Eur. J. Biochem. 204, 531 -537 (1992) J? FEBS 1992

Arginine catabolism in the phototrophic bacterium Rhodobactev capsulatus ElFl Purification and properties of arginase Conrad0 MORENO-VIVIAN I , German SOLER2 and Francisco CASTILLO '

'

*

Departamento de Bioquimica, Biologia Molecular y Fisiologia, Facultad de Ciencias, Universidad de Cordoba, Spain Departamento dc Bioquimica y Biologia Molecular, Facultad de Veterinaria, Universidad de Extremadura, Spain

(Received June 27, 1991) - EJB 91 0838

The phototrophic bacterium Rhodobacter capsulatus E l F l grew with L-arginine or L-homoarginine as nitrogen source under lightlanaerobiosis. However, when L-arginine was used as the only source of both carbon and nitrogen, the bacterium exhibited weak growth levels and the excess of nitrogen was excreted to the medium as ammonia. By contrast, L-ornithine was used under phototrophic conditions as either nitrogen or carbon source. Other compounds of the arginine catabolic pathways, such as putrescine or proline, also supported phototrophic growth of this bacterium. Under heterotrophic/dark conditions, R . capsulatus always showed a low growth rate with those nitrogen compounds. Cells growing on media containing L-arginine, L-homoarginine or L-ornithine induced an Mn'+-dependent arginase activity regardless of the presence of ammonium ions and other readily utilizable nitrogen sources. Arginase activity was strongly inhibited by Zn2 ,Cu2 ,borate, L-cysteine, L-ornithine and y-guanidinobutyrate. Mercurials also inactivated arginase, the activity being partially restored by the presence of thiols. Arginase was purified to electrophoretic homogeneity and found to consist of four identical subunits of 31 kDa. The molecular parameters and kinetic constants of arginase from R . capsulatus E l F l resembled those previously described for the Saccharomyces cerevisiae enzyme rather than those of bacterial arginases. +

L-Arginine can serve as nitrogen source for growth of many microorganisms including fungi which couple arginine catabolism to synthesis of proline [l] and bacteria which degrade this amino acid through several metabolic pathways [2, 31. The arginase pathway is the best known enzymatic process of arginine catabolism in living organisms; the first enzyme of the route, arginase, has been exhaustively characterized in fungi [4 - 71. Nevertheless, biochemical studies on bacterial arginase are scarce [8,9]. The metabolic fate of arginase reaction products may differ from one organism to another. Some cyanobacteria cannot metabolize ornithine and thus only use the arginase pathway as a source of nitrogen by coupling urea production to urease activity [lo]. In contrast, many bacilli are unable to utilize urea, but can use ornithine as a source of both nitrogen and carbon through the glutamate semialdehyde pathway [3]. Urease and ornithine aminotransferase have been described in Agrobacterium turnefaciens strains [l 11. Arginase activity is mainly regulated at level of enzyme synthesis. In yeasts, arginase is induced by L-arginine in the absence of ammonium or in nitrogen-starved cells [12- 141. In Neurospora crassa, an overproduction and compartmentation of metabolites have been proposed to regulate arginine metabolism [15]. Regulation of the arginase pathway by the nitrogen source has been also reported in bacteria. Thus, in Correspondence to F. Castillo, Departamento de Bioquimica, Biologia Molecular y Fisiologia, Facultad de Ciencias, Universidad de Cordoba, E-14071 Cordoba, Spain Enzymes. Arginase (EC 3.5.3.1); ornithine aminotransferase (EC 2.6.1.1 3); ornithine carbamoyltransferase (EC 2.1.3.3); putrescine aminotransferase (EC 2.6.1 .-); pyrroline dehydrogenase (EC 1.2.1.19); urease(EC 3.5.1.5).

+

Streptomyces clavuligerus arginase is repressed by ammonia and induced by L-arginine [16]. Other bacterial enzymes of arginine catabolism such as ornithine aminotransferase, pyrroline dehydrogenase or putrescine aminotransferase are also regulated by the nitrogen supply [17-191. A role of arginase as inhibitor of ornithine carbamoyltransferase in the presence of L-arginine and L-ornithine, thus preventing a futile urea cycle, has been reported in yeasts [20,21] and bacilli [22]. Arginase has been purified and characterized at the molecular level in some microorganisms such as Neurospora crassa [5, 61, Saccharomyces cerevisiae [7, 21, 231, Bacillus licheniformis [8], Bacillus anthracis [9], and Bacillus subtilis [22]. On the basis of the quaternary structure and kinetic properties of the enzyme, two classes of arginases have been proposed: arginase of ureotelic type, found in S. cerevisiae, which has a low molecular mass (140 kDa) and a low apparent K,,, for Larginine (1 O mM), and arginase of uricotelic type, found in fungi and Bacillus, which has a high molecular mass (260 kDa) and a high apparent K , for L-arginine (100 mM) [5]. In the present paper we report, for the first time in a phototrophic bacterium, the purification and molecular properties of an arginase. The multimeric structure of the enzyme and the influence of the nitrogen source on activity levels have also been studied.

