The Effects of Chilling in the Light on Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase Activation in Tomato (Lycopersicon esculentum Mill.).

Photosynthesis rate, ribulsoe-1,5-bisphosphate carboxylase/oxygenase (Rubisco) activation state, and ribulose bisphosphate concentration were reduced after exposing tomato (Lycopersicon esculentum Mill.) plants to light at 4[deg]C for 6 h. Analysis of lysed and reconsituted chloroplasts showed that activity of the thylakoid membrane was inhibited and that Rubisco, Rubisco activase, and other soluble factors were not affected. Leaf photosynthesis rates and the ability of chilled thylakoid membranes to promote Rubisco activation recovered after 24 h at 25[deg]C. Thylakoid membranes from control tomato plants were as effective as spinach thylakoids in activating spinach Rubisco in the presence of spinach Rubisco activase. This observation is in sharp contrast to the poor ability of spinach Rubisco activase to activate tomato Rubisco (Z.-Y. Wang, G.W. Snyder, B.D. Esau, A.R. Portis, and W.L. Ogren [1992] Plant Physiol 100: 1858-1862). The ability of thylakoids from chilled tomato plants to activate Rubisco in the assay system was greatly inhibited compared to control plants. These experiments indicate that chilling tomato plants at 4[deg]C interferes with photosynthetic carbon metabolism at two sites, thioredoxin/ferredoxin reduction (G.F. Sassenrath, D.R. Ort, and A.R. Portis, Jr. [1990] Arch Biochem Biophys 282: 302-308), which limits bisphosphatase activity, and Rubisco activase, which reduces Rubisco activation state.

Localization of light-inducible and tissue-specific regions of the spinach ribulose bisphosphate carboxylase/oxygenase (rubisco) activase promoter in transgenic tobacco plants.

Deletions in the spinach rubisco activase (Rca) promoter in transgenic tobacco were analyzed to define the regions necessary for conferring light-inducible and tissue-specific expression. Transgenic plants were constructed with Bal 31 deletions of the Rca promoter fused to the coding region of the bacterial reporter gene beta-glucuronidase (GUS). Analysis of the Rca deletion mutants localized the region conferring normal expression downstream from -294 relative to the Rca transcription start site. A second set of transgenic plants containing the cauliflower mosaic virus (CaMV) 35S enhancer fused to the 3' end of the Rca/GUS constructs demonstrated the presence of a light-responsive element between -150 and -78 active in leaves. Regions 10 bp long within the light-responsive region, which included putative G box and GT elements, were removed by recombinant polymerase chain reaction. Deletion of the G box element resulted in a loss of gene expression in the leaves of transgenic tobacco, while deletion of the GT motif caused a 10-100-fold increase in expression in roots. However, site-directed mutagenesis of the GT motif resulted in expression patterns identical to the normal promoter. These experiments demonstrated that light-inducible and tissue-specific expression of the Rca promoter involves multiple cis elements proximal to the transcription start site, and that interactions between these elements are essential for regulating expression.

Phosphoglycolate phosphatase: purification and preparation of antibodies.

Phosphoglycolate phosphatase was partially purified from leaves of Nicotiana rustica using ion exchange and chromatofocusing columns. The native molecular weight of the enzyme was determined to be about 58 kD from Ferguson plots, with a subunit size of about 32 kD. The native enzyme is thus likely to be a dimer. A polyclonal antibody prepared against the LDS denatured enzyme cross reacted with proteins from Nicotiana tabacum, Glycine max, Spinacea oleracea and Arabidopsis thaniana. There was little or no reaction with an Arabidopsis mutant lacking phosphoglycolate phosphatase activity, indicating a much reduced level of phosphoglycolate phosphatase protein in the mutant.

Alteration of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase activase activities by site-directed mutagenesis.

