Eur. J. Biochem. 224, 853-861 (1994) 0 FEBS 1994

Purification and properties of a succinyltransferase from Pseudomonas aeruginosa specific for both arginine and ornithine Catherine TRICOT ’, Corinne VANDER WAUVEN’.’, Ruddy WATTIEZ3, Paul FALMAGNE3 and Victor STALON’ ’ Laboratoire de Microbiologie, FacultC des Sciences, UniversitC Libre de Bruxelles, Belgique

Institut de Recherches du Centre d’Enseignement et de Recherches des Industries Alimentaires et Chimiques, Bruxelles, Belgique Service de Chimie biologique, DCpartement de Biologie et de Pathologie Cellulaire, FacultC des Sciences, UniversitC de Mons-Hainaut, Belgique

(Received April 29/July 7, 1994) - EJB 94 0612/4

The arginine and ornithine succinyltransferase from Pseudomonas aeruginosa, a bifunctional enzyme involved in the aerobic utilization of arginine and ornithine, has been purified to homo- geneity. The apparent molecular mass of the native enzyme was 150 kDa by gel filtration and 140 kDa by polyacrylamide gel electrophoresis under non-denaturing conditions. After SDSPAGE two subunits of 35 kDa and 37 kDa were evident, indicating that the enzyme is a heterotetramer. Microsequence analysis of the electroblotted protein bands gave two different but well-conserved N-terminal amino acid sequences.

The L-arginine saturation curve followed Henri-Michaelis kinetics with an apparent K,, value of 0.5 mM. The sigmoidal saturation curve for L-ornithine indicated allosteric behaviour. D-Arginine, a competitive inhibitor with respect to L-arginine, reduced L-ornithine cooperativity. In the presence of spermidine, the L-ornithine saturation curve became increasingly sigmoidal, the Hill coefficient shifting from 2.5 in the absence of the inhibitor, to 3.5 in the presence of 20 mM spermidine. The L-arginine analog, L-homoarginine, was also a substrate of the succinyltransferase, and the saturation of the enzyme by this substrate was also cooperative. All these data confirmed the allosteric nature of the enzyme. Moreover, a mutant growing faster on L-ornithine than the parent strain had a modified succinyltransferase with a reduced L-ornithine cooperativity.

The fate of L-homoarginine was different depending on whether the succinyltransferase was induced or not ; excreted succinylhomoarginine was found in cultures induced for the transferase activity whereas guanidinovalerate was excreted in non-induced cultures. The ‘waste’ of succinyl CoA, which could not be regenerated from the excreted succinylhomoarginine, explained the inhibi- tion exerted by L-homoarginine on growth when ornithine or arginine was used as the growth medium.

Aerobic degradation of L-arginine as the sole carbon source by a variety of Pseudomonas, Aeromonas or Kleb- siella species, proceeds through the succinylation of the a amino group of L-arginine by succinyl CoA (Stalon et al., 1987). The succinyltransferase pathway comprises the following sequence of reactions : L-arginine - succinylargin-

Correspondence to V. Stalon, Laboratoire de Microbiologie, UniversitC Libre de Bruxelles, c/o Centre d’Enseignement et de Re- cherches des Industries Alimentaires et Chimiques, 1 avenue Emile Gryson, B-1070 Brussels, Belgium

~~

Fax: +32 2 526 72 73. Abbreviations. AST, L-arginine succinyltransferase ; OST, L-or-

nithine succinyltransferase. Enzymes. L-Arginine N-succinyltransferase (EC 2.3.1.109) ; NL-

succinylarginine dihydrolase (EC 3.5.3.-) ; W-succinylornithine aminotransferase (EC 2.6.1 .-) ; NL-succinylglutamate semialdehyde dehydrogenase (EC 1.2.1 .-); W-succinylglutamate desuccinylase (EC 3.5.1.-); L-ornithine succinyltransferase (EC 2.3.1.-); L-homo- arginine succinyltransferase (EC 2.3.1 .-) ; L-homoarginine racemase (EC 5.1.1 .-); D-homoarginine dehydrogenase (EC 1.4.99.-); oxoho- moarginine decarboxylase (EC 4.1 .l .-) ; guanidinovaleraldehyde de- hydrogenase (EC 1.1.1 .-) ; L-ornithine carbamoyltransferase (EC 2.1.3.3).

ine - succinylornithine - succinylglutamate semialdehyde - succinylglutamate (Vander Wauven and Stalon, 1985; Jann et al., 1986; see Fig. 1).

In Pseudomonas aeruginosa the catabolism of L-ornith- ine converges with L-arginine catabolism at the level of suc- cinylornithine (Vander Wauven et al., 1988). Arginine and ornithine succinyltransferase activities are induced during growth on L-arginine and L-ornithine (Jann et al., 1986; Vander Wauven et al., 1988). A single mutation abolishing both the arginine and ornithine succinyltransferase activities suggests that the same enzyme is involved in the catabolism of both amino acids (Vander Wauven et al., 1988).

In the present study, we report the copurification to homogeneity of the arginine and ornithine succinyltransfer- ase activities, thus showing that a multispecific enzyme is involved in the utilization of the two amino acids. The results presented in this study also establish the unique kinetic prop- erties of the succinyltransferase activity which exhibits allo- steric properties with L-ornithine as substrate and with the alternative substrate L-homoarginine, but Henri-Michaelis ki- netics with L-arginine.