MATERIALS AND METHODS Organism and growth conditions Rhodobacter capsulatus strain E l F l was cultured phototrophically under anaerobic conditions in a culture chamber

532 at 30°C under continuous illumination (40 W . m-2). Dark/ aerobic conditions were achieved by culturing the cells at 30°C in conical flasks capped with sterile cotton stoppers in a Gallenkamp orbital incubator at 120 rev./min. Cells were cultured in the RCV medium [24] with DL-malate (4 g . 1-') as carbon source and each of the following nitrogen sources (1 g . 1- '): NH4CI, L-arginine, L-ornithine, L-homoarginine, L-proline, L-lysine, L-glutamate, L-glutamine or putrescine. Where indicated, m-malate was omitted and L-arginine, Lornithine or putrescine (2 g . 1-') was used as both carbon and nitrogen sources. Growth was monitored by following the absorbance at 680 nm, and the purity of cultures was checked by plating them on agar solid media containing 1.5% Difco-Bacto agar and 1YOyeast extract.

Step IV. L-Arginine -agarose chromatograph??

Preparation of crude extracts

Enzyme assays

Cells were harvested at the end of the exponential phase of growth by centrifugation at 20000 x g for 15 min and, after washing with 50 mM Tris/HCl pH 7.5, resuspended in 50 mM Tris/HC1 pH 7.5, 1 mM MnClz (buffer A). Cells were stored by freezing at -20 'C without loss of arginase activity and broken, after 1 mM phenylmethylsulfonyl fluoride addition, by ultrasonic disruption at 90 W for 5 rnin (ten periods of 30-s cavitation, each followed by a 30-s pause) in a Vibracell sonifier (Sonics & Materials Inc., Danbury, Connecticut, USA). Broken cells were centrifuged at 36000 x g for 30 rnin and the resulting supernatant was centrifuged at 60000 x g for 4 h to remove chromatophores. The high-speed supernatant was used as starting material for enzyme purification.

Arginase (L-arginine amidinohydrolase) was assayed colorimetrically by measuring either the urea or the ornithine formed in the reaction. The incubation mixture contained 10 pmol L-arginine pH 9.0, 0.5 pmol MnCI2, 100 pmol carbonate/bicarbonate pH 9.0, the enzyme preparation and distilled water to a final volume of 0.5 mi. After a 10-min incubation at 30"C, the reaction was stopped by adding either 0.5 ml 0.5 M HC104 (urea method) or 1.5 ml glacial acetic acid (ornithine method). One unit (U) of enzyme activity is defined as the amount of enzyme which catalyzes the formation of 1 pmol product/ min .

Purification procedure

Urea was measured by the method of Archibald [25]. Ornithine was measured at 515 nm by the ninhydrin reaction in the presence of acetic acid [26]. Extracellular ammonia was determined colorimetrically at 410 nm by the Conway microdiffusion procedure (271. Protein was estimated by the methods of Lowry et al. [28] and Bradford [29] using bovine serum albumin as a standard.

Step I . Ammonium sulfate precipitation and ge1,filtrution chromatography

The 60000 x g supernatant was adjusted to 30% saturation with solid (NH4)2S04 and stirred gently at 4°C for 30 min. The solution was centrifuged at 20000 x g for 15 min and the pellet was discarded. The supernatant was brought to 50% saturation with ammonium sulfate, stirred and centrifuged as above. The pellet was resuspended in an appropriate volume of buffer A and filtered through a Sephacryl S-300 HR column (1.6 x 100 cm) previously equilibrated with the same buffer at a flow rate of 60 ml . h-'. Step 11. Amicon green und Cibacron blue F3GA ugurose chromatographies

Fractions containing arginase from gel filtration chromatography were pooled and passed through a column of Amicon green (1.1 x 11 cm), previously equilibrated with buffer A at a flow rate of 20 ml . h-'. As the arginase activity was not retained, the effluent was passed at the same flow rate through a Cibacron blue F3GA agarose column (1.1 x 6 cm) equilibrated with the same buffer. Arginase was not retained in this column, either. Step III. DEAE-Sephacel chromutography

The effluent from step I1 was applied to a DEAE-Sephacel column (2.5 x 11.5 cm) equilibrated with buffer A. After washing with the same buffer, the enzyme was eluted by using a linear gradient of 0 - 1.O M KC1 at a flow rate of 20 ml . h- '. Fractions containing arginase activity were pooled and diluted 20-fold with buffer A to decrease ionic strength.

