10
Eur. J. Biochem. 247, 1046-1055 (1997) 0 FEBS 1991 Biochemical characterisation of ornithine carbamoyltransferase from Pyrococcus furiosus Christianne LEGRAIN', Vincent VILLERET'.', Martine ROOVERS4.5, Daniel GIGOT6, Otto DIDEBERG3, Andrt PIERARD1.6 and Nicolas GLANSDORFF'.'.' ' Institut de Recherches du Centre d'Enseignement et de Recherches des Industries Alimentaires, Commission de la Communautt FranGaise, * Laboratorium voor Eiwitbiochemie en Eiwitengineering,Universiteit Gent, Belgium ' Laboratoire de Cristallographie Macromoltculaire, Institut de Biologie Structurale Jean-Pierre Ebel (CEA-CNRS), Grenoble, France Laboratorium voor Erfelijkheidsleer en Microbiologie, Vrije Universiteit Brussel, Belgium ' Vlaams Interuniversitair Instituut voor Biotechnologie, Bruxelles, Belgium ' Laboratoire de Microbiologie, Facult6 des Sciences, Universite Libre de Bruxelles, Belgium (Received 21 February 1997) - EJB 97 029814 Bruxelles, Belgium Ornithine carbamoyltransferase (OTCase) was purified to homogeneity from the hyperthermophilic archaeon Pyrococcus furiosus. The enzyme is a 400 2 20-kDa polymer of a 35-kDa subunit, in keeping with the corresponding gene sequence [Roovers, M., Hethke, C., Legrain, C., Thomm, M. & Glansdorff, N. (1997) Isolation of the gene encoding Pyrococcus furiosus ornithine cabamoyltransferase and study of its expression profile in vivo and in vitro, Eur: J. Biochem. 247, 1038-10451. In contrast with the dode- cameric catabolic OTCase of Pseudomonas aeruginosa, l? furiosus OTCase exhibits no substrate coopera- tivity. In keeping with other data discussed in the text, this suggests that the enzyme serves an anabolic function. Half-life estimates for the purified enzyme ranged over 21-65 min at 100°C according to the experimental conditions and reached several hours in the presence of ornithine and phosphate. The sta- bility was not markedly influenced by the protein concentration. Whereas comparative examination of OTCase sequences did not point to any outstanding feature possibly related to thermophily, modelling the enzyme on the X-ray structure of I? aeruginosa OTCase (constituted by four trimers assembled in a tetrahedral manner) suggests that the molecule is stabilized, at least in part, by a set of hydrophobic interactions at the interfaces between the trimers. The comparison between f? aeruginosa and l? furiosus OTCases suggests that two different properties, allostery and thermostability, have been engineered start- ing from a similar quaternary structure of high internal symmetry. Recombinant l? furiosus OTCase synthesised by Escherichia coli proved less stable than the native enzyme. In Saccharomyces cerevisiae, however, an enzyme apparently identical to the native one could be obtained. Keywords: ornithine carbamoyltransferase; Pyrococcus furiosus; hyperthermophiles; thermostability; Archaea. Ornithine carbamoyltransferase catalyses the sixth step of the de novo biosynthesis of arginine. This enzyme occupies a key position in nitrogen metabolism since one of its substrates, carbamoylphosphate, is also converted by the collateral aspartate carbamoyltransferase (ATCase) into carbamoylaspartate, precur- sor of uridine monophosphate, while the other substrate, orni- thine, is a direct source of putrescine (Fig. 1). Changes in OTCase activity can therefore influence the fluxes of carbon and nitrogen in the related pyrimidine and polyamine pathways. When the arginine deiminase pathway is also present, a catabolic OTCase converts citrulline into ornithine and carbamoylphos- phate which can serve as source of energy. Since the pioneering Correspondence to N. Glansdorff, Research Institute, CERIA- COOVI, avenue E. Gryson 1, B-1070 Bruxelles, Belgium Phone: +33 2 5261275. Fax: +32 2 5261273. Abbreviations. ATCase, aspartate carbamoyltransferase; CPSase, carbamoylphosphate synthetase; Om[AcPO(OH,)], d-N-phosphonoace- tyl-L-ornithine; OTCase, ornithine carbamoyltransferase. Enzyme. Aspartate carbamoyltransferase (EC 2.1.3.2) ; carbamoyl- phosphate synthetase (EC 6.3.5.5); ornithine carbamoyltransferase (EC 2.1.3.3). work of Gorini and Maas (1957), studies of OTCase in various bacteria and yeast have shown that these metabolic functions are exquisitely regulated (reviewed in Cunin et al., 1986; Davis, 1986; Glansdorff, 1996). Carbamoyltransferases have become a model system for the analysis of structure function relationships and molecular evolu- tion. Since the last OTCase survey (15 sequences, Ruepp et al., 1995), our recent analysis of 64 carbamoyltransferase sequences originating from the three kingdoms of life (LabCdan et al., un- published results) confirmed that all known OTCases are similar and that they are related to ATCases. Crystallographic studies of the catabolic OTCase from Pseudomonas aeruginosa have provided the first detailed three-dimensional OTCase structure (Villeret et al., 1995). This context makes the study of OTCases from extremophilic organisms particularly attractive. In this pa- per we report the characterisation of an OTCase from Pyrococ- cus furiosus, an hyperthermophilic archaeon with an optimal growth temperature of 102°C (Fiala and Stetter, 1986). Several features call attention to thermophilic carbamoyltransferases and to OTCases in particular. First, carbamoylphosphate is a very thermolabile and potentially toxic substance at high temper-

Biochemical Characterisation of Ornithine Carbamoyltransferase from Pyrococcus Furiosus

Embed Size (px)

Citation preview

Eur. J. Biochem. 247, 1046-1055 (1997) 0 FEBS 1991

Biochemical characterisation of ornithine carbamoyltransferase from Pyrococcus furiosus Christianne LEGRAIN', Vincent VILLERET'.', Martine ROOVERS4.5, Daniel GIGOT6, Otto DIDEBERG3, Andrt PIERARD1.6 and Nicolas GLANSDORFF'.'.' ' Institut de Recherches du Centre d'Enseignement et de Recherches des Industries Alimentaires, Commission de la Communautt FranGaise,

* Laboratorium voor Eiwitbiochemie en Eiwitengineering, Universiteit Gent, Belgium ' Laboratoire de Cristallographie Macromoltculaire, Institut de Biologie Structurale Jean-Pierre Ebel (CEA-CNRS), Grenoble, France

Laboratorium voor Erfelijkheidsleer en Microbiologie, Vrije Universiteit Brussel, Belgium ' Vlaams Interuniversitair Instituut voor Biotechnologie, Bruxelles, Belgium ' Laboratoire de Microbiologie, Facult6 des Sciences, Universite Libre de Bruxelles, Belgium