854

2 N - succinylarginine

D-hornoarginine I

I Succinyl CoA

g uanidinovaleraldehyde

o r + NADH + H+

N f succinylornithine oxoglutarate

glutamate

N 5 succinylglutarnate sernialdehyde

N f succinylglutarnate

9 glutarnate + succinate

guanidinovalerate

Fig. 1. Catabolism of L-arginine and L-ornithine and possible pathway of L-homoarginine utilization by R aeruginosa. (1) Arginine succinyltransferase; (2) Nz-succinylarginine dihydrolase; (3) Nz-succinylornthine aminotransferase; (4) W-succinylglutamate semialtiehyde dehydrogenase ; ( 5 ) W-succinylglutamate desuccinylase; (6) L-ornithine succinyltransferase ; (7) L-homoarginine succinyltransferase : (8) L- homoarginine racemase; (9) D-homoarginine dehydrogenase; (10) oxohomoarginine decarboxylase; (1 1) guanidinovaleraldehyde dehydro- genase.

Table 1. Bacterial strains. The genotype symbols are as follows: nmi, acetamide utilization; aru, arginine utilization ; met, auxotrophy for methionine; om, ornithine utilization.

Strains Genotype and relevant characteristics References

PA01 wild-type Holloway et al., 1987 PA0977 oru-180 Van der Wauven et al., 1984 PA061 SO met-9001, amiE 2000aru-77 Jann et al., 1986 PA0505 met-9001, amiE200 Voelmy and Leisinger, 1976 PA061 90 met-9001, amiE200, om+ this study PA06191 met-9001, amiE200, om+ this study PA06192 met-9001, amiE200, oru+ this study

MATERIALS AND METHODS

Bacterial strains

Table 1 lists the P. aeruginosa strains used in this study. The generalized transducing phage GlOl was used for map- ping (Holloway et al., 1979).

Isolation of ornithine-utilizing strains Mutant strains growing faster on ornithine were derived

from strain PA0505 by N-methyl-N’-nitro-N-nitrosoguanid- ine mutagenesis (Fruh et al., 1985) by selecting the largest colonies appearing on mineral medium 154 supplemented with 20 mM ornithine as the sole carbon and nitrogen \ource.

Media and growth conditions

F! aeruginosa was grown at 37°C in nitrogen-free mini- mal medium 154 as described by Stalon et al. (1967). Sub- strates used as carbon and nitrogen sources were added after separate sterilization by filtration. The concentration of sub- strates used as carbon sources was 20mM. When used as a nitrogen source, substrates were usually added at a final concentration of 10 mM. When needed, this medium was supplemented with 1 mM methionine.

Growth experiments For measurements of the mass doubling time and for the

determination of enzyme activities, strains were grown in mineral medium 154 containing different carbon and nitro- gen sources. The cells were grown at 37°C with shaking in 1-1 flasks containing 100 ml medium. The doubling time was calculated from the absorbance at 660 nm during the expo- nential growth phase. For enzyme determination, cells were harvested in exponential phase of growth by centrifugation at 20000 g for 20 min, washed twice with NaCl (0.9%) and,

855

if not used immediately, kept as a frozen pellet at -20°C. For the isolation of succinyltransferase activities, the cells from a culture in a 10-1 fermentor were used.

Enzyme assays Unless otherwise stated, assays were performed on cell

extracts prepared by sonication of cells suspended in 25 mM potassium phosphate, pH 7.5, containing 3 mM 2-mercapto- ethanol. Cell debris were removed by centrifugation at 20000 g for 15 min. Extracts were kept at 4°C. Protein con- centrations were determined by the Lowry method. Enzyme activities were assayed at 37°C. 1 U of enzyme is defined as the amount of enzyme catalysing the utilization of 1 pmol substrate/h. Specific activities are expressed as U . mg pro- tein-'. Succinyltransferase activities were assayed by moni- toring the disappearance of succinyl CoA in a 1.0-ml cuvette at 232 nm using a Philips spectrophotometer. A reference containing all the components of the reaction mixture with the exception of succinyl CoA was used. Unless otherwise stated, assays were initiated by the addition of succinyl CoA (0.2 pmol) to a reaction mixture which contained 100 pmol Tris/HCl, pH 8.9, and either 10 pmol arginine, 40 pmol ho- moarginine or 100 pmol ornithine and in a total volume of 1 .O ml. Linear rates could be obtained provided that no more than 30% of the substrate was consumed. The data presented in the figures are averages obtained from assays performed in duplicate. The steady-state kinetic data were analyzed ac- cording to Lineweaver-Burk (1934) and Hanes (1932) to de- rive maximal velocities (V,,,) and substrate concentrations required for half-maximal velocity [S], 5. Cooperativity was analyzed by standard graphical methods and the Hill coeffi- cient (h) was determined in terms of the Hill equation (1910). W-succinylarginine dihydrolase, W-succinylornithine ami- notransferase and N--succinylglutamate desuccinylase were measured as described previously (Stalon et al., 1987).