Diluted fractions from step 111 were passed at a flow rate of 20 ml . h - ' through a column (1.1 x 8 cm) of L-arginineagarose (Sigma Chemical Co.) previously equilibrated with buffer A. The enzyme was eluted with 50 mM L-arginine dissolved in the same buffer and the fractions containing arginase activity were pooled. Step V . Heating

Pooled fractions from step IV were heated at 65°C for 15 min and centrifuged at 20000 x g for 15 min. The supernatant contained highly purified arginase as deduced from electrophoretic criteria.

Analytical determinations

Determination of molecular parameters The molecular mass of the arginase subunit was determined electrophoretically under denaturing conditions in the presence of SDS by using the following standard proteins: rabbit muscle myosin (205 kDa), Escherichiu coli /I-galactosidase (116 kDa), rabbit muscle phosphorylase b (97.4 kDa), bovine serum albumin (66 kDa), egg albumin (45 kDa), rabbit muscle glyceraldehyde-3-phosphatedehydrogenase (36 kDa), bovine erythrocytes carbonic anhydrase (29 kDa), bovine pancreas trypsinogen (24 kDa) and soybean trypsin inhibitor (20.1 kDa). Sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE) was performed according to Laemmli [30] using gel slabs with a 12% resolving gel. Samples were treated with 30 mM dithiothreitol, 2% (masslvol.) SDS, 10% (byvol.) glycerol and 1YO (massivol.) bromophenol blue without or with thermal treatment at 100°C for 10 min to denature the proteins. Proteins in gels were stained with 1% (mass/vol.) Coomassie brilliant blue R-250 in 7% (by vol.) acetic acid, the gels were subsequently destained with a mixture 10% acetic acid and 10% isopropyl alcohol for 12 h. The molecular mass of the native enzyme was calculated from chromatographic and sedimentation data by the methods of Siege1 and Monty I311 and Pundak and Eisenberg [32], respectively. The sedimentation coefficient s20,wwas de-

533 1.5 200

/

0-.-•

+' -*

A

.

400

/ ..

1 .o

h

4

-

300

0

u 4

I

100

0

0.5

3

e,

H 2

.'

/

z

A

0

E

Y

4m-o-o-o-

0

200

h

0.0

sfj

I

U

c

W

100 I

o~-o-o-o

z

fi

!i

1.0

g

100

0.5

0.0

0

10

20

30 TIME (h)

40

50

0

12

TIME (h)

2I 0

0

' 8

4

0

b I

Induction of arginase by L-arginine, L-ornithine and Lhomoarginine in R. cupsulutus E l F l . Cells from L-glutamate/DI.-malate cultures were washed and grown phototrophically in the presence of L-arginine ( A ) , L-arginine plus ammonia (0),L-ornithine (0),I.homoarginine ( O ) , minus nitrogen ),(. and I*-arginine plus SO pg . ml-' chloramphenicol or SO pg . ml-' rifampicin ( 0 ) .The activity was measured as indicated in Materials and Methods at the times indicated in the figure

60

Fig. 1. Time course of growth, arginase activity and ammonia excretion to the medium by R. cupsulutus E l F l cells grown under phototrophic and heterotrophic conditions. R. capsulutus El F1 were inoculated in media with L-arginine as the nitrogen source and m-malate as the carbon source under light anaerobic (A) and dark acrobic (B) conditions. Cell growth ),(. arginase activity (A)and ammonium ions in the medium (0) were determined as indicated in Materials and Methods a t the times indicated in the figure.

1.5

1 .o

0.5

termined by the method of Martin and Ames [33], as previously described [34,35], by centrifugation in a linear sucrose gradient (5 - 20%) of arginase and the following standard proteins : bovine thyroglobulin (19.45 S), horse spleen apoferritin (17.65 S ) , bovine liver catalase (1 1.35 S), yeast alcohol dehydrogenase (7.65 S), bovine serum albumin (4.65 S) and bovine erythrocytes carbonic anhydrase (3.25 S). The Stokes' radius (a) [31] and the diffusion coefficient ( D 2 0 . w )[36] were estimated by gel filtration in a column (2.4 x 36 cm) of Sephacryl S-300 HR with the following standards: bovine pancreas ribonuclease A (1.80 nm; 11.9 x lo-' cm-2 . s- I ) , bovine serum albumin (3.55 nm; 5.90 x lo-' cm2 . s- '), yeast alcohol dehydrogenase (4.61 nm; 4.70 x lo-' cm2 . s-'), bovine liver catalase (5.22 nm; 4.10 x cmz . s-') and horse spleen apoferritin (7.80 nm; 3.60 x cm2 . s-'). The relative frictional quotient (f/fo)was calculated according to the method of Brewer et al. [37], using the value of 0.731 cm3/g for the partial specific volume of arginase subunit calculated by Duong et al. [7]. Western blots were performed as described by Towin and Gordon [38] by using polyclonal antibodies obtained from rabbits previously immunized against rat liver arginase purified according to Soler and coworkers (unpublished work).

