Comparative affinities of the epimeric reaction-intermediate analogs 2- and 4-carboxy-D-arabinitol 1,5-bisphosphate for spinach ribulose 1,5-bisphosphate carboxylase

An Insight of RuBisCO Evolution through a Multilevel Approach

RuBisCO is the most abundant enzyme on earth; it regulates the organic carbon cycle in the biosphere. Studying its structural evolution will help to develop new strategies of genetic improvement in order to increase food production and mitigate CO₂ emissions. In the present work, we evaluate how the evolution of sequence and structure among isoforms I, II and III of RuBisCO defines their intrinsic flexibility and residue-residue interactions. To do this, we used a multilevel approach based on phylogenetic inferences, multiple sequence alignment, normal mode analysis, and molecular dynamics. Our results show that the three isoforms exhibit greater fluctuation in the loop between αB and βC, and also present a positive correlation with loop 6, an important region for enzymatic activity because it regulates RuBisCO conformational states. Likewise, an increase in the flexibility of the loop structure between αB and βC, as well as Lys330 (form II) and Lys322 (form III) of loop 6, is important to increase photosynthetic efficiency. Thus, the cross-correlation dynamics analysis showed changes in the direction of movement of the secondary structures in the three isoforms. Finally, key amino acid residues related to the flexibility of the RuBisCO structure were indicated, providing important information for its enzymatic engineering.

Identification and characterization of second-generation EZH2 inhibitors with extended residence times and improved biological activity

The histone methyltransferase EZH2 has been the target of numerous small-molecule inhibitor discovery efforts over the last 10+ years. Emerging clinical data have provided early evidence for single agent activity with acceptable safety profiles for first-generation inhibitors. We have developed kinetic methodologies for studying EZH2-inhibitor-binding kinetics that have allowed us to identify a unique structural modification that results in significant increases in the drug-target residence times of all EZH2 inhibitor scaffolds we have studied. The unexpected residence time enhancement bestowed by this modification has enabled us to create a series of second-generation EZH2 inhibitors with sub-pM binding affinities. We provide both biophysical evidence validating this sub-pM potency and biological evidence demonstrating the utility and relevance of such high-affinity interactions with EZH2.

Mechanism of Enzyme Repair by the AAA+ Chaperone Rubisco Activase

How AAA+ chaperones conformationally remodel specific target proteins in an ATP-dependent manner is not well understood. Here, we investigated the mechanism of the AAA+ protein Rubisco activase (Rca) in metabolic repair of the photosynthetic enzyme Rubisco, a complex of eight large (RbcL) and eight small (RbcS) subunits containing eight catalytic sites. Rubisco is prone to inhibition by tight-binding sugar phosphates, whose removal is catalyzed by Rca. We engineered a stable Rca hexamer ring and analyzed its functional interaction with Rubisco. Hydrogen/deuterium exchange and chemical crosslinking showed that Rca structurally destabilizes elements of the Rubisco active site with remarkable selectivity. Cryo-electron microscopy revealed that Rca docks onto Rubisco over one active site at a time, positioning the C-terminal strand of RbcL, which stabilizes the catalytic center, for access to the Rca hexamer pore. The pulling force of Rca is fine-tuned to avoid global destabilization and allow for precise enzyme repair.

Rubisco oligomers composed of linked small and large subunits assemble in tobacco plastids and have higher affinities for CO2 and O2

Manipulation of Rubisco within higher plants is complicated by the different genomic locations of the large (L; rbcL) and small (S; RbcS) subunit genes. Although rbcL can be accurately modified by plastome transformation, directed genetic manipulation of the multiple nuclear-encoded RbcS genes is more challenging. Here we demonstrate the viability of linking the S and L subunits of tobacco (Nicotiana tabacum) Rubisco using a flexible 40-amino acid tether. By replacing the rbcL in tobacco plastids with an artificial gene coding for a S40L fusion peptide, we found that the fusions readily assemble into catalytic (S40L)8 and (S40L)16 oligomers that are devoid of unlinked S subunits. While there was little or no change in CO2/O2 specificity or carboxylation rate of the Rubisco oligomers, their Kms for CO2 and O2 were reduced 10% to 20% and 45%, respectively. In young maturing leaves of the plastome transformants (called ANtS40L), the S40L-Rubisco levels were approximately 20% that of wild-type controls despite turnover of the S40L-Rubisco oligomers being only slightly enhanced relative to wild type. The reduced Rubisco content in ANtS40L leaves is partly attributed to problems with folding and assembly of the S40L peptides in tobacco plastids that relegate approximately 30% to 50% of the S40L pool to the insoluble protein fraction. Leaf CO2-assimilation rates in ANtS40L at varying pCO2 corresponded with the kinetics and reduced content of the Rubisco oligomers. This fusion strategy provides a novel platform to begin simultaneously engineering Rubisco L and S subunits in tobacco plastids.