Enzyme purification l? aeruginosa PA0 cells were grown aerobically in 10 1

mineral medium 154 supplemented with 25 mM arginine as previously described (Mercenier et al., 1980). Cells were col- lected by centrifugation at 20000 g for 20 min in the expo- nential phase of growth (approximately 1.2X lo9 cell/ml) and were sonicated in 65 ml 50 mM potassium phosphate, pH 7.5 (Stalon, 1972). Unless otherwise stated, all purification pro- cedures were performed at 4°C and in the presence of 3 mM 2-mercaptoethanol. After centrifugation at 20000 g for 20 min, the crude extract was fractionated by ammonium- sulfate precipitation. The ornithine and arginine succi- nyltransferase activities were found in the 25-50% satura- tion fraction. This fraction was dialysed against the extrac- tion buffer and applied to a DEAE Sepharose column (2.5 cmX40 cm) equilibrated with the same buffer. Enzymes were eluted with a linear gradient of KC1 (0-600mM, 500 ml) in 50 mM potassium phosphate at pH 7.5. The or- nithine succinyltransferase and arginine succinyltransferase activity peaks were eluted together. The active fractions were pooled and concentrated by ultrafiltration in a Diaflow sys- tem on a PM30 membrane (Amicon) and were dialysed against 10 mM potassium phosphate, pH 7.5. The succi- nyltransferase preparation was loaded onto an arginine - Sepharose 4B (Pharmacia) column (2 cmX40 cm) equili- brated with this buffer, washed and eluted with a linear KC1 gradient (0-250 mM, 500 ml) in 10 mM potassium phos-

phate, pH 7.5. Again, a single peak of succinyltransferase activity was associated with L-arginine and 1,-ornithine. Electrophoretic analysis of the active fractions revealed the presence of contaminants. The succinyltransferase containing fractions were pooled, dialysed and further purified on a se- cond arginine- Sepharose 4B column (1.5 cm X 10 cm) equil- ibrated as described above. The succinyltransferase activities were desorbed together by a linear ornithine gradient (0- 200mM, 400ml) in 1OmM potassium phosphate, pH7.5. The active fractions were pooled and concentrated by ultraf- iltration in a Diaflow system as decribed above and dialyzed against 10 mM potassium phosphate, pH 7.5, and stored at -20°C.

Molecular mass was estimated by Sephacryl HR300 gel filtration (2.6 cmX60 cm) equilibrated in 10 mM potassium phosphate, pH 7.5, containing 3 mM dithiothreitol and cal- ibrated with Pseudomonas catabolic ornithine carbamoyl- transferase (456 kDa), catalase (232 kDa), Escherichia coli ornithine carbamoyltransferase (140 kDa), bovine serum al- bumin (67 kDa), ovalbumin (43 kDa), and blue dextran 2000 as a void-volume indicator.

Polyacrylamide gel electrophoresis Analytical gel electrophoresis was performed with the

Phast system (Pharmacia) following the instructions given by the manufacturer. Gels of PI 3 -9 were used for isoelectrofo- cusing and gradient gels (acrylamide concentration 8 -25 %) for electrophoresis either in native or in denaturing condi- tions.

Automated microsequence analysis of the electroblotted proteins

The proteins were prepared by SDS/polyacrylamide gel electrophoresis. The protein bands were electroblotted onto a polyvinylidene difluoride membrane (Bio Rad) by the method of Matsudaira (1987), using a semi-dry blotting ap- paratus (Biolyon); the blot was stained with Coomassie bril- liant blue R-250.

Amino acid microsequence analysis of the electroblotted proteins was performed by automated Edman degradation of 1 - 10 pmol protein on a Beckman LF 3400 protein sequencer equipped with an on-line system gold 126 microgradient HPLC and a model 168 diode array detector (Beckman In- struments). All samples were sequenced using standard Beckman sequencer procedure 4. The phenylthiohydantoin derivatives were quantitatively identified by reverse-phase HPLC on an Ultrasphere column (5-pm diameter particles, 2.0 mmX250.0 mm, Beckman Instruments). All sequencing reagents were from Beckman.

Analytical methods Arginine or homoarginine concentrations in the culture

supernatant were determined by the method of Miklus and Stein (1973). The presence of guanidino or amino derivatives was also identified by high-voltage electrophoresis (Vander Wauven and Stalon, 1985).

Chemicals The chemicals employed were obtained from commercial

sources, with the following exceptions : succinylarginine and succinylhomoarginine were synthesized chemically as de-

856

Table 2. Summary of the purification of the arginine and ornithine succinyltransferase activities from PAO1. Assays were perf( Irmed in 100 mM Tris/HCl, pH 8.9, containing 0.2 mM succinyl CoA with either 10 mM L-arginine or 100 mM L-ornithine. AST, L-arginine succinyltransferase; OST, L-ornthine succinyltransferase.

Step Total protein Activity Total activity Specific activity AST!OST

m& U Cell-free extract 2830 AST 178 000

OST 67 920 Ammonium-sulfate precipitation 589 AST 69 500

OST 24 820 DEAE Sepharose 86 AST 71 900

OST 25 550

OST 10 363 Arginine-Sepharose (ornithine gradient) I .2 AST 21 430

OST 7 762

Arginine-Sepharose (KC1 gradient) 6 AST 28 200

U . mg protein-'

63 2.62 24

118 2.80 42

832 2.80 297

4 698 2.72 1727

17 860 2.76 6 470

scribed in Vander Wauven and Stalon (1985), and guanidino- valerate was synthesized according to Tricot et al. (1990).