RESULTS Rhodohucter cupsulatus El F1 cells grew phototrophically with L-arginine as nitrogen source by inducing an arginase

0.0 A

B

C

D

E

Fig.3. Effect of Mn" on cell growth and arginase activity of R. cupsulutus E l F l . Cells were grown phototrophically with L-arginine as nitrogen source and DL-malate as carbon source in the absence (A) or in the presence of 1 pM Mn2+ (B), 10 pM MnZi (C), 100 pM Mn2+ (D)and 1000 pM Mn2+ (E).Growthandarginaseactivitywere determined as indicated in Materials and Methods in cells harvested a t thc end of the logarithmic phase of growth.

activity which paralleled the logarithmic growth phase (Fig. 1A). The appearance of this activity was due to de n o w protein synthesis since it did not take place in the presence of transcription or translation inhibitors rifampicin and chloramphenicol (Fig. 2). Induction of arginase required the presence of L-arginine and was independent of the presence of ammonium ions in the culture medium (Fig. 2 ) . The enzyme was also induced in cells growing with L-ornithine or Lhomoarginine as nitrogen sources. As for cells growing in Larginine, the presence of other nitrogen compounds such as ammonium ions, urea, L-glutamine, L-glutamate or L-proline did not affect arginase induction in L-ornithine cultures (results not shown). Induction of arginase took place under phototrophic anaerobic conditions either with L-arginine as nitrogen source

534 in the presence of DL-malate (Fig. 1A) or with L-arginine as the source of both carbon and nitrogen (not shown). The ability of the cells to use L-arginine as N source was limited under dark/aerobic conditions, where a high excretion of ammonia and a continuous induction of the enzyme were observed (Fig. 1B). When R . cupsulutus ElFl was grown with

Table 1. Effect of divalent cations on the arginase activity of R. cupsulutus ElFZ. Arginase activity was determined as indicated in Materials and Methods in the presence of the divalent cations listed. Ba2 ', Caz+,Mgz+, Snz+ and NiZ+were without effect.

Addition

Concn

Relative arginase activity

None Fez Fe2 CoZ coz cu2+ c u z+ Zn2+ ZnZ+

mM 0.1 1 .o 0.1 1.o 0.1 1 .o 0.1 1.o

Yo 100 112 115 73 59 47 16 31 6

+

+

+

L-arginine as carbon and nitrogen sources under phototrophic conditions, patterns of growth, arginase induction and ammonia excretion were similar to those observed under dark/ aerobic conditions (Fig. 1B). Neither cell growth, arginase induction nor ammonia excretion was observed under dark/ aerobic conditions with L-arginine as the sole carbon and nitrogen source (results not shown). Arginase activity required Mn2'. Maximal growth and arginase activity were obtained in cells from media supplied with 1 pM MnC12, whereas higher concentrations of this cation showed little inhibitory effect (Fig. 3). R. cupsulutus E l F l seems to have some capability to store Mn2', since the requirement of the cation in vitro was about 1.0 mM. Addition of several divalent cations to the arginase assay had no effect except for Fe2+, which caused a slight activation of the enzyme, and Cu2+or Zn2+,which strongly inhibited arginase (Table 1). Arginase was purified from R . capsulatus E l F l by using a procedure that combined anion-exchange, gel-filtration and affinity chromatographies with thermal treatment in the presence of 50 mM L-arginine (Table 2). A highly purified preparation, exhibiting a single protein band of 31 kDa on SDS/

L

4.9

I 0,

0 -

G3P DH 4.5

4.1

'

0.0

Fig. 4. SDS/PAGE of fractions from the purification steps. Samples from fractions containing arginase activity of the indicated purification procedure steps were analyzed by SDSjPAGE performed as indicated in Materials and Methods. Lane 1 , 2 p1 crude extract; lane 2, 5 pl effluent of Sephacryl S-300 HR chromatography (step I); lane 3 , 5 pI effluent of DEAE-Sephacel chromatography (step 111); lane 4, 1 pl supernatant of heated (65"C, 15 min) cffluent of 1.-arginineagarose chromatography (stcp IV) without treatment at 100°C; lane 5 , the same as lanc 4 but with a thermal treatment at 100°C for 10 min to dissociate native protein; lane 6, the same as lane 4 but with 10-pl sample; lanc 7, the same as lane 5 but with 10-pl sample.

