Proc. Natl. Acad. Sci. USA Vol. 93, pp. 9414-9419, September 1996 Biochemistry

Use of a designed fusion protein dissociates allosteric properties from the dodecameric state of Pseudomonas aeruginosa catabolic ornithine carbamoyltransferase (allostery/assembly/glutathione S-transferase/thermostability) NICOLAs Mouz*, CATHERINE TRICOTtt, CHRISTINE EBEL, YVES PETILLOTI, VICTOR STALONt, AND Orro DIDEBERG*11 *Laboratoire de Cristallographie Macromoleculaire, §Laboratoire de Spectrometrie de Masse des Proteines, and ILaboratoire de Biophysique Moleculaire, Institut de Biologie Structurale Jean-Pierre EBEL, Commissariat a l'Energie Atomique-Centre National de la Recherche Scientifique, 41 avenue des Martyrs, F-38027 Grenoble Cedex 1, France; and tLaboratoire de Microbiologie, Universite Libre de Bruxelles, and tInstitut de Recherches du Centre de Recherches des Industries Alimentaires, Avenue E. Gryson, 1, B-1070 Brussels, Belgium Communicated by William N. Lipscomb, Harvard University, Cambridge, MA, June 5, 1996 (received for review April 22, 1996)

ABSTRACT The catabolic ornithine carbamoyltrans- homotropic cooperativity with respect to the substrate, car- ferase from Pseudomonas aeruginosa, an enzyme consisting of bamoylphosphate, and heterotropic activation by AMP and 12 identical 38-kDa subunits, displays allosteric properties, other nucleoside monophosphates and inhibition by poly- namely carbamoylphosphate homotropic cooperativity and amines (9). Homotropic cooperativity and low affinity for CP heterotropic activation by AMP and other nucleoside mono- prevent the catabolic activity from performing the anabolic phosphates and inhibition by polyamines. To shed light on the reaction in vivo (10, 11). The structure of the catabolic OTCase effect of the oligomeric organization on the enzyme's activity is known at 3-A resolution (7). and/or allosteric behavior, a hybrid ornithine carbamoyl- In the Escherichia coli trimeric anabolic OTCase, substitu- transferase/glutathione S-transferase (OTCase-GST) mole- tion of Arg-106 by Gly creates an enzyme with reduced cule was constructed by fusing the 3' end of the P. aeruginosa cooperativity for both substrates (12). Moreover, the aspartate arcB gene (OTCase) to the 5' end ofthe cDNA encoding Musca carbamoyltransferase of Bacillus subtilis, a trimeric enzyme domestica GST by using a polyglycine encoding sequence as a structurally related to P. aeruginosa anabolic OTCase, nor- linker. The fusion protein was overexpressed in Escherichia mally shows Michaelis-Menten kinetics for aspartate, but can coli and purified from cell extracts by affinity chromatogra- also be transformed by an Arg/Ala substitution at position 107 phy, making use of the GST domain. It was found to exist as (equivalent to Arg-106 in E. coli OTCase) into an enzyme with a trimer and to retain both the homotropic and heterotropic cooperative kinetics (13). characteristic interactions of the wild-type catabolic OTCase To understand the role of the dodecameric structure in both but to a lower extent as compared with the wild-type OTCase. the homotropic and heterotropic cooperativity, we decided to The dodecameric organization of catabolic P. aeruginosa OT- interfere with the assembly process by adjunction of a protein Case may therefore be related to an enhancement of the with a different oligomeric organization by fusing two genes. substrate cooperativity already present in its trimers (and This paper describes the design, using molecular modeling, the perhaps also to the thermostability of the enzyme). construction of the fused gene, and the production and characterization of the OTCase-glutathione S-transferase Protein oligomerization is a very important process in biology. (GST) protein. Although the fusion protein could have been Since the discovery of oligomeric proteins (1), much effort has dimeric (as GST), trimeric (as in most OTCases), or dodecam- been devoted to the study of the assembly and disassembly of eric, it was found to be trimeric and to retain both the polymeric proteins and the effect of this process on their homotropic and heterotropic properties of wild-type P. aerugi- biological