RESULTS The various steps of a typical purification of arginine and

ornithine succinyltransferase activities are summarized (Table 2). An almost 280-fold purification of arginine and ornithine succinyltransferases was accomplished in this way (Table 2). Throughout the purification the L-arginine and L- ornithine succinyltransferase activities remained associated with the same fraction, an observation consistent with the multispecific nature of the enzyme. The ratio of arginine to ornithine succinyltransferase activities remained approxi- mately constant throughout the purification procedure with an average value of 2.7.

Characterization of the purified succinyltransferase activity

The purifed enzyme preparation gave only one major protein band on native polyacrylamide gel electrophoresis. This band accounted for more than 95% of the total proteins and corresponded to a molecular mass of 140 kDa. A similar molecular mass (150 kDa) was deduced from gel filtration. A single protein band was also observed on an isoelectrofo- cusing gel, with a gradient of pH 3-9. The isoelectric point of the enzyme was pH4.1-4.3. SDS polyacrylamide gel electrophoresis resolved the protein into two components in the same proportion with respective molecular masses of 35 kDa and 37 kDa. Compared to the molecular mass of the native enzyme, this would suggest that the succinyltransfer- ase has a tetrameric structure consisting of four subunits with two different sizes.

Microsequence analysis of the electroblotted protein bands gave two different but quite similar N-terminal amino acid sequences again indicating the existence of two distinct subunits, namely M L V M R P A Q A A D (37 kDa) and M I V R P V T S A D L P A L I (35 kDa).

Stability The final enzyme preparation retained full activity with

both substrates for several weeks when stored at -20°C.

However, the enzyme lost 10-40% of its activity following repeated freezing and thawing. Therefore, the preparation was stored as small aliquots. The enzyme was stable at 4°C for 48 h in 10 mM potassium phosphate, pH 7.5, even at a protein concentrations of less than 1 mg . ml-' but lost 25% of its activity when kept for 1 h in the assay buffer at 4°C.

Demonstration of the multispecific character of the enzyme

The preceding results clearly demonstrated that a single protein was endowed with both the arginine and ornithine succinyltransferase activities. Additional experiments were performed to confirm these results. Heat inactivation of the enzyme in 50 mM potassium phosphate, pH 7.5, revealed a simultaneous decrease of both catalytic activities. 50% inac- tivation was reached after 10 min of incubation at 50°C. However, the presence of either 10 mM arginine or 100 mM ornithine effectively protected the enzyme against thermal inactivation in the same way. Furthermore, both enzymic activities were sensitive to sulfhydryl reagents such as tlithio- nitrosobenzoic acid (0.5 mM) or 2-dithiopyridine (0.2 mM). Again, the enzyme was less sensitive to these reagents in the presence of either 10 mM arginine or 100 mM ornithine.

pH dependence of arginine and ornithine succinyltransferase activities

The optimal pH of the purified enzyme was measured in Tris/HCl buffer in the presence of 0.2 mM succinyl CoA and either 10 mM arginine or 100 mM ornithine. The succi- nyltransferase reaction exhibited optimal activity at pH 8.7 with arginine, at pH 8.5 with homoarginine and at pH 9.0 with ornithine used as substrates.

Arginine saturation curves The plots of reaction velocity of the purified arginine

succinyltransferase as a function of either succinyl CoA or arginine concentrations were hyperbolic and gave linear double reciprocal plots. An apparent K,,, value of 0.5 mM was obtained for arginine in the presence of 0.2 mM succinyl CoA and a value of 20 pM was obtained for succinyl CoA in the presence of 1 O m M arginine. The specificity of the succinyltransferase was tested by incubating the enzyme with

857

Table 3. Effectors of the arginine and ornithine succinyltransfer- ase activities. The enzyme was assayed in the presence of 100 mM Tris/HCI, pH 8.9, 0.2 mM succinyl CoA and 2 mM arginine or 25 mM ornithine. The relative activity is expressed as the ratio of activities with the effector versus activities in its absence. Results are mean values of at least two independent experiments. n.t., not tested. AST, L-arginine succinyltransferase ; OST, L-ornithine succi- nyltransferase.

Effector Concentrations Relative activity of

AST OST

mM %

None D- Arginine 10 D-Omithine 100 L-Citrulline 20 L-Lysine 10

100 Agmatine 10 Putrescine 10 Cadaverine 10 Spermidine 10 2-Aminovalerate 100 5- Aminovalerate 50 2,4-Aminobutyrate 50

100 59 2 4 56 2 5

100 2 2 97 2 3 56 2 5 59 2 5 68 2 6 75 2 7 56 2 5 83 2 7 60 2 5 60 2 6

100 360 -t 20 2 0 2 4

1002 2 105 2 4 6 6 2 6 5 5 2 4

n.t. n.t.

2 7 2 3 7 7 2 7 4 7 2 5 2 0 2 3

succinyl CoA and various amino derivatives and substrate analogs. The enzyme was inactive when L-arginine was sub- stituted by D-arginine, D-ornithine, L-lysine, L-citrulline, 2,4- aminobutyrate, 5-aminovalerate, 2-aminovalerate, agmatine or the polyamines propanediamine, putrescine, cadaverine and spermidine. Most of these compounds were effective in- hibitors of the arginine succinyltransferase reaction (Table 3). Double reciprocal plots of velocity versus L-arginine concen- tration at several spermidine, D-arginine or agmatine concen- trations showed that they act as competitive inhibitors with respect to L-arginine with K, values of 0.920.2, 2.620.4 and 3.4 2 0.5 mM, respectively (data not shown).

