Strategic protein target analysis for developing drugs to stop dental caries.

Dental caries is the most common disease to cause irreversible damage in humans. Several therapeutic agents are available to treat or prevent dental caries, but none besides fluoride has significantly influenced the disease burden globally. Etiologic mechanisms of the mutans group streptococci and specific Lactobacillus species have been characterized to various degrees of detail, from identification of physiologic processes to specific proteins. Here, we analyze the entire Streptococcus mutans proteome for potential drug targets by investigating their uniqueness with respect to non-cariogenic dental plaque bacteria, quality of protein structure models, and the likelihood of finding a drug for the active site. Our results suggest specific targets for rational drug discovery, including 15 known virulence factors, 16 proteins for which crystallographic structures are available, and 84 previously uncharacterized proteins, with various levels of similarity to homologs in dental plaque bacteria. This analysis provides a map to streamline the process of clinical development of effective multispecies pharmacologic interventions for dental caries.

Structural genomics of enzymes involved in sterol/isoprenoid biosynthesis.

X-ray structures of two enzymes in the sterol/isoprenoid biosynthesis pathway have been determined in a structural genomics pilot study. Mevalonate-5-diphosphate decarboxylase (MDD) is a single-domain alpha/beta protein that catalyzes the last of three sequential ATP-dependent reactions which convert mevalonate to isopentenyl diphosphate. Isopentenyl disphosphate isomerase (IDI) is an alpha/beta metalloenzyme that catalyzes interconversion of isopentenyl diphosphate and dimethylallyl diphosphate, which condense in the next step toward synthesis of sterols and a host of natural products. Homology modeling of related proteins and comparisons of the MDD and IDI structures with two other experimentally determined structures have shown that MDD is a member of the GHMP superfamily of small-molecule kinases and IDI is similar to the nudix hydrolases, which act on nucleotide diphosphatecontaining substrates. Structural models were produced for 379 proteins, encompassing a substantial fraction of both protein superfamilies. All three enzymes responsible for synthesis of isopentenyl diphosphate from mevalonate (mevalonate kinase, phosphomevalonate kinase, and MDD) share the same fold, catalyze phosphorylation of chemically similar substrates (MDD decarboxylation involves phosphorylation of mevalonate diphosphate), and seem to have evolved from a common ancestor. These structures and the structural models derived from them provide a framework for interpreting biochemical function and evolutionary relationships.

Homology-based annotation yields 1,042 new candidate genes in the Drosophila melanogaster genome.

The approach to annotating a genome critically affects the number and accuracy of genes identified in the genome sequence. Genome annotation based on stringent gene identification is prone to underestimate the complement of genes encoded in a genome. In contrast, over-prediction of putative genes followed by exhaustive computational sequence, motif and structural homology search will find rarely expressed, possibly unique, new genes at the risk of including non-functional genes. We developed a two-stage approach that combines the merits of stringent genome annotation with the benefits of over-prediction. First we identify plausible genes regardless of matches with EST, cDNA or protein sequences from the organism (stage 1). In the second stage, proteins predicted from the plausible genes are compared at the protein level with EST, cDNA and protein sequences, and protein structures from other organisms (stage 2). Remote but biologically meaningful protein sequence or structure homologies provide supporting evidence for genuine genes. The method, applied to the Drosophila melanogaster genome, validated 1,042 novel candidate genes after filtering 19,410 plausible genes, of which 12,124 matched the original 13,601 annotated genes. This annotation strategy is applicable to genomes of all organisms, including human.

Protein structure modeling for structural genomics.

The shapes of most protein sequences will be modeled based on their similarity to experimentally determined protein structures. The current role, limitations, challenges and prospects for protein structure modeling (using information about genes and genomes) are discussed in the context of structural genomics.

Site-directed mutagenesis of the active site serine290 in flavanone 3beta-hydroxylase from Petunia hybrida.