function (2, 3). nosa catabolic OTCase. We are currently using ornithine carbamoyltransferase (OT- Case) as a model system to study the role of assembly and MATERIALS AND METHODS disassembly on the biological activity of the enzyme. OTCase catalyzes the conversion of ornithine and carbamoylphosphate Molecular Design of the Fusion Protein. All carbamoyl- (CP) into citrulline and phosphate. The anabolic OTCases transferases form functional trimers. However, the catabolic from various sources have similar properties. They are gener- OTCase from P. aeruginosa has a higher symmetry: a 23-point ally trimeric molecules, made up of identical 35- to 40-kDa group (7). The oligomeric molecule consists of four trimers, subunits, and display Michaelis-Menten kinetics with both organized like in a tetrahedron (Fig. 1). In the 23 symmetry, substrates. Pseudomonas aeruginosa possesses another OT- there are three two-fold and four three-fold axes. The fusion Case, involved in the arginine deiminase pathway, which protein was generated by gene fusion between the catabolic P. under anaerobic con- aeruginosa OTCase and the Musca domestica GST, the latter supplies energy derived from arginine enzyme being s-elected for four reasons. (i) Native GST forms ditions (4). The catabolic enzyme is a dodecamer, made up of very stable homodimers. (ii) The crystal structures of homol- identical 38-kDa subunits organized as four trimers (5-7). In ogous proteins are known (for a review, see ref. 14). (iii) The vivo, it catalyzes the phosphorolysis of citrulline, the reverse of enzyme assays are very easy to perform. (iv) The fusion protein the anabolic reaction. Since the thermodynamic equilibrium of can be a chroma- the reaction strongly favors citrulline synthesis, in vitro assays purified by single-step glutathione affinity are performed in this direction (8). The catabolic OTCase tography. exhibits the characteristics of an allosteric enzyme, such as Abbreviations: OTCase, ornithine transcarbamylase(carbamoyl- transferase); GST, glutathione S-transferase; OTCase-GST, fusion The publication costs of this article were defrayed in part by page charge protein; [S]gc, concentration of carbamoylphosphate giving half- payment. This article must therefore be hereby marked "advertisement" in maximum velocity; CP, carbamoylphosphate. accordance with 18 U.S.C. §1734 solely to indicate this fact. I"To whom reprint requests should be addressed. 9414 Downloaded by guest on September 26, 2021 Biochemistry: Mouz et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9415

-.1tfz

:'fV

FIG. 1. Scattered stereoview of an OTCase dodecamer showing the packing of trimers. Ca backbone of each trimer is color-coded. The tetrahedron has a 23-point symmetry: the three two-fold axes link the middle of opposite edges. The N and C termini of one monomer are marked by A and *, respectively. The GSTs (EC 2.5.1.18) are a group of isoenzymes that play chromosomal copy of the T7 RNA polymerase gene under an important role in the detoxification and metabolism of lacUV5 control. The plasmids used in this work were pCR- many xenobiotic and endobiotic compounds. Cytosolic GSTs Script SK(+) (Stratagene) and pET5a (Novagen). are dimeric enzymes that catalyze the reaction between the Construction of the OTCase-GST Protein. The OTCase- tripeptide glutathione and a wide variety of electrophilic encoding arcB gene and the cDNA encoding GST were alkylating agents. These proteins are encoded by at least four modified by PCR at the 5' and 3' ends. The primers used for distinct gene classes: a, ,u, rr, and cr. To model M. domestica the construction were designed as follows. Primer A (5'- GST, we first aligned the sequence of this protein with CATATGGCTTTCAACATGCAC-3') incorporates a NdeI published amino acid sequences for classes a (human Al-1), , restriction site at the fMet of arcB gene. Primer B (5'- (rat 3-3), and rr (pig) (see figure 1 in ref. 15), using the program ATCGATGGCGAGGGTCGAGACGAG-3') introduces a HCA-PLOT (16). The length of each region of secondary Clal restriction site before the TAA stop codon arcB gene. structure in the parent protein, in the present case, rat 