(Received 21 February 1997) - EJB 97 029814

Bruxelles, Belgium

Ornithine carbamoyltransferase (OTCase) was purified to homogeneity from the hyperthermophilic archaeon Pyrococcus furiosus. The enzyme is a 400 2 20-kDa polymer of a 35-kDa subunit, in keeping with the corresponding gene sequence [Roovers, M., Hethke, C., Legrain, C., Thomm, M. & Glansdorff, N. (1997) Isolation of the gene encoding Pyrococcus furiosus ornithine cabamoyltransferase and study of its expression profile in vivo and in vitro, Eur: J . Biochem. 247, 1038-10451. In contrast with the dode- cameric catabolic OTCase of Pseudomonas aeruginosa, l? furiosus OTCase exhibits no substrate coopera- tivity. In keeping with other data discussed in the text, this suggests that the enzyme serves an anabolic function. Half-life estimates for the purified enzyme ranged over 21-65 min at 100°C according to the experimental conditions and reached several hours in the presence of ornithine and phosphate. The sta- bility was not markedly influenced by the protein concentration. Whereas comparative examination of OTCase sequences did not point to any outstanding feature possibly related to thermophily, modelling the enzyme on the X-ray structure of I? aeruginosa OTCase (constituted by four trimers assembled in a tetrahedral manner) suggests that the molecule is stabilized, at least in part, by a set of hydrophobic interactions at the interfaces between the trimers. The comparison between f? aeruginosa and l? furiosus OTCases suggests that two different properties, allostery and thermostability, have been engineered start- ing from a similar quaternary structure of high internal symmetry. Recombinant l? furiosus OTCase synthesised by Escherichia coli proved less stable than the native enzyme. In Saccharomyces cerevisiae, however, an enzyme apparently identical to the native one could be obtained.

Keywords: ornithine carbamoyltransferase; Pyrococcus furiosus; hyperthermophiles; thermostability; Archaea.

Ornithine carbamoyltransferase catalyses the sixth step of the de novo biosynthesis of arginine. This enzyme occupies a key position in nitrogen metabolism since one of its substrates, carbamoylphosphate, is also converted by the collateral aspartate carbamoyltransferase (ATCase) into carbamoylaspartate, precur- sor of uridine monophosphate, while the other substrate, orni- thine, is a direct source of putrescine (Fig. 1). Changes in OTCase activity can therefore influence the fluxes of carbon and nitrogen in the related pyrimidine and polyamine pathways. When the arginine deiminase pathway is also present, a catabolic OTCase converts citrulline into ornithine and carbamoylphos- phate which can serve as source of energy. Since the pioneering

Correspondence to N. Glansdorff, Research Institute, CERIA- COOVI, avenue E. Gryson 1, B-1070 Bruxelles, Belgium

Phone: +33 2 5261275. Fax: +32 2 5261273. Abbreviations. ATCase, aspartate carbamoyltransferase; CPSase,

carbamoylphosphate synthetase; Om[AcPO(OH,)], d-N-phosphonoace- tyl-L-ornithine; OTCase, ornithine carbamoyltransferase.

Enzyme. Aspartate carbamoyltransferase (EC 2.1.3.2) ; carbamoyl- phosphate synthetase (EC 6.3.5.5); ornithine carbamoyltransferase (EC 2.1.3.3).

work of Gorini and Maas (1957), studies of OTCase in various bacteria and yeast have shown that these metabolic functions are exquisitely regulated (reviewed in Cunin et al., 1986; Davis, 1986; Glansdorff, 1996).

Carbamoyltransferases have become a model system for the analysis of structure function relationships and molecular evolu- tion. Since the last OTCase survey (15 sequences, Ruepp et al., 1995), our recent analysis of 64 carbamoyltransferase sequences originating from the three kingdoms of life (LabCdan et al., un- published results) confirmed that all known OTCases are similar and that they are related to ATCases. Crystallographic studies of the catabolic OTCase from Pseudomonas aeruginosa have provided the first detailed three-dimensional OTCase structure (Villeret et al., 1995). This context makes the study of OTCases from extremophilic organisms particularly attractive. In this pa- per we report the characterisation of an OTCase from Pyrococ- cus furiosus, an hyperthermophilic archaeon with an optimal growth temperature of 102°C (Fiala and Stetter, 1986). Several features call attention to thermophilic carbamoyltransferases and to OTCases in particular. First, carbamoylphosphate is a very thermolabile and potentially toxic substance at high temper-

Legrain et al. (Eur J. Biochem. 247) 1047

atures; in l? furiosus (Legrain et al., 1995) and Thermus aquat- icus Z05 (Van de Casteele et al., 1997) carbamoylphosphate actually appears to be protected from the bulk of the aqueous phase by channelling towards carbamoylaspartate or citrulline. The molecular analysis of the protein-protein interactions as- sembling the cognate enzymatic complex requires a detailed study of the partner proteins. Second, the mode of synthesis of carbamoylphosphate in l? furiosus appears original in not being glutamine-dependent (as in all other micro-organisms investi- gated) but ammonia-dependent (Legrain et al., 1995; see also Purcarea et al., 1996, for the related species Pyrococcus abyssi). As the same carbamoylphosphate synthetase (CPSase) appears to channel carbamoylphosphate towards citrulline and carbamo- ylaspartate (Durbecq, V. and Legrain, C., unpublished work), understanding the partitioning of this substance between the ar- ginine and pyrimidine pathways also requires detailed knowl- edge of the carbamoyltransferases involved. Third, we have re- ported that the intrinsic thermostability of Thermus aquaticus Z05 ATCase (Van de Casteele et al., 1994) and Thermotoga maritima OTCase (Van de Casteele et al., 1997) seem incom- patible with the optimal growth temperature of their host. In the case of l? furiosus the thermostability of OTCase in crude extracts at first sight appeared sufficient (a half-life of 30 min was observed at 100°C; Van de Casteele et al., 1997) but the temperature profile of the purified enzyme remained to be deter- mined.

As the genetic analysis of structure function relationships in P. furiosus OTCase requires expression of the corresponding gene in a foreign host, we have compared recombinant OTCases produced in Escherichia coli and in Saccharomyces cerevisiae and found, somewhat to our surprise, that yeast, but not E. coli, produces a protein with properties identical to those of the native enzyme. Recently, Roovers et al. (1997) reported the isolation and characterisation of the l? furiosus OTCase gene (argF) and analysed its expression profile in vitro and in vivo.

EXPERIMENTAL PROCEDURES

Bacterial strains and culture conditions. Pyrococcus furio- sus Vcl (DSM 3638) was kindly provided by Professor K. 0. Stetter (Regensburg). Bacteria were grown in a complex medium based on artificial sea water (Fiala et al., 1986) supplemented with 0.1 % yeast extract and 0.5 % peptone. Cells were grown at 95°C in a Braun Biostat U fermenter in 60-1 batch, with con- tinuous nitrogen sparging at 400 ml min-' to maintain anaerobic conditions and agitation at 200rpm. Cells were harvested by centrifugation at the end of the exponential growth phase (about 10' cells ml-') and washed in 3% (mass/vol.) NaC1.

E. coli recombinant strain C600 OTC-/pPFl and S. cerevi- siae recombinant strain S S l/pPF2, and their growth conditions have been described previously by Roovers et al. (1997). The E. coli protease-deficient strain L798 was kindly provided by Dr C. J. Lusty (Guillou et al., 1989).