Flavanone 3beta-hydroxylase (FHT) catalyzes a pivotal reaction in the formation of flavonoids, catechins, proanthocyanidins and anthocyanidins. In the presence of oxygen and ferrous ions the enzyme couples the oxidative decarboxylation of 2-oxoglutarate, releasing carbon dioxide and succinate, with the oxidation of flavanones to produce dihydroflavonols. The hydroxylase had been cloned from Petunia hybrida and expressed in Escherichia coli, and a rapid isolation method for the highly active, recombinant enzyme had been developed. Sequence alignments of the Petunia hydroxylase with various hydroxylating 2-oxoglutarate-dependent dioxygenases revealed few conserved amino acids, including a strictly conserved serine residue (Ser290). This serine was mutated to threonine, alanine or valine, which represent amino acids found at the corresponding sequence position in other 2-oxoglutarate-dependent enzymes. The mutant enzymes were expressed in E. coli and purified to homogeneity. The catalytic activities of [Thr290]FHT and [Ala290]FHT were still significant, albeit greatly reduced to 20 and 8%, respectively, in comparison to the wild-type enzyme, whereas the activity of [Val290]FHT was negligible (about 1%). Kinetic analyses of purified wild-type and mutant enzymes revealed the functional significance of Ser290 for 2-oxoglutarate-binding. The spatial configurations of the related Fe(II)-dependent isopenicillin N and deacetoxycephalosporin C synthases have been reported recently and provide the lead structures for the conformation of other dioxygenases. Circular dichroism spectroscopy was employed to compare the conformation of pure flavanone 3beta-hydroxylase with that of isopenicillin N synthase. A double minimum in the far ultraviolet region at 222 nm and 208-210 nm and a maximum at 191-193 nm which are characteristic for alpha-helical regions were observed, and the spectra of the two dioxygenases fully matched revealing their close structural relationship. Furthermore, the spectrum remained unchanged after addition of either ferrous ions, 2-oxoglutarate or both of these cofactors, ruling out a significant conformational change of the enzyme on cofactor-binding.

MODBASE, a database of annotated comparative protein structure models.

MODBASE is a queryable database of annotated comparative protein structure models. The models are derived by MODPIPE, an automated modeling pipeline relying on the programs PSI-BLAST and MODELLER. The database currently contains 3D models for substantial portions of approximately 17 000 proteins from 10 complete genomes, including those of Caenorhabditis elegans, Saccharomyces cerevisiae and Escherichia coli, as well as all the available sequences from Arabidopsis thaliana and Homo sapiens. The database also includes fold assignments and alignments on which the models were based. In addition, special care is taken to assess the quality of the models. ModBase is accessible through a web interface at http://guitar.rockefeller.edu/modbase/

Defining the location and function of domains of McrB by deletion mutagenesis.

The GTP-dependent restriction endonuclease McrBC of E. coli K12, which recognizes cytosine-methylated DNA, consists of two protein subunits, McrB and McrC. We have investigated the structural assignment and interdependence of the McrB subunit functions, namely (i) specific DNA recognition and (ii) GTP binding and hydrolysis. Extending earlier work, we have produced McrB variants comprising N- and C-terminal fragments. The variants McrB1-162 and McrB1-170 are still capable of specific DNA binding. McrB169-465 shows GTP binding and hydrolysis characteristics indistinguishable from full-length McrB as well as wild-type like interaction with McrC. Thus, DNA and GTP binding are spatially separated on the McrB molecule, and the respective domains function quite independently.

The GTP-binding domain of McrB: more than just a variation on a common theme?

The methylation-dependent restriction endonuclease McrBC from Escherichia coli K12 cleaves DNA containing two R(m)C dinucleotides separated by about 40 to 2000 base-pairs. McrBC is unique in that cleavage is totally dependent on GTP hydrolysis. McrB is the GTP binding and hydrolyzing subunit, whereas MrC stimulates its GTP hydrolysis. The C-terminal part of McrB contains the sequences characteristic for GTP-binding proteins, consisting of the GxxxxGK(S/T) motif (position 201-208), followed by the DxxG motif (position 300-303). The third motif (NKxD) is present only in a non-canonical form (NTAD 333-336). Here we report a mutational analysis of the putative GTP-binding domain of McrB. Amino acid substitutions were initially performed in the three proposed GTP-binding motifs. Whereas substitutions in motif 1 (P203V) and 2 (D300N) show the expected, albeit modest effects, mutation in the motif 3 is at variance with the expectations. Unlike the corresponding EF-Tu and ras -p21 variants, the D336N mutation in McrB does not change the nucleotide specificity from GTP to XTP, but results in a lack of GTPase stimulation by McrC. The finding that McrB is not a typical G protein motivated us to perform a search for similar sequences in DNA databases. Eight microbial sequences were found, mainly from unfinished sequencing projects, with highly conserved sequence blocks within a presumptive GTP-binding domain. From the five sequences showing the highest homology, 17 invariant charged or polar residues outside the classical three GTP-binding motifs were identified and subsequently exchanged to alanine. Several mutations specifically affect GTP affinity and/or GTPase activity. Our data allow us to conclude that McrB is not a typical member of the superfamily of GTP-binding proteins, but defines a new subfamily within the superfamily of GTP-binding proteins, together with similar prokaryotic proteins of as yet unidentified function.