3-3 This modification replaces the last 2 amino acid residues of (class ,t), was retained (17). For the purpose of the present OTCase: Asp-334/Ile-335 by Ile-334/Asp-335. The GST work, a fly GST model was constructed by replacing the side cDNA was modified by introducing a Sall restriction site at the chains of the residues involved in secondary structures in rat 5' end by using primer C (5'-GTCGACATGGATTTCTAC- GST by their equivalents in the fly sequence and by modeling TACTTG-3'). Primer D (5'-CCAGATCTTTAACCGAAG- loop regions. This model was then improved by energy mini- TATTTCTT-3') contains a BglII site after the TAA stop mization (18). For the graphical visualization we used the codon of the GST cDNA. program TURBO-FRODO (19). The PCR conditions used for the modification of arcB gene To design the fusion protein, we constructed a dodecameric and the GST cDNA were as follows: 95°C for 1 min, 55°C for protein with 23-point group symmetry. The monomer was 1 min, and 72°C for 1.5 min for 30 cycles. Each PCR mixture duplicated to form six GST dimers. The two-fold axis of each (50 ,l total volume) contained 2 ng template DNA, 50 pmol dimer was superimposed on either side of each of the three primers, 0.2 mmol 2'-deoxynucleotide 5'-triphosphates, and 1 two-fold axes of the OTCase dodecameric molecule. Each unit Vent DNA polymerase (New England Biolabs). The dimer was translated and rotated along this axis to yield a resulting two DNA fragments were purified by gel electro- compact oligomer with a 23-point group symmetry. On the phoresis on a 1% agarose gel by using GeneClean kits (Bio basis of this crude model, we decided to fuse the OTCase C 101), and the isolated fragments cloned into the pCR-Script terminus to the GST N terminus by using a 10-glycine residue SK(+) plasmid. The plasmids containing the modified arcB linker. The order chosen for the fusion protein was OTCase- gene or the modified GST cDNA were digested, respectively, linker-GST since the glutathione binding site is near the C with the restriction enzymes NdeI/ClaI or SalI/BglII. terminus of GST (17). However, in practice, such a designed The synthetic linker encoding the polyglycine sequence was protein may not be expressed in the cell or expressed with obtained by annealing two oligonucleotides as follows: another oligomeric organization. A clear advantage of our 5'CGATGGCGGTGGTGGTGGTGGCGGCGGTGGCGGCG-3' design is the possibility of following, by enzymatic assays, the 3'-TACCGCCACCACCACCACCGCCGCCACCGCCGCAGCT-5' relationship between enzyme activity (GST, OTCase, or both) ClaI SalI and protein oligomerization, for the dimer, trimer, or dodec- amer, respectively. The NdeI-ClaI fragment, the ClaI-SalI linker, and the Sall- Bacterial Strains and Plasmids. E. coli XL1-Blue MRF' BglII fragment were ligated simultaneously into the expression [A(mcrA) 183, A(mcrCB-hsdSMR-mrr)173, supE44, thi-1, plasmid pET5a, previously treated with NdeI and BamHI (Fig. recAl, gyrA96, endAl, lac, I-/F', proAB, lacIq ZAM15, 2). The resulting expression plasmid, carrying the new gene TnlO,(Tetr)] and JM110 [rpsL, thr, leu, thi-1, lacY, galK galT, encoding the fusion protein OTCase-GST, was named pNM1. ara, tonA, tsx, dam, dcm, supE44, A(lac, proAB),/F' traD36, The constructs were checked by sequencing the double- proAB, lacIqZAM15/] were used for cloning. OTCase-GST stranded DNA, using the dideoxy-chain termination method (20). was expressed in E. coli BL21(DE3) [hsd S gal (Ac Its 857 ind Protein Purification. E. coli strain BL21(DE3) carrying the 1 Sam 7 nin 5 lacUV5-T7 gene 1)]; this strain contains a pNM1 plasmid was routinely grown at 37°C on Luria-Bertani Downloaded by guest on September 26, 2021 9416 Biochemistry: Mouz et aL Proc. Natl. Acad. Sci. USA 93 (1996)