Ornithine carbamoyltransferase assay. The enzyme was assayed by measuring the formation of citrulline as described by Stalon et al. (1972). Ornithine carbamoyltransferase activity was measured at 55°C because of the thermal lability of carbamoylphosphate (Legrain et al., 1995). The reaction mixture contained, in a final volume of 2.0 ml : 200 mM Tris/HCl pH 7.3 (at 25 "C), 2 mM ornithine, 10 mM carbamoylphosphate and ex- tract. Unless otherwise specified, the incubation time was 5 min. Formation of citrulline was linear during this time. When still greater sensitivity was required, a ['4C]carbam~ylpho~phate assay was used (Legrain and Stalon, 1976). The composition of the reaction mixture was the same as described above except

that the final volume was 1.0 ml. Assays at higher temperatures were performed by measuring the appearance of carbon dioxide formed in the reverse reaction, from ~-[carbamoyl-'~C]citrulline, as described by Legrain and Stalon (1976). The reaction mixture contained, in a final volume of 2.0 ml: 400 mM sodium arsenate pH 6.0, 20 mM ~-[carbamoyl-'~C]citrulline (0.1 pCi pmol-') and extract. Protein concentration was about 1 mg ml-' in the assay and the incubation time was 10 min.

Determination of protein. Protein was measured by the Lowry method, or in the last purification steps by ultraviolet absorption.

Preparation of 6-N-phosphonoacetyl-L-ornithine coupled to epoxy-activated Sepharose 6B. S-N-Phosphonoacetyl-L-or- nithine { Orn[AcPO(OH),]} was prepared as described by Pen- ninckx and Gigot (1978) and reacted with epoxy-activated Se- pharose 6B as described in the Pharmacia handbook on affinity chromatography. Approximately 5 pmol Orn[AcPO(OH),]/ml resin was coupled.

Enzyme purification. Cell disruption. Frozen cells (40 g wet mass) were thawed in 160 ml 50 mM Tris/HCl pH 7.3 and disrupted by sonication for 15 min in Raytheon sonic oscillator (250 W, 10 kHz). Following sonication, the suspension was centrifuged at 20000 g for 30 min. The supernatant was col- lected.

Ammonium sulphate precipitation. Solid ammonium sul- phate was added to the supernatant described above to 40% sat- uration and stirred for 1 h. The solution was centrifuged at 20000 g for 30 min. The supernatant was brought to 80% satu- ration, stirred for 1 h, and then centrifuged at 20000 g for 30min. The pellet was suspended in 50mM TrisMCl pH 7.3 and extensively dialysed against this buffer.

Ion-exchange chromatography. The dialysed solution was applied to a DEAE-Sepharose CL6B column (1.6 X 25 cm) equil- ibrated with 50 mM Tris/HCl pH 7.3. The ornithine carbamoyl- transferase activity was eluted with a linear gradient of 0-0.5 M KCl, dissolved in 50 mM TrisMCl pH 7.3. Active fractions were pooled and extensively dialysed against 50 mM Tris/HCI pH 7.3.

AfSinity chromatography. The dialysed solution was run through a column (1.6X 10 cm) of Orn[AcPO(OH),-Sepharose equilibrated with 50 mM TrisMCl pH 7.3 supplemented with 1 O O m M KCI. The column was washed with the same buffer until no more protein eluted from the column (100 ml). Orni- thine carbamoyltransferase activity was eluted with 100 mM carbamoylphosphate dissolved in the same buffer. Active frac- tions were pooled and dialysed against 50 mM Tris/HCl pH 7.3.

Molecular mass determination. The molecular mass of the native enzyme was determined by gel filtration on a Pharmacia FPLC system fitted with a Superose PI2 HR 10/30 column (eluted with 50 mM-Tris/HCl pH 7.3 containing 100 mM NaCI). The gel-filtration column was calibrated using the following standards : thyroglobulin (669 kDa), ferritin (440 m a ) , catalase (232 kDa), aldolase (158 kDa). The elution volume (V,) of each standard was determined from the absorbance at 280nm. OTCase was detected by activity measurements. The void vol- ume (V,) was determined by blue dextran exclusion.

PAGE. Electrophoresis under non-denaturing conditions and in the presence of SDS were performed on 8-25 % gradient gels in the Phast System (Pharmacia). Protein bands were visualised by staining with Coomassie brilliant blue.

Determination of PI by isoelectric focusing. Analytical isoelectric focusing was performed in the pH range 3.0-9.0 on a polyacrylamide gel slab in a Phast System apparatus (Phar- macia). Gels were stained for proteins with Coomassie brillant blue.

pH optimum, optimal temperature and kinetic constants. The pH optimum for the forward reaction was determined at

1048

Argininosuccina te

Putrescine

Legrain et al. (Em J. Biochem. 247)

Ornithine carbamoyl transferase

Ornithine

Carbarnoyl- pi phophate . .

Aspartate

ATP+ AMP Aspartate + PPi

Y

Fumarate

Carbamoylaspartate Pyrimidines Aspartate carbamoyl transferase

Fig. 1. Metabolic reactions involving carbamoylphosphate, a precursor common to arginine and the pyrimidines. Hatched lines indicate anabolic reactions, thin black lines refer to the catabolic arginine deiminase pathway.

Table 1. Purification of Pyrococcus furiosus ornithine carbamoyltransferase.

Purification step Protein Total activity Specific activity Purification Yield

mg U U/mg -fold %

Crude extract 3510 141 876 40 100 Ammonium sulphate precipitation 2400 119627 50 1.3 84 DEAE-Sepharose chromatography 478 97 980 205 5.1 69 Orn [AcPO(OH),] - Sepharose chromatography 27 97 600 3642 91 69

55°C in 250 mM Tris/HC1 pH 6.0-8.0 and in 250 mM potas- sium phosphate pH 6.0-8.0. The pH was measured at the assay temperature.

The temperature dependence of the reaction was studied un- der standard assay conditions over 30-65°C for the forward reaction and 40-90°C for the reverse reaction. Kinetic con- stants for substrates were determined at 55°C in 200 mM Tris/ HCI pH 7.3 (at 25OC).

Heat stability tests. The thermal stability of ornithine car- bamoyltransferase was investigated by incubating enzyme solu- tions either in sealed Eppendorf tubes or in sealed glass capillary tubes. At each desired time, samples were cooled in ice and assayed after centrifugation.

Model building. The TURBO-FRODO software (Roussel and Cambillau, 1989) was used to build the three-dimensional model, starting from the FI aeruginosa OTCase coordinates (Vil- leret et al., 1995). Side chains conserved between the P. aerugi- nma and t? furiosus sequences were kept at the orientations ob- served in the crystal structure. Other side chains were positioned in their most probable conformation following Summers et al. (1987). The model was energy-minimised using the simulated annealing molecular dynamics procedure implemented in Xplor 3.1 (Briinger, 1992). During the minimisation, Cu positions were strictly constrained and the exact 23 point group symmetry of the dodecamer imposed (strict non-crystallographic symmetry option Xplor) in order to avoid model divergence from the crys- tal structure. Conserved side chains between the thermophilic and catabolic enzyme were also constrained in their observed conformation. Following the molecular dynamic simulation 100 cycles of least-squares minimisation were applied. After energy minimisation, the empirical energy term representing electro- static interactions equals 25 338 kJ molP and the term represent- ing Van der Waal's interactions equals -61 488 kJ mol -'.

Fig.2. Native-PAGE (A) and SDSA'AGE (B) of purified I! furiosus ornithine carbamoyltransferase. Electrophorese were performed as de- scribed in Experimental Procedures. Marker proteins : (A), thyroglobulin, 669 kDa; ferritin, 440 kDa; catalase, 232 kDa; lactic dehydrogenase, 140 kDa and albumin, 67 kDa; (B), phosphorylase b, 94 m a ; albumin, 67 kDa; ovalbumin, 43 kDa; carbonic anhydrase, 30 kDa; trypsin inhibi- tor, 20 kDa and lactalbumin, 14 kDa.