Role of the omega-loop in the activity, substrate specificity, and structure of class A beta-lactamase.

The structure of class A beta-lactamases contains an omega-loop associated with the active site, which carries a key catalytic residue, Glu166. A 16-residue omega-loop deletion mutant of beta-lactamase from Staphylococcus aureus PC1, encompassing residues 163-178, was produced in order to examine the functional and structural role of the loop. The crystal structure was determined and refined at 2.3 A, and the kinetics of the mutant enzyme was characterized with a variety of beta-lactam antibiotics. In general, the wild-type beta-lactamase hydrolyzes penicillin compounds better than cephalosporins. In contrast, the deletion of the omega-loop led to a variant enzyme that acts only on cephalosporins, including third generation compounds. Kinetic measurements and electrospray mass spectrometry revealed that the first and third generation cephalosporins form stable acyl-enzyme complexes, except for the chromogenic cephalosporin, nitrocefin, which after acylating the enzyme undergoes hydrolysis at a 1000-fold slower rate than that with wild-type beta-lactamase. Hydrolysis of the acyl-enzyme adducts is prevented because the deletion of the omega-loop eliminates the deacylation apparatus comprising Glu166 and its associated nucleophilic water site. The crystal structure reveals that while the overall fold of the mutant enzyme is similar to that of the native beta-lactamase, local adjustments in the vicinity of the missing loop occurred. The altered beta-lactam specificity is attributed to these structural changes. In the native structure, the omega-loop restricts the conformation of a beta-strand at the edge of the active site depression. Removal of the loop provides the beta-strand with a new degree of conformational flexibility, such that it is displaced inward toward the active site space. Modeled Michaelis complexes with benzylpenicillin and cephaloridine show that the perturbed conformation of the beta-strand is inconsistent with penicillin binding because of steric clashes between the beta-lactam side chain substituent and the beta-strand. In contrast, no clashes occur upon cephalosporin binding. Recognition of third generation cephalosporins is possible because the bulky side chain substituents of the beta-lactam ring typical of these compounds can be accommodated in the space freed by the deletion of the omega-loop.

Structural features of halophilicity derived from the crystal structure of dihydrofolate reductase from the Dead Sea halophilic archaeon, Haloferax volcanii.

BACKGROUND: The proteins of halophilic archaea require high salt concentrations both for stability and for activity, whereas they denature at low ionic strength. The structural basis for this phenomenon is not yet well understood. The crystal structure of dihydrofolate reductase (DHFR) from Haloferax volcanii (hv-DHFR) reported here provides the third example of a structure of a protein from a halophilic organism. The enzyme is considered moderately halophilic, as it retains activity and secondary structure at monovalent salt concentrations as low as 0.5 M. RESULTS: The crystal structure of hv-DHFR has been determined at 2.6 A resolution and reveals the same overall fold as that of other DHFRs. The structure is in the apo state, with an open conformation of the active-site gully different from the open conformation seen in other DHFR structures. The unique feature of hv-DHFR is a shift of the alpha helix encompassing residues 46-51 and an accompanied altered conformation of the ensuing loop relative to other DHFRs. Analysis of the charge distribution, amino acid composition, packing and hydrogen-bonding pattern in hv-DHFR and its non-halophilic homologs has been performed. CONCLUSIONS: The moderately halophilic behavior of hv-DHFR is consistent with the lack of striking structural features expected to occur in extremely halophilic proteins. The most notable feature of halophilicity is the presence of clusters of non-interacting negatively charged residues. Such clusters are associated with unfavorable electrostatic energy at low salt concentrations, and may account for the instability of hv-DHFR at salt concentrations lower than 0.5 M. With respect to catalysis, the open conformation seen here is indicative of a conformational transition not reported previously. The impact of this conformation on function and/or halophilicity is unknown.

The recognition of methylated DNA by the GTP-dependent restriction endonuclease McrBC resides in the N-terminal domain of McrB.