1 .o

0.5

Rf

Fig. 5. Determination of molecularmasses of arginase subunit and native enzyme by SDS/PAGE. SDSjPAGE was performed in slab gels as indicated in Materials and Methods. Standards used were: myosin (myo), b-galactosidase @-gal), phosphorylase h (phos b), bovine serum albumin (SA), egg albumin (EA), glyceraldehyde-3-phosphate dehydrogenase (G3P DH), carbonic anhydrase (CA), trypsinogen (TRP) and trypsin inhibitor (TI). (A) Tetrameric arginase (unheated sample); (B) arginase monomer (sample heated at IOOT, 10 min)

Table 2. Purification of arginase from R. cupsulutus ElF1.

Fraction

Crude extract Ammonium sulfate/Sephacryl S-300 HR Amicon green/Cibacron blue F 3GA DEAE-Sephacel L-Arginine-agarose Heating (15 min, 65°C) supernatant

Volume

Protein

Specific activity

Total activity

Yield

Purification

ml 150 24 30 8 2

mg 398 56.9 7.2 1.9

U/mg

U 158 73.4 59.4 12.2 18.4 18.3

YO 100 46.5 37.6 7.7 11.6 11.6

-fold 1 3.2 20.5 16.0 65.7 451.5

2

0.7 0.1

0.4

1.3 8.2 6.4 26.3 183.0

535 Table 3. Molecular and kinetic parameters of arginase from R. capsulatus ElF1.

Parameter

Value

Stokes' radius, x Sedimentation coeffcicnt, s20,w Diffusion coeflicient, D20,w Frictional coefficient,f Frictional quotient.flfo Molecular mass of native enzyme: gel filtration gel filtration and sucrose gradient centrifugation polyacrylamide gel electrophoresis Number of subunits Type of monomers Optimal temperature Activation energy Q1o (25-35°C) Optimum pH for activity for stability Apparent K,,, for L-arginine Apparent Ki for L-ornithine Optimal MnZ+concentration (for the in v i m assay)

4.5 nm 6.2 S 4.6x10-7cm* . s - l 0.849 x 10- ' g . s 1.39

1

138 kDa 117 kDa 126 kDa 4 1 35°C 25.0 kJ . mol-' 1.56 9.0 8.5 16 mM 3 mM 1 mM

-0.04

0.00

.

0.04 1/[Arg]

0.08

Relative arginasc activity

YO 100 113 125 81 113 69 137 59 38

Table 5. Effect of different thiol compounds on the arginase activity of R. cupsulutus ElF1. Enzyme preparations were preincubated for 10 min in the presence of the indicated concentrations of thiols. Arginase activity was determined as indicated in Materials and Methods. For the final addition, arginase was previously inactivated by incubating reaction mixtures with 0.05 M p-hydroxymercuribenzoate (pHMB) for 10 min and then activity was assayed in the presence of 0.1 mM dithiothreitol. Other thiols (dithioerythritol, reduced glutathione, 2-mercaptoethanol) showed a similar effect on the reactivation of arginase activity

~

-0.08

Addition

None L-Alanine L-Lysine L-Orni thine y- Aminobutyrate y-Guanidinobutyrate L-Argininic acid Borate Ethanol (10% by vol.)

4

3

Table 4. Effect of several amino acids and other effectors on the arginase activity from R. capsulatus EIF1. Arginase activity was determined in the presence of the compounds listed in the table at a final concentration of 2 mM. Other compounds such as citrate, guanidoacetate, succinate, 1.-asparagine, L-aspartate, L-glutamate, 1.-glutamine, glycine, L-histidine and hi pro line were without effect.

0.1 2

(rnM-')

Fig. 6. Kinetic properties of arginase from R. capsulatus E l F l . Doublereciprocal plots were obtained by assaying the activity in the presence of L-arginine at the concentrations indicated in the figure, and L-ornithine at the following concentrations: ( 0 )zero; (0) 1 mM; ( 3 )5 mM; ( A ) 10 mM.

