 50.0 mM Tris at pH 8.0, 5.0 mM MgC12, 0.1 mM ribulose-1 ,5-di-P, and 0.05 mM EDTA in a final volume of 1 .O ml. The vessels were flushed with Nz, COz-free air or O2 and shaken for 15 min, then were sealed and 20.0 mM NaH14C03 (1.0 PCi) injected via a serum cap. Cot-free air was used to avoid isotope dilution by atmospheric COZ. The reaction was initiated by injection of the enzyme, or alternatively ribulose-1 ,5di-P, and stopped with 0.1 ml of 6.0 M acetic acid after 4 min at 25". Aliquots were then taken and dried at 90", dissolved in a modified Bray's solution (21)  The extracts were centrifuged at 35,000 x g for 15 min in capped tubes under Nz, and then assayed immediately for P-enolpyruvate carboxylase activity (22). The P-enolpyruvate carboxylase assay contained, in a final volume of 1.0 ml, 50.0 mM Tris at pH 8.0, 10.0 mM MgC&, 0.1 mrvr EDTA, 5.0 mM sodium glutamate, 5.0 mM NaH14C03 (2.0 PCi), and 2.0 mM P-enolpyruvate. The crude ribulose-1 ,5-di-P carboxylase assay mixture was similar except sodium glutamate was omitted and 0.2 mM ribulose-1,5-di-P substituted for P-enolpyruvate.
Both the P-enolpyruvate and ribulose-1 ,5-di-P carboxylase assays were initiated with 0.1 ml of the extract and stopped with 0.1 ml of 6.0 M acetic acid after 3 min at 25" under atmospheres of N:! or OZ.
In crude and purified enzyme assays, reaction rates were linear over the assay period employed.
No prior incubation of ribulose-1,5-di-P carboxylase with Mg2+ and HCO;-was required for linear rates.
Kinetic Studies of O2 Inhibition-Further investigation showed the 0% inhibition was fully and rapidly reversible (Table I). Prior to initiation of the reaction with ribulose-1 ,5-di-P, flushing the reaction mixture and enzyme with Nn for 6 min followed by O2 for 6 min, or vice versa, produced inhibition only if the last gas treatment was OZ. Thus the inhibitory effect of O2 was not due to permanent inactivation of the enzyme. A double reciprocal plot of the inhibitory effect of 02 showed that inhibition was competitive with respect to CO2 (Fig. 2). The K, value was 0.13 mM in Ns, 0.19 mM in 21yG 02 (Con-free air), and 0.35 mM in 1007, OZ. The K, in COz-free air is within the range of values reported for spinach carboxylase (I), when all values are calculated on the basis of CO2 rather than HC03 concentration.
The Ki for O2 was 0.8 mM, which is slightly lower than previously reported (4). The competitive nature of With respect to ribulose-1 ,5-d&P, O2 produced mixed inhibition of the uncompetitive type (Fig. 3). The K, for ribulose-1,5-di-P was 0.19 rnnt in Ne, which compares with reported spinach carboxylase values in air ranging from 0.12 to 0.7 mM (1). One enzyme preparation, stored in an intermediate stage of purification, exhibited variation from these K, values, although in other respects its properties were unaltered.
The reduced rate of 14C02 incorporation observed in COz-free air could be largely overcome by GSH. The rate of 14C02 incorporation at 20 mM HC03, measured as nanomoles per min per mg of protein, was reduced from 25.5 in Nz to 21.9 in COz-free air, and the addition of 3.0 mM GSH increased the rate to 24.3.

Measurements
with an O2 electrode showed that GSH reduced the 02 concentration in a reaction mixture under COz-free air. Thus the protection afforded by GSH can be attributed to a decrease in the 02 concentration of the reaction mixture, through the aut'oxidation of GSH, and not to an effect on enzyme sulfhy-dry1 groups.
E$ect of O2 on Carboxylases in Corn and Soybean Leaf Extracts-14C02 incorporation by P-enolpyruvate carboxylase in extracts of both corn and soybean was unaffected by 02 (Table II).
Repeating the experiment with 20.0 mM NaHW03 produced similar results, except the inhibition of ribulose-I ,5-di-P carboxylase was less at the higher CO2 level.
