Mutagenesis at Two Distinct Phosphate-Binding Sites Unravels Their Differential Roles in Regulation of Rubisco Activation and Catalysis

Abstract

Orthophosphate (P_i) has two antagonistic effects on ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), stimulation of activation and inhibition of catalysis by competition with the substrate RuBP. The enzyme binds P_i at three distinct sites, two within the catalytic site (where 1P and 5P of ribulose 1,5-bisphosphate [RuBP] bind), and the third at the latch site (a positively charged pocket involved in active-site closure during catalysis). We examined the role of the latch and 5P sites in regulation of Rubisco activation and catalysis by introducing specific mutations in the enzyme of the cyanobacterium Synechocystis sp. strain PCC 6803. Whereas mutations at both sites abolished the P_i-stimulated Rubisco activation, substitution of residues at the 5P site, but not at the latch site, affected the P_i inhibition of Rubisco catalysis. Although some of these mutations substantially reduced the catalytic turnover of Rubisco and increased the K_m(RuBP), they had little to moderate effect on the rate of photosynthesis and no effect on photoautotrophic growth. These findings suggest that in cyanobacteria, Rubisco does not limit photosynthesis to the extent previously estimated. These results indicate that both the latch and 5P sites participate in regulation of Rubisco activation, whereas P_i binding only at the 5P site inhibits catalysis in a competitive manner.

Activation of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco, EC4.1.1.39) is essential for catalysis and occurs by carbamylation of Lys-201 and binding of Mg²⁺ by the carbanion (14). Several effectors, mostly anions such as orthophosphate (P_i), sugar-phosphates, sulfate, and NADPH, modulate the activation process via a mechanism that is still controversial (2, 3, 5, 11, 12, 16, 18, 20, 21, 26). It is accepted that these compounds evoke activation by elevating the carbamylation level, presumably by slowing the rate of enzyme deactivation (3, 12, 18). Nonetheless, evidence was brought that in addition to stimulation of Rubisco carbamylation, P_i also enhances the activation of the enzyme without a parallel increase in the carbamylation level (16, 21). Whereas McCurry et al. (18) reported monophasic, hyperbolic P_i concentration-dependent activation of Rubisco, nonmonotonously biphasic kinetics was obtained in other, more recent studies (2, 16), indicating that P_i stimulates enzyme activation via a mechanism that involves multiple interacting sites (28).

Paradoxically, in addition to the stimulatory effect on activation, the effectors also inhibit the catalytic activity of the enzyme by competing with the substrate ribulose 1,5-bisphosphate (RuBP) (3, 16). Based on the observation of McCurry et al. (18) and the ability of the transition state analog carboxy-arabinitol-bisphosphate to prevent 6-phosphogluconate from binding, Badger and Lorimer (3) suggested that inhibition of activity and stimulation of activation are both induced by effector binding to the catalytic site. Although this would imply that RuBP competitively inhibits the stimulation of Rubisco activation by P_i, we have previously found that saturating RuBP concentrations decreased only slightly in a noncompetitive manner the stimulatory effect on activation of the cyanobacterial enzyme (16). This unexpected difference raised the question of whether binding of P_i to the catalytic site stimulates the activation of Rubisco.

X-ray analysis of a tobacco Rubisco in complex with P_i revealed that the anion binds at three sites: two sites are those occupied by 1P and 5P of RuBP, and the third site is a positively charged pocket on the surface of the enzyme (Fig. 1) (7). As implied from the X-ray data, the last site is involved in structural transitions of the enzyme during catalysis. Binding of RuBP leads to closure of the catalytic site by movement of loop 6 across the substrate. The C-tail, which is randomly oriented in the open state, pushes against loop 6 and binds tightly via strong ionic interactions with a positively charged site on the surface of the enzyme. This charged site is formed by Arg-134 and His-310, which are surrounded by Arg-41, Lys-305, and Arg-312 (Fig. 1) (7, 27). For that reason, this site was named ‘the latch’ (Fig. 1). As P_i competes with Asp-473 of the C-tail on binding to this site, it may prevent enzyme closure and hence, inhibit catalysis (Fig. 1B and C) (7).

FIG. 1.

