Isolation and characterization of a new strain of Achromobacter sp ...

2 downloads 0 Views 1020KB Size Report
Jun 24, 2003 - Isolation and characterization of a new strain of Achromobacter sp. with b-lactam antibiotic acylase activity. Received: 25 April 2003 / Accepted: ...
Appl Microbiol Biotechnol (2003) 62:507–516 DOI 10.1007/s00253-003-1353-0

ORIGINAL PAPER

K. Plhcˇkov · S. Becˇka · F. krob · P. Kyslk

Isolation and characterization of a new strain of Achromobacter sp. with b-lactam antibiotic acylase activity Received: 25 April 2003 / Accepted: 25 April 2003 / Published online: 24 June 2003  Springer-Verlag 2003

Abstract A bacterial strain producing a b-lactam antibiotic acylase, able to hydrolyze ampicillin to 6-aminopenicillanic acid more efficiently than penicillin G, was isolated from soil and characterized. The isolate was identified as Achromobacter sp. using the phenotypic characteristics, composition of cellular fatty acids and 16S rRNA gene sequence. The enzyme synthesis was fully induced by phenylacetic acid (PAA) at a concentration of 2 g l1. PAA at concentrations up to 12 g l1 had no negative effect on the specific activity of acylase and biomass production, but slowed down the specific growth rate. Benzoic or 4-hydroxyphenylacetic acids can also induce synthesis of the enzyme. The inducers were metabolized in all cases. Acylase activity in cell-free extracts was determined with various substrates; ampicillin, cephalexin and amoxicillin were hydrolyzed 1.5- and 2-times faster than penicillin G. A high stability of acylase activity was observed over a wide range of pH (5.0–8.5) and at temperatures above 55C.

Introduction b-Lactam antibiotic acylases have been described in a broad spectrum of procaryotic and eucaryotic microorganisms. Penicillin acylases are particularly involved in biotechnological applications (Arroyo et al. 2003), mainly in the production of 6-aminopenicillanic acid (6-APA) and 7-aminodeacetoxycephalosporanic acid (7-ADCA), key precursors of semi-synthetic b-lactam antibiotics (e.g., ampicillin, amoxicillin, and cephalexin). According to their substrate specificity, penicillin acylases are classified as penicillin G acylases (PGA), penicillin V acylases (PVA) and ampicillin acylases. PGA, described mainly in bacterial strains, differ in substrate specificity K. Plhcˇkov ()) · S. Becˇka · F. krob · P. Kyslk Laboratory of Enzyme Technology, Institute of Microbiology, Academy of Sciences of the Czech Republic, Vdenˇsk 1083, 142 20 Prague 4, Czech Republic e-mail: [email protected] Fax: +420-2-41727021

but all of them hydrolyze penicillin G faster than ampicillin. PGA from Escherichia coli (Kutzbach and Rauenbush 1974), Alcaligenes faecalis (Quax 1991, Baker 1992), Kluyvera citrophila (Alvaro et al. 1992) and Proteus (Providencia) rettgeri (Robak and Szewczuk 1981) are localized in the periplasmic space. Arthrobacter viscosus (Ohashi et al. 1989) and Bacillus megaterium (Chiang and Bennet 1967; Illanes et al.1994) produce the acylase extracellularly. In addition to penicillin acylases, enzymes from Acetobacter turbidans and Xanthomonas citri capable of hydrolysis and synthesis of ampicillin and cephalexin have been described. Since only a-amino acid derivatives could act as substrates, these enzymes were named aamino acid ester hydrolases (AEH; Takahashi et al. 1974). Ampicillin acylases from Pseudomonas melanogenum with a rather narrow substrate specificity (Nara et al. 1971; Okachi et al. 1973) relative to AEH, can both hydrolyze and synthesize ampicillin, amoxicillin and cephalexin, but they have no activity towards penicillins G and V (Kawamori et al. 1983; Kim and Byun 1990). DNA sequence analysis showed that AEH are members of a new class of b-lactam antibiotic acylases (PoldermanTijmes et al. 2002). The physiological role of b-lactam antibiotic acylases has not yet been elucidated. The production of PGA is subject to a complex control mechanism. At the level of transcription, the enzyme is inducible by phenylacetic acid (PAA). Although PAA is a common source of carbon and energy for a wide variety of microorganisms, a detailed study of its bacterial catabolism has been carried out only recently. PAA and 4-hydroxyphenylacetic acid (4-OH-PAA) are aerobically catabolized in Pseudomonas putida using two different unrelated pathways. The new catabolic pathway for PAA involves a specific PAA transport system that is not induced or inhibited by 4-OHPAA (Olivera et al. 1994). Another pathway was described for the aerobic catabolism of PAA in E. coli (Ferrndez et al. 1998). Different degradative pathways of 4-OH-PAA through 3,4-dihydroxyphenylacetic acid (3,4diOH-PAA) have been described in E. coli W producing