PAGE, was obtained after heating the effluent of the affinity column at 65°C for 15 min (Figs 4 and 5). Heating was made at the end of the purification procedure because arginase co-precipitated with other proteins in crude extracts heated at 50°C. Although ion-exchange chromatography showed a negative yield, the step was absolutely necessary to eliminate contaminating proteins which appeared in the final SDS/ PAGE if the DEAE-Sephacel step was omitted. The quaternary structure of the enzyme, as deduced from its hydrodynamic parameters (Table 3), corresponded to a tetramer of identical subunits, with a molecular mass of 117 kDa. A tetrameric structure of native arginase was also deduced from SDSjPAGE and gel-filtration data (Table 3 ) . The kinetic constants of the enzyme are also listed in Table 3. Arginase from

Addition

Concn

Relative arginase activity

mM 1.o 1.o 1.o 1 .o 0.1 1.o 0.001 0.05 0.1 0.05jO.l

%

None Dithioerythritol Dithiothreitol Reduced glutathione L-Cysteine 2-Mercaptoethanol 2-Mercaptoethanol pHMB pHMB pHMB pHMB/dithiothreitol

100 87 96 99 64 79 34 78 8 0 71

R. capsulatus E l F l exhibited hyperbolic saturation kinetics with an apparent K, value for L-arginine of 16 mM and was inhibited by excess of substrate and competitively by L-ornithine (Fig. 6). The enzyme was stable at 50°C and at a pH range of 8.5-9.5; it could be stored at -20°C for several months without loss of activity. Purified arginase was not affected by most amino acids but was slightly activated by L-lysine and L-argininic acid. A weak inhibition by y-guanidinobutyrate was observed, whereas the enzyme was strongly inhibited by borate and ethanol (Table 4). Neither y-guanidinobutyrate, guanidoacetate nor L-argininic acid could act as substrates in the arginase reaction. In contrast, arginase from R . capsulatus E l F l used L-canavanine as an alternative substrate with an efficiency of 1 :3 with respect to L-arginine (not shown). Sulphydryl-group reagents such as p-hydroxymercuribenzoate strongly inhibited arginase activity from R. cupsulutus E l F l (Table 5 ) . The enzyme was also inhibited in the presence of high concentrations of L-cysteine or 2-

536 mercaptoethanol, whereas other thiols were without effect (Table 5). In addition, p-hydroxymercuribenzoate-inactivated arginase was partially reactivated upon treatment with thiols (Table 5). Purified arginase from R. cupsulutus E l F l showed no antigenic similarities with eucaryotic arginases as deduced from Western blotting analysis performed with polyclonal antibodies obtained from rabbits previously immunized against rat liver arginase (not shown).

DISCUSSION Riiodobacter cupsulutus E l F l is a versatile microorganism capable of utilizing several nitrogen sources for growth, including nitrate and nitrite [39, 401, as well as L-amino acids as nitrogen and carbon sources [41]. Early steps in amino acid biosynthesis such as ammonia incorporation to carbon skeletons have been well studied in phototrophic bacteria [34, 35, 39, 42, 431 but the enzymatic steps of amino acid catabolism are poorly understood in these microorganisms. We have focused our attention on the arginase pathway due to the high growth levels observed in arginine media, which suggested the existence of an active catabolic pathway. On the other hand, there are no previous reports about the molecular properties of arginase from phototrophic bacteria which, from an evolutionary point of view, are placed far enough from other bacterial groups where the enzyme has been characterized (for reviews see [2, 31). In R.capsulutu.~EIF1, as in other bacteria [3], arginine is metabolized through the arginase pathway. The enzyme was induced in the presence of L-arginine since the activity was only detectable in cells growing with L-arginine and the appearance of the activity was due to de n o w protein synthesis (Figs 1 and 2). In baker's yeast, arginase is induced by r-arginine and repressed by ammonia whereas nitrogen starvation derepresses arginase [I 2 - 141. Bacterial arginase is also induced by L-arginine and repressed by ammonia 1161. In some bacilli, arginase is not subject to carbon catabolite repression but the enzyme is under nitrogen catabolite control. Actually, ammonia represses induction of arginase by c-proline whereas L-glutamine represses induction of the enzyme by L-proline or r-arginine [18]. In R. cupsulatus, arginase was induced by L-arginine, L-ornithine and L-homoarginine but, in contrast to results reported in other microorganisms, the enzyme was not controlled by nitrogen catabolite repression because neither nitrogen starvation nor ammonia, L-glutamine or Lasparagine seemed to affect arginase induction by L-ornithine or L-arginine (Fig. 2). The induction of the enzyme in the presence of L-ornithine and L-homoarginine could be due to structural similarities with L-arginine. In Succhuronzyces cerevisiue, L-homoarginine is a non-metabolizable inducer of arginase [44] whereas in R . cupsulatus this compound, as well as L-lysine, were used as nitrogen source under phototrophic growth conditions. Other intermediates of the arginase pathway such as L-proline, L-glutamate or putrescine neither induced nor repressed arginase in R. cupsulatus ElF1. Therefore, in this organism, regulation of arginase in vivo can be achieved at the level of enzyme synthesis by induction of arginase in the presence of L-arginine and at the level of enzyme activity by competitive inhibition by L-ornithine (Fig. 6). Arginase has been purified and well characterized in eucaryotic microorganisms. In Neurosporu crassa, the native enzyme has an hexameric structure and a molecular mass of