A, Diagram of the three P_i-binding sites on Rubisco. The three-dimensional structure of a large and small subunit protomer of a nonactivated tobacco enzyme in complex with P_i (PDB code 1EJ7) (7). The three P_i ions are in red, residues directly associated with P_i are in light green, and residues that contribute indirectly to P_i binding are in dark green; activator Lys is in magenta. Moving elements (C-tail and loop 6) that participate in closure of the catalytic site are in blue. Thr-65 of the paired large subunit, which also participates in P_i binding, is not labeled. B, Space-filled model of P_i bound to the latch site of tobacco Rubisco, a top view. Basic residues, blue; acidic residues, red; neutral residues, green; P_i, yellow; water molecules, white. C, Space-filled model of the latch site in Synechoccocus Rubisco in the closed state (PDB code 1RBL). The last four residues of the C-tail (Met-472, Asp-473, Lys-474, and Leu-475) are shown as balls and sticks. Note that Asp-473 occupies the region where P_i binds. Colors are as in B. The three models were drawn using RasMol.

In this study we used a mutant of the cyanobacterium Synechocystis sp. strain PCC 6803, Syn6803Δrbc, which enables site-directed mutagenesis of Rubisco and analysis of the effects in vitro on the isolated enzyme as well as in vivo on the photosynthetic performance and growth of the cells (1). We introduced mutations at two P_i binding sites (5P and the latch) in order to examine their role in stimulation of activation and inhibition of Rubisco activity.

MATERIALS AND METHODS

Growth conditions and photosynthesis assays.

Synechocystis sp. strain PCC 6803 wild type and mutants were grown on BG-11 medium as was described (17). Photosynthesis rate was determined using a Clark-type O₂ electrode (Rank Brothers, United Kingdom) according to Marcus et al. (15). Rubisco was partially purified by two steps of ammonium sulfate fractionation followed by dialysis (cutoff, 12,000 Da), which eliminated low-molecular-weight effectors. All steps were performed in the presence of protease inhibitors (P2714, Sigma) to prevent proteolysis.

The partially purified enzyme was compared to Rubisco further purified by gel filtration chromatography (Sepharose 6B; 100 by 2 cm column). The kinetic properties of both types of enzyme preparations were identical. Examination of enzyme integrity using native and sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis followed by immunochemical analysis (for details, see reference 17) did not reveal any dissociation or degradation products (Fig. 2). Since the yield of purified enzyme decreased substantially with further purification steps, all experiments were performed using the ammonium sulfate preparation.

FIG. 2.

PAGE and immunochemical analysis of Rubisco integrity. A, electrophoresis on native 7% gel and a Western blot. B, electrophoresis on 12% SDS-PAGE and a Western blot. Aliquots of 16 μg total soluble protein from wild-type and H298A and H327Q mutant cells were loaded in each lane and separated under constant current (15 mA). The gels were stained with Coomassie blue or transferred to nitrocellulose. Immunochemical detection was performed using rabbit anti-Rubisco serum, goat anti-rabbit immunoglobulin G linked to alkaline phosphatase, and colorization by nitro blue tetrazolium and 5-bromo-4-chloro-3-indolylphosphate. Rubisco of tobacco was added as a reference. Note that the polyclonal antibodies did not bind the enzyme small subunit in detectable amounts. PB, phycobilliproteins.

The molar concentration of Rubisco catalytic sites was determined by binding of [¹⁴C]CPBP (carboxy-pentitol-bisphosphate) as was described in detail (16). Briefly, the enzyme was incubated for 10 min with 15 μM ¹⁴CPBP (66.6 × 10⁷ Bq μmol⁻¹) at 30°C. ¹⁴CPBP-bound Rubisco was separated from the free ligand by gel filtration using a 10-cm Sephadex G-100 column. The reliability of measuring catalytic site concentration by CPBP in mutants substituted at the RuBP binding site (H298A and H327Q) was validated by comparing the ratio of CPBP data of the wild type and the mutants to the ratio obtained following densitometry of enzyme bands separated on a native polyacrylamide gel electrophoresis (Fig. 2A). The similar ratio obtained, irrespective of the mutation, indicated that determination of the catalytic site number by CPBP binding was reliable. Determination of carbamylation level, enzyme activation, and assays have been described previously (16).

Mutagenesis and genetic analysis.