Ndel arcB Clal Sail gst Bgul (Perkin-Elmer Sciex). Horse heart myoglobin (16951.5 Da) W.." '.. '.. . . was used for calibration. Molecular mass is given as isotope- Bhmt-ended PCR average In products values. flow injection analyses, the sample dissolved + in ammonium acetate buffer (pH 7) was introduced by means pCR-Script SK(+) of a Harvard 22 syringe pump at a flow rate of 5 ,ul/min on a + NdeI SrfITr4 DNA ligase Valco C6W injector equipped with a 1 ,u internal loop. The solvent used for infusion consisted of methanol/water/acetic acid (25/74/1; vol/vol), and 30 pmol of protein was used to obtain the spectrum. Enzyme Assays. OTCase activity was determined by quan- tifying citrulline production by the method of Prescott and Jones (21). The standard assay mixture consisted of 2.0 ml of 150 mM imidazole/HCl (pH 6.8), 10 mM L-ornithine and varying amounts of carbamoylphosphate (and effectors, where NdeI/Clal appropriate) (22). The reaction was started by the addition of carbamoylphosphate, allowed to continue for 10 min at 37°C, Ndel Chi Sail Bguil then stopped by the addition of 1.0 ml of antipyrine diacetyl- Us IxW... - Clal Sail monoxime, used for the colorimetric assay of citrulline. The were a Synthetic linker tubes placed in boiling water bath for 20 min, and the absorbance was measured at 466 nm. Assays were performed pEr5a(NqdcI/BamHI) in duplicate and the data are the average of at least two independent determinations. Specific activities are expressed as mmol citrulline formed/h/mg protein. The steady-state kinetic data were analyzed by Lineweaver-Burk, Hanes, and Eadie plots to determine the maximal velocity. Cooperativity 17 promoter was analyzed by standard graphical methods and the Hill coefficient (nH) was determined by the Hill equation as described (9). GST catalyzes the conjugation of 1-chloro-2,4-dinitroben- pNM1 zene (CDNB) to glutathione to form a CDNB-glutathione 7w3as bP I arcB-gst product with a strong molar absorptivity at 340 nm (s = 9.6 mM-1'cm-1). The standard assay mixture consisted of 1 mM CDNB, 5 mM glutathione and 100 mM Tris-HCl (pH 7.5). Optical density changes were recorded at 340 nm (23). FIG. 2. Construction of the expression plasmid pNM1. The blunt- Heat Inactivation Assays. The OTCase-GST and native ended PCR products, designed as described in Materials and Methods, OTCase were incubated at 60°C in 0.02 M potassium phos- were cloned into Srfl-digested pCR-Script SK(+), giving, respectively, phate buffer (pH 6.8). Samples were withdrawn after various pCR-arcB and pCR-gst. The fragments NdeI/ClaI, SalI/BglII, and the periods of time and the residual activities were determined in synthetic linker ClaIlSalI were ligated simultaneously in pETSa which presence of 150 mM imidazole (pH 6.8), 10 mM L-ornithine, has been digested by NdeI and BamHI, finally giving pNM1. 10 mM phosphate, and 20 mM carbamoylphosphate at 37°C. Molecular Weight of the Native Enzyme. The molecular with 100 of per broth, supplemented ,ug ampicillin ml. Cultures weight of the fusion protein was determined by three different were incubated until the cell density reaches 109 cells/ml, then methods: (i) analytical centrifugation, (ii) gel filtration on a expression of the OTCase-GST protein was induced by the Sephacryl column, and (iii) native gel electrophoresis. addition of 400 ,uM isopropyl ,B-D-thiogalactopyranoside and The ultracentrifugation experiments were performed at culture continued for a further 3 h. The cells were harvested 10°C in 50 mM Tris HCl (pH 8.0) in a Beckman XL-A by centrifugation (10,000 x g for 10 minutes at 4°C) and analytical ultracentrifuge, equipped with a UV scanning sys- washed with PBS buffer (150 mM NaCl/16 mM Na2HPO4/4 tem, using a four holes An-60 Ti rotor with double-sector mM NaH2PO4, pH 7.3). Cell extracts were prepared by sus- centerpieces of 1.2 cm path-length. The S value was deter- pending the cells in PBS buffer and disrupting them by mined at 30,000 rpm from the position of the point of sonication using a Rayteon sonic oscillator at 10 kHz. Cell inflection. The equilibrium sedimentation data obtained after debris was removed by centrifugation at 25,000 x g for 30 min. 22 h at 10,000 rpm were analyzed by using the Optima XL-A The cell extract was loaded onto a glutathione-Sepharose data analysis software, version 2.0, from Beckman, using a 4B affinity column (Pharmacia) (1 x 10 cm), previously value for the partial specific volume for OTCase-GST of 0.73 equilibrated with PBS buffer. The column was washed with the cm3/g, determined from its amino acid