Site-directed mutagenesis was performed on the 1.6 and 1.9 kilobase spinach (Spinacea oleracea) ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) activase cDNAs, encoding the 41 and 45 kilodalton (kD) isoforms of the enzyme, to create single amino acid changes in the putative ATP-binding site of Rubisco activase (Lys-107, Gln-109, and Ser-112) and in an unrelated cysteine residue (Cys-256). Replacement of Lys-107 with Met produced soluble protein with reduced Rubisco activase and ATPase activities in both isoforms. Substituting Ala or Arg for Lys-107 produced insoluble proteins. Rubisco activase activity increased in the 41-kD isoform when Gln-109 was changed to Glu, but activity in the 45-kD isoform was similar to the wild-type enzyme. ATPase activity in the Glu-109 mutations did not parallel the changes in Rubisco activase activity. Rather, a higher ratio of Rubisco activase to ATPase activity occurred in both isoforms. The mutation of Gln-109 to Lys inactivated Rubisco activase activity. Replacement of Ser-112 with Pro created an inactive protein, whereas attempts to replace Ser-112 with Thr were not successful. The mutation of Cys-256 to Ser in the 45-kD isoform reduced both Rubisco activase and ATPase activities. The results indicate that the two activities of Rubisco activase are not tightly coupled and that variations in photosynthetic efficiency may occur in vivo by replacing the wild-type enzyme with mutant enzymes.

The mechanism of Rubisco activase: Insights from studies of the properties and structure of the enzyme.

Rubisco, the primary carboxylating enzyme in photosynthesis, must be activated to catalyze CO2 fixation. The concept of an 'activase', a specific protein for activating Rubisco, was first introduced in 1985 based largely on biochemical and genetic studies of a high CO2-requiring mutant of Arabidopsis (Salvucci et al. (1985) Photosynth Res 7: 193-201). Over the past ten years, details about the occurrence, structure, and properties of Rubisco activase have been elucidated. However, the mechanism of action of Rubisco activase remains elusive. This review discusses the need for and function of Rubisco activase and summarizes information about the properties and structure of Rubisco activase. The information is evaluated in the context of the mechanism of Rubisco activase.

Species variation in kinetic properties of ribulose 1,5-bisphosphate carboxylase/oxygenase.

Several kinetic parameters of ribulose-1,5-bisphosphate (RuBP) carboxylase/oxygenase from different species were measured and compared. The CO2/O2 specificity (VcKo/VoKc) was found to be about 80 in the enzymes from several C3 species and two C4 species. Specificity values of 58 and 70, respectively, were found in enzymes from the C4 plants Setaria italica and Sorghum bicolor. Two enzymes from cyanobacteria had values of about 50. Substitution of Mn2+ for Mg2+ reduced the CO2/O2 specificity by a factor of about 20 for all enzymes except that of Rhodospirillum rubrum, which was reduced by a factor of 10. Values for KMg2+(apparent) measured at 102 microM CO2 were found to vary by a factor of 8 between different RuBP carboxylase/oxygenase enzymes. Enzymes with high KMg2+(apparent) values generally had high Michaelis constants for CO2. The rate of CO2/Mg2+ activation was inhibited by RuBP in all enzymes, although the concentration of RuBP required to inhibit activation in the enzyme from the cyanobacterium Aphanizomenon flos-aquae was increased by an order of magnitude compared to other higher plant structural-type enzymes. The wide variation found in the kinetic properties of RuBP carboxylase/oxygenase isolated from diverse species appears to be determined in part by past evolutionary pressures and the present physicochemical environment in which the enzyme functions.

Photorespiration mutants of Arabidopsis thaliana deficient in serine-glyoxylate aminotransferase activity.

Three mutants of the crucifer Arabidopsis thaliana (Linnaeus) Heynhold were isolated that are completely lacking in activity catalyzed by serine-glyoxylate aminotransferase (EC 2.6.1.45), a peroxisomal enzyme involved in photorespiratory carbon metabolism. These mutants were viable and exhibited normal photosynthesis under conditions that suppressed photorespiration, but they were inviable and photosynthesized at greatly reduced rates under conditions that promoted photorespiration. Serine and glycine accumulated as end products of photosynthesis in the mutants, mostly at the expense of starch and sucrose. The mutants are allelic, and the segregation patterns of plant viability, photosynthetic activity, and enzyme activity in the F(1) and F(2) generations indicated that all the observed effects were caused by a single recessive nuclear mutation. This conclusion was confirmed by the isolation of seven revertants in which viability, photosynthetic capacity, and enzyme activity were simultaneously restored. Mutants of the type described here, in which photorespiration is changed from a merely wasteful process into one that is lethal, may permit the direct selection of secondary mutations that reduce photorespiration.