Plasmid pSynR4.0, a pBluescript derivative containing the entire rbc operon and flanking sequences from Synechocystis sp. strain PCC 6803 (1), was used for mutagenesis. Mutations were introduced by PCR using three oligonucleotide primers (Table 1). A primer designed with a mutation was reacted with either an upstream or a downstream primer encompassing the restriction sites used for subcloning. Plasmid pSynR4.0 (1) was the DNA template, and POW DNA polymerase (Roche) was used for amplification. The resulting DNA fragment was used in a second PCR step with a complementary primer encompassing a restriction site. For mutagenesis of Arg-134 we used a 488-bp XbaI-NsiI fragment of rbcL. All other mutations were carried out using an 516-bp NsiI -NruI fragment (Table 1). The final PCR product was cleaved by the appropriate enzymes and inserted back into the pSynR4.0 vector. The final construct was used to transform the Syn6803Δrbc recipient (1). Transformed cells were inoculated onto solid BG-11 medium at ambient CO₂ levels. Air-grown colonies appeared within 14 to 21 days. The mutations were verified in the cyanobacterial genome by DNA sequencing.

TABLE 1.

Oligonucleotide primers used for mutagenesis^a

Purpose	Mutation or cloning site	Sequence (5′ to 3′)
Mutagenesis	R134A	GGAAAACGAATATCTTCTAAAGCGAGGGCCCGCAGAGCC
	R134E	GGAAAACGAATATCTTCTAATTCGAGGGCCCGCAGAGCC
	H298A	CGGTCAACTACGGCGGCCATTGCCCGGTG
	K305A	GTTGACCGTCAGGCAAACCACGGGATC
	K305E	GTTGACCGTCAGGAAAACCACGGG
	H310A	CCACGGGATCGCCTTCCGGGTTTTGG
	H310D	CCACGGGATCGACTTCCGGGTTTTGG
	H327Q	GGTGACCACCTCCAGTCCGGTACCGTG
Subcloning	XbaI upstream	CTGTTCCCAACGAAGATAACC
	NsiI upstream	CACCCCCATCATCATGCAT
	NsiI downstream	CCGCCGGTGAAGAAGTC
	NruI downstream	CCCGGATAACGTCATTACCTTCGC

^aSubstituted nucleotides are in bold, and altered codons are underlined.

RESULTS AND DISCUSSION

The role of amino acid residues in the phosphate binding sites of Rubisco of the cyanobacterium Synechocystis sp. strain PCC 6803 was examined using site-directed mutagenesis. Either P_i or 5P of RuBP interacts with the 5P binding site via hydrogen bonds with Arg-295 and His-298 in the open state of the enzyme, whereas 5P of RuBP interacts with Arg-295 and His-327 in its closed state (6, 7). The highly conserved His-298 and His-327 were substituted by Ala and Gln, respectively (mutants H298A and H327Q, respectively). Arg-134 and His-310 of the latch site, which directly bind P_i and Asp-473 as well as Lys-305 at the other side of this domain (Fig. 1) (7), were substituted by neutral or negatively charged residues (mutations R134A/E, K305A/E, and H310A/D). In addition, a double mutant, K305A/H310A, was constructed. All these mutants, including H327Q (at the catalytic site), grew photoautotrophically at ambient CO₂ levels. PCR analysis and the inability of the mutants to grow on chloramphenicol indicated that the mutations segregated fully among the identical Synechocystis chromosomes (data not shown).

Identification of the P_i binding site involved in inhibition of Rubisco activity.

To identify the P_i binding site that competitively inhibits the catalytic activity of Rubisco, we determined carboxylase activity at various P_i and RuBP concentrations of enzyme mutants R134A and K305E (latch site) as well as H327Q and H298A (5P site). Dixon analysis (25) of the data revealed that RuBP carboxylation by all four mutant enzymes was competitively inhibited by P_i. The inhibition constant of P_i, K_i(P_i), was not affected by any of the mutations at the latch site or by H298A at the 5P site (Table 2), although mutations K305E and H298A hindered the interaction with P_i, as was evident from the inability of P_i to stimulate enzyme activation (Fig. 3A and B). In contrast, mutation H327Q reduced the K_i(P_i) 10-fold (Table 2). This suggests that inhibition of the catalytic activity of Rubisco occurs when P_i and 5P of RuBP compete for binding to Arg-295 and His-327 of the 5P subsite during catalysis but not to the Arg-295 and His-298 5P subsites at the open state.