Ornithine saturation curves

In the presence of 100 mM L-ornithine, the velocity with respect to succinyl CoA concentration appeared hyperbolic and an apparent K,, value of 40 pM. In contrast, the ornithine saturation curve was clearly sigmoidal (Fig. 2) with a Hill index of 2.5 and an ornithine concentration at half-maximal velocity, [S], 5 , of 25 mM (Table 4). A similar curve was ob- tained with a crude cell extract indicating that the properties of this enzyme were not altered during the purification pro- cedure. Among the effectors of the arginine succinyltransfer- ase activity, D-arginine was shown to markedly increase the ornithine succinyltransferase activity (Fig. 3). At the ornith- ine concentration giving the half-maximal velocity, a maxi- mum activation of approximately threefold was observed in the presence of 10 mM D-arginine (Table 3). Upon further increase of the effector concentration, inhibition became do- minant and the activity decreased slowly (Fig. 3).

10 mM D-arginine almost entirely abolished the ornithine cooperativity, as indicated by the reduction of the Hill coeffi- cient h from 2.7 to 1.0, and decreased the ornithine concen- tration required to reach the half-maximum velocity from

0 20 40 60 80 100 120 140 [ Ornithine ] (mM)

Fig. 2. Ornithine saturation curve of the wild-type L-ornithine succinyltransferase activity. The reaction was carried out in 100 mM Tris/HCI, pH 8.9, in the presence of 0.2 mM succinyl CoA. 0, without effector; 0, in the presence of 10 mM D-arginine; 0, in the presence of 20 mM spermidine. The maximum velocity in the absence of effector was set equal to 100%. The standard deviations are 5-10%.

25 mM to 6.9 mM (Table 4, Fig. 2). Although f! aeruginosa has an arginine racemase, in vivo conversion of L-arginine into D-arginine does not seem to occur in the wild-type strain (Jann et al., 1988). Therefore, it is unlikely that D-arginine could be the physiological effector of the ornithine succi- nyltransferase activity. The purified enzyme showed the same velocity in the presence of 25 mM L-ornithine or 0.2 mM L-arginine. In the presence of both substrates, the enzyme activity was 2.5-fold higher than the sum of the indi- vidual reactions. These observations suggested that L-argi- nine is not only a substrate, but also an allosteric activator of the ornithine succinyltransferase activity.

L-Lysine also behaved as an activator of the ornithine succinyltransferase activity. Stimulation of the enzyme activ- ity was observed albeit at a much lower ornithine concentra- tion than with D-arginine (Fig. 4). Ornithine cooperativity was also reduced in the presence of 20 mM L-lysine, whereas the [S],, value for ornithine was shifted from 25 mM in the absence of the effector to 35 mM in its presence (Table 4). In contrast the polyamine, spermidine affected the ornithine saturation curve by increasing the cooperativity index and the ornithine concentration required to reach half maximal activity (Table 4 ; Fig. 2). Consequently, the ornithine succi- nyltransferase activity clearly showed all the manifestations of an allosteric enzyme, namely cooperativity with respect to its substrate, activation and inhibition of its activity by effectors.

Saturation curves for L-homoarginine

L-Homoarginine was found to be an alternative substrate of the purified enzyme. The pH optimum for the reaction was nearly identical to the optima found when L-arginine or L-ornithine were used a substrates. In contrast to L-arginine, but like L-ornithine, L-homoarginine appeared to be a cooper- ative substrate (Fig. 5). Activation by D-arginine and inhibi- tion by spermidine were also observed, although in the latter

858

Table 4. Values of the half-saturating substrate concentrations and the cooperativity coefficient h for the saturation curves of the wild-type and mutant succinyltransferases. Assays were performed in 100 mM Tris/HCl, pH 8.9, containing 0.2 mM succinyl CoA. [S],, is the concentration of substrate at half-maximal velocity; h, the Hill coefficient. The results are means of at least two independent expriments.

Strain Substrate Effector Concentration [Sl" 5 h

mM

PA0 1.-arginine L-arginine L-arginine L-arginine L-ornithine L-ornithine L-omithine L-ornithine L-homoarginine L-homoarginine L-homoarginine

L-ornithine L-ornithine

PA061 90 L-arginine

none D-arginine spermidine L-lysine none D-arginine spermidine L-lysine none D-arginine spermidine none none D-Xginine

0 10 20 20 0

10 20 20 0

10 20 0 0

10

0.5 ? 0.1 2.4 ? 0.3 3.9 ? 0.4 1.2 ? 0.1

6.9 ? 0.8 25 ? 3

52 ? 5 35 ? 3

11 ? 1 18 ? 2 1.4 ? 0.2

15 2 2 8.5 ? 1

7.4 ? 0.8

1 +- 0.05 1 -+ 0.05 1 ? 0.05 1 +- 0.05 2.5 ? 0.2 1.1 ? 0.05 3.5 2 0.3 1 ? 0.05 2.7 ? 0.2 1.2 ? 0.1 2.7 ? 0.2 1 ? 0.05 1.5 t- 0.2 1 ? 0.05

a, > m a, [r

.- c -

I I I

0 5 10 15 20

[ Effector] (mM)

Fig. 3. Influence of D-arginine and spermidine on L-ornithine succinyltransferase activity. The enzyme was assayed in the pres- ence of 100mM TrisMC1, pH8.9, 0.2mM succinyl CoA and 25 mM L-ornithine. 0, D-arginine; 0, spermidine. The relative velocity is the ratio of the activity in the absence of effector to that in its presence multiplied by 100. The standard deviations are 5- 10%.

case, the inhibitor did not seem to significantly affect the L- homoarginine cooperativity (Table 4).