The catalytic properties of hybrid Rubisco comprising tobacco small and sunflower large subunits mirror the kinetically equivalent source Rubiscos and can support tobacco growth

Plastomic replacement of the tobacco (Nicotiana tabacum) Rubisco large subunit gene (rbcL) with that from sunflower (Helianthus annuus; rbcL(S)) produced tobacco(Rst) transformants that produced a hybrid Rubisco consisting of sunflower large and tobacco small subunits (L(s)S(t)). The tobacco(Rst) plants required CO(2) (0.5% v/v) supplementation to grow autotrophically from seed despite the substrate saturated carboxylation rate, K(m), for CO(2) and CO(2)/O(2) selectivity of the L(s)S(t) enzyme mirroring the kinetically equivalent tobacco and sunflower Rubiscos. Consequently, at the onset of exponential growth when the source strength and leaf L(s)S(t) content were sufficient, tobacco(Rst) plants grew to maturity without CO(2) supplementation. When grown under a high pCO(2), the tobacco(Rst) seedlings grew slower than tobacco and exhibited unique growth phenotypes: Juvenile plants formed clusters of 10 to 20 structurally simple oblanceolate leaves, developed multiple apical meristems, and the mature leaves displayed marginal curling and dimpling. Depending on developmental stage, the L(s)S(t) content in tobacco(Rst) leaves was 4- to 7-fold less than tobacco, and gas exchange coupled with chlorophyll fluorescence showed that at 2 mbar pCO(2) and growth illumination CO(2) assimilation in mature tobacco(Rst) leaves remained limited by Rubisco activity and its rate (approximately 11 micromol m(-2) s(-1)) was half that of tobacco controls. (35)S-methionine labeling showed the stability of assembled L(s)S(t) was similar to tobacco Rubisco and measurements of light transient CO(2) assimilation rates showed L(s)S(t) was adequately regulated by tobacco Rubisco activase. We conclude limitations to tobacco(Rst) growth primarily stem from reduced rbcL(S) mRNA levels and the translation and/or assembly of sunflower large with the tobacco small subunits that restricted L(s)S(t) synthesis.

Artificially evolved Synechococcus PCC6301 Rubisco variants exhibit improvements in folding and catalytic efficiency

The photosynthetic CO2-fixing enzyme, Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase), is responsible for most of the world's biomass, but is a slow non-specific catalyst. We seek to identify and overcome the chemical and biological constraints that limit the evolutionary potential of Rubisco in Nature. Recently, the horizontal transfer of Calvin cycle genes (rbcL, rbcS and prkA) from cyanobacteria (Synechococcus PCC6301) to gamma-proteobacteria (Escherichia coli) was emulated in the laboratory. Three unique Rubisco variants containing single (M259T) and double (M259T/A8S, M259T/F342S) amino acid substitutions in the L (large) subunit were identified after three rounds of random mutagenesis and selection in E. coli. Here we show that the M259T mutation did not increase steady-state levels of rbcL mRNA or L protein. It instead improved the yield of properly folded L subunit in E. coli 4-9-fold by decreasing its natural propensity to misfold in vivo and/or by enhancing its interaction with the GroES-GroEL chaperonins. The addition of osmolites to the growth media enhanced productive folding of the M259T L subunit relative to the wild-type L subunit, while overexpression of the trigger factor and DnaK/DnaJ/GrpE chaperones impeded Rubisco assembly. The evolved enzymes showed improvement in their kinetic properties with the M259T variant showing a 12% increase in carboxylation turnover rate (k(c)cat), a 15% improvement in its K(M) for CO2 and no change in its K(M) for ribulose-1,5-bisphosphate or its CO2/O2 selectivity. The results of the present study show that the directed evolution of the Synechococcus Rubisco in E. coli can elicit improvements in folding and catalytic efficiency.