composition (24). buffer until the optical density at 280 nm reached the baseline, Gel filtration studies were carried out as described (25) using then the fusion protein was eluted with 50 mM Tris HCl (pH a column (80 x 2.6 cm) of Pharmacia Sephacryl S300, equil- 8), containing 10 mM glutathione. ibrated and eluted with 150 mM imidazole-HCl (pH 6.8) Criteria of Purity. The purity of the enzyme was assessed by without effector, or with either 10 mM AMP or 10 mM analytical gel electrophoresis under native and denatured spermidine. OTCase-GST (200 ,ug) was loaded onto the conditions as described (5). Mass spectrum was obtained on an column and the elution profile followed by the absorbance at API III+ triple-quadrupole mass spectrometer (Perkin-Elmer 280 nm. A standard curve of partition coefficient, Kav = (ve - Sciex) equipped with a nebulizer-assisted electrospray source vo)/(vt - vo), versus molecular mass was prepared using blue (ionspray) operating at atmospheric pressure. A 5-kV voltage dextran 2000 and proteins of 25, 43, 158, 232, 440, and 669 kDa. was applied to the electrospray needle. The mass spectrometer Electrophoresis was performed in native conditions in was scanned from m/z 1100 to 2000, with steps of 1 m/z unit Tris-Bicine (pH 8.6) (Phast system; Pharmacia) as described (26). with a 2-ms dwell time. An average of 10 scans were obtained in the flow injection analyses mode. The molecular species RESULTS AND DISCUSSION produced a series of multiply charged protonated molecular ions. The reconstructed molecular mass profile was deter- Production of OTCase-GST. After designing monomeric mined with a data system by using a deconvolution algorithm OTCase-GST by molecular modeling, the fusion gene was Downloaded by guest on September 26, 2021 Biochemistry: Mouz et aL Proc. Natl. Acad. Sci. USA 93 (1996) 9417 constructed (Fig. 2) (see Materials and Methods). The fusion protein was then overexpressed in E. coli and purified to homogeneity by affinity chromatography on glutathione- Sepharose. Analysis of purity by SDS/polyacrylamide gels 18 showed a single band with an electrophoretic mobility corre- sponding to an apparent molecular weight of -64 kDa. Mass 67 spectrometry analysis gave a value of 62,698 ± 6 Da, including one adduct of glutathione molecule consistent with that pre- dicted from the DNA sequence (62,396 Da). The yield was 14 mg per liter of culture medium. - 140 Characterization of the Oligomeric State. Ultracentrifuga- - 232 tion. The sedimentation velocity profiles showed a single boundary. From the experimental sedimentation coefficient of - 440 5.7 + 0.1 S, we can estimate a S20,. value of the order of 7.5 S, compatible with a homogenous globular protein with a mo- - 669 lecular weight of about 150 kDa (27). Equilibrium sedimen- tation is an absolute procedure for determining the molar mass of macromolecules in solution. The data, fitted to a model with one single solute, gave a molecular weight of 180 kDa. Gel filtration and polyacrylamide gel electrophoresis. On gel filtration, OTCase-GST eluted as a single symmetrical peak, indicating sample homogeneity. Its partition coefficient com- 1 2 3 pared with those of standard proteins gave an apparent molecular size of 191 kDa (Fig. 3) in the absence of effector FIG. 4. Polyacrylamide gel electrophoresis of purified OTCase- or in the presence of AMP or spermidine. The ultracentrifu- GST. Electrophoresis in native conditions on a 8-25% polyacrylamide gation and gel filtration experiments show that the OTCase- gel. Lanes: 1, wild-type OTCase; 2, OTCase-GST; 3, marker proteins: GST is a trimer of 62-kDa subunits in both 50 mM Tris HCl thyroglobulin (Mr 669,000), ferritin (Mr 440,000), catalase (Mr buffer (pH 8) and enzyme assay buffer (150 mM 232,000), lactate dehydrogenase (Mr 140,000), bovine serum albumin imidazole-HCl, pH 6.8). (Mr 67,000), ferritin subunit (Mr 18,000). A molecular mass of -190 kDa was observed for the to 50 OTCase-GST enzyme on polyacrylamide gel electrophoresis there is no effect of ornithine concentration between 0.5 in native conditions (Fig. 4). mM on the cooperative carbamoylphosphate saturation curve All methods of