266 kDa as deduced from gel-filtration data, with an apparent K, of 131 mM for L-arginine [ 5 ] . Therefore, this enzyme can be described as the uricotelic type, whereas the ureotelic arginase from S. cerevisiue has a lower molecular mass (120 kDa) and K, (12.5 mM) [23]. Arginases from Bacillus unthracis [9] and B. licheniformis [8] are composed of four and six identical subunits corresponding to molecular masses of 160 kDa and 260 kDa, respectively. Our purification procedure for arginase from R. capsulatus E l F l included a step of affinity chromatography in L-arginine- agarose and a precipitation at 65 'C which yielded a single protein band of 31 kDa in SDSjPAGE, corresponding to the subunit of denatured arginase. However, when the samples for SDS electrophoresis were not heated at 100"'C, a single band of 126 kDa, corresponding to a tetrameric oligomer, was observed in SDSjPACE (Figs 4 and 5). Hydrodynamic properties of the purified enzyme and the corresponding parameters were deduced from gel-filtration and sedimentation studies and revealed a molecular mass of 117 kDa for the native protein, thus indicating a tetrameric structure. These results, taken together with kinetic data and the apparent K, value of 16 mM, strongly suggest that arginase from R. cupsulutus E l F l is of the ureotelic type. Kinetic parameters of arginases from B. anthracis [9] and B. .rubtili.r [22]are of the same order of magnitude as those found in R. crrpsulatus ElF1. The calculated frictional quotient of 1.39 for the enzyme from R. cupsulatus E l F l is the same as that previously obtained in the S . cerevisiue enzyme, which has a trimeric structure as deduced from electron microscopy as well as a low molecular mass and K,,, [7]. The discrepancy in the molecular mass of native protein between arginases from R. cupsulutus and Bacillus strains can be explained by a significant deviation from the spherical shape in the native protein molecule which can cause anomalous behaviour in gel filtration, thus yielding an over-estimation of the molecular mass, which needs to be corrected by introducing the s20,w value obtained in sedimentation studies [33]. Yeast arginase is a metallo-enzyme which shows maximal activity in the presence of divalent cations [45]. The bacterial cnzyme is strongly activated by Mn2 that in addition protects arginase from thermal denaturation and inhibitors [9]. Arginase from R. cupsulatus E l F l also required Mn 2 + and was strongly inhibited by Zn2+ and Cu2+ (Table 1). Like B . anthracis arginase [9], the enzyme from R. capsulatus E l F1 was reversibly inhibited by mercurials and, to a lesser extent, by excess of thiols such as L-cysteine and 2-mercaptoethanol (Table 5). The inhibition of arginase byp-hydroxymercuribenzoate was partially reversed by a low concentration of thiols (Table 5), thus suggesting that the enzyme requires -SH groups for activity. +

The authors wish to thank Dr. J. Cardenas and Dr. E. Fernandez for helpful advice and encouragement. The financial support of the Con?isi6n de Investigaci6n Cientifica I; T6cnica (Grant PB 89-0336; Spain) and the Alexander von Humboldt Foundation (FRG), as well as thc secretarial assistance of C. Santos and 1. Molina, are also gratefully acknowledged.

REFERENCES 1 . Davis, R. H. (1986) Microbiol. Rev. 50, 280-313. 2. Abdelal, A. T. (1979) Annu. Rev. Microhiol. 33. 139-168. 3. Cunin, R., Glansdorff, N . , Pikrard, A. & Stalon, V. (1986) Microbiol. Rev. 50, 314-352. 4. Sumrada, K. A. &Cooper, T. G. (1984) J . Bucteriol. 160. 10781087.