Catalytic by-product formation and ligand binding by ribulose bisphosphate carboxylases from different phylogenies

During catalysis, all Rubisco (D-ribulose-1,5-bisphosphate carboxylase/oxygenase) enzymes produce traces of several by-products. Some of these by-products are released slowly from the active site of Rubisco from higher plants, thus progressively inhibiting turnover. Prompted by observations that Form I Rubisco enzymes from cyanobacteria and red algae, and the Form II Rubisco enzyme from bacteria, do not show inhibition over time, the production and binding of catalytic by-products was measured to ascertain the underlying differences. In the present study we show that the Form IB Rubisco from the cyanobacterium Synechococcus PCC6301, the Form ID enzyme from the red alga Galdieria sulfuraria and the low-specificity Form II type from the bacterium Rhodospirillum rubrum all catalyse formation of by-products to varying degrees; however, the by-products are not inhibitory under substrate-saturated conditions. Study of the binding and release of phosphorylated analogues of the substrate or reaction intermediates revealed diverse strategies for avoiding inhibition. Rubisco from Synechococcus and R. rubrum have an increased rate of inhibitor release. G. sulfuraria Rubisco releases inhibitors very slowly, but has an increased binding constant and maintains the enzyme in an activated state. These strategies may provide information about enzyme dynamics, and the degree of enzyme flexibility. Our observations also illustrate the phylogenetic diversity of mechanisms for regulating Rubisco and raise questions about whether an activase-like mechanism should be expected outside the green-algal/higher-plant lineage.

'Irreversible' slow-onset inhibition of orotate phosphoribosyltransferase by an amidrazone phosphate transition-state mimic

A mimic of the putative transition-state intermediate has been synthesized and found to be a very slow-onset inhibitor of yeast orotate phosphoribosyltransferase. The mechanism of inhibition may involve a rate-determining isomerization of the enzyme to a form receptive to the inhibitor, which then remains tightly bound.

Despite slow catalysis and confused substrate specificity, all ribulose bisphosphate carboxylases may be nearly perfectly optimized

The cornerstone of autotrophy, the CO(2)-fixing enzyme, d-ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), is hamstrung by slow catalysis and confusion between CO(2) and O(2) as substrates, an "abominably perplexing" puzzle, in Darwin's parlance. Here we argue that these characteristics stem from difficulty in binding the featureless CO(2) molecule, which forces specificity for the gaseous substrate to be determined largely or completely in the transition state. We hypothesize that natural selection for greater CO(2)/O(2) specificity, in response to reducing atmospheric CO(2):O(2) ratios, has resulted in a transition state for CO(2) addition in which the CO(2) moiety closely resembles a carboxylate group. This maximizes the structural difference between the transition states for carboxylation and the competing oxygenation, allowing better differentiation between them. However, increasing structural similarity between the carboxylation transition state and its carboxyketone product exposes the carboxyketone to the strong binding required to stabilize the transition state and causes the carboxyketone intermediate to bind so tightly that its cleavage to products is slowed. We assert that all Rubiscos may be nearly perfectly adapted to the differing CO(2), O(2), and thermal conditions in their subcellular environments, optimizing this compromise between CO(2)/O(2) specificity and the maximum rate of catalytic turnover. Our hypothesis explains the feeble rate enhancement displayed by Rubisco in processing the exogenously supplied carboxyketone intermediate, compared with its nonenzymatic hydrolysis, and the positive correlation between CO(2)/O(2) specificity and (12)C/(13)C fractionation. It further predicts that, because a more product-like transition state is more ordered (decreased entropy), the effectiveness of this strategy will deteriorate with increasing temperature.