McrBC is a GTP-dependent restriction endonuclease of E. coli K12, selectively directed against DNA containing modified cytosine residues. McrB, one of its components, is responsible for the binding and, together with McrC, for the cleavage of DNAs containing two 5'-Pu(m)C sites separated by 40-80 base pairs. Gel retardation assays with wild-type and mutant McrB reveal that (i) single 5'-Pu(m)C sites in DNA can be sufficient to elicite binding by McrB. Binding to such substrates is, however, weak and strongly dependent on the sequence context of Pu(m)C sites. (ii) Strong DNA binding (K(ass) approximately 10(7)M[-1]) is dependent on the presence of at least two Pu(m)C sites, even if they are separated by less than 40 bp, and is modulated by the sequence context (-A(m)CCGGT- --> -A(m)CT(C/G)AGT- --> -AGG(m)CCT- --> -AAG(m)CTT-). (iii) DNA binding by McrB is accompanied by formation of distinct multiple complexes whose distribution is modulated by GTP. (iv) McrC, which cannot bind DNA by itself, moderately stimulates the DNA binding of McrB and converts McrB-DNA complexes to large aggregates. (v) Deletion of the C-terminal half of McrB, which harbors the three consensus sequences characteristic for guanine nucleotide binding proteins, leads to protein inactive in GTP binding and/or hydrolysis and in McrC-assisted DNA cleavage; the protein, however, remains fully competent in DNA binding. (vi) Mutations in McrB which lead to a reduction in GTP binding and/or hydrolysis can affect DNA binding, suggesting that the two activities are coupled in the full-length protein.

Characterization of the interaction between the restriction endonuclease McrBC from E. coli and its cofactor GTP.

McrBC, a GTP-dependent restriction enzyme from E. coli K-12, cleaves DNA containing methylated cytosine residues 40 to 80 residues apart and 3'-adjacent to a purine residue (PumCN40-80PumC). The presence of the three consensus sequences characteristic for guanine nucleotide binding proteins in one of the two subunits of McrBC suggests that this subunit is responsible for GTP binding and hydrolysis. We show here that (i) McrB binds GTP with an affinity of 10(6) M-1 and that GTP binding stabilizes McrB against thermal denaturation. (ii) McrB binds GDP about 50-fold and ATP at least three orders of magnitude more weakly than GTP. (iii) McrB hydrolyzes GTP in the presence of Mg2+ with a steady-state rate of approximately 0.5 min-1. (iv) McrC stimulates GTP hydrolysis 30-fold, but substrate DNA has no detectable effect on the GTPase activity of McrB, neither by itself nor in the presence of McrC. (v) Substitution of N339 and N376 with alanine allowed us to identify NTAD (339 to 342) rather than NKKA (376 to 379) as the equivalent of the third consensus sequence motif characteristic for guanine nucleotide binding proteins, NKXD.

Circularly permuted beta-lactamase from Staphylococcus aureus PC1.

The role that domain flexibility plays in the enzymatic activity of beta-lactamase from Staphylococcus aureus PC1 was investigated by producing two circularly permuted molecules. The C- and N-termini of the wild-type enzyme are adjacent to each other and remote from the active site, which is located between two domains. The polypeptide chain crosses over from one domain to the other twice. For the circularly permuted molecules, the termini were joined by an eight amino acid residue insertion, and new termini were introduced elsewhere. The first construct, termed cp254, was cleaved in a loop remote from the domain interface. The crystal structure of cp254 has been determined and refined at 1.8 A resolution, revealing essentially the same structure as that of the native protein. The activity profile with a representative sample of beta-lactam antibiotics is also very similar to that of wild-type beta-lactamase. The termini of the second circularly permuted mutant, cp228, occur within the second crossover region and therefore may enhance the flexibility of the molecule. Cp228 beta-lactamase shows a large decrease in enzymatic activity toward the sample of beta-lactam antibiotics, with catalytic rates that are 0.5-1% of those of the wild-type enzyme. One exception is the hydrolysis of the third generation cephalosporin, cefotaxime, which is hydrolyzed by the cp228 enzyme 10-fold faster than by wild-type beta-lactamase. Cp228 has not been crystallized. However, the circular dichroism spectra of the two circularly permuted proteins are very similar, indicating that, by analogy to cp254, cp228 adopts a global folded state. Thermal denaturation experiments reveal that cp254 is somewhat less stable than the wild-type enzyme, whereas cp228 is substantially less stable. Together, the data highlight the profound consequences that introducing flexibility at the domain interface has on both enzyme activity and protein stability.

Probing phosphatidylinositolphosphates and adenosinenucleotides on talin nucleated actin polymerization.