RESULTS Enzyme purification and physical properties. Table 1 summa- rises the results of a typical purification procedure. Affinity chromatography on Orn[AcPO(OH),] -Sepharose (see Experi- mental Procedures) represents the most effective purification

Legrain et al. (Eur J. Biochern. 247) 1049

0 10 20 30 40 50 [ornithine](mM)

0.3 B l

0.2 -

0.1 -

0 1 [ornithine](mM)

0.0 0.5 1.0 1.5 2.0 [ornithine]( m M)

Fig. 3. Saturation curves of f! furiosus ornithine carhamoyltransferase by ornithine. The activity assays were performed under the conditions described in Experimental Procedures. Insert: Hanes-Woolfs reciprocal plot.

step. OTCase was purified 390-fold over the activity in the cell extract, with a final yield of 69%. Native-PAGE of the purified OTCase revealed a single protein band (Fig. 2A). SDSPAGE gave three bands corresponding to 35, 71 and 148kDa (Fig. 2B). As there are three Cys residuedmonomer, we per- formed the same experiment after having submitted the purified enzyme to conditions expected to reduce and alkylate putative S-S bridges (Wdxdal et al., 1968); the same three-band pattern was obtained (data not shown). Treating the enzyme for 10 min at temperatures between 100-120°C did not change the pattern either. At 122°C the protein was progressively hydrolysed. Complete monomerization of P: furiosus OTCase could there- fore not be achieved. However, as the electrophoreses following the heat treatment were performed at 15"C, we can not exclude that monomerization actually occurred but was followed by reas- sociation at lower temperature. The N-terminal amino acid se- quences determined for the three bands were identical (VVSLAGRDLLC) and correspond to the N-terminus predicted from the DNA sequence (Roovers et al., 1997).

The molecular mass of the purified OTCase was determined by gel-filtration on a calibrated Superose column (two indepen- dent runs, see Experimental Procedures for marker protein) ; the enzyme was eluted at a position corresponding to a molecular mass of 400 2 20 kDa.

Isoelectric focusing of the purified enzyme showed a band corresponding to a PI of 6.1 (data not shown). The value pre- dicted from the sequence is 5.9.

pH optimum, kinetics constants and temperature depen- dence. The optimal pH of native I? furiosus OTCase at 55°C was found close to 6.5, whether the assay was performed in Trisl HC1 or in potassium phosphate buffer (data not shown).

The apparent Michaelis constants (KZp) were determined for both substrates. When ornithine was the variable substrate, at a saturating level of carbamoylphosphate, the saturation curve displayed a slight substrate inhibition (Fig. 3). It is known from other OTCases that only the zwitterionic species of ornithine binds the enzyme productively, while the cationic species acts as an inhibitor (Snodgrass, 1968; Legrain and Stalon, 1976; Kuo et al., 1985). From velocities measured at low ornithine concen- trations (up to 2 mM, Fig. 3) and in the presence of 10 mM carbamoylphosphate, using the Lineweaver-Burk and Hanes- Woolfs plots, we derived a PZp for ornithine of 0.13 ? 0.03 mM. When carbamoylphosphate was the variable substrate, the satu- ration curve appeared michaelian; F! furiosus OTCase does not display the strong cooperativity towards carbamoylphosphate

Table 2. Thermal stability of Pyrococcusfuriosus ornithine carbamo- yltransferase. All buffers were pH7.0 at 100°C. All buffers were pH 7.0 at 100°C.

Buffer Half-life Enzyme Reactant concn

mg/ml

20 mM TridHCl 0.025 20 mM Tris/HCI 0.1

100 mM TridHC1 0.1 20 mM Pipes 0.1 20 mM Tris/HCI 0.025 20 mM Tris/HCI 0.025 20 mM TridHCl 0.025

(mM) none none none none phosphate (100) ornithine (100) phosphate (100) + ornithine (100)

~

inin

19-21 38-42 44-50 58-64

>120 >500

>I200

which is characteristic of the catabolic OTCase of I? aeruginosa and actually restricts the functioning of this enzyme to the phos- phorolysis of citrulline at physiological carbamoylphosphate concentrations (Stalon et al., 1972). The value of @' of I? furio- sus OTCase for carbamoylphosphate determined in the presence of the optimal ornithine concentration (4mM) was 0.13 -C 0.02 mM.

The enzyme was strongly inhibited by the bisubstrate ana- logue Orn[AcPO(OH)], with an apparent K, of about 0.1 pM. This fact was taken advantage of in the purification described in the previous section.

The effect of temperature on the velocity of citrulline synthe- sis was studied between 30-65 "C, carbamoylphosphate thermo- lability (Legrain et al., 1995) preventing the use of higher tem- peratures; the factor by which the velocity was increased on raising the temperature 10°C (QIJ was 1.7 and the activation energy (EJ 48.1 kJ mol-' (Fig. 4). When the reaction was studied in the other direction, by monitoring citrulline arseno- lysis up to 90°C, a linear Arrhenius relationship was obtained, with an E, of 70.3 kJ mol-' and a Qlo of 2. With a Qlo of 1.7, citrulline synthesis would be more than 30 times faster in the optimal range of temperature o f f ? furiosus than at 37°C.

Thermal stability. In 1.0-ml samples containing about 1.0 mg crude extract protein, half-lives of F! furiosus OTCase amounted to 30-40 min at 100°C and 200-225 min at 90°C (Van de Cas- teele et al., 1997 and more recent unpublished experiments). These values were not significantly affected by dialysing the

1050 Legrain et al. IEui: J. Biochern. 247l

0' 30 40 50 60

Temperature ("C)

2.7 2.8 2.9 3.0 3.1 3.2 lrr(K) x 1000 i i

40 50 60 70 80 90 Temperature ("C)

Fig. 4. Temperature dependence of I? furiosus ornithine carbamoyltransferase activity for the forward (A) and reverse (B) reactions. The activity assays were performed under the conditions described in Experimental Procedures. Insert: Arrhenius diagram.

P-aer P - f u r

P-aer P-fur

P-aer P-fur

P-aer P-fur

P-aer P-fur

B1 H1 82 H2 - _ _ . . . . . . . . . . . . . . . . . . . . _ _ _ _ _ _ _ _ * ************

~ L S L ~ L Q L ~ D ~ ~ ~ T E ~ H L K ~ - ~ - I F ~ T S ~ ~ C ~ ~ Y - ~ 7 0 W ~ G ~ L L C ~ Y ~ E E ~ ~ L ~ T ~ ~ K I W Q K I G K P ~ L L E G K T - ~ I F Q K P S ~ ~ V ~ F ~ 70

H 9 810 B10 * * * * * * * * * * * * * * * -_---

E l 0 ' Ell H12 * * * * * * * * * * **** *******************

E T K V G K Q I A E Q Y P N L A N G I E V T ~ ~ S P ~ A F ~ ~ ~ I K A ~ L V S T ~ I _-----__-________ GEEVTQWDSPNS_vvW_DgAENRL - U Q K A W m B I K F

335 3 1 3

Fig. 5. Alignment of €! aeruginosa catabolic ornithine carbamoyltransferase and P. furiosus ornithine carbamoyltransferase. Residues strictly conserved between the two sequences are indicated in bold and similar residues are underlined. Above the alignments are the secondary structure regions as defined by Villeret et al. (1995); regions of a-helices are indicated by an asterisk and regions with P-strands are indicated with broken line.

extracts, suggesting that the enzyme was not being stabilised by substances of small molecular mass to any appreciable extent (data not shown). When pure OTCase was submitted to heat treatment, the intrinsic thermostability of the protein was con- firmed and did not appear markedly concentration-dependent : at 100°C, and for the higher enzyme concentration (0.1 mg ml-l in the sample treated) half-lives varied over 40-65 min, depend- ing on the buffer used (Table 2); at a quarter of this concentra- tion the half-life was still 20 min i n Tris/HCI buffer. The pres- ence of ornithine andor phosphate during the heat treatment had a striking protective effect on the activity of the purified enzyme (Table 2).