The relationship between side reactions and slow inhibition of ribulose-bisphosphate carboxylase revealed by a loop 6 mutant of the tobacco enzyme

The first directed mutant of a higher plant ribulose-bisphosphate carboxylase/oxygenase (Rubisco), constructed by chloroplast transformation, is catalytically impaired but still able to support the plant's photosynthesis and growth (Whitney, S. M., von Caemmerer, S., Hudson, G. S., and Andrews, T. J. (1999) Plant Physiol. 121, 579-588). This mutant enzyme has a Leu to Val substitution at residue 335 in the flexible loop 6 of the large subunit, which closes over the substrate during catalysis. Its active site was intact, as judged by its barely impaired competency in the initial enolization step of the reaction sequence, and its ability to bind tightly the intermediate analog, 2'-carboxy-D-arabinitol-1,5-bisphosphate. Prompted by observations that the mutant enzyme displayed much less slow inhibition during catalysis in vitro than the wild type, its tendency to catalyze side reactions and its response to the slow inhibitor D-xylulose-1,5-bisphosphate were studied. The lessening in slow inhibition was not caused by reduced production of inhibitory side products. Except for pyruvate production, these reactions were strongly enhanced by the mutation, as was the ability to catalyze the carboxylation of D-xylulose-1,5-bisphosphate. Rather, reduced inhibition was the result of lessened sensitivity to these inhibitors. The slow isomerization phase that characterizes inhibition of the wild-type enzyme by D-xylulose-1,5-bisphosphate was completely eliminated by the mutation, and the mutant was more adept than the wild type in catalyzing the benzylic acid-type rearrangement of D-glycero-2,3-pentodiulose-1,5-bisphosphate (produced by oxidation of the substrate, D-ribulose-1,5-bisphosphate). These observations are consistent with increased flexibility of loop 6 induced by the mutation, and they reveal the underlying mechanisms by which the side reactions cause slow inhibition.

The role of ecotin dimerization in protease inhibition

Ecotin is a homodimeric protein from Escherichia coli that inhibits many serine proteases of the chymotrypsin fold, often with little effect from the character or extent of enzyme substrate specificity. This pan-specificity of inhibition is believed to derive from formation of a heterotetrameric complex with target proteases involving three types of interface: the dimerization interface, a primary substrate-like interaction, and a smaller secondary interaction between the partner ecotin subunit and the protease. A monomeric ecotin variant (mEcotin) and a single-chain ecotin dimer (scEcotin) were constructed to study the effect of a network of protein interactions on binding affinity and the role of dimerization in broad inhibitor specificity. mEcotin was produced by inserting a beta-turn into the C-terminal arm, which normally exchanges with the other subunit. While the dimerization constant (K(dim)) of wild-type (WT) ecotin was found to be picomolar by subunit exchange experiments using FRET and by association kinetics, mEcotin was monomeric up to 1 mM as judged by gel filtration and analytical centrifugation. A crystal structure of uncomplexed mEcotin to 2.0 A resolution verifies the design, showing a monomeric protein in which the C-terminal arm folds back onto itself to form a beta-barrel structure nearly identical to its dimeric counterpart. The kinetic rate constants and equilibrium dissociation constants for monomeric and dimeric ecotin variants were determined with both trypsin and chymotrypsin. The effect of the secondary binding site on affinity was found to vary inversely with the strength of the interaction at the primary site. This compensatory effect yields a nonadditivity of up to 5 kcal/mol and can be explained in terms of the optimization of binding orientation. Such a mechanism of adaptability allows femtomolar affinities for two proteases with very different specificities.

The transition between the open and closed states of rubisco is triggered by the inter-phosphate distance of the bound bisphosphate

d-Ribulose-1,5-bisphosphate carboxylase/oxygenase (rubisco) catalyses the central CO(2)-fixing reaction of photosynthesis in a complex, multiple-step process. Several structures of rubisco complexed with substrate analogues, inhibitors and products have been determined by X-ray crystallography. The structures fall into two well-defined and distinct states. The active site is either "open" or "closed". The timing and mechanism of the transition between these two states have been uncertain. We solved the crystal structure of unactivated (metal-free) rubisco from tobacco with only inorganic phosphate bound and conclude that phosphate binding per se does not trigger closure, as it does in the similarly structured enzyme, triosephosphate isomerase. Comparison of all available rubisco structures suggests that, instead, the distance between the terminal phosphates (P1 and P2) of the bisphosphate ligand is the trigger: if that distance is less than 9.1 A, then the active site closes; if it is greater than 9.4 A then the enzyme remains open. Shortening of the inter-phosphate distance results from the ligand binding in a more curved conformation when O atoms of the ligand's sugar backbone interact either with the metal, if it is present, or with charged groups in the metal-binding site, if the metal is absent. This shortening brings the P1 phosphate into hydrogen bonding contact with Thr65. Thr65 exists in two discrete states related by a rotation of the backbone psi torsion angle. This rotation is coupled to domain rotation and hence to active site closure. Rotation of the side-chain of Thr65 also affects the C-terminal strand of large subunit which packs against Loop 6 after closure. The position of the C-terminal strand in the closed state is stabilised by multiple polar interactions with a distinctive highly-charged latch site involving the side-chain of Asp473. In the open state, this latch site may be occupied instead by phosphorylated anions.