A Sensitive Assay Procedure for Simultaneous Determination of Ribulose-1,5-bisphosphate Carboxylase and Oxygenase Activities.

A sensitive assay procedure is described for the simultaneous determination of ribulose-1,5-bisphosphate (RuBP) carboxylase and oxygenase activities. In this assay, [1-(3)H]RuBP is incubated with (14)CO(2) and O(2). Carboxylation rate is determined from (14)CO(2) incorporation and oxygenation rate is determined from [2-(3)H]glycolate-phosphate production. The assay was found to be suitable at all CO(2) and O(2) concentrations examined, which ranged from 0 to 300 micromolar CO(2) (20 millimolar NaHCO(3)) and 0 to 1.15 millimolar (100%) O(2). In combination with a polarographic assay, the stoichiometry of the RuBP oxygenase reaction was found to be RuBP-O(2)-glycolate phosphate-glycerate phosphate (1:1:1:1).The ratio of soybean (Merr.) RuBP carboxylase and oxygenase activities was measured at various CO(2) and O(2) concentrations and in the presence of several compounds which have been reported to alter differentially the rate of the two reactions. The ratio of the two activities was a linear function of the ratio of the CO(2) and O(2) concentrations. The substrate specificity factor V(c)K(o)/V(o)K(c), which determines relative carboxylase and oxygenase reaction rates as a function of the CO(2) and O(2) concentrations, was found to be 73 in the presence of Mg(2+). Of the several compounds which have been reported to alter differentially the two reactions, we found that only Mn(2+) substitution for Mg(2+) was effective. Compared to Mg(2+), Mn(2+) reduced the K(m)(O(2)) from 690 to 36 micromolar O(2) and reduced the specificity factor to about 4. The K(m)(CO(2)) was about 20 micromolar CO(2) in the presence of both Mg(2+) and Mn(2+). Comparison of reaction rates in the presence of activated and inactivated enzyme allowed a direct determination that both carboxylase and oxygenase activities are similarly activated by CO(2), with an activation equilibrium constant of about 1.3 millimolar NaHCO(3) (27 micromolar CO(2)) at pH 7.85 and 10 millimolar Mg(2+).

Photorespiration-deficient Mutants of Arabidopsis thaliana Lacking Mitochondrial Serine Transhydroxymethylase Activity.

Three allelic mutants of Arabidopsis thaliana which lack mitochondrial serine transhydroxymethylase activity due to a recessive nuclear mutation have been characterized. The mutants were shown to be deficient both in glycine decarboxylation and in the conversion of glycine to serine. Glycine accumulated as an end product of photosynthesis in the mutants, largely at the expense of serine, starch, and sucrose formation. The mutants photorespired CO(2) at low rates in the light, but this evolution of photorespiratory CO(2) was abolished by provision of exogenous NH(3). Exogenous NH(3) was required by the mutants for continued synthesis of glycine under photorespiratory conditions. These and related results with wild-type Arabidopsis suggested that glycine decarboxylation is the sole site of photorespiratory CO(2) release in wild-type plants but that depletion of the amino donors required for glyoxylate amination may lead to CO(2) release from direct decarboxylation of glyoxylate. Photosynthetic CO(2) fixation was inhibited in the mutants under atmospheric conditions which promote photorespiration but could be partially restored by exogenous NH(3). The magnitude of the NH(3) stimulation of photosynthesis indicated that the increase was due to the suppression of glyoxylate decarboxylation. The normal growth of the mutants under nonphotorespiratory atmospheric conditions indicates that mitochondrial serine transhydroxymethylase is not required in C(3) plants for any function unrelated to photorespiration.

An Arabidopsis thaliana mutant defective in chloroplast dicarboxylate transport.

Reactions of the photorespiratory pathway of C(3) plants are found in three subcellular organelles. Transport processes are, therefore, particularly important for maintaining the uninterrupted flow of carbon through this pathway. We describe here the isolation and characterization of a photorespiratory mutant of Arabidopsis thaliana defective in chloroplast dicarboxylate transport. Genetic analysis indicates the defect is due to a simple, recessive, nuclear mutation. Glutamine and inorganic phosphate transport are unaffected by the mutation. Thus, in contrast to previous reports for pea and spinach, glutamine uptake by Arabidopsis chloroplasts is mediated by a transporter distinct from the dicarboxylate transporter. Both the inviability and the disruption of amino-group metabolism of the mutant under photorespiratory conditions suggest that the primary function of the dicarboxylate transporter in vivo is the transfer of 2-oxoglutarate and glutamate across the chloroplast envelope in conjunction with photorespiratory nitrogen metabolism. The role commonly ascribed to this transporter, conducting malate-aspartate exchanges for the indirect export of reducing equivalents from the chloroplast, appears to be a minor one.