We have investigated the binding of PI, PIP and PIP2 to talin and the effect of phosphoinositides and adenosinenucleotides on talin-induced actin polymerization. At physiological salt concentrations, talin coprecipitates with liposomes when containing phosphoinositides but not when containing PI. The nucleating effect of talin as reflected by a twofold increase of fluorescence during the polymerization of actin labelled with NBD is not inhibited by phosphoinositides. The polymerization of ADP-actin versus ATP-actin was investigated in the presence and absence of talin by NBD fluorescence. ADP-actin nucleation induced by talin is comparably efficient as with ATP-actin. These experimental findings in summary have implications when evaluating the role of talin during cell activation.

Kinetic analysis of the cleavage of natural and synthetic substrates by the Serratia nuclease.

The extracellular nuclease from Serratia marcescens is a non-specific endonuclease that hydrolyzes double-stranded and single-stranded DNA and RNA with high specific activity. Steady-state and presteady-state kinetic cleavage experiments were performed with natural and synthetic DNA and RNA substrates to understand the mechanism of action of the Serratia nuclease. Most of the natural substrates are cleaved with similar Kcat and K(m) values, the Kcat/K(m) ratios being comparable to that of staphylococcal nuclease. Substrates with extreme structural features, like poly(dA).poly(dT) or poly(dG).poly(dC), are cleaved by the Serratia nuclease with a 50 times higher or 10 times lower K(m), respectively, as salmon testis DNA. Neither with natural DNA or RNA nor synthetic oligodeoxynucleotide substrates did we observe substrate inhibition for the Serratia nuclease as reported recently. Experiments with short oligodeoxynucleotides confirmed previous results that for moderately good cleavage activity the substrate should contain at least five phosphate residues. Shorter substrates are still cleaved by the Serratia nuclease, albeit at a rate reduced by a factor of more than 100. Cleavage experiments with oligodeoxynucleotides substituted by a single phosphorothioate group showed that the negative charge of the pro-Rp-oxygen of the phosphate group 3' adjacent to the scissile phosphodiester bond is essential for cleavage, as only the Rp-phosphorothioate supports cleavage at the 5' adjacent phosphodiester bond. Furthermore, the modified bond itself is only cleaved in the Rp-diastereomer, albeit 1000 times more slowly than the corresponding unmodified phosphodiester bond, which offers the possibility to determine the stereochemical outcome of cleavage. Pre-steady-state cleavage experiments demonstrate that it is not dissociation of products but association of enzyme and substrate or the cleavage of the phosphodiester bond that is the rate-limiting step of the reaction. Finally, it is shown that Serratia nuclease accepts thymidine 3',5'-bis(p-nitrophenyl)phosphate as a substrate and cleaves it at its 5'-end to produce nitrophenol and thymidine 3'-(p-nitrophenylphosphate) 5-phosphate. The rate of cleavage of this artificial substrate, however, is 6-7 orders of magnitude smaller than the rate of cleavage of macromolecular DNA or RNA.

The end of a polymerizing actin filament contains numerous ATP-subunit segments that are disconnected by ADP-subunits resulting from ATP hydrolysis.

ATP hydrolysis by copolymers of ATP-actin and ADP-actin was investigated in order to analyze the effect of interfaces between ATP-subunits and ADP-subunits on hydrolysis of actin-bound ATP. Copolymers of ATP- and ADP-subunits were formed by polymerization of ATP- and ADP-actin monomers onto filaments. By changing the ratio of polymerizing ATP-actin monomers to ADP-actin monomers, the number of interfaces between ATP- and ADP-subunits and of ATP-subunits only surrounded by further ATP-subunits was varied. The rate of actin polymerization and of ATP hydrolysis was measured simultaneously on the same samples. The lag time between incorporation of actin monomers into filaments and subsequent ATP hydrolysis was found to be similar both for polymerized ATP-actin and for copolymers formed by various ratios of ATP- to ADP-actin. The experiments were performed in the presence of 1 mM MgCl2, 0.05 mM CaCl2, and 100 mM KCl or of 1 mM MgCl2, and 0.4 mM EGTA. The type of cations was found to have no major effect on the rate of ATP hydrolysis. A quantitative evaluation of the experimental data suggests that ATP at interfaces between ATP- and ADP-subunits is hydrolyzed not more than 10 times faster than ATP of subunits surrounded by further ATP-subunits. On the basis of these results, one can conclude that an actin filament onto which ATP-actin monoMers polymerize contains numerous segments of ATP-subunits that are disconnected by ADP-subunits resulting from ATP hydrolysis. The average length of the numerous ATP segments of a steadily polymerizing filament is in the range of 10 ATP-subunits or below.