TABLE 2.

Kinetic properties of Rubisco mutants^a

Strain	Site	Kcat (min−1)	Km(RuBP) (μM)	Kmapp(CO2) (μM)b	Ki(Pi) (mM)
Wild type		545	140	181	5.8
R134A	Latch	178	212	161	6.0
K305E	Latch	185	242	190	5.7
K305A	Latch	177	199	185	NM
H310D	Latch	185	387	NM	NM
H310A	Latch	540	179	NM	NM
H298A	5P	43	267	NM	6.3
H327Q	5P	92	490	175	0.67

^aEnzyme was activated for 30 min by 280 μM CO₂ and 10 mM MgCl₂ at 30°C. Aliquots of 10μl (1.25 to 1.5 μM activated Rubisco) were added to 290-μl reaction mixtures containing either various concentrations of ¹⁴CO₂ and 0.5 mM RuBP or various concentrations of RuBP and P_i and 660 μM ¹⁴CO₂ (4 Bq nmol⁻¹), 5 mM DTT, 10 mM MgCl₂, and 7.5 Wilbur-Anderson units carbonic anhydrase in 50 mM HEPES, pH 8.0. The reaction was terminated after 2 min with 200 μl 6 N acetic acid, and the acid-stable material was counted after 48 h. Catalytic site concentration was determined by [¹⁴C]CPBP binding as described (16, 17). Each assay was performed in triplicate. The data were analyzed according to Hanes-Woolf or Dixon (25), and the best-fitting lines were calculated using the least squares method. The correlation coefficients for the regression lines were higher than 0.95.

^bApparent K_m(CO₂) was determined at 21% O₂. NM, not measured.

FIG. 3.

Effect of P_i concentration on the kinetics of Rubisco activation (A and B). Rubisco from Synechocystis sp. strain PCC 6803 (wild type) and mutants R134A, H298A and H327Q (A) and K305E, H310A/D, and K305A/H310A (B) was activated for 30 min at 30°C with 1 mM dithiothreitol, 70 μM CO₂, 10 mM MgCl₂ and various concentrations of potassium phosphate (phosphate was added 5 min after the other substances). Aliquots of 12 μl activated enzyme were added to 288 μl reaction mixture containing 0.5 mM RuBP, 70 μM ¹⁴CO₂ (11.3 Bq nmol⁻¹), 10 mM MgCl₂, 1 mM dithiothreitol, and 7.5 Wilbur-Anderson units carbonic anhydrase in 50 mM HEPES buffer, pH 8.0. The reaction was terminated after 2 min with 200 μl 6 N acetic acid. Activation was defined as the ratio between the activity of the enzyme activated at the indicated P_i concentration and that in the absence of P_i during activation.

An alternative interpretation of the competitive inhibition of Rubisco activity by P_i with respect to RuBP was raised by Duff et al. (7). They have suggested that competition between P_i and the C-terminal Asp-473 on binding to the latch site inhibits the closure of the catalytic site and consequently the catalytic activity of the enzyme. If this hypothesis is correct, mutagenesis of residues in the latch site that differentially affect the binding of P_i and Asp-473 should influence the inhibitory effect of P_i. However, the inhibitory effect of P_i remained competitive with respect to RuBP (data not shown) and the K_i(P_i) was not changed by mutations R134A and K305E (Table 2). This implies that competition between P_i and the C-terminal Asp-473 on binding to the latch site does not prevent the closure of the catalytic site. It is noteworthy that although P_i may also compete with RuBP on binding to the 1P site, probing of this site by mutagenesis seems more difficult and has not been reported thus far.

Analysis of the P_i-binding sites involved in stimulation of Rubisco activation.

All Rubisco mutants modified at the P_i-binding sites (Fig. 1) were activated in vitro by CO₂ and Mg²⁺. Moreover, their activation in vivo was evident from the ability of the mutant cells to grow photoautotrophically. Interestingly, the activation of R134A mutant enzyme, determined by carboxylase activity after activation with 285 μM CO₂ and 20 mM MgCl₂, was apparently higher than that of the wild type and the other mutants.