Rapid recovery of chloroplast mutations affecting ribulosebisphosphate carboxylase/oxygenase in Chlamydomonas reinhardtii.

Based on the unique ability of chloroplast genes to recombine in Chlamydomonas reinhardtii, a collection of acetate-requiring mutants was screened for recombination with a mutation affecting ribulose-1,5-bisphosphate carboxylase/oxygenase [Rbu-1,5-P(2) carboxylase/oxygenase; 3-phospho-D-glycerate carboxy-lyase (dimerizing), EC 4.1.1.39]. This chloroplast mutation, rcl-u-1-10-6C, causes the absence of Rbu-1,5-P(2) carboxylase/oxygenase activities and alters the isoelectric point of the larger subunit. Several mutants that displayed little or no recombination with 10-6C were recovered, and two lacked carboxylase activity. These new chloroplast mutants lack both large and small Rbu-1,5-P(2) carboxylase/oxygenase subunits. The approach demonstrated here permits the routine recovery of chloroplast mutations affecting this enzyme. Multiple mutations in the Rbu-1,5-P(2) carboxylase/oxygenase large-subunit gene can be used to investigate the function and regulation of this enzyme and the regulation of chloroplast genes in general.

Nuclear Suppressors of the Photosensitivity Associated with Defective Photosynthesis in Chlamydomonas reinhardii.

Several nuclear mutations were recovered that suppress the photosensitivity associated with the Chamydomonas reinhardii chloroplast mutant rcl-u-1-10-6C, which is defective in ribulose-1,5-bisphosphate carboxylase/oxygenase. Two of the suppressor mutations affect other components of photosynthesis. These results show that suppressors of photosensitivity are sufficiently common to permit the recovery of photosensitive, photosynthesis-deficient mutants in bright light, and indicate that photosynthesis-deficient mutants selected and maintained in the light may accumulate suppressors which can confuse the biochemical analysis of lesions in photosynthesis. One of the suppressor mutations inhibits photosystem II activity, indicating that photosensitivity can be mediated by partial reactions of the photosynthetic electron transport chain.

Photorespiratory-induced senescence of plants under conditions of low carbon dioxide.

A light-induced senescence of soybeans at low carbon dioxide concentration has been characterized and ascribed to photorespiration. It is suggested that this system can be used to screen large numbers of plants for photorespiratory-deficient individuals of photorespiratory species. Optimal conditions for this senescence include a low CO(2) concentration, a temperature near 34 degrees , high light intensity, and a high oxygen concentration.

Mutants of the cruciferous plant Arabidopsis thaliana lacking glycine decarboxylase activity.

A mutant of Arabidopsis thaliana (L.) Heyn. (a small plant in the crucifer family) that lacks glycine decarboxylase activity owing to a recessive nuclear mutation has been isolated on the basis of a growth requirement for high concentrations of atmospheric CO2. Mitochondria isolated from leaves of the mutant did not exhibit glycine-dependent O2 consumption, did not release 14CO2 from [14C]glycine, and did not catalyse the glycine-bicarbonate exchange reaction that is considered to be the first partial reaction associated with glycine cleavage. Photosynthesis in the mutant was decreased after illumination under atmospheric conditions that promote partitioning of carbon into intermediates of the photorespiratory pathway, but was not impaired under non-photorespiratory conditions. Thus glycine decarboxylase activity is not required for any essential function unrelated to photorespiration. The photosynthetic response of the mutant in photorespiratory conditions is probably caused by an increased rate of glyoxylate oxidation, which results from the sequestering of all readily transferable amino groups in a metabolically inactive glycine pool, and by a depletion of intermediates from the photosynthesis cycle. The rate of release of 14CO2 from exogenously applied [14C]glycollate was 14-fold lower in the mutant than in the wild type, suggesting that glycine decarboxylation is the only significant source of photorespiratory CO2.