Structural evidence for the evolutionary divergence of mycoplasma from gram-positive bacteria: the histidine-containing phosphocarrier protein.

BACKGROUND: The three-dimensional structures of histidine-containing phosphocarrier protein (HPr), a member of the phosphoenolpyruvate:sugar phosphotransferase system (PTS), have been determined from Gram-negative and Gram-positive bacteria. The structure of HPr reported here for Mycoplasma capricolum is the first protein structure to be determined for this class of organism. Comparative structural studies with the bacterial proteins highlight sequence-structure correlations relevant to proposals about the evolutionary origin of mycoplasmas. RESULTS: The crystal structure of HPr from M. capricolum has been determined and refined at 1.8 A resolution, revealing the same overall fold as that of other HPrs of known structure. However, mycoplasma HPr resembles HPrs from Gram-positive bacteria more closely than those from Gram-negative bacteria. As in HPrs from Bacillus subtilis and Escherichia coli, the phosphoryl group carrier (His15) forms the N-terminal cap of a helix, but in contrast to the other crystal structures, the side chain of the adjacent Arg17 is conformationally disordered. A sulfate ion interacts with Ser46, a residue known to be phosphorylated in a regulatory manner. CONCLUSIONS: The greater degree of structural similarity of the M. capricolum HPr to HPrs from Gram-positive rather than Gram-negative bacteria is consistent with the proposal that mycoplasma evolved from Gram-positive bacteria. The proposal that no major conformational transition is required for phosphorylation of the active-site histidine is reinforced by comparing the crystal structures with and without an anion in the active site. The conformational disorder of the Arg17 side chain suggests that its guanidinium group does not have to form specific interactions with other protein groups before phosphorylation at His15. The association of a sulfate ion with Ser46 serves as a model for HPr(Ser46-P). As there is no evidence of a conformational change accompanying Ser46 phosphorylation, the inhibitory effect of this event may be attributable to altered surface electrostatics.

Molecular basis for biosynthesis and accumulation of polyhydroxyalkanoic acids in bacteria.

The current knowledge on the structure and on the organization of polyhydroxyalkanoic acid (PHA)-biosynthetic genes from a wide range of different bacteria, which rely on different pathways for biosynthesis of this storage polyesters, is provided. Molecular data will be shown for genes of Alcaligenes eutrophus, purple non-sulfur bacteria, such as Rhodospirillum rubrum, purple sulfur bacteria, such as Chromatium vinosum, pseudomonads belonging to rRNA homology group I, such as Pseudomonas aeruginosa, Methylobacterium extorquens, and for the Gram-positive bacterium Rhodococcus ruber. Three different types of PHA synthases can be distinguished with respect to their substrate specificity and structure. Strategies for the cloning of PHA synthase structural genes will be outlined which are based on the knowledge of conserved regions of PHA synthase structural genes and of the PHA-biosynthetic routes in bacteria as well as on the heterologous expression of these genes and on the availability of mutants impaired in the accumulation of PHA. In addition, a terminology for the designation of PHAs and of proteins and genes relevant for the metabolism of PHA is suggested.

Identification, cloning and sequence analysis of the poly(3-hydroxyalkanoic acid) synthase gene of the gram-positive bacterium Rhodococcus ruber.

The first polyhydroxyalkanoic acid (PHA) synthase gene (phbCRr) of a Gram-positive bacterium was cloned from a genomic library of Rhodococcus ruber in the broad-host-range plasmid vector pRK404. The hybrid plasmid harboring phbCRr allowed the expression of polyhydroxybutyric acid (PHB) synthase activity and restored the ability of PHB synthesis in a PHB-negative mutant of Alcaligenes eutrophus. Nucleotide sequence analysis of phbCRr revealed an open reading frame of 1686 bp starting with the rare codon TTG and encoding a protein of relative molecular mass 61,371. The deduced amino acid sequence of phbCRr exhibited homologies to the primary structures of the PHA synthases of A. eutrophus and Pseudomonas oleovorans. Preparation of PHA granules by discontinuous density gradient centrifugation of crude cellular extracts revealed four major bands in an SDS polyacrylamide gel. A Mr 61,000 protein was identified as the PHA synthase of R. ruber by N-terminal amino acid sequence determination.