508

PGA (Prieto et al. 1996), and in P. putida (O’Connor et al. 2001). In connection with the degradation of PAA, the term phenylacetyl-CoA catabolon was introduced as a complex functional unit integrated by several catabolic pathways, which can be coordinately regulated and which catalyze the transformation of structurally related molecules into a common catabolite. The degradation of PAA constitutes the common part (core) of this complex functional unit (Olivera et al. 1998, Luengo et al. 2001) and phenylacetyl-CoA is a true inducer of this route (Garca et al. 2000). The work reported here was aimed at the isolation of a bacterial strain producing a b-lactam antibiotic acylase with distinct properties in terms of a specificity for substrates containing a d-a-aminophenylacetyl group in the acyl moiety, and better stability to acidic pH and temperature. The paper describes the screening, isolation, identification and properties of a new soil-borne strain of Achromobacter sp. showing a b-lactam antibiotic acylase activity with preferential hydrolysis of ampicillin, amoxycillin and cephalexin. The induction of this enzyme activity by different aromatic compounds and its tolerance to PAA were studied. The relationship between the synthesis of the acylase and utilization of inducers is discussed.

aminobenzaldehyde according to Balasingham et al. (1972) and the positive strains were tested again in 100-ml batch cultures in a medium containing 30 g yeast extract per liter (cultivation time 24 h, temperature 30C, pH 7.2). Characterization and identification of the isolate CCM 4824 The physiological characteristics of the strain were obtained by conventional methods. Cellular fatty acids were analyzed using a gas chromatograph according to the Sherlock Microbial Identification System (MIDI, Newark, Del.) in Regional Hygiene Station (Ostrava, Czech Republic). Cells grown on NA medium with or without 0.2% PAA were used for the analysis. The 16S rRNA gene sequence of the strain was determined at the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany) by direct sequencing of PCRamplified 16S rDNA. The purified PCR products were sequenced using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Applera Deutschland, Darmstadt, Germany). The resulting sequence data of our strain were compared with representative 16S rRNA gene sequences of organisms belonging to the b-proteobacteria (Maidak et al 1999). The operations of the PHYLIP package (Felsenstein 1993) were used to construct a phylogenetic dendrogram. Pairwise evolutionary distances were computed from percent similarities by the correction of Jukes and Cantor (1996). Based on the evolutionary distance values, the phylogenetic tree was constructed by the neighborjoining method. 16S rRNA sequences were obtained from the EMBL database or Ribosomal Database Project (Maidak et al. 1999) for comparison. The 16S rRNA gene similarity values were calculated by pairwise comparison of the sequences within the alignment.

Materials and methods Chemicals

Media and culture conditions

Soil samples (1 g) from various ecosystems and polluted industrial areas in the Czech Republic were shaken in saline solution (10 ml) for 10 min. The solids were removed by filtration and the diluted filtrate was plated on a solid nutrient agar (NA) medium containing (g l1): nutrient broth 8, yeast extract 10, bacto casitone 5, NaCl 8 and agar 15. The pH was adjusted to 7.2. Single colony isolates with distinct morphologies were checked for the hydrolysis of ampicillin to 6-APA.