Analysis of the amino acid sequence. The nucleotide sequence of the argF gene was determined as described by Roovers et al. (1997). The derived amino acid sequence is presented in Fig. 5 along with that of urcB protein of P. aeruginosa. The first 11 amino-terminal residues were determined by Edman degradation

and found to be identical to the codon-predicted ones (starting with the Va12). The full alignment and the phylogenetic analysis of the 33 known OTCase sequences is being discussed else- where (LabCdan et al., unpublished results). The gene presents extensive overall similarity with all other OTCase determinants isolated up to now. From the functional point of view, it is sig- nificant that the motifs most strongly conserved among OTCases and containing residues involved in interactions with carbamoylphosphate (Ser56, Thr57, Arg58, Thr59, Arg60, Argl07, His134; I? aeruginosu numbering) or ornithine (His272, Cys273, Leu274, Pro275) are integrally present in I? furiosus OTCase (see Lerner and Switzer, 1986; Kuo et al., 1988, 1990; Villeret et al., 1995). Moreover, Asp231, recently postulated to interact with the a-amino group of ornithine (Vil- leret et al., 1995; Murata and Schachman, 1996) is conserved as well. However, comparative examination of OTCase sequences did not bring to light any outstanding feature that could be re- lated to the thermal properties of P furiosus OTCase. Attempts

Legrain et al. ( E m J. Biochem. 247) 1051

CP domain Ornithine domain Fig. 6. Ribbon representation of P. furiosus ornithine carhamoyl- transferase monomer. The secondary structures are defined in Fig. 5. The figure was generated using the MOLSCRIPT program (Kraulis, 1991).

to crystallise the enzyme have not provided suitably diffracting crystals. Consequently we resorted to model building for further analysis of the sequence.

Modelling of E! furiosus OTCase. The model of l? furiosus OTCase has been elaborated from the X-ray structure of Po aeru- ginosa OTCase (gene arc@ reported at a resolution of 3 A (Vil- leret et al., 1995). This oligomer is composed of four trimers assembled in a structure of 23 point group symmetry; therefore the enzyme has a high internal symmetry, with three twofold symmetry axes and four threefold symmetry axes. This catabolic OTCase is also characterised by a high molecular mass similar to that determined for f! furiosus OTCase, suggesting a common

Fig.8. Stereo view of three helices H1 (from three different mono- mers) forming the interface around a threefold symmetry axis at the top of the tetrahedron. Hydrophobic residues are shown in ball-and- stick models. For clarity, Ile23 has been omitted. This figure was gener- ated using the MOLSCRIPT program (Kraulis, 1991).

quaternary structure for the two enzymes. With a sequence iden- tity of 40 % between the thermophilic and catabolic OTCases their sequences could be aligned manually (Fig. 5). The align- ment was checked with the hydrophobic cluster analysis method HCA (Gaboriaud et al., 1987). No insertions were observed in the thermophilic enzyme when compared to the catabolic one. However, five regions showed deletions (Fig. 5). One a-helix (H10’) is missing in the thermophilic enzyme. From the crystal structure no function can be attributed to this a helix in P. aeru- ginosa OTCase, as it is not involved in the quaternary structure organisation or in the catalytic site, and points towards the solvent; moreover this helix is also absent in most of the ana- bolic OTCases sequenced so far. Other deletions are located in loops (the numbering corresponds to the catabolic sequence) : Arg165 (loop B6-H6), Gly209 (loop H7-B8), Trp243 (loop B9- H9) and Gly264 (loop H10-B10). The monomer (313 amino acids) is composed of two domains both organised around one b-pleated sheet of five parallel strands (Fig. 6). The first domain involved in carbamoylphosphate binding comprises residues 1 -

Fig. 7. Stereo illustration of the dodecameric ornithine carbamoyltransferase symmetry. The Cn backbones of the four trimers are shown using different colours. One trimer (in red) was translated along its threefold symmetry axis. The four trimers are arranged in a tetrahedral manner, with the threefold symmetry axes passing through the middle of the faces and the twofold symmetry axes passing through the middle of opposite edges. From Villeret (1994).

1052 Legrain et al. (Eur: J. Biochem. 247)

120, n I " n 0

100 . A -

0 60 65 70 75 80 85 90

Temperature ("C)

Fig. 9. Effect of temperature on the stability of R furiosus ornithine carbamoyltansferase expressed in E. coli. and in S. cerevisiue. The thermal stability of the enzyme was determined by measuring residual activity with the standard assay after a 15-min preincubation in the pres- ence of 100 mM potassium phosphate pH 7.5 at the temperatures indi- cated. (0) crude protein extact from P. furiosus; (0) OTCase purified from P. f'rio.sus; (A) crude protein extract from recombinant E. coli (strain C600 OTC-/pPFl); (Aj OTCase purified from recombinant E. coli (strain C600 OTC-/pPFl); (U) crude protein extract from recombi- nant S. cerevisiae (strain SSl/pPF2).

150. The second domain, comprising residues 151 -313, is in- volved in ornithine binding. Finally, a C-terminal a helix (H12, residues 289-307) joins the two domains. Three monomers form a trimer in which the interfaces between monomers involve the N-terminal domains (carbamoylphosphate domains). The complete oligomer is built up of four trimers and displays the 23 point group symmetry, with four threefold symmetry axes and three twofold symmetry axes. The structure can be seen as a tetrahedron in which the threefold symmetry axes pass through the middle of the faces and the twofold symmetry axes pass through the edge of opposites faces (Fig. 7). Striking features are observed when analysing the E furiosus OTCase model. The interface occurring between three different trimers around a threefold symmetry axis (at the top of the tetrahedron) displays an interesting pattern : in contrast to the corresponding interface in I? aeruginosa OTCase it involves three helices (Hl) which display a strong hydrophobic character, containing two Trp (po- sition 21 and 33), three Ile (23, 32, 36), one Phe (30) and one Met (29). Therefore each interface around a threefold symmetry axis contains 6 Trp, 9 Ile, 3 Phe and 3 Met (Fig. 8). The interface located around a twofold symmetry axis comprises residues from the C-terminus and helix H3. This interface is also modi- fied when compared to that of I? aeruginosu OTCase because of a two-residue insertion at the C-terminus (Lys336 and Phe337). From the model Lys312 is located close to Asp102 from another trimer and Phe313 is facing another aromatic resi- due, Tyrl 00 (from another trimer). The comparison between this model and the structure of P. aeruginosa OTCase is presented in the Discussion

Expression in E. coli and S. cerevisiae. Incubation of crude protein extract, prepared from recombinant E. coli expressing the I? furiosus OTCase gene (strain C600 OTC-IpPF1; see Roovers et al., 1997, for the details of the genetic construction) for 15 min at 60°C resulted in 60% loss of activity (Fig. 9). Only after purification using the protocol described in Experi- mental Procedure (but with 45% loss of activity in the ammo- nium sulphate precipitation step) could we recover an enzyme

displaying a thermal profile similar to that of the native one (Fig. 9). Even though the same pattern was obtained when using strain L798 (lon mutant, Guillou et al., 1989) as a host, we sus- pect that proteases (possibly activated by the high temperature treatment) are responsible for this phenomenon. We therefore examined the properties of recombinant OTCase produced in yeast (strain SSl/pPF2, see Roovers et al., 1997, for the recom- binant construction). The thermodenaturation profile obtained from yeast recombinant cells was this time very similar to that obtained with the native enzyme (Fig. 9) and an almost identical activation energy (49.8 kJ mol-') was measured (data not shown).