Mutants growing at higher rates on L-ornithine

The growth of R aeruginosa on L-ornithine medium was distinctly slower than on L-arginine medium (Table 5). Nev- ertheless, mutant strains that grew at higher rates than the parent strain on L-ornithine could be obtained. After N-meth- yl-N'-nitro-N-nitrosoguanidine mutagenesis of strain PA0505, twelve mutants were selected for their ability to

o ! . , . , . I , , . , 0 20 40 60 80 1 00

[ Lysine] (mM)

Fig. 4. Influence of L-lysine on L-ornithine succinyltransferase activity. The conditions of the experiment were as described in the legend of Fig. 2 except that omithine was maintained at 25 m M (0) or 10 mM (0). The standard deviations are 5 - 10%.

form larger colonies on L-ornithine. Eight of these mutants had normal levels of the enzymes of the arginine succi- nyltransferase pathway. Over 92% of the recombinants reco- vered after transduction of strain PA0977 with phage GlOl propagated on each of the mutants, and selected for growth on L-arginine, had also gained the capacity to grow faster on L-ornithine. This indicated that all of these mutations were localized in the vicinity of the previously described aru clus- ter (Friih et al., 1985; Vander Wauven et al., 1988) in the 65 -70 min region of the revised R aeruginosa chromosome map (O'Hoy and Krishnapillai, 1987). The kinetic properties of the succinyltransferase from the mutant PA06190 in a crude cell extract were compared to those of the wild-type enzyme. The mutant enzyme had a strongly reduced coopera-

859

' O 0 1

0 10 20 30 40 [ Homoarginine ] (mM)

Fig. 5. L-Homoarginine saturation curve. The reaction was carried out in 100 mM Tris/HCI, pH 8.9, in the presence of 0.2 mM succinyl CoA. 0, without effector; 0, in the presence of 10 mM D-arginine, 0, in the presence of 20 mM spermidine. The maximum velocity in the absence of effector was set equal to 100%. The standard devia- tions are 5-10%.

Table 5. Regulation of succinyltransferase synthesis. Cells were grown at 37°C in minimal medium 154 supplemented with sub- strates as indicated. Concentrations were as described in the Materi- als and Methods section. Enzyme assays were performed in dupli- cate on two independently grown cultures. Values are means, the standard deviations were approximately 5 - 10%. AST, L-arginine succinyltransferase

Strain Growth medium Doubling Specific time activity

AST

PA01 glutamate L-arginine L-ornithine D-arginine L-lysine glutamate + L-arginine succinate + L-arginine L-ornithine + D-arginine L-ornithine + L-lysine glutamate + homoarginine L-arginine + homoarginine L-ornithine + homoarginine succinate + homoarginine

PA0505 L-arginine L-ornithine

PA061 90 L-arginine L-ornithine

PA061 50 L-ornithine

min U . m g

60 < 1 66 75

220 28 162 82 7 80 3

54 39 54 16 99 58

138 21 100 < I 330 62 515 13

3800 < 1

60 80 245 26

60 56 120 20 180 20

protein-'

tivity for L-ornithine as expressed by the Hill coefficient and a lower L-ornithine concentration at half-maximal velocity (Table 4). Consequently, the mutant was much more efficient in succinylornithine synthesis than the wild-type enzyme and this was consistent with the increased growth rate observed on ornithine as the sole carbon source. Mutant strains

PA06191 and PA06192, also growing at higher rates on L- ornithine than the parent strain, had a succinyltransferase with a similarly reduced L-ornithine cooperativity as strain PA06190 (data not shown).

The four remaining mutant strains that showed some increase in growth rate on L-ornithine medium were unable to use L-arginine as the sole carbon source. This defect turned out to result from a deficiency in the succinylarginine dihy- drolase activity, the second enzyme of the arginine succi- nyltransferase pathway. Other dihydrolase mutants, of which PA06150 is a good representative, had the same growth rate increase on L-ornithine (Vander Wauven et al., 1988). Cells of strain PA06150 growing on glutamate-arginine medium, excreted urea in addition to succinylarginine. This strongly suggested that, beside the arginine succinyltransferase activ- ity, the D-arginine dehydrogenase route (Jann et al., 1988) leading to guanidinobutyrate and urea participated in the utilization of arginine in this mutant strain. When PA06150 was cultivated on L-ornithine as the sole carbon and nitrogen source, some succinylarginine and urea were also found in the culture fluid, indicating that a substantial part of ornithine was converted into L-arginine and D-arginine through the ar- ginine biosynthetic pathway, both compounds being activa- tors of the ornithine succinyltransferase activity. This obser- vation probably accounts for the decrease in generation time observed with succinylarginine dihydrolase mutants on L-or- nithine medium.