The maximal affinity of ligands

We explore the question of what are the best ligands for macromolecular targets. A survey of experimental data on a large number of the strongest-binding ligands indicates that the free energy of binding increases with the number of nonhydrogen atoms with an initial slope of approximately -1.5 kcal/mol (1 cal = 4.18 J) per atom. For ligands that contain more than 15 nonhydrogen atoms, the free energy of binding increases very little with relative molecular mass. This nonlinearity is largely ascribed to nonthermodynamic factors. An analysis of the dominant interactions suggests that van der Waals interactions and hydrophobic effects provide a reasonable basis for understanding binding affinities across the entire set of ligands. Interesting outliers that bind unusually strongly on a per atom basis include metal ions, covalently attached ligands, and a few well known complexes such as biotin-avidin.

A common structural basis for the inhibition of ribulose 1,5-bisphosphate carboxylase by 4-carboxyarabinitol 1,5-bisphosphate and xylulose 1,5-bisphosphate

Ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) catalyzes the carboxylation of ribulose 1,5-bisphosphate. The reaction catalyzed by Rubisco involves several steps, some of which can occur as partial reactions, forming intermediates that can be isolated. Analogues of these intermediates are potent inhibitors of the enzyme. We have studied the interactions with the enzyme of two inhibitors, xylulose 1,5-bisphosphate and 4-carboxyarabinitol 1,5-bisphosphate, by x-ray crystallography. Crystals of the complexes were formed by cocrystallization under activating conditions. In addition, 4-carboxyarabinitol 1,5-bisphosphate was soaked into preformed activated crystals of the enzyme. The result of these experiments was the release of the activating CO2 molecule as well as the metal ion from the active site when the inhibitors bound to the enzyme. Comparison with the structure of an activated complex of the enzyme indicates that the structural basis for the release of the activator groups is a distortion of the metal binding site due to the different geometry of the C-3 hydroxyl of the inhibitors. Both inhibitors induce closure of active site loops despite the inactivated state of the enzyme. Xylulose 1,5-bisphosphate binds in a hydrated form at the active site.

Transition-state characterization: a new approach combining inhibitor analogues and variation in enzyme structure

A new strategy of potentially broad application for probing transition-state (TS) analogy in enzymatic systems is described in this paper. The degree to which a series of phosphonate inhibitors act as TS analogues of rat carboxypeptidase A1 has been determined for the wild-type enzyme, for the R127K, R127M, and R127A mutants, and for the R127A mutant in the presence of 0.5 M guanidine hydrochloride. The impact that the mutations have on the inverse second-order rate constants (Km/kcat) for substrate hydrolysis is mirrored by the effect on the inhibition constants (Ki) for the corresponding phosphonate inhibitors. These results demonstrate that the phosphonate moiety mimics some of the electronic as well as the geometric characteristics of the TS. A similar but distinctly separate correlation is observed for tripeptide analogues in comparison to analogues of the dipeptide Cbz-Gly-Phe, reflecting an anomalous mode of binding for the latter system. The selective rate increases and corresponding enhancement in inhibitor binding observed on addition of 0.5 M guanidine hydrochloride to the R127A mutant indicate that the exogenous cation can assume the role played by Arg-127 in stabilizing the TS and in providing substrate selectivity at the P2 position.