The detection of flagella was accomplished using flagella broth medium containing (g l1): nutrient broth 8, agar 3, pH 7.2. Production of the acylase by the strain CCM 4824 was studied in medium A, consisting of (g l1): yeast extract 10, KH2PO4 1, NaCl 2, and MgCl2 0.2, pH 7.2. The medium was supplemented with PAA at a concentration of between 1 and 12 g l1 to study the induction of acylase synthesis and utilization of PAA. Medium M 465, consisting of (g l1): Na2HPO4·2H2O 3.5, KH2PO4 1, (NH4)2SO4 0.5, MgCl2·6H2O 0.1, Ca(NO3)2·4H2O, trace element solution SL-4 1 ml, pH 7.25, was used to study the inducing effect of related aromatic compounds. SL-4 solution consists of (g l1) EDTA 0.5, FeSO4·7H2O 0.2, ZnSO4·7H2O 0.01, MnCl2·4H2O 0.003, H3BO3 0.03, CoCl2·6H2O 0.02, CuCl2·2H2O 0.001, NiCl2·6H2O 0.002, Na2MoO4·2H2O 0.003. This medium was also used to prove the utilization of various organic compounds as the only source of carbon and energy. Inoculum cultures of 50 ml in 100 ml flasks were shaken on an orbital incubator (210 rpm) at 28C until the stationary phase of growth; the length of the cultivation time (20–98 h) was dependent on the concentration of PAA. The inoculum was diluted 50 times with fresh medium and the production batch cultures were grown under the same conditions. The cultures were grown in medium A supplemented with PAA (4 g l1) in a 10-l bioreactor Biostat MD (working volume 7 l; Braun, Melsungen, Germany). The bioreactor was operated at 28C for 15 h, with initial stirring at 400 rpm and aeration rate of 0.6 vvm. The pH was maintained at 7.5 and the dissolved oxygen tension in the medium (PO2) was in the range of 5–10%.

Screening for acylase-producing strains

Assay of the acylase activity

Each colony was transferred into 50 ml ampicillin solution (0.5% in 0.1 M phosphate buffer, pH 7.5) in a well of a microtiter plate. The plates were incubated at 37C for 4 h. 6-APA formed by the enzyme was assayed by a chromogenic reaction with p-dimethyl-

The activity of acylase was assayed using whole cells (washed with 0.1 M phosphate buffer, pH 7.5) or a cell-free extract prepared by disintegration of a cell suspension by sonication (Microson XL 2000, Clevedon, England). For sonication, the biomass was washed

Penicillin G (sodium salt), ampicillin, 6-APA and p-dimethylaminobenzaldehyde were purchased from Fluka (Buchs, Switzerland). Penicillin V, cephalosporin C (sodium salt) and deacetoxycephalosporin V (DAOC V) were obtained from Biochemie Kundl (Austria), and 7-aminodeacetoxycephalosporanic acid (7-ADCA) and deacetoxycephalosporin G from Galena (Opava, Czech Republic). Amoxicillin, cephalexin, PAA and 6nitro-3-phenylamido-benzoic acid (NIPAB) were obtained from Sigma (St. Louis, Mo.). The complex components of solid media were purchased from Difco (Detroit, Mich.) or Oxoid (Hampshire, England). Yeast extract for liquid media was obtained from Ohly (DHW, Hamburg, Germany). Mineral salts were purchased from Lachema (Brno, Czech Republic). All reagents were of analytical or microbiological grade. Isolation of microorganisms from soil

509 with 0.1 M phosphate buffer (pH 7.5) and the suspension of cells (5–20 g dry weight l1) sonicated in the same buffer (8 W, 41min with 30 s rests) under cooling. Penicillin G or ampicillin (12.5 mM solution in 0.1 M sodium phosphate buffer, pH 7.5) was used as the enzyme substrate and the amount of 6-APA produced was assayed spectrophotometrically at 415 nm after reaction with p-dimethylaminobenzaldehyde according to Balasingham et al. (1972). One unit of enzyme activity was defined as the amount of acylase required to produce 1 mmol 6-APA min1 at 37C. The activity of acylase with other b-lactams as the reaction substrate (e.g., amoxicillin, cephalexin, penicillin V, cephalosporin C, DAOC G and DAOC V) was determined in an analogous manner using the appropriate compound for calibration (6-APA, 7-ACA and 7ADCA). Hydrolysis by the enzyme of the substrate NIPAB in 0.1 M phosphate buffer (pH 7.5) at 37C was monitored at 405 nm. The molar absorbance coefficient for 5-amino-2-nitrobenzoic acid (e405) was 9.09 mM1 cm1. One unit of enzyme activity was defined as the amount of acylase producing 1 mmol p-nitroaminobenzoic acid min1 at 37C. Determination of 6-APA, 7-ADCA and PAA by HPLC Samples were analyzed by HPLC using an LKB 2150 pump (flow rate 1 ml min1), a LiChrospher 100 RP-18 column (1254 mm), an LKB 2151 UV detector at 225 nm, and integration software (3000 Series Chromatography Data System, Nelson Analytical, PE Nelson, San Jose, Calif.). The eluent was prepared by mixing solutions of KH2PO4 (0.525 g) and Na2HPO4 (1.6 g) in 1 l of water with 0.66 l methanol (HPLC grade). The flow rate was 0.5 ml min1. Elution times for 6-APA, phenylglycine, PAA, ampicillin and penicillin G were 2.05, 2.42, 2.57, 3.72 and 6.36 min, respectively. Preparation of cell-free extract Cells from the early stationary phase of growth in a bioreactor were harvested by centrifugation and diluted in 0.025 M phosphate buffer (pH 7.5) to prepare a suspension containing 140 g cell dry wt l1. The cell suspension was subjected to three cycles of disintegration in a Manton Gaulin homogenizer (type 15 M-8 TA) at a pressure of 50–60 MPa and a temperature of 15C. The homogenate was centrifuged (14,000 g, 20 min) and the supernatant used for stability and substrate specificity assays.

0

genesis with N-methyl-N -nitro-N-nitrosoguanidine yielding a b-lactamase negative mutant. On the basis of the relevant morphological and phenotypic characteristics (Table 1) the strain was classified as Comamonas testosteroni and deposited in the Czech Collection of Microorganisms, Masaryk University, Brno under the collection number CCM 4824 (Plhcˇkov et al. 2002). Analysis of the cellular fatty acid profile showed a Similarity Index of the MIS of 0.701 for Achromobacter xylosoxidans subsp. xylosoxidans; the value for Alcaligenes faecalis was 0.449. The major cellular fatty acids assayed in the strain cultured in PAA-free medium were 16:0 (39.6%) and 17:0 cyclo (26.4%) acids. The minor fatty acids were 14:0 (3.3%); 18:1 w7c (2.85%); 12:0 2OH (2.6%); 18:0 (2.18%); the sum of 16:1 ISO I/14:0 3OH (7.05%); the sum of 16:1 w7c/15 iso 2OH (9.46%); and other minor fatty acids (less then 2.0%). Similar results were obtained from a cell culture of the strain grown in medium supplemented with PAA. The results of 16S rRNA gene sequencing (1,502 bases) are presented as a similarity matrix (Table 2) and as a phylogenetic tree (Fig. 1). As the strain CCM 4824 was originally identified as C. testosteroni, the nucleotide sequences of the 16S rRNA genes of C. testosteroni and Comamonas terrigena were included in the phylogenetic tree. It is evident that the complete 16S rRNA gene sequence of strain CCM 4824 shows the highest similarity (99.6%) to the corresponding sequence of the strain A. xylosoxidans ssp. xylosoxidans. There is a high similarity (99.4%) also with A. ruhlandii. However, since there are some discrepancies in chemotaxonomic characteristics, it might be possible that the strain represents a new species belonging to the genus Achromobacter. The strain CCM 4824 was classified as Achromobacter sp. The apolar location of flagella shown by electron microscopy was in accordance with the classification of the strain in the genus Achromobacter.

Thermal and pH stability of the acylase activity The pH in aliquots of cell-free extract was adjusted to values from 4.0 to 9.0 with 0.2 M acetic acid or 1.25 M NaOH. The aliquots were incubated at a given pH and a temperature of 50C for 60 min. The vials were then cooled down to room temperature before adjusting the pH in each vial back to a value of 7.5 and assaying for acylase activity. The change in volume of each sample during pH adjustment was taken into consideration when calculating the enzyme activity.

Results Isolation and characterization of the bacterial isolate About 2,000 bacterial culture collection strains and isolates from nature were screened for the activity of ampicillin acylase. A soil isolate able to hydrolyze ampicillin more efficiently than penicillin G was further characterized. Since the strain exhibited a b-lactamase activity, the bacterium was subjected to chemical muta-

Induction of acylase synthesis To determine whether the acylase activity was localized in the periplasmic space or in cytoplasm, the activity of acylase (measured with ampicillin) was assayed with whole cells and in a cell-free extract prepared by sonication of a cell suspension obtained from a 50 ml batch culture grown in medium A supplemented with 2 g PAA per liter. The activity in cell-free extract was nearly 3-fold higher than that found in intact cells, indicating the presence of an active, soluble enzyme in the cytoplasm. The basal activity in inducer-free medium was about 4 U/ g cell dry wt. The effect of PAA supplementation of the medium at concentrations of between 1 and 12 g l1 on the activity of acylase is shown in Fig. 2. The biomass concentration reached at the beginning of the stationary phase of growth increased with increasing concentration of the inducer. As expected, the increasing concentration of PAA reduced the specific growth rate (m) of the culture (Fig. 2a) and prolonged the lag time. Secretion of the

510 Table 1 Morphological and physiological characteristics of Achromobacter sp. strain CCM 4824 Morphology Cell shape Motility Gram-staining Colony morphology (nutrient agar) Growth conditions Temperature pH Relation to oxygen Growth in the presence of 6.5% NaCl Physiological characteristics Ox-ferm test Negative

Positive Utilization of organic compounds Positive

Cocco-rods, rods, individual cells or in irregular clusters Flagella >1 Negative Circular, smooth, glossy, slightly convex, margin entire, on average 1–4 mm Good growth at 37–42C 6.5–8.5 Strictly aerobic Negative Alkalization (glucose) Indole test, Voges-Proskauer test, methyl-red test, H2S production, starch hydrolysis, o-nitrophenyl-b-galactoside, Tween 80 hydrolysis, aeskulin hydrolysis, lecithine and DNA hydrolysis, gelatin liquefaction, nitrite reduction, fluorescein production, urease, lysine decarboxylase, ornithine decarboxylase, arginine dihydrolase, hemolysis Tyrosine hydrolysis, nitrate reduction, catalase, oxidase

Glycerol, acetate, oxalacetate, malonate, succinate, glutarate, adipate, sebacate, fumarate, maleate, malate, citrate, d-gluconate, pyruvate, benzoate, phenylacetate, 4-hydroxyphenylacetate, acetamide, phenylacetamide, l-alanine, b-alanine, l-glutamate, l-aspartate, glycerol, l-lysine, l-serine, l-tryptophan, l-b-phenylalanine, glutamine, testosterone Negative d-Glucose, fructose, maltose, l-arabinose, lactose, trehalose, d-xylose, d-ribose, mannitol, sorbitol, oxalate, tartrate, salicylate, 4-hydroxybenzoate, 3-sulfobenzoate, 2-hydroxyphenylacetate, phthalate, norleucine, l-threonine, l-arginine, l(+)-rhamnose, polyethyleneglycol Sensitivity to 6-aminopenicillanic acid 100 mg ml1

Fig. 2 a Effect of phenylacetic acid (PAA) supplement on the maximum biomass yield of Achromobacter sp. CCM 4824 (s) and specific growth rate (o) of batch cultures of growing in medium A. b Effect of PAA supplement on the specific activity of penicillin acylase (PA) in sonicated cells of batch cultures of Achromobacter sp. CCM 4824 growing in medium A. Ampicillin (l) or penicillin G () were used as the substrates for the activity assay Fig. 1 Phylogenetic tree based on 16S rRNA gene sequences of Achromobacter strain CCM 4824 and related taxa. The scale bar below represents 5 nucleotide substitutions per 100 nucleotides

23

22

21

20

19

18

17

16

15

14

13

12

11

10

9

8

7

6

5

4

3

1 2

1

CCM 4824  Achromobacter 99.0 piechaudii DSM 10342 Achromobacter 99.4 ruhlandii DSM 653 Achromobacter 99.1 xylosooxidans ssp denitrificans ATCC 5173 Achromobacter 99.6 xylosooxidans ssp xylosoxidans DSM 10346 Alcaligenes 94.5 faecalis IAM 12369 Alcaligenes 96.2 defragrans 54PIN Bordetella avium 98.0 ATCC 35086 Bordetella 97.9 pertussisATCC 9797 Bordetella 98.1 bronchiseptica ATCC 19395 Bordetella holmesii 97.4 NCBI gi-517413 Bordetella trematum 98.1 DSM 11334 Bordetella petrii 97.9 DSM 12804 Bordetella hinzii LMG98.4 13501 Pigmentiphaga 96.7 kullae DSM 13608 Brackiella oedipodis 92.8 LMG 1945 Lautropia mirabilis 91.9 NCTC 12852 Taylorella asini93.4 genitalis ATCC 700933 Taylorella equi93.3 genitalis NCTC 11184 Pelistega 94.6 europaea LMG 10982 Sutterella 89.4 wadsworthensis ATCC 51579 Comamonas 88.3 testosteroni ATCC 11996 Comamonas 89.7 terrigena DSM 7099

Strain

89.4

88.5

89.5

94.1

93.4

93.6

89.7

88.3

89.3

94.4

93.4

93.6

91.9

92.8

92.8

91.8

96.4

98.4

97.6

98.1

97.4

98.1

97.9

98.0

95.9

94.6

99.6

99.2



3

96.2

98.2

97.5

97.8

97.6

98.2

97.9

98.2

95.8

94.8

99.1

98.8

99.3



2

90.0

88.4

89.4

94.3

93.8

93.9

92.2

93.1

96.7

98.2

97.3

98.0

97.3

97.9

97.7

97.9

95.9

94.6

99.1



4

89.6

88.2

89.4

94.5

93.4

93.6

91.9

92.8

96.6

98.6

97.9

98.3

97.5

98.2

97.9

98.2

95.9

94.7



5

88.2

87.0

89.7

93.2

92.7

93.0

90.0

92.4

94.5

94.2

94.3

94.6

93.8

94.3

94.1

94.4

95.0



6

89.3

88.0

89.2

94.8

94.2

94.5

90.9

93.4

95.8

95.9

95.8

96.2

95.9

96.4

96.3

96.1



7

89.4

88.8

89.0

93.6

93.3

93.4

92.4

92.8

96.3

98.9

98.2

98.7

98.2

98.7

98.5



8

89.4

88.8

89.0

93.6

93.3

93.6

92.6

93.1

96.2

99.2

98.4

98.4

99.6

99.8



9

89.3

88.8

89.1

93.6

93.4

93.5

92.7

93.2

96.3

99.4

98.5

98.5

99.3



10

89.1

88.6

88.7

93.1

93.0

93.3

92.2

92.7

95.9

98.7

98.0

98.0



11

89.2

88.4

88.8

93.6

93.1

93.2

92.1

92.8

96.2

99.0

98.2



12

88.9

88.3

89.4

93.8

93.1

93.2

92.7

92.7

96.6

98.3



13

89.0

88.5

88.9

93.3

93.1

93.2

92.3

93.1

96.2



14

Table 2 Percent sequence similarity values of 16S rRNA gene obtained for the strain CCM 4824 and related taxa

89.1

87.4

89.6

94.2

93.8

93.9

91.9

93.0



15

87.3

86.2

88.9

92.5

92.7

93.0

89.4



16

89.8

88.8

89.5

89.7

89.7

89.6



17

88.1

86.6

88.5

95.7

98.0



18

87.8

85.9

88.2

95.6



19

89.1

87.1

88.6



20

87.4

86.4



21

95.2



22



23

511

512 Table 3 Effect of different aromatic compounds on the induction of acylase activity of Achromobacter sp. CCM 4824.RAA/P Ratio of specific relative activities (RA) determined with the substrates ampicillin and penicillin G. The medium was supplemented with Inducer

Medium A

Medium M 465 1

PAA 4-OH-PAA Phenoxyacetate b-Phenylpropionate Benzoate d,l-Phenylglycine a

the aromatic compounds at a concentration of 14.69 mM, corresponding to 0.2% of phenylacetic acid (PAA). The activity was assayed with ampicillin

Cell dry wt (g l )

U (g dry wt)

3.5 3.4 2.2 2.2 3.5 2.2

70.0 44.4 11.9 10.9 49.2 4.4

1

RAA/P

Cell dry wt (g l1)

U (g dry wt)1

RAA/P

1.46 1.06 1.05 1.09 1.11 1.14

0.7 0.7 No growth No growth 0.3 No growth

60.0 19.0 NDa ND 53.9 ND

1.43 0.97 ND ND 1.14 ND

Not determined

conditions was studied in a bioreactor. The medium was supplemented with 4 g PAA l1 to reach a higher biomass concentration. The culture was sampled at regular intervals and the biomass concentration, inducer concentration and the activity of acylase (with ampicillin as the substrate) were determined (Fig. 3). The total activity of acylase and the biomass concentration increased simultaneously, which indicated a growth-associated production of acylase. PAA was depleted after 16 h of growth and the specific growth rate equaled 0.3 h1. The maximum specific activity of acylase reached a value of 40 U/g cell dry wt. Fig. 3 PA production in a batch culture of Achromobacter sp. CCM 4824 growing in medium A supplemented with PAA (4 g l1) in a bioreactor MD: l total activity of PA, s cell dry weight, o concentration of PAA in the medium. The activity was assayed in sonicated cells with ampicillin as the substrate

enzyme to the medium was not observed. Enzyme synthesis was fully induced at a PAA concentration of 2 g l1 (Fig. 2b). A further increase in inducer concentration had no effect on the specific activity. The amount of acylase synthesized in the cell could be induced as much as 17-fold. The ratio of specific activities determined with the substrates ampicillin and penicillin G (relative activity, RAA/P) differed at concentrations of PAA of 1 and 2 g l1. Aromatic compounds structurally related to PAA were studied with respect to their inducing effect and their use as a sole source of carbon and energy. In addition to PAA, 4-OH-PAA or benzoic acids were also able to induce the synthesis of PGA (Table 3). The induction effect of phenoxyacetic acid, b-phenylpropionic acid and phenylglycine was negligible and these compounds were not utilized as the only source of carbon. Compared to PAA and 4-OH-PAA, benzoate was more toxic, and a lower concentration of biomass was reached in mineral medium M 465 in its presence. In this medium, the inducing effect of 4-OH-PAA was lower. The relative activity with ampicillin and penicillin G (RAA/P) was higher when the enzyme was induced by PAA compared to other aromatic compounds. The production of acylase by Achromobacter sp. growing in the complex medium A under controlled

Acylase activity of Achromobacter sp. with various substrates The activity of acylase in sonicated cells of Achromobacter sp. was higher when assayed with ampicillin than with penicillin G (Fig. 2b). The ratio of the two specific activities (RAA/P) in cells from a stationary culture grown in medium A supplemented with 2 g PAA per liter was about 1.5. The natural and derived b-lactam antibiotics and NIPAB, a structurally analogous chromogenic substrate, were tested as substrates to characterize the hydrolytic capabilities of the acylase (Table 4). For this purpose, a cell-free extract prepared by mechanical disintegration of cells cultured in a bioreactor was used. The relative activities (Table 4) indicated that b-lactam substrates with a hydroxy- or amino-group in the aposition of the side chain were rapidly hydrolyzed by the cell-free extract. Among these substrates, amoxicillin was hydrolyzed fastest (RA=2.15). It is evident that substrates with a phenylacetyl side chain moiety (penicillin G, DAOC G) are hydrolyzed at almost the same rate. Also, the rate of hydrolysis of substrates with phenoxyacetyl (penicillin V, DAOC V) or aminophenylacetyl (ampicillin, cephalexin) side chain moieties was similar. Regarding activity towards penicillins and their counterpart cephalosporins, there was no difference between antibiotics with penam or cephem rings.

513 Table 4 Acylase activity ofAchromobacter sp. CCM 4824 with various substrates. Cell-free extract was used for activity assays (12.5 mmol substrates per ml, 0.1 M sodium phosphate buffer, pH 7.5, 37C). The value of the relative activity (%) was calculated taking the rate of hydrolysis of penicillin G as 100%

Fig. 4 pH stability of PA of Achromobacter sp. CCM 4824. The cell-free extract was incubated at a given pH and temperature of 50C for 1 h. The activity was assayed with ampicillin (l) and penicillin G () as substrates

Fig. 5 Thermostability of PA of Achromobacter sp. CCM 4824. The cell-free extract was incubated at various temperatures and pH 7.0 for 1 h. The activity was assayed with ampicillin (l) and penicillin G () as substrates

pH 7.0 occurred above 55C (Fig. 5) and led to a complete loss of activity at 65C.

Discussion

pH and temperature stability The optimum pH for the acylase hydrolytic activity lay between 7.0 and 8.0. In order to investigate the effect of pH and temperature on the stability of the enzyme activity, the cell-free extract was incubated for 1 h at different temperatures or pH and the residual activity was estimated. The acylase from Achromobacter sp. was very stable in the form of a cell-free extract over a wide range of pH of 5.0–8.5 (50C) (Fig. 4). Thermal inactivation at

The bacteria belonging to genus Achromobacter lack unique biochemical activities that could be used for reliable classification. One of the most obvious characteristics of Achromobacter and Comamonas is their extremely limited action on carbohydrates (Krieg and Holt 1984). Analyses of 16S rRNA gene sequences and the cellular fatty acid profile revealed a high degree of similarity of the strain CCM 4824 with A. xylosoxidans ssp. xylosoxidans, with the next closest species being A. ruhlandii. On the other hand, no growth in medium with xylose as the carbon source, and acid formation in medium with xylose or glucose were observed with strain CCM 4824. However, the oxidative degradation of xylose or glucose, an important biochemical characteristic of these strains, has been described in A. xylosoxidans and A. ruhlandii (Yabuuchi et al. 1998). These differences indicate the possibility that strain CCM 4824 may represent a new species.

514

Among the producers of PGA, the bacterium Alcaligenes faecalis is the only strain that is closely related to Achromobacter sp. CCM 4824 from the taxonomical point of view. However, only relatively low similarity (94.5%) of our strain to A. faecalis was found based on the 16S rRNA gene sequence. An even lower similarity was found in the cellular fatty acid profile. Within the genus, only Achromobacter has been described to have PVA activity (Savidge and Cole 1975) and an unspecific enzyme synthesizing cephalexin (Fujii et al. 1976). The strain referred to in this study is the only bacterium so far described that is able to hydrolyze ampicillin faster than penicillin G. The ampicillin acylase from Pseudomonas melanogenum (Okachi et al. 1973; Kim and Byun 1990) did not hydrolyze penicillin G at all. The synthesis of acylase in Achromobacter sp. CCM 4824 was also induced by PAA as has been described for PGA produced by various bacteria. In E. coli and B. megaterium, concentrations of PAA higher than those providing maximum induction (1–2 g l1) suppressed both the specific activity of PGA and the specific growth rate (Illanes et al. 1994; Parmar et al. 2000). A higher tolerance to PAA was described for Arthrobacter viscosus (Ohashi at al.1989) and Proteus rettgeri (Robak and Szewczuk 1981): 5 and 7 g l1 PAA induced synthesis of PGA to the maximum level in the two bacteria, respectively. Concentrations of PAA exceeding 7 g l1 significantly suppressed both PGA activity and biomass production in these strains. Achromobacter sp. CCM 4824 exhibits the highest tolerance to PAA described so far: 12 g l1 PAA supplemented to medium A had no negative effect on the specific acylase activity or biomass production. However, there was a negative effect of the increasing concentration of PAA on the specific growth rate of the culture. The relationship between synthesis of acylase and PAA utilization, as well as the effect of related aromatic compounds during cultivation of strain CCM 4824 suggest a close relationship between enzyme induction and inducer utilization. In the case of CCM 4824, phenoxyacetic acid, b-phenylpropionic acid and phenylglycine are not utilized, and exhibit a negligible induction effect that may not even be always observed. On the other hand, 4-OH-PAA or benzoic acid can serve as sole sources of carbon and energy for growth and, simultaneously, induce the synthesis of acylase. There is little information concerning the effect of aromatic compounds on the induction of PGA synthesis in production strains. Contrary to our results, production of PGA by E. coli (isolate Ny I/3–67) has been found, induced to the same extent by phenoxyacetic acid, phenylglycine and PAA (Szentirmai 1964, 1965). An 8times lower synthesis of PGA in E. coli in the presence of phenoxyacetic acid in a PAA-