Since the wild-type and R134A enzymes were fully carbamylated (data not shown), the apparent higher activation ability of the R134A enzyme could be the result of a lower activation level at ambient CO₂ concentrations, presumably due to a higher deactivation rate. This assumption was examined by comparing the kinetics of deactivation of the wild-type and R134A enzymes. Following activation (in the presence of 285 μM CO₂ and 20 mM MgCl₂), enzyme deactivation was monitored by carboxylase activity following 10-fold dilution with HEPES buffer, pH 8.0, or with buffer, pH 8.0, that contained identical CO₂ and MgCl₂ concentrations (control). As shown in Fig. 4, the activation level of the wild-type enzyme, under both conditions, and that of the R134A enzyme, at high CO₂ and MgCl₂ concentrations, hardly changed over 5 h. In contrast, the activation level of the R134A enzyme at low CO₂ and MgCl₂ concentrations declined continuously and after 5 h was half that of the wild-type enzyme (Fig. 4). These findings imply that the mutations investigated here did not hinder the formation of the ternary complex, namely, carbamylation of Lys-201 and binding of Mg²⁺ to the carbamate. However, substitution R134A seems to impose a long-distance effect which accelerates decomposition of the activated complex.

FIG. 4.

Rubisco deactivation rate. Rubisco from the wild type and mutant R134A was activated by 285 μM CO₂ and 10 mM MgCl₂ for 30 min at 30°C. The activated enzyme was diluted 1:9 with 20 mM HEPES, pH 8.0, with no change in the concentration of MgCl₂ and CO₂ or in the absence of these substances (final concentrations were 28 μM CO₂ and 1 mM MgCl₂) and incubated at 18°C and at the indicated time. Rubisco activity was determined as described for Fig. 2.

The effect of the various mutations at the P_i-binding sites on stimulation of enzyme activation by phosphate was examined. The kinetics of P_i-stimulated activation in the presence of suboptimal CO₂ and Mg²⁺ concentrations is biphasic, i.e., two distinct concentration-dependent phases separated from one another by an intermediary plateau (Fig. 3A and B), a phenomenon previously reported by Anwaruzzaman et al. (2) and Marcus and Gurevitz (16). Under these conditions P_i has two effects on Rubisco activation: it raises the carbamylation level, but also elevates the activity of the carbamylated enzyme (16, 21). Whereas neutralization of a single positive charge at the latch site (substitution K305A or H310A) had no effect, the double mutation K305A/H310A had a clear inhibitory effect, especially at low P_i concentrations (Fig. 3B).

The lack of effect upon single-charge elimination could be due to the fact that the latch site is composed of five positively charged residues (see Fig. 1). The effect of the double mutation could be due to perturbation and, hence, reduced affinity of P_i for the latch site. Substitution of Arg-134, which is also involved in P_i binding (mutant R134A; Fig. 1), apparently elevated the activation level (Fig. 3A). We found that this unprecedented effect resulted from a higher decarbamylation rate at suboptimal CO₂ concentrations (Fig. 4), leading to an extremely low level of activation at the onset of the experiment. Charge inversion of Lys-305 and His-310 (mutants K305E and H310D, respectively) abolished the P_i effect on activation (Fig. 3A and B), probably due to P_i exclusion from this site. These results suggest that although these residues bind P_i, they are not involved in its stimulatory effect.

Inhibition of P_i-stimulated Rubisco activation was obtained upon H327Q and H298A substitutions at the two subsites of the 5P domain made of Arg-295 with His-298 or Arg-295 with His-327, in the open and closed states of the catalytic site, respectively (Fig. 3A). This may be explained in a number of ways. Interaction of P_i with one of these subsites initiates the stimulatory effect, yet substitution at the other subsite hinders this effect. Assuming that Arg-295, which is shared by both subsites, induces the stimulation of Rubisco activation upon P_i binding, substitution of either His-298 or His-327, involved in these subsites, hinders this interaction. One of these subsites is the receptor site for P_i, whereas the other subsite transduces the P_i effect to the target site. As the stimulation of activation occurs in general at the open state of the enzyme, it is likely that P_i associates with His-298, while the effect on the target site, which is most likely Lys-201, is mediated by the adjacent His-327 (Fig. 1). Despite the crystallographic data suggesting that P_i interacts with His-298 (7), perhaps P_i also interacts with His-327 in the open state of the catalytic site, which stimulates the activation.

All the above results indicate that at least two P_i ions are involved in P_i stimulation of activation by interaction with the 5P site and with the latch site.