Effect of different effectors on growth The addition of D-arginine improved the growth rate on

L-ornithine (Table 5). D-Arginine, however, was a relatively good carbon source for I? aeruginosa. More remarkably, L- lysine, a poor carbon source for the wild-type strain, reduced significantly the doubling time on L-ornithine medium, in spite of the fact that the succinyltransferase specific activity was lower than in the L-ornithine-grown cells (Table 5) . This observation suggests that the kinetic and allosteric properties of the succinyltransferase may have an influence on the growth rate.

In contrast, L-homoarginine, a poor nitrogen source for I? aeruginosa, led to growth inhibition on L-arginine or L- ornithine medium (Table 5). All the L-homoarginine that had been added to either L-ornithine or L-arginine media was re- covered as succinylhomoarginine. Consequently, the large increase in doubling time produced by L-homoarginine on L- arginine or L-ornithine media was most probably caused by the L-homoarginine consumption of succinyl CoA which could not be regenerated from the excreted succinylhomoar- ginine. In cells where the succinyltransferase activity was not induced, L-homoarginine utilization seemed to follow an- other pathway (Fig. 1). The homoarginine which had been consumed in succinate or glutamate medium was recovered under the form of excreted guanidinovalerate, a compound not used by I? aeruginosa (Tricot et al., 1990). This observa- tion is reminiscent of the D-arginine dehydrogenase pathway occurring in I? aeruginosa (Jann et al., 1988). Resting cells first grown on glutamate and L-homoarginine and subse- quently incubated with 10 mM L-homoarginine in the pres- ence of 10 mM aminooxyacetic acid, an inhibitor of pyri- doxal-phosphate-dependent enzymes, did not accumulate gu- anidinovalerate any more. This is expected if a pyridoxal- phosphate-dependent racemase is involved in the intercon- version of the enantiomers and if D-homoarginine is an inter- mediate in the utilization of L-homoarginine.

860

DISCUSSION

The arginine and ornithine succinyltransferases activities from l? aeruginosa are coordinately induced by L-arginine or L-ornithine and both activities are absent in mutants defective in the aerobic utilization of L-arginine (Jann et al., 1986; Vander Wauven et al., 1988). While this observation suggests the existence of an enzyme showing substrate multiplicity, definitive evidence for this fact came only from the complete purification of the aforementioned activities, as described in this study.

The arginine and ornithine succinyltransferase activities were copurified from l? aeruginosa cells grown on L-argi- nine. The ratio of these activities remained approximately constant during the different chromatographic steps used in the course of the purification procedure.

The enzyme has a molecular mass of 140 kDa and is probably composed of two kinds of subunits, 35 kDa and 37 kDa in equivalent amounts, as indicated by SDSPAGE and microsequence analysis. The latter showed two different but albeit quite similar N-terminal amino acid sequences. A definite proof that two kinds of polypeptides are involved in succinyltransferase activity should also come from character- ization of the corresponding genes. Interestingly, the P. put- ida P-ketoadipate succinyl CoA transferase was also formed by the association of non-identical subunits bearing signifi- cant sequence resemblances (Yeh and Ornston, 1982)

Henri-Michaelis kinetics were observed for the saturation by L-arginine whereas sigmoidal saturation curves were ob- served for L-ornithine and for the L-arginine analog, L-homo- arginine. The allosteric behaviour of L-ornithine succi- nyltransferase was also sugested by the multimodulation of its activity by positive and negative effectors. D-Arginine and L-arginine stimulated the succinylation of L-ornithine by re- ducing the cooperativity and the substrate concentration re- quired to reach-half maximal velocity [S],, + Spermidine acted as an inhibitor by increasing the cooperativity index and the [S],,, value.

A striking feature of the kinetic behaviour of allosteric enzymes is the increase of enzymic activity upon the addition of a low concentration of substrate analogs. This stimulation of activity by compounds that compete with the substrate at the active sites represents one of the earliest clues that partial saturation of the enzyme by the analog leads to a conforma- tional transition resulting in enhanced affinity of the unoccu- pied site (Monod et al., 1965). In the case of ornithine succi- nyltransferase, the introduction of L-lysine into the reaction mixture resulted in the disappearance of the sigmoidal nature of the velocity plot with respect to L-ornithine and, conse- quently, in the activation of the enzyme at low concentration of the substrate. D-Arginine and L-arginine also promoted extensive stimulation of ornithine succinyltransferase. The fact that the enzyme displayed a sigmoidal dependence of activity versus L-homoarginine and followed Henri-Michae- lis kinetics with respect to L-arginine led to the hypothesis that the enzyme may possess two different sites, an active site and an allosteric site for D-arginine and L-arginine. Addi- tional data are however needed to solve this question.

The growth of strain PA01 was slower on L-ornithine than on L-arginine, possibly because the level of the succi- nyltransferase pathway enzymes was limiting during growth on L-ornithine. Nevertheless, a mutant of the ornithine succi- nyltransferase with enzyme levels similar to that of the wild- type strain on L-ornithine medium, but with reduced coopera- tivity towards L-ornithine, showed an increased growth rate

as compared to the wild-type strain. This led us to conclude that the kinetic properties of the ornithine succinyltransfrrase enzyme determine, beside the level of enzyme, the growth rate of l? aeruginosa on L-ornithine medium.