Synthesis and evaluation of an inhibitor of carboxypeptidase A with a Ki value in the femtomolar range

Comparative studies among a series of tripeptide phosphonate inhibitors of the zinc peptidase carboxypeptidase A indicate that incorporation of the phosphonic acid analogue of valine at the P1 position results in significantly higher affinity than the glycine, alanine, or phenylalanine analogues. When applied to the tripeptide analogue Cbz-Phe-ValP-(O)Phe [ZFVP(O)F], determination of the inhibition constant Ki was complicated by the very slow rate of dissociation. The rate of exchange of [3H]ZFVP(O)F with enzyme-bound [14C]ZFVP(O)F was followed for periods of 3-4 months to measure dissociation rate constants in the range of (1.7-4.4) x 10(-9) s-1, corresponding to half-lives of 5-13 years. Although the on- and off-rate constants differ for different carboxypeptidase isozymes, their ratios, corresponding to the inhibition constants Ki, are consistently in the range of 10-27 fM. Both the inhibition constants and the dissociation rate constants appear to be the lowest values yet determined for an enzyme-small inhibitor interaction.

Modulation of the tight binding of carboxyarabinitol 1,5-bisphosphate to the large subunit of ribulose 1,5-bisphosphate carboxylase/oxygenase

The large subunit (L) of ribulose 1,5-bisphosphate carboxylase/oxygenase (rubisco) from Synechococcus PCC 6301 was expressed in Escherichia coli, purified as the octamer L8, and analyzed for its ability to tightly bind the transition state analog, 2-carboxyarabinitol 1,5-bisphosphate (CABP). [14C]CABP remained tightly bound to L8 after challenging with [12C]CABP and gel filtration, indicating that L8 alone without the small subunit (S) could tightly bind CABP. Binding of CABP to L8 induced a shift in the gel filtration profile due to apparent aggregation of L8. Aggregation did not occur with the L8S8-CABP complex nor with L8-CABP in the presence of 150 mM MgCl2. If ionic strength was increased with either KCl or MgCl2 during or after the binding of [14C]CABP to L8, [14C]CABP in the complex exchanged with [12C]CABP and was lost from the protein. Ionic strength strongly affected the rate constant (k4) for [14C]CABP dissociation from the L8-[14C]CABP complex, but had little effect on k4 for the L8S8-CABP complex. The differences in CABP binding characteristics between the L8-CABP and L8S8-CABP complexes demonstrate that S is intimately involved in maintaining the stability of the tight binding of CABP to the active site. These are the same interactions stabilizing the intermediate, 3-keto-2-carboxyarabinitol 1,5-bisphosphate, to native rubisco during CO2 fixation.

Oxalyl hydroxamates as reaction-intermediate analogues for ketol-acid reductoisomerase

N-Hydroxy-N-isopropyloxamate (IpOHA) is an exceptionally potent inhibitor of the Escherichia coli ketol-acid reductoisomerase. In the presence of Mg2+ or Mn2+, IpOHA inhibits the enzyme in a time-dependent manner, forming a nearly irreversible complex. Nucleotide, which is essential for catalysis, greatly enhances the binding of IpOHA by the reductoisomerase, with NADPH (normally present during the enzyme's rearrangement step, i.e., conversion of a beta-keto acid into an alpha-keto acid, in either the forward or reverse physiological reactions) being more effective than NADP. In the presence of Mg2+ and NADPH, IpOHA appears to bind to the enzyme in a two-step mechanism, with an initial inhibition constant of 160 nM and a maximum rate of formation of the tight, slowly reversible complex of 0.57 min-1 (values that give an association rate of IpOHA, at low concentration, of 5.9 X 10(4) M-1 s-1). The rate of exchange of [14C]IpOHA from an enzyme-[14C]IpOHA-Mg2(+)-NADPH complex with exogenous, unlabeled IpOHA has a half-time of 6 days (150 h). This dissociation rate (1.3 X 10(-6) s-1) and the association rate determined by inactivation kinetics define an overall dissociation constant of 22 pM. By contrast, in the presence of Mn2+ and NADPH, the corresponding association and dissociation rates for IpOHA are 8.2 X 10(4) M-1 s-1 and 3.2 X 10(-6) s-1 (half-time = 2.5 days), respectively, which define an overall dissociation constant of 38 pM. In the presence of NADP or in the absence of nucleotide (both in the presence of Mg2+), the enzyme-IpOHA complex is far more labile, with dissociation half-times of 28 and 2 h, respectively. In the absence of Mg2+ or Mn2+, IpOHA does not exhibit time-dependent inhibition of the reductoisomerase.(ABSTRACT TRUNCATED AT 250 WORDS)