molecular mass determination indicate that (9). The apparent Km for ornithine determined at saturation of OTCase-GST is a trimer in the absence of effector or in the carbamoylphosphate for the OTCase-GST enzyme is 1.4 mM, presence of the positive effector AMP or of the negative the same as that determined for the wild-type enzyme (data effector spermidine. The comparison of the biochemical prop- not shown). The carbamoylphosphate saturation curves of erties of wild-type dodecameric OTCase and trimeric OTCa- wild-type OTCase and OTCase-GST are shown in Fig. 6, and se-GST should provide information relating the degree of a summary of the kinetic parameters is given in Table 1. The oligomerization with the enzymatic properties of OTCase. maximal velocity of the fusion enzyme is decreased by 65% as Thermostability of Wild-Type OTCase and OTCase-GST. compared with wild-type OTCase. The cooperativity with Following incubation of the enzymes for various lengths of respect to carbamoylphosphate is reduced and the concentra- time at 60°C in 0.02 M potassium phosphate buffer (pH 6.8), the residual activity was assayed (Fig. 5). Wild-type OTCase was found to be very stable, retaining full activity for at least 2 h, 11 while the OTCase-GST lost 50% of its activity within 10 min. Kinetic Properties of the Trimeric OTCase-GST Enzyme. The maximum enzymatic activities of the fusion enzyme was I-- 50% of that of wild-type GST and 35% of that of wild-type OTCase. Table 1 shows the kinetic and allosteric properties of 0 the fusion enzyme as compared with wild-type dodecameric ._a catabolic OTCase. It will be noted that for wild-type OTCase a1) 4- 10-

0,8-

0,6-I ./ 0,4-1 aOcm.-OST mocictd mus: 191 kDa 0,2- Time (minutes) FIG. 5. Thermostability of the OTCase and OTCase-GST. The 4.5 S 5.5 enzyme was incubated at 60°C for each time period at a protein log MM concentration of 50 ,ug/ml in 0.02 M potassium phosphate buffer (pH 6.8). The remaining activity was assayed in 150 mM imidazole*HCI (pH FIG. 3. Gel filtration of OTCase-GST. Partition coefficient K., = 6.8), 10 mM L-ornithine, 10 mM phosphate, and 20 mM car- (ve - vo)/(vt - vo) is plotted against log of molecular mass. Markers bamoylphosphate at 370C. o, Wild-type OTCase; 0, OTCase-GST have a molecular mass of 25, 43, 158, 232, 440, and 669 kDa. enzyme. Downloaded by guest on September 26, 2021 9418 Biochemistry: Mouz et aL Proc. Natl. Acad. Sci. USA 93 (1996) Table 1. Allosteric properties of the wild-type OTCase and chimeric OTCase-GST Wild-type OTCase OTCase-GST enzyme Effector mM [S]C5fH Vma [S]i Vmax None 4.0 (0.2) 4.1 (0.2) 13.5 (0.5) 9.6 (0.3) 2.4 (0.1) 4.4 (0.3) Pi 10 1.2 (0.1) 1.0 (0.05) 2.7 (0.1) 1.7 (0.1) Pi 25 2.4 (0.1) 1.1 (0.1) AMP 10 1.2 (0.1) 2.5 (0.1) 4.8 (0.3) 2.0 (0.1) Spermidine 20 11.0 (0.5) 4.3 (0.2) 18.0 (1.0) 2.4 (0.1) The assays were _performed in 150 mM imidazole HCl buffer (pH 6.8) containing 10 mM ornithine at variable CP concentrations. [S]c., concentration of substrate at half maximum velocity; nH, Hill index. The results are the mean values from at least two independent experiments. Vmax is expressed in mmol citrulline formed h-1 x (mg-prot)-1 in the presence of 10 mM L-ornithine and at CP saturation concentration. Vm.x of the enzymes is not changed by the various effectors reported. tion of carbamoylphosphate giving half maximum velocity, AMP, an allosteric effector of the wild-type enzyme, acts [S]., is increased by -2-fold by comparison with wild-type primarily by promoting a local conformational change that OTCase. increases the affinity of the catalytic sites for CP and has a Forwild-type OTCase, inorganic phosphate Pi, an analog for secondary effect on the allosteric transition resulting from the the CP sites (or PP1 and phosphono acetyl-L-ornithine), is able increase in occupancy of the catalytic sites by CP (V.S., to activate the enzyme at low concentration of carbamoylphos- unpublished data). A value of 10 mM AMP also reduces the phate, where the wild-type enzyme is in the low-activity-low- [S]c value for the OTCase-GST as well as the Hill index, affinity state. Under these conditions the addition of phos- indicating that the