531 5. Borkovich, K. A. & Weiss, R. L. (1987)J. B i d . Chem. 262,7081 7086. 6. Borkovich. K. A. 8~Weiss, R. L. ( 1 ’ W J. Bacterial. 169, 55105517. 7. Duong, L. T.. Eisenstein, E., Green, S. M., Ornberg, R. L. & Hensley, P. (1986) J . Biol.Chenz. 261, 12807-12813. 8. Simon, J. P. & Stalon, V. (1976) Biochimir58. 1419-1421. 9. Soru, E. (1983) Mol. Cell. Biochem. 50, 173-183. 10. Weathers. P. S., Chcc, H. L. & Allen, M. M. (1978) Arch. Microhiol. IIN, 1 -6. 11. Dessaux. Y.. Petit, A., Tempe, J., Demarez. M., Legrain, C. & Wiame. J . M. (1986) J. Bacferiol. 166.44-50. 12. Middelhoven. W. J. (1968) Biochim. Biophjs. Acfa 156, 440443. 13. Middelhoven, W. J. (1 970) Anronie Leeuwenhoek J . Microhiol. 36, 1-19. 14. Benitez, T. & Farrar, L. (1980) J . Bacteriol. 144, 836-839. 15. Davis. R . H. & Weiss, R. L. (1988) Trends Biochem. Sci.13,101 104. 16. Bascarin, V.. Hardisson, C. & Braiia, A. F. (1989) J . Gen. Microbiol. 135. 2465 -2414. 17. Friedrich. B.. Friedrich, C. G. & Magasanik, B. (1978) J. Bucteriol. 133, 686 691. 18. Baumberg. S . & Harwood, C. R. (1979) J . Bucteriol. 137, 189196. 19. Shaibe, E., Metzer, E. & Halpern, Y. S . (1985) J . Bucteriol. 163, 933 -937. 20. Messenguy, F. & Wiame, J. M. (1969) FEBS Lett. 3, 47-49. 21. Eiscnstein, E., Duong, L. T., Omberg, R. L., Osborne, J. C., Jr. & Hensley, P. (1986) J . Bid. Chem. 261, 12814-12819. 22. Issaly. I. M. & Issaly, A. S. (1974) Eur. J . Biochem. 49,485 -495. 23. Chan, P. Y. & Cossins, E. A. (1973) Plmr Cell. Physiol. 14,641 651. 24. Weaver, P. F., Wall, J. D. & Gest, H. (1975) Arch. Microhiol. 105, 207 - 21 6. 25. Archibald, R. M. (1945) J . Biol. Chem. 157, 507-518. 26. Colombo, J.-P. & Konarska, L. (1984) in A4erhod.P of enzymatic N ~ N I ~ :3rd T ~ edn . T . (Bergmeycr, H. V., ed.) vol. IV, pp. 285-294, Verlag Chemie, Weinheim - Basel. ~

27. Conway, E. J. (1962) in Microdffusion unulj.si.y und volunierrii error. 5th edn, pp. 105-110, Crosby Lockwood, London. 28. Lowry, 0. H., Rosebrough, N. J.. Farr. A. L. & Randall. R. .I. (1951) J . Biol. Chtw,. 153, 265-275. 29. Bradford, M. M. (1976) Anul. Biocheni. 72, 248-2254, 30. Laemmli, U. K. (1970) Nuture227, 680-685. 31. Siegel, L. M. & Monty, K . J. (1966) Biochim. Biuphj,.,. A m 112, 346 - 362. 32. Pundak, S. & Eiscnberg, H . (1981) Eur. J . Biochem. 118. 463470. 33. Martin, R . G. & Ames, B. N. (1961) J . Biol. Chem. 236; 13721379. 34. Caballero, F. J., Cejudo, F. J., Florencio, F. J.: Cardenas, J. & Castillo, F. (1985) J . Bucteriol. 162, 804-809. 35. Caballero, F. J., Cardenas, J. & Castillo, F. (1989) J . Bucrerid 171, 3205-3210. 36. Andrcws. P. (1965) Biochem. J . 96, 595-606. 37. Brewer, J. M., Pesce, A. J. & Spencer, T. E. (1974) in E-uperinientul techniques in biochemistry (Brewer, J. M., Pesce, A . J. & Ashworth, R. B., eds) pp. 161 -215, Prentice-Hall, Englewood Cliffs, NJ. 38. Towin, H. & Gordon, J. (1984) J . fmmunol. Methods 72, 313340. 39. Moreno-Vivian, C., Cejudo. F. J., Cardenas, J. & Castillo, F. (1983) Arch. Microbiol. 136, 147-151. 40. Caballero, F. J.. Moreno-Vivian, C., Castillo, F. & Cardenas, J. (1986) Biochini. Biophys. Actu 848, 16-23. 41. Moreno-Vivian, C., Caballero, F. J., Cbrdenas, J. & Castillo, F. (1989) Biochim. Biophys. Actu 977. 297 - 300. 42. Johansson, B. C. & Gest, H. (1976) J . Bacteriol. 128. 683688. 43. Alef, K. & Zumft, W. G. (1981) Z . Nuturfursch. 36, 784-789. 44. Whitney, P. A. & Magasanik, B. (1973) J . Bid. Chetn. 248,6197 6202. 45. Middelhoven, W. J.: De Waard, M. A. & Mulder, E. G. (1969) Biocliim. Biophys. Actu 191. 122- 129.