DISCUSSION

An earlier paper by Roovers et al. (1997) analysed a P: furio- sus gene able to complement E. coli and S. cerevisiae mutants lacking a functional OTCase. The gene is similar to all known OTCases (LabCdan et al., unpublished results) and contains the motifs involved in the molecular interactions of the cognate en- zyme with the two substrates, carbamoylphosphate and orni- thine.

The enzyme was purified to homogeneity from cells grown at 98 "C by a procedure which included affinity chromatography on the immobilised bisubstrate analogue B-N-phosphonoacetyl- L-ornithine. Only one form of OTCase could be detected, in contrast to those organisms where both anabolic and catabolic OTCases were found to occur (Legrain et al., 1977). SDSRAGE experiments indicated that the enzyme is a homopolymer of a subunit of a 35 kDa subunit, in agreement with the value estimated from the sequence (i.e. 34.97 kDa). A value of 400 ? 20 kDa was determined by gel-filtration of the purified OTCase, indicating that the enzyme could be a dodecamer. OTCases are either trimeric, as most anabolic OTCases (see Cunin et al., 1986), hexameric as in Streptococcus faecalis (Mar- shall and Cohen, 1972) or dodecameric as the catabolic OTCase of I? aeruginosa which is constituted of four trimers arranged in a tetrahedral manner (Villeret et al., 1995). Before the actual three-dimensional structure of l? aeruginosa OTCase couId be determined by crystallography, some uncertainty remained as re- gards the exact arrangement of subunits, i.e. nonameric or dodecameric (Baur et al., 1987). Considering that a nonameric structure would be surprising from the topological point of view we consider, in the absence of crystallographic evidence, that I? furiosus OTCase is likely to be a dodecamer, an assumption which has been included in the model discussed below, where the three-dimensional structure of P aeruginosa OTCase was taken as frame of reference.

Apart from the temperature dependence, the kinetic proper- ties of P furiosus OTCase are in keeping with those reported for most anabolic OTCases. The R:Yp for ornithine and carbamoylphosphate and the substrate inhibition by ornithine are reminiscent of Thermus OTCase (Van de Casteele et al., 1997), but somewhat lower than for E. coli OTCase (Legrain and Sta- Ion, 1976). F1 furiosus OTCase also appears devoid of the coop- erative interactions towards carbamoylphosphate which restrict the catabolic OTCase of l? aeruginosa to the phosphorolysis of citrulline at low physiological concentrations of carbamoylphos- phate, i.e. the second step of the arginine deiminase pathway (Stalon et al., 1972). In keeping with this, we were unable to detect any arginine deiminase activity in extracts of l? furiosus cells cultivated in a medium containing 200 pg arginine ml-'; on the other hand both argininosuccinate synthetase and argi- ninosuccinate lyase were shown to be present (Van de Casteele et al., 1990). Moreover carbamoylphosphate appears efficiently

Legrain et al. (Eur J. Biochem. 247) 1053

channelled in vitro towards citrulline by J? furiosus OTCase (Le- grain et al., 1995) and J? furiosus could be grown in a defined medium containing ornithine as arginine precursor (unpublished experiments from this laboratory). It appears therefore likely that the enzyme fulfils an anabolic function.

Arrhenius plots of the curves relating enzyme activity to temperature gave activation energies of 48.1 kJ mol-I and 70.3 kJ mol-' for the forward and backward reactions, respec- tively. No breaks in the Arrhenius curves were observed (in con- trast to Thermus OTCase, Van de Casteele et al., 1997), even for the backward reaction which could be monitored up to 90°C. The relatively high values obtained for l? furiosus OTCase are in keeping with its thermophilic nature. However, even though citrulline synthesis catalysed by I? furiosus OTCase would be about 40 times faster in the optimal range of temperatures of this organism than at 30°C (see Results), it was still efficient enough to allow complementation of both E. coli and yeast OTCase-less auxotrophs (Roovers et al., 1997).

The thermal stability of P. furiosus OTCase in crude extracts or in the pure state is considerable: half-lives ranging between 30-65 min at 100°C were obtained depending on the condi- tions. These values were not affected by dialysing the extracts, suggesting that the latter did not contain thermostabilizing factors of small molecular mass in appreciable concentrations. When pure enzyme was submitted to thermodenaturation, the intrinsic stability of the protein was confirmed and did not show any striking concentration dependence. In the presence of orni- thine and/or phosphate, the half-life was increased considerably. Such protective effects of ligands on OTCase were observed previously in E. coli and in other thermophiles (Legrain et al., 1977; Van de Casteele et al., 1997); they indicate the forniation of a ternary enzyme-phosphate-ornithine complex. The concen- tration of ornithine may be too low in the cell to contribute to thermal protection in vivo, but a role of phosphate (in possible synergy with ornithine) or of an organic ion playing the same role (Scholz et al., 1992) is not excluded. CPSase itself may contribute to OTCase stability in J? furiosus; indeed, isotopic competition experiments strongly suggest that CPSase and OTCase associate to form a complex channelling carbamoyl- phosphate (Legrain et al., 1995).

Despite some uncertainty on the actual thermal behaviour of J? furiosus OTCase around 100°C in the intracellular environ- ment, the profile of the enzyme appears highly thermophilic, both in activity and in stability, when compared to that of meso- philic OTCases and even to Thermus aquaticus Z05, Thermo- toga maritima and Sulfolobus solfataricus OTCases (Van de Casteele et al., 1997). The OTCase from I? furiosus is actually the most thermophilic one identified up to now. As the compara- tive analysis of J? furiosus OTCase did not indicate any relevant features that could explain its thermal profile, we elaborated a molecular model from the X-ray structure of the l? aeruginosa catabolic OTCase. It is also a high-molecular-mass OTCase, be- ing composed of four trimeric subunits assembled in a structure of 23 point group symmetry (Villeret et al., 1995). The model was investigated on the basis of criteria recognised as playing a potential role for the thermostability of proteins : hydrophobic and ionic interactions, hydrogen bonds patterns, disulphide brid- ges, potential further stabilisation of a helices and aromatic in- teractions. The interface occurring between three different tri- mers around a threefold symmetry axis (at the top of the tetrahe- dron) displays a characteristic pattern: in contrast to the corre- sponding structure in F! aeruginosa OTCase which is mainly hydrophilic, this interface involves three helices (Hl) which dis- play a strong hydrophobic character, containing two Trp, three Ile, one Phe and one Met. This unusual feature suggests that the thermostability of P. furiosus OTCase is due in part to the as-

sembly into a dodecameric structure of four trimers which in- teract through interfaces presenting a hydrophobic character. In J? aeruginosa Trp21 is replaced by Arg, and Ile28-Try29 by Arg28-Ala29. These two Arg have been shown to play a crucial role in the allosteric regulation of l? aeruginctsa catabolic OTCase (Villeret, 1994; Mouz et al., 1996) suggesting that among OTCases two very different properties (pronounced allo- steric interactions and high thermostability) have been engi- neered from a similar quaternary structure of high internal sym- metry.