Photosynthetic performance of the mutants.

Substitution of Arg-134, Lys-305, and His-310 at the latch site had only a marginal effect on the rate of photosynthesis of the mutant cells. The substitutions that had the most effect on the rate of CO₂ fixation (20% reduction) at saturating light intensity and CO₂ concentration were R134A (Fig. 5) and R134E (not shown). Yet, at limiting light intensity (Fig. 5) or CO₂ concentration, the rate of photosynthesis was unaffected in either of these mutants (data not shown). Unlike the mutations at the latch site, substitutions H298A and H327Q at the RuBP binding site reduced the maximal rate of photosynthesis by 40% and 70%, respectively (Fig. 5).

FIG. 5.

O₂ exchange rate of Synechocystis sp. strain PCC 6803 (wild type) and mutants R134A, H298A, K305E, and H327Q. Measurements were taken at various light intensities and 2 mM inorganic carbon in 25 mM HEPES buffer, pH 7.0, using a Clark-type O₂ electrode (Rank Brothers, United Kingdom). Each assay was performed in triplicate. Standard deviations were less than 10%.

Despite the minor effect on the rate of photosynthesis, in most instances the mutations at the latch site reduced up to two-thirds of the catalytic turnover of the enzyme and increased the K_m(RuBP) (Table 2). On the other hand, these mutations did not significantly alter the affinity of Rubisco for CO₂ (Table 2). Whereas similar findings were obtained upon truncation of the last eight residues of the large subunit of Rubisco from Synechococcus sp. (9), substitution of the C-tail Asp-473 in Chlamydomonas Rubisco reduced the affinity of the enzyme for CO₂ (24). These results suggest that the mutations at the latch site or at the C-terminal region affect the K_cat and K_m(RuBP) of Rubisco by hindering the mechanism of the catalytic site closure.

It is noteworthy that His-291 and His-321 of the dimeric Rhodospirillum rubrum enzyme, which are the equivalents of His-298 and His-327, respectively, in the hexadecameric enzyme, were mutagenized and the enzymes were expressed in Escherichia coli (10, 19). Whereas substitution of His-291 in the R. rubrum enzyme had a moderate effect and therefore this residue was considered nonessential (19), the K_cat of the cyanobacterial mutant enzyme, H298A, was only 8% of that of the wild type, and the K_m(RuBP) increased 90% (Table 2). This suggests that His-298 in the cyanobacterial enzyme is important for RuBP binding and hence for catalysis, which is in accordance with the findings of Duff et al. (7).

Moreover, we were able to evaluate how these changes affect the photosynthetic and growth abilities of the mutant cells. We found that the cyanobacterial mutants K305E, H310D, H327Q, and H298A partially compensated for the impaired kinetic properties of Rubisco by increasing its cellular content. For example, in mutant H327Q, in which the K_cat of the enzyme declined 84%, the K_m(RuBP) increased 3.5-fold, K_i(P_i) decreased 10-fold (Table 2), and stimulation of enzyme activation by P_i was abolished (Fig. 3), and the cellular Rubisco content doubled compared to that in the wild type (Fig. 6). Similarly, in mutant H298A, in which the K_cat of the enzyme was only 8% that of the wild type (Table 2) and the P_i-stimulated activation was abolished as well (Fig. 3A), a threefold elevation in Rubisco content was observed (Fig. 6). Consequently, Rubisco activity per cell in mutants H298A, H327Q, and K305E was still 0.24-, 0.3-, and 0.5-fold, respectively, the level measured in the wild type (Fig. 6). Such levels of Rubisco activity were sufficient to support 60, 30, and 100%, respectively, of the photosynthesis rate measured in the wild type (Fig. 5). On the other hand, the two-thirds decrease in K_cat and 50% increase in K_m(RuBP) of Rubisco in mutant R134A (Table 2), without any compensating elevation in Rubisco content (Fig. 6), had a marked effect (65% decrease) on cellular Rubisco activity. Notably, despite the prominent reduction in cellular Rubisco activity, the photosynthesis rate of this mutant declined only 20% (Fig. 5). Similarly, we have recently shown that a C172A mutation in Rubisco of Synechocystis sp. reduced enzyme activity 60% with a minor effect on the carbon assimilation rate (17).

FIG. 6.