Temperature Response of Ribulose-i ,5-&P Carboxylase Activify in Nz and O-Over a X-min assay period purified ribulose-1,5-di-P carboxylase showed maximum activity in both N, and O2 at 55" (Fig. 4). Activity was greatly reduced above 60", due to rapid inactivation of the enzyme. The percentage inhibition by O2 when calculated on a millimolar O2 basis, to allow for lower O2 solubility at higher temperatures, varied only 6% over the temperature range 15-60" (Fig. 4) 2. Double reciprocal plot of the rate of '4CO2 incorporation by soybean ribulose-1,5-di-P carboxylase as a function of COT concentration in NP, COz-free air, and 0,. The general assay condi- 3. Double reciprocal plot of the rate of '4CO2 incorporation by soybean ribulose-l,5-di-P carboxylase as a function of ribulose-1,5-di-P concentration, under atmospheres of Nz, 02, and under NP with 1.0 mM fructose-1,64-P present.
The general assay conditions were as described under "Experimental Procedure" except the addition of 0.05 to 0.60 mM ribulose-1,5-di-P initiated the react,ion. 4. Activity and O:! inhibition of soybean rihldosc-1,5-di-1' carboxylase as a function of temperature.
The general assay COW ditions were as described under "ISxperimental Procedure." The reaction flasks were incubated at temperatures ranging from 15~65" while being shaken and flushed \vith N? or O2 prior to sealing and initiation of the reaction with enzyme. To compensate for reduced O2 solubility at higher temperntllres, the percentage inhibition of activity in OZ, as compared to N,, was plotted on the basis of the calculated 0, concentration in the reaction mixture.

in OZ. The figure for C&free
air is comparable to 16.9 kcal per mole calculated for spinach carboxylase (2).
I;nhibitor &u&es-The product of the carboxylnsc react,ion, 3-P-glycerate, is an inhibitor of spinach carboxylase (10). The results for soybean carboxylase, in Table III, confirm the inhibitory nature of 3-P-glycerate.
All phosphate compounds tested produced inhibition, but fructose-1,6-di-I' \Tas the most effective. A double reciprocal plot of inhibition by 1.0 mM fructose-1,6&P with respect to ribulose-1 , M-P showed competitive inhibit'ion ( Fig.  3), with a Ki of 0.88 mniI in Nz. This supports t,he proposal that ribulose-1 ,5-di-P binds to the enzyme through one or both phosphate groups (10). As fructose-1,5-di-P inhibit,ed to a greater extent than the monophosphates fructose-l-P, fructose-6-I', and ribulose-5-P (Table III) it is likely that both phosphate groups are involved in the binding.
Sucrose, glucose, fructose, ribulosc, glycolate, and glyoxylate had no appreciable inhibitory effect, thus it would seem that the phosphate groups of ribulosc-1 ,5-di-P are the principal groups involved in its binding to the enzyme. None of the sugar phosphates tested for inhibition were able to substitute for ribulose-1 ,5&P as a 14C02 acceptor. NAD', NADH, NADP+, NADPH, ADP, and ATP all caused inhibition of the reaction, probably due to the phosphate groups (Table III).
The products of the photosynthetic light reactions, ATP and NADPH, were the most effective nucleotide inhibitors. Acetate, formate, and CSZ, substances resembling COZ, had no inhibitory effect. The photorespiration inhibitor a-hydroxy-2-pyridinemethanesulfonate inhibited soybean carboxylase by 300/', at 10.0 mM.
Incubation for 5 min with 0.01 mM HgC12 reduced the rate of W02 incorporation from 22.6 (without HgC1.J to 10.9 nmoles of COa per min per mg of protein, while 15.min incubation furbher reduced the rate to 8.3. The The general assay conditions were AS described under "Experimental Procedure." Letters in parentheses refer to different experiments. Contjrol rates of 'Y302 incorporation (rranomolrs of COP per min per mg of protein) for the different experiments were: 5. Rate of W02 incorporation by soybean riblllosc-1,5di-P carboxylase as a function of the logarithm of HgCl, concerltration.
The assay conditions were as described under "Experimental Procedure." Assays were initiated with enzyme, withollt prior incubation in HgCl?.
This suggests that the effect of the Hg2+ ions is on sulfhydryl groups of the enzyme.