A novel role for light in the activation of ribulosebisphosphate carboxylase/oxygenase.

Light stimulated the activation of ribulosebisphosphate carboxylase/oxygenase (rubisco) in a buffered lysed chloroplast system in the presence of saturating concentrations of ATP. This indicates a role for light in the rubisco activase activation system in addition to the previously identified requirement for the synthesis of ATP. Rubisco activation was nearly as great at low irradiance (10 micromoles of photons per square meter per second) as at high irradiance (1000 micromoles of photons per square meter per second). Light stimulation of activation occurred at both low bicarbonate (equivalent to air levels of CO(2)) and high bicarbonate (10 mm) concentrations. Light activation was inhibited by DCMU and glyoxylate. Methyl viologen did not inhibit light activation, and dithiothreitol did not stimulate activation in the dark, indicating that the ferredoxin/thioredoxin system was not involved. Following a transition of the lysed chloroplasts from light to dark, the light-dependent increase in activation ceased immediately. The experiments were conducted with chloroplasts from spinach (Spinacea oleracea L.), a species which was previously shown not to contain the endogenous inhibitor of rubisco, 2-carboxyarabinitol 1-phosphate. Assays of total rubisco activity in the light and dark confirmed the absence of such a tight binding inhibitor of activity. The observations reported here cannot be explained by current hypotheses of the role of light in rubisco activation and demonstrate that in addition to providing ATP needed for rubisco activase activity, at least one other light-dependent reaction is required for regulating the activation state of rubisco in vivo.

Chlamydomonas reinhardtii Phosphoribulokinase : Sequence, Purification, and Kinetics.

The sequence and kinetic properties of phosphoribulokinase purified from Chlamydomonas reinhardtii were determined and compared with the spinach (Spinacea oleracea) enzyme. Chlamydomonas phosphoribulokinase was purified to apparent homogeneity, with a specific activity of 410 micromoles per minute per milligram. Polyclonal antibodies to the purified protein were used to isolate a Chlamydomonas cDNA clone, which, upon sequencing, was found to contain the entire coding region. The transit peptide cleavage site was determined by Edman analysis of the mature protein. The precursor protein consists of a 31 amino acid transit peptide and a 344 amino acid mature polypeptide. The mature polypeptide has a calculated molecular weight of 38.5 kilodaltons and a pl of 5.75. The V(max) of the purified enzyme was 465 micromoles per minute per milligram, with apparent K(m) values of 62 micromolar ATP and 56 micromolar ribulose 5-phosphate. Immunoblot analysis indicated antigenic similarity and a similar subunit size for the enzyme from five higher plant species and Chlamydomonas. Southern blot analysis of Chlamydomonas genomic DNA indicated the presence of a single phosphoribulokinase gene. Comparison of the mature proteins from Chlamydomonas and spinach revealed 86 amino acid differences in primary structure (25% of the total) without a major difference in kinetic properties. The transit peptides of the spinach and Chlamydomonas proteins possessed little sequence homology.

Electron Transport through photosystem I Stimulates Light Activation of Ribulose Bisphosphate Carboxylase/Oxygenase (Rubisco) by Rubisco Activase.

The activation state of ribulose bisphosphate carboxylase/oxygenase (rubisco) in a lysed chloroplast system is increased by light in the presence of a saturating concentration of ATP and a physiological concentration of CO(2) (10 micromolar). Electron transport inhibitors and artificial electron donors and acceptors were used to determine in which region of the photosynthetic electron transport chain this light-dependent reaction occurred. In the presence of DCMU and methyl viologen, the artificial donors durohydroquinone and 2,6-dichlorophenolindophenol (DCPIP) plus ascorbate both supported light activation of rubisco at saturating ATP concentrations. No light activation occurred when DCPIP was used as an acceptor with water as electron donor in the presence of ATP and dibromothymoquinone, even though photosynthetic electron transport was observed. Nigericin completely inhibited the light-dependent activation of rubisco. Based on these results, we conclude that stimulation of light activation of rubisco by rubisco activase requires electron transport through PSI but not PSII, and that this light requirement is not to supply the ATP needed by the rubisco activase reaction. Furthermore, a pH gradient across the thylakoid membrane appears necessary for maximum light activation of rubisco even when ATP is provided exogenously.