We have reported above that a complete separation into mo- nomers could not be achieved, even after treatment such as ex- posure to temperatures up to 120°C. One may wonder whether the oligomeric forms observed after such treatments are not arte- factual (for example, some monomers could reassemble by the putative hydrophobic interactions which the model highlights at the interfaces between trimeric subunits), or if the monomeriza- tion process is not obscured by partial reassociation occurring at low temperature after the high-temperature treatment. Indeed, investigating interfaces between monomers inside a trimer did not indicate any sign of strong thermoresistance. The interface located between monomers inside a trimer displays similar char- acteristics when compared to that of J? aeruginosa OTCase. Three salt bridges involving residues from two monomers are present in both OTCases: Arg58-Glu88, Glu318-Arg95 and Asp307-Arg99 (J? aeruginosa residue numbers). No unusual features are observed at this interface in J? furiosus OTCase. Arg84, which is replaced by Gly in all other OTCases sequenced so far, except in Halobacterium where it is a Ser (Labtdan et a]., 1997; unpublished results) is not involved in subunit con- tacts but is solvent exposed. However, the replacement of a very well conserved Gly by Arg is a striking feature, and it may be that this Arg plays a potential role during the association be- tween OTCase and CPSase.

Possible thermophilic precedents for the stabilisation by hy- drophobic interactions occurring at subunit interfaces are the Thermus thermophilus 3-isopropylmalate dehydrogenase and the Thermotoga maritima glyceraldehyde-3-phosphate dehydroge- nase, for which the three-dimensional structure suggests that the thermostability can be conferred by increased hydrophobic in- teractions at the subunit interfaces (Kirino et al., 1994; Korndorfer et al., 1995). By contrast, structural studies of P. furiosus glutamate dehydrogenase indicate that extended ion- pair networks present on the surface of the protein and at the subunit interfaces may play an important role in the stabi- lisation of proteins (Britton et al., 1995; Yip et al., 1995; Rice et al., 1996). It would appear therefore that different strate- gies may have evolved to stabilise proteins at very high tem- peratures.

Testing an hypothesis on the molecular basis for l? furiosus OTCase thermophily requires faithful expression of the cognate gene in a host amenable to genetic engineering. The complemen- tation experiments described in an earlier paper (Roovers et al., 1997) and analysed at the biochemical level in the present one show that I? furiosus OTCase synthesised by E. coli is less stable than the native enzyme. This appears to be due to proteolysis since, after purification, a certain fraction of the enzyme mole- cules are recovered in a stable state. In S. cerevisiae, however, P. furiosus OTCase appears as stable as in the native host and displays the same activation energy. Other thermophilic proteins which we found to be unstable in E. coli are being tested in yeast to see whether the present observations open the perspective of applications.

We thank F. Ronvaux for her help during the first stages of this work, M. Demarez for the production of large amounts of P. furiosus

1054 Legrain et al. (Eur:

cells, and P. Falmagne and R. Wattiez (University of Mons-Hainaut) for N-terminal amino acid sequence determination. This work was supported by grants from the Belgian Fund for Joint Basic Research, by the Euro- pean Community Programme Biotechnology, by the Research Council of the Universitt Libre de Bruxelles and by the Vlaamse Actieprogrummn Biotechnologie.

REFERENCES Baur, H., Stalon, V., Falmagne, P., Luethi, E. & Haas, D. (1987) Primary

and quaternary structure of the catabolic ornithine carbamoyl- transferase from Pseudomonas aeruginosa. Extensive sequence ho- mology with the anabolic ornithine carbamoyltransferases of E.sche- richia coli, Eur: J. Biochem. 166, 11 1 - 117.

Britton, K. L., Baker, P. J . , Borges, K. M. M., Engel, P. C., Pasquo, A,, Rice, D. W., Robb, F. T., Scandurra, R., Stillman, T. J. & Yip, K. S. P. (1995) Insights into thermal stability from a comparison of the glutamate dehydrogenase from Pyrococcus furiosus and Thermococ- cus litoralis, Eut: J. Binchem. 229, 688-695.

Brunger, A. T. (1992) X-PLOR: a system,for X-ray crystallography and NMR, Yale University Press, New Haven CT.

Cunin, R., Glansdorff, N., PiCrard, A. & Stalon, V. (1986) Biosynthesis and metabolism of arginine in bacteria, Microbinl. Rev. 50, 314- 352.

Davis, R. H. (1986) Compartmental and regulatory mechanisms in the arginine pathway of Neurosporu crmsa and Saccharomyces cerevi- siae, Microbinl. Rev. 50, 280-313.

Fiala, G. & Stetter, K. 0. (1986) Pyrococcus furiosus sp. nov. represents a novel genus of marine heterotrophic archaebacteria growing opti- mally at 100"C, Arch. Microbiol. 145, 56-61.

Fiala, G., Stetter, K. O., Jannasch, H. W., Langworthy, T. A. & Madon, J. (1986) Staphylntherrnu.r marinus sp. nov. represents a novel genus of extremely thermophilic submarine heterotrophic archaebacteria growing up to 98"C, Syst. Appl. Microbiol. 8, 106-113.

Gaboriaud, C., Bissery, V., Benchetrit, T. & Mornon, J. P. (1987) Hy- drophobic cluster analysis: an efficient new way to compare and analyse amino acid sequences, FEBS Lett. 224, 149-155.

Glansdorff, N. (1 996) Biosynthesis of arginine and polyamines, in Escherichia coli and Saimonella. Cellular und molecular biology (Neidhardt, F., ed. in chief) pp. 408-433, American Society for Mi- crobiology, Washington DC.

Gorini, L. & Maas, W. K. (1957) The potential for the formation of a biosynthetic enzyme i n E.ccherichia coli, Biochirn. Biophys. Acta 25,

Guillou, F., Rubino, S. D., Markovitz, R. S., Kinney, D. M. & Lusty, C. J. (1989) Escherichia coli carbamoyl-phosphate synthetase: Domains of glutaminase and synthetase subunit interaction, Proc. Nut/ Acud. Sci. USA 86, 8304-8308.

Kirino, H., Aoki, M., Aoshima, M., Hayashi, Y., Ohba, M., Yamagishi, A,, Wakagi, T. & Oshima, T. (1994) Hydrophobic interaction at the subunit interface contributes to the thermostability of 3-isopropylma- late dehydrogenase from an extreme thermophile, Thermus thermo- philus, Eur: J , Biochem. 220, 275-281.

Korndorfer, I., Steipe, B., Huber, R., Tomschy, A. & Jaenicke, R. (1995) The crystal structure of holo-glyceraldehyde-3-phosphate dehydroge- nase- from the hyperthermophilic bacterium Thermotoga inaritima at 2.5 A resolution, J . Mol. Biol. 246, 511-521.