heterotropic activation by the nucleoside phate induced the allosteric transition to the conformation phosphate is conserved also in the trimeric enzyme form. The having the highest affinity for CP (9). The influence of Pi on allosteric effector AMP was however less effective as an the activity of the OTCase-GST enzyme was measured in the activator since a 5-fold reduction in binding affinity was presence of 5 mM CP and compared with its influence on observed as compared with the wild-type enzyme (Table 2). wild-type OTCase in the presence of 2 mM CP, thus taking into Spermidine, an allosteric inhibitor of the wild-type enzyme, account the different affinities of the two enzymes for CP. The also promotes a conformational change that is distinct from OTCase-GST was also activated by phosphate but its sensi- those involved in the homotropic cooperativity. When sper- tivity to this effector differed from the wild-type enzyme. The midine binds to its regulatory site, it promotes a local confor- binding affinity indicated a 16-fold increase in the [A]o.5 value mational change that, by contrast with the effect of AMP, as compared with the wild-type enzyme, in spite of the fact that primarily decreases the affinity for CP and secondarily in- the OTCase-GST has a reduced cooperativity toward the creases the carbamoylphosphate homotropic cooperativity (9). allosteric substrate by comparison with the wild-type catabolic Like the wild-type dodecameric OTCase, the OTCase-GST OTCase (Table 2). is inhibited by spermidine as shown by the increase in the [Sicg A value of 10 mM Pi is sufficient to convert the sigmoidal CP value but within experimental error the Hill index does not saturation curve into a Michaelis-Menten kinetics for wild- change, indicating that the spermidine acts primarily on the type OTCase. The same concentration only caused a reduction affinity of the CP site for CP and has no effect on the in the [S]cp value and in the Hill index for the OTCase-GST carbamoylphosphate homotropic cooperativity. The affinity of enzyme. Higher concentrations of phosphate however abolish the OTCase-GST enzyme for spermidine is apparently re- the carbamoylphosphate cooperativity. duced 2.5-fold by comparison with the wild-type OTCase.

1.0 CONCLUSIONS To elucidate the relationships between the structure and function of the allosteric OTCase we had previously selected 0.8- variants of the catabolic OTCase able to carry out the anabolic reaction in vivo (22, 28). In three different mutant catabolic OTCases, a point mutation converting Glu-105 into either Ala, E 0.6- Gly, or Lys largely abolishes carbamoylphosphate cooperativ- ity. Because these mutants remain dodecameric, we engi- neered hybrid genes which consist of the N-terminal part ofE. Table 2. Activation and inhibition of the wild-type OTCase and OTCase-GST Wild-type OTCase-GST OTCase enzyme [A]P'5 0.5 (0.05) 7.8 (0.3) AMP 1.3 7.0 [I]spermidine[Al 3.2 (0.1)(0.2) 8.0 (0.2)(0.5) 20 30 The assays were performed in 150 mM imidazole HCl buffer (pH [ Carbamoylphosphate ] (mM) 6.8) in the presence of 10 mM ornithine. The carbamoylphosphate concentration was chosen for each enzyme at its respective [S].P value FIG. 6. Carbamoylphosphate saturation curves for wild-type OT- in order to maximize the effect of the effectors. [A]o.5 is the concen- Case and OTCase-GST. Relative velocities are plotted as a function tration of a positive effector required to reach half maximum activa- of the carbamoylphosphate concentration, and were determined in the tion. [110.5 is the concentration of spermidine required to reach 50% presence of 10 mM ornithine and 150 mM imidazole HCl (pH 6.8). of the velocity observed in the absence of the inhibitor. The concen- V.. values are given in Table 1. 0, Wild-type OTCase; *, OTCase- trations are expressed in mM. The results are the means of values from GST enzyme. at least two independent experiments. Downloaded by guest on September 26, 2021 Biochemistry: Mouz et al. Proc. Natl. Acad. Sci. USA 93 (1996) 9419 coli argF gene encoding a trimeric and nonallosteric enzyme 8. Stalon, V., Legrain, C. & Wiame, J. M. (1977) Eur. J. Biochem. combined with the C-terminal part of the catabolic P. aerugi- 74, 319-327. nosa arcB gene (26). This study indicated that the dodecameric 9. Tricot, C., Nguyen, V. T. & Stalon, V. (1993) Eur. J. Biochem. structure correlates with carbamoylphosphate cooperativity. 