Primary Structure of Chlamydomonas reinhardtii Ribulose 1,5-Bisphosphate Carboxylase/Oxygenase Activase and Evidence for a Single Polypeptide.

Immunoblot analysis of ribulose 1,5-bisphosphate carboxylase/oxygenase (rubisco) activase from the green alga Chlamydomonas reinhardtii indicated the presence of a single polypeptide. This observation contrasts with the Spinacea oleracea (spinach) and Arabidopsis thaliana proteins, in which two polypeptide species are generated by alternative pre-mRNA splicing. A Chlamydomonas rubisco activase cDNA clone containing the entire coding region was isolated and sequenced. The open reading frame encoded a 408 amino acid, 45 kilodalton polypeptide that included a chloroplast transit peptide. The presumptive mature polypeptide possessed 62% and 65% amino acid sequence identity, respectively, with the spinach and Arabidopsis mature polypeptides. The Chlamydomonas rubisco activase transit peptide possessed almost no amino acid sequence identity with the higher plant transit peptides. The nucleotide sequence of Chlamydomonas rubisco activase cDNA provided no evidence for alternative mRNA splicing, consistent with the immunoblot evidence for only one polypeptide. Genomic DNA blot analysis indicated the presence of a single Chlamydomonas rubisco activase gene. In the presence of spinach rubisco activase, a lower extent and rate of activation were obtained in vitro with Chlamydomonas rubisco than with spinach rubisco. We conclude Chlamydomonas rubisco activase comprises a single polypeptide which differs considerably from the higher plant polypeptides with respect to primary structure.

Glyoxylate inhibition of ribulosebisphosphate carboxylase/oxygenase activation in intact, lysed, and reconstituted chloroplasts.

At bicarbonate concentrations equivalent to air levels of CO2, activation of ribulosebisphosphate carboxylase/oxygenase (rubisco) was inhibited by micromolar concentrations of glyoxylate in intact, lysed, and reconstituted chloroplasts and in stromal extracts. The concentration of glyoxylate required for 50% inhibition of light activation in intact chloroplasts was estimated to be 35 micromolar. No direct inhibition by glyoxylate was observed with purified rubisco or rubisco activase at micromolar concentrations. Levels of ribulose 1,5-bisphosphate and ATP increased in intact chloroplasts following glyoxylate treatment. Results from experiments with well-buffered lysed and reconstituted chloroplast systems ruled out lowering of pH as the cause of inhibition. With intact chloroplasts, micromolar glyoxylate did not prevent activation of rubisco at high (10 mM) concentrations of bicarbonate, indicating that rubisco could be spontaneously activated in the presence of glyoxylate. These results suggest the existence of a component of the in vivo rubisco activation system that is not yet identified and which is inhibited by glyoxylate.

The CO2/O 2 specificity of ribulose 1,5-bisphosphate carboxylase/oxygenase : Dependence on ribulosebisphosphate concentration, pH and temperature.

The substrate specificity factor, V cKo/VoKc, of spinach (Spinacia oleracea L.) ribulose 1,5-bisphosphate carboxylase/oxygenase was determined at ribulosebisphosphate concentrations between 0.63 and 200 µM, at pH values between 7.4 and 8.9, and at temperatures in the range of 5° C to 40° C. The CO2/O2 specificity was the same at all ribulosebisphosphate concentrations and largely independent of pH. With increasing temperature, the specificity decreased from values of about 160 at 5° C to about 50 at 40° C. The primary effects of temperature were on K c [Km(CO2)] and V c [Vmax (CO2)], which increased by factors of about 10 and 20, respectively, over the temperature range examined. In contrast, K o [Ki (O2)] was unchanged and V o [Vmax (O2)] increased by a factor of 5 over these temperatures. The CO2 compensation concentrations (Γ) were calculated from specificity values obtained at temperatures between 5° C and 40° C, and were compared with literature values of Γ. Quantitative agreement was found for the calculated and measured Γ values. The observations reported here indicate that the temperature response of ribulose 1,5-bisphosphate carboxylase/oxygenase kinetic parameters accounts for two-thirds of the temperature dependence of the photorespiration/photosynthesis ratio in C3 plants, with the remaining one-third the consequence of differential temperature effects on the solubilities of CO2 and O2.