With respect t,o COZ, HgClz was a noncompetitive inhibitor of Procedure" and the legend to Fig. 6. The addit,ion of 0.01 to 0.50 mM ribnlose-1,5-di-P init,iated t,he reaction.
soybean carboxglnse (Fig. 6)) indicating that sulfhydryl groups are not at the site of CO2 binding. With respect to ribulose-l,5di-I', IIgCl? showed mixed inhibition (Fig. 7) suggesting that Hg* acts on sulfhydryl groups adjacent to the site of ribulose-1,5-di-I' binding.

Oxygen E@cts-The
Warburg effect is manifested by 02 inhibition of phot,osynthesis in Cs species and a concomitant 02 stimulation of glycolate production. In leaves of higher plants, the glycolate produced is subsequently metabolized in part to CO, by the photorespiratory pathway (23). These 0, effects are rapidly reversed by removing O2 or by increasing the COZ level (5, 6, 24). Soybean ribulose-1 ,5-di-P carboxylase is reversibly and competitively inhibited by 02 with respect to CO2 (Table 1, Fig. 2) and this enzyme catalyzes an oxygen-dependent oxidation of ribulose-1 ,5-di-P to 2-I'-glycolate (25), a glycolate precursor (23). Thus the effects of 02 on purified ribulose-1,5di-P carboxylase mimic the effects of 02 in the Warburg effect, strongly suggesting that ribulose-1 ,5-di-P carboxylase is the specific site of action of 02 in the Warburg effect.
It has been suggested (5) that 02 may reversibly inactive sulfhydryl groups in enzymes of the Calvin cycle, with reactivation being induced by natural reductants.
There is no evidence that the oxidation of these enzymes is reversible, and sulfhydryl oxidation cannot explain the competitive nature of O2 and COz in the Warburg effect. In contrast, 02 inhibition of ribulose-1,5-di-P carboxylase is not an oxidation of sulfhydryl groups. This can be deduced from the observations that HgZ+, a sulfhydryl inhibitor, is a noncompetitive inhibitor of carboxylase with respect to COZ, whereas 02 is a competitive inhibitor with respect to COz. Unlike NAnI?+-glyceraldehyde-3-P dehydrogenase and ribulose-5-P kinase, extractable ribulose-1 ,5-d&P carboxylase activity is not in excess of the photosynthetic rate in the plant,.
Any contribution to the Warburg effect by the dehydrogenase and kinase appears to be small, especially if ribulose-1 ,5-di-P carboxylase limits photosynthetic rate (3,4). In contrast to the marked inhibition of Ca photosynthesis by O2 in air, Cq photosynthesis is affected only by 02 concentrations considerably higher than atmospheric (22,28,29). In species with Cq photosynthesis, including corn, the initial photosynthetic carboxylation is catalyzed by P-enolpyruvate carboxylase in the mesophyll.
The oxalacetate produced is reduced to malate, transported to the bundle sheath, and decarboxylated. The COZ released is refixed by ribulose-1,5-di-P carboxylase and enters the CB cycle (8, 9). P-Enolpyruvate carboxylase may function as a COz pump, increasing the CO2 concentration in the bundle sheath to a level higher than atmospheric (30).
In corn extracts, P-enolpyruvate carboxylase is unaffected by oxygen while ribulose-1 ,5-di-P carboxylase is inhibited (Table  TT), Thus in C4 photosynthwis the initial carhoxylation is insensitive to 02, while the second carboxylation, transfer of CO2 into the Calvin cycle, is inhibited by Oz. At atmospheric 02 levels, cycling of CA photosynthesis may proceed unhindered since the initial carboxylation is not inhibited and the increased COz level in the bundle sheath will allow COa to compete more successfully with O2 for ribulose-1,5-di-P carboxylase. At 02 levels greater than atmospheric, inhibition of ribulose-1 ,5-di-P carboxylase would restrict the carbon fl.ow from C4 acids to the sugar phosphates of the Calvin cycle. Since P-enolpyruvate carboxylase is not affected by 02, Cd acids may accumulate at high O2 in concentrations sufficient to inhibit P-enolpyruvate carboxylase (31), and thus inhibit photosynthesis.
Since the mesophyll layers surround the bundle sheath, P-enolpyruvate carboxylase in the mesophyll would rapidly refix any photorespiratory CO2 and prevent leakage to the atmosphere (9).
In Mimulus, energy of activation values for photosynthesis of