215, 833-839. Besides, in some trimeric hybrid enzymes that retained CP 10. Stalon, V., Ramos, F., Pierard, A. & Wiame, J.-M. (1972) Eur. cooperativity, heterotropic interactions are disconnected from J. Biochem. 29, 25-35. ones 11. Haas, D., Evans, R., Mercenier, A., Simon, J.-P. & Stalon V. homotropic (26). (1979) J. Bacteriol. 139, 713-720. The results presented here show that it is possible to prepare 12. Kuo, L. C., Zambidis, I. & Caron, C. (1989) Science 245,522-524. a trimeric OTCase-GST that retains both the homotropic and 13. Stebbins, J. W. & Kantrowitz, R. (1992) Biochemistry 31, 2326- heterotropic properties of the wild-type enzyme, clear evi- 2332. dence that allosteric interactions are intrinsic characteristics of 14. Dirr, H., Reinemer, P. & Huber, R. (1994) Eur. J. Biochem. 220, the trimer. 645-661. As the result of differences in the oligomeric state of 15. Sinning, I., Kleywegt, G. J., Cowan, S. W., Reinemer, P., Dirr, OTCase and GST, the monomers associate to yield a func- H. W., Huber, R., Gilliland, G. L., Armstrong, R. N., Ji, X. & tional trimer for the OTCase and one functional dimer and one Board, P. G. (1993) J. Mol. Biol. 232, 192-212. nonfunctional monomer for the GST. Since the fusion protein 16. Gaboriaud, C., Bissery, V., Bencherit, T. & Mornon, J.-P. (1987) retains both homotropic and heterotropic activation and in- FEBS Lett. 224, 149-155. hibition, the evolutionary advantage for OTCase to pass from 17. Ji, X., Zhang, P., Armstrong, R. N. & Gilliland, G. L. (1992) a trimeric to a dodecameric structure results in an increase in Biochemistry 31, 10169-10184. substrate cooperativity and in thermostability as already noted 18. Brunger, A. T. (1992) X-PLOR: A System for X-Ray Crystallog- for hybrid trimeric OTCase (26). It should be noted that raphy and NMR (Yale Univ. Press, New Haven, CT). another dodecameric OTCase displaying Michaelis-Menten 19. Roussel, A. & Cambillau, C. (1989) Silicon Graphics Geometry Partner Directory (Silicon Graphics, Mountain View, CA), pp. kinetics has been characterized in a hyperthermophilic bacte- 77-78. rium, Pyrococcus furiosus. 20. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. This is publication No. 334 of the Institut de Biologie Structurale 21. Prescott, L. M. & Jones, M. E. (1969) Anal. Biochem. 32, 408- Jean-Pierre EBEL. This project was supported by the Belgian Fund for 409. Basic Research (Grant 2.4507.91) and by the French Ingenierie des 22. Baur, H., Tricot, C., Stalon, V. & Haas, D. (1990) J. Biol. Chem. Macromolecules Biologiques program. 265, 14728-14731. 1. Svedberg, T. (1929) Nature (London) 123, 871. 23. Habig, W. H., Pabst, M. J. & Jakoby, W. B. (1974) J. Biol. Chem. 2. Monod, J., Wyman, J. & Changeux, J. P. (1965) J. Mol. Biol. 12, 249, 7130-7139. 88-118. 24. Harpaz, Y., Gerstein, M. & Chothia, C. (1994) Structure 2, 3. Koshland, D. E., Nemethy, G. & Filmer, D. (1966) Biochemistry 641-649. 5, 365-385. 25. Tricot, C., Schmidt, S., Baur, H., Villeret, V., Dideberg, 0. & 4. Vander Wauven, C., Pierard, A., Klay-Raymann, M. & Haas, D. Stalon, V. (1994) Eur. J. Biochem. 215, 833-839. (1985) J. Bacteriol. 160, 923-934. 26. Nguyen, V. T., Baker, D., Tricot, C., Stalon, V., Dideberg, O., 5. Baur, H., Stalon V., Falmagne P., Luthi E. & Haas D. (1987) Eur. Villeret, V. & Haas, D. (1996) Eur. J. Biochem. 236, 283-293. J. Biochem. 116, 111-117. 27. Van Holde, K. E. (1975) in The Proteins, eds. Neurath H. & Hill, 6. Marcq, S., Diaz-Ruano, A., Charlier, P., Dideberg, O., Tricot, C., R. L. (Academic, New York), 3rd Ed., Vol. 1, pp. 232-233. Pierard, N. & Stalon, V. (1991) J. Mol. Biol. 220, 9-12. 28. Nguyen, V. T., Baker, D., Tricot, C., Baur, H., Villeret, V., 7. Villeret, V., Tricot, C., Stalon, V. & Dideberg, 0. (1995) Proc. Dideberg, O., Gigot, D., Stalon, V. & Haas, D. (1994) FEMS Natl. Acad. Sci. USA 92, 10762-10766. Microbiol. Lett. 124, 411-418. Downloaded by guest on September 26, 2021