Rubisco content (white bars) and activity (black bars) in Synechocystis sp. strain PCC 6803 (wild type) and mutants R134A, H298A, K305E, and H327Q. Cells were grown on BG-11 medium at ambient CO₂ under continuous illumination (20 μmol × m⁻² × s⁻¹). The cells were spun down and resuspended in 50 mM HEPES buffer, pH 8.0, containing 285 μM CO₂, 10 mM MgCl₂, 1 mM dithiothreitol, and a protease inhibitor cocktail (P2714, Sigma). The cells were broken under 24,000 lb/in² using a French press. The crude extract was centrifuged and the supernatant containing the enzyme was activated by incubation for 30 min at 30°C. The content of Rubisco catalytic sites was determined by incubation of 200 μl activated enzyme with 10 μM [¹⁴C]CPBP (185 Bq nmol⁻¹). The CPBP-bound enzyme was separated from the free ligand using a 10-cm Sephadex-G50 column. Rubisco activity was determined by incubation of 50 μl activated enzyme with 250 μl reaction mixture containing 1 mM RuBP, 640 μM ¹⁴CO₂ (1.3 Bq nmol⁻¹), 10 mM MgCl₂, 1 mM dithiothreitol, and 7.5 Wilbur-Anderson units carbonic anhydrase in 50 mM HEPES buffer, pH 8.0. The reaction was terminated after 2 min by addition of 200 μl 6 N acetic acid. Each assay was performed in triplicate. Standard deviations were less than 10%.

Even though cyanobacteria concentrate CO₂ at the carboxylation site (13, 15), thereby compensating for the low affinity of Rubisco for CO₂ (Table 2), the carboxylation rate cannot rise above that dictated by all other kinetic properties of the enzyme. Therefore, the CO₂ concentrating mechanism cannot compensate for a substantial loss in the catalytic turnover and K_m(RuBP) of a mutated enzyme. For that reason, the lack of effect of the mutant enzymes on the CO₂-limited rate of photosynthesis implies that, unlike in higher plants (8, 22, 23), the constraint imposed by the cyanobacterial Rubisco on the photosynthetic flux is low, presumably due to its higher catalytic turnover (4) and the ability to regulate its concentration in the cell (Fig. 6). Since the most pronounced effect on the photosynthesis rate was observed at saturating CO₂ and light intensities, we suggest that RuBP recycling is the bottleneck of photosynthesis in cyanobacteria (Fig. 5) (29).

Although Rubisco activity per cell in mutants H327Q, H298A, and R134A was similar (Fig. 6), mutation H327Q prominently inhibited the rate of photosynthesis, whereas R134A had little effect and H298A had a moderate effect (Fig. 5). This difference could result from the elevated P_i inhibitory effect of mutation H327Q (Table 2). The lack of effect of mutations K305E and H310D on the photosynthesis rate (Fig. 5), despite the elimination of P_i-stimulated activation (Fig. 3A and B), was due to a high activation level of Rubisco, most likely achieved by the CO₂ concentrating mechanism (13, 15).

Conclusions.

Our mutational analysis at the P_i binding sites of Rubisco indicates that binding of P_i to Arg-295 and His-327 during catalysis competitively inhibits carboxylation. P_i binding to the latch site or to His-298 of the 5P site had no such effect (2). P_i stimulates Rubisco activation via at least two distinct sites, the 5P site, which involves His-298, His-327, and presumably Arg-295, and the latch site, in which the residues that bind P_i differ from those mediating the P_i stimulatory effect (3). Mutations R134A and K305E at the latch site decrease catalytic turnover and increase the K_m(RuBP) of Rubisco, most likely due to obstruction of the catalytic-site closure (4). Arg-134 at the latch site is distant from the activation site and yet has a stabilizing effect on the carbamate, as indicated by the accelerated rate of Rubisco deactivation upon its substitution to Ala (5). Although the limitation imposed by Rubisco on the photosynthetic flux is lower than that observed in higher plants, Synechocystis cells partially compensate for the inferior kinetic properties of Rubisco mutants by elevating the enzyme content per cell.

This study not only provides further molecular details on P_i binding sites and their differential role in regulation of cyanobacterial Rubisco, it also demonstrates that, unlike in higher plants, the cyanobacterial enzyme is apparently not the limiting factor of light-saturated photosynthesis.