Although L-homoarginine is a poor nitrogen source, its utilization by P: aeruginosa leads to the appearance of gua- nidinovalerate, a compound not used by the wild-type strain (Tricot et al., 1990). However, cells induced for succi- nyltransferase activity converted homoarginine to excreted succinylhomoarginine. Hence, homoarginine-dependent con- sumption of succinyl CoA, which cannot be regenerated from succinylhomoarginine, probably accounts for the lowering of the growth rate on L-ornithine and L-arginine medium in the presence of L-homoarginine.

We thank Prof. A. PiCrard for reading the manuscript, Prof. D. Haas for stimulating discussions, and C. Verhulst and W. Akli for their help in some experiments. This study was supported by grants from the Belgian Fund for Joint Basic Research (grant numbers 2.4507.91 and 2.9007.92) and by the Research Council from the UniversitC Libre de Bruxelles. V. S. is Research Associate of the National Fund for Scientific Research (Belgium).

REFERENCES Friih, R., Haas, D. & Leisinger, T. (1985) Altered control of gluta-

mate dehydrogenase in ornithine utilization mutants of Pseudo- monas aeruginosa, Arch. Microbiol. 141, 170-176.

Hanes, C. S. (1932) LXVII: Studies on plant amylases. I. The effect of starch concentration upon the velocity of hydrolysis b! amy- lase of germinated barley, Biochem. J. 26, 1406-1421.

Hill, A. V. (1910) The possible effect of disaggregation of the mole- cules of hemoglobins on its dissociation curves, J. Physiol. 40,

Holloway, B. W., Krishnapillai, V. & Morgan, A. F. (1979) Chromo- somal genetics of Pseudomonas, Microbiol. Rev. 43, 73- 102.

Holloway, B. W., O’Hoy, K. & Matsumoto, H. (1987) Pseudomonas aeruginosa PAO, in Genetic maps (O’Brien, S . J., ed.) vol. 4, pp. 213-221, Cold Spring Harbor Laboratory, Cold Spring Har- bor, NY.

Jann, A,, Stalon, V., Vander Wauven, C., Leisinger, T. & Haas, D. (1 986) Succinylated intermediates in an arginine catabolic path- way of Pseudomonas aeruginosa, Proc. Nut1 Acad. Sci. USA 83, 4937-4941.

Jann, A,, Matsumoto, H. & Haas, D. (1988) The fourth arginine catabolic pathway of Pseudomonas aeruginosa, J. Gen. .Micro- biol. 134, 1043-1053.

Lineweaver, H. & Burk, D. (1934) The determination of enzyme dissociation constants, J. Am. Chem. Soc. 56, 658-666.

Matsudaira, P. (1 987) Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membranes, .I. Biol. Chem. 262, 10035-10040.

Mercenier, A., Simon, J. P., Haas, D. & Stalon, V. (1980) Catabolism of L-arginine by Pseudomonas aeruginosa, J. Gen. Microbiol.

Miklus, J. M. & Stein, I. (1973) The colorimetric determination of mono and disubstituted guanidines, Anal. Biochern. 54, 545 - 553.

Monod, J., Wyman, J. & Changeux, J. P. (1965) On the nature of allosteric transitions, J. Mol. Biol. 12, 88-108.

O’Hoy, K. & Krishnapillai, V. (1987) Recalibration of the Pseudo- monas aeruginosa strain PA0 chromosome map in time units using high-frequency of recombination donors, Genetics 115, 611-618.

Stalon, V., Ramos, F., PiCrard, A. &Wiame, J. M. (1967) The occur- rence of a catabolic and an anabolic omithine carbamoyltransfer- ase, Biochim. Biophys. Acta 139, 91 -97.

Stalon, V. (1972) Regulation of the catabolic omithine carbamoyl- transferase of Pseudomonas fluorescens. A study of the a1 losteric properties, Eur: J. Biochem. 29, 36-46.

4-8.

116, 381 -389.

861

Stalon, V., Vander Wauven, C., Momin, P. & Legrain, C. (1987) Catabolism of arginine, citrulline and ornithine by Pseudomonas and related bacteria, J. Gen. Microbiol. 133, 2487-2495.

Tricot, C., PiCrard, A. & Stalon, V. (1990) Comparative studies on the degradation of guanidino and ureido compounds by Pseu- domonas, J. Gen. Microbiol. 136, 2307-2317.

Vander Wauven, C., PiCrard, A., Kley-Raymann, M. & Haas, D. (1 984) Pseudomonas aeruginosa mutants affected in anaerobic growth on arginine. Evidence for a four-gene cluster encoding the arginine deiminase pathway, J. Bacteriol. 160, 928-934.

Vander Wauven, C. & Stalon, V. (1985) Occurrence of succinyl de- rivatives in the catabolism of arginine in Pseudomonas cepacia, J. Bacteriol. 164, 882-886.

Vander Wauven, C., Jann, A,, Haas, D., Leisinger, T. & Stalon, V. (1988) W-Succinylornithine in ornithine catabolism of Pseudo- rnonas aeruginosa, Arch. Microbiol. 150, 400-404.

Voellmy, R. & Leisinger, T. (1976) Role of 4-aminobutyrate amino- transferase in the arginine metabnolism of Pseudomonas aerugi- nosa, J. Bacteriol. 128, 722-729.

Yeh, W. K. & Ornston, L. N. (1982) Similar structures in y-carboxy- muconolactone decarboxylase and P-ketoadipate succinyl coen- zyme A transferase, J. Bacteriol. 149, 374-377.