Kraulis, P. J. (1991) MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures, J. Appl. Crystallogr: 24,

Kuo, L. C., Herzberg, W. & Lipscomb, W. N. (1985) Substrate specific- ity and protonation state of ornithine transcarbamoylase as deter- mined by pH studies, Biochemistry 24, 4754-4761.

Kuo, L. C., Miller, A. W., Lee, S. & Kozuma, C. (1988) Site-directed mutagenesis of Escherichiu coli ornithine transcarbamoylase : role of arginine-57 in substrate binding and catalysis, Biochemistry 27,

Kuo, L. C., Caron, C., Lee, S. & Herzberg, W. (1990) Zn2+ regulation of ornithine transcarbamoylase. 11. Metal binding site, J. Mol. B id .

Legrain, C. & Stalon: V. (1 976) Ornithine carbamoyltransferase from Escherichin coli W. Purification, structure and steady-state kinetic analysis. Eur: J . Biochem. 63, 289- 301.

208 -209.

946 - 950.

8823-8832.

211, 27 1-280.

J. Biochem. 247)

Legrain, C., Stalon, V., Noullez, J. P., Mercenier, A., Simon, J. P., Bro- man, K. & Wiame, J. M. (1977) Structure and function of ornithine carbamoyltransferases, Eul: J. Biochem. 80, 401 -409.

Legrain, C., Demarez, M., Glansdorff, N. & Pitrard, A. (1995) Ammo- nia-dependent synthesis and metabolic channeling of carbamoyl- phosphate in the hyperthermophilic archaeon Pyrococcus furiosus, Microbiology 141, 1093-1099.

Lerner, C. G. & Switzer, R. L. (1986) Cloning and structure of the Bacil- lus subtilis aspartate transcarbamylase gene (PyrB), J. Binl. Chem.

Marshall, M. & Cohen, P. P. (1972) Ornithine transcarbamylase from Streptococcus ,faeculis and bovine liver. I. Isolation and subunit structure, J. B id . Chem. 247, 1641 - 1653.

Murata, L. B. & Schachman, H. K. (1996) Structural similarity between ornithine and aspartate transcarbamoylases of Escherichia coli: char- acterization of the active site and evidence for an interdomain car- boxy-terminal helix in ornithine transcarbamoylase, Protein Sci. 5,

Mouz, N., Tricot, C., Ebel, C., Petillot, Y., Stalon, V. & Dideberg, 0. (1996) Use of a designed fusion protein dissociates allosteric proper- ties from the dodecameric state of Pseudomonas aeruginosa cata- bolic omithine carbamoyltransferase, P roc. Natl Acad. Sci. USA 93,

Penninckx, M. & Gigot, D. (1978) Synthesis and interaction with Esche- richia coli L-ornithine carbamoyltransferase of two potential transi- tion-state analogues, FEBS Lett. 88, 94-96.

Purcarea, C., Simon, V., Prieur, D. & HervC, G. (1996) Purification and characterization of carbamoyl-phosphate synthetase from the deep- sea hyperthermophilic archaebacterium Pyrococcus abyssi, Eur: J. Biochem. 236, 189-199.

Rice, D. W., Yip, K. S. P., Stillman, T. J., Britton, K. L., Fuentes, A,, Connerton, I., Pasquo, A,, Scandurra, R. & Engel, P. C. (1996) In- sights into the molecular basis of thermal stability from the structure determination of PyrococcuJ furiosus glutamate dehydrogenase, FEMS Microbiol. Rev. 18, 105-117.

Roovers, M., Hethke, C., Thornrn, M. & Glansdorff, N. (1997) Isolation of the gene encoding Pyrococc~is furiosus ornithine carbamoyl- transferase and study of its expression profile in vivo and it? vitro, Eut: J. Biochem. 247, 1038- 1045.

Roussel, A. & Cambillau, C. (1989) Silicon graphics geometry partner directory, pp. 77-78, Silicon Graphics, Mountain View CA.

Ruepp, A,, Muller, H. N., Lottspeich, F. & Soppa, J. (1995) Catabolic ornithine carbamoyltransferase of Halobacteirum halobium (sali- nariuni): purification, characterization, sequence determination, and evolution, J. Bacteriol. 177, 1 129- 1136.

Scholz, S., Sonnenbichler, J., Schafer, W. & Hensel, R. (1992) Di-myo- inositol-1,l'-phosphate: a new inositol phosphate isolated from Pyrn- coccus woesei, FEBS Lett. 306, 239-242.

Snodgrass, P. J. (1968) The effects of pH on the kinetics of human liver ornithine carbamylphosphate transferase, Biochemistry 7, 3047 - 305 1.

Stalon, V., Ramos, F., Pitrard, A. & Wiame, J. M. (1972) Regulation of the catabolic ornithine carbamoyltransferase of Pseudomonas jiho- rescens. A comparison with the anabolic transferase and with a muta- tionally modified catabolic transferase, Eur: J. Biochem. 29, 25 - 35.

Summers, N. L., Carlson, W. D. & Karplus. M. (1987) Analysis of side- chain orientations in homologous proteins, J. Mol. Bid. 196, 175- 198.

Van de Casteele, M., Demarez, M., Legrain, C., Glansdorff, N. & Pitrard, A. (1990) Pathways of arginine biosynthesis in extreme ther- mophilic archaeo- and eubacteria, J. Gen. Microbiol. 136, 1177- 1183.

Van de Casteele, M., Desrnarez, L., Legrain, C., Chen, P. G., Van Lierde, K., Pitrard, A. & Glansdorff, N. (1994) Genes encoding thermophilic aspartate carbarnoyltransferases of Thermus uquuticus ZO5 and Thermotogu muritima MSB8: modes of expression in E. coli and properties of their products, Biocaralyis 112, 165- 179.

Van de Casteele, M., Legrain, C., Desmarez, L., Chen, P. G., Pitrard, A. & Glansdorff, N. (1997) Molecular physiology of carbamoylation under extreme conditions: what can we learn from extreme thermo- philic microorganisms ? Comp. Physiol. Biochem., in the press.

261, 11 156-11 165.

709-718.

941 4- 9419.

Legrain et al. (Eur: J. Biochem. 247) 1055

Isolation and characterization of the cyanogen bromide fragments, Biochemistry 7, 1959- 1966.

Yip, K. S. P., Stillman, T. J., Britton, K. L., Artymiuk, P. J., Baker, P. J., Sedelnikova, S. E., Engel, P. C., Pasquo, A,, Chiaraluce, R., Con- salvi, v,, scanduma, R, & ~ i ~ ~ , D, w, (1995) The of pyre- coccus fLlriosus glutamate dehydrogenase reveals a key role for ion. pair networks in maintaining enzyme stability at extreme temper- atures, Structure 3, 1147- 1158,

Villeret, V. (1994) Etudes structurales de l’ornithine carbamoyltransfCr- ase catabolique de Pseudomonas aeruginosa, PhD thesis, UniversitC de Libge.

Villeret, V., Tricot, C., Stalon, V. & Dideberg, 0. (1995) Crystal StNctUre of Pseyhnonas aeruginosa catabolic ornithine transcarbamoylase at 3.0-A resolution: a different oligomeric organization in the trans- carbamoylase family, Proc. Nut1 Acad. Sci. USA 92, 10762- 10766.

Waxdal, M. J., Konigsberg, W. H., Henley, W. L. & Edelman, G. M. (1968) The covalent strncture of a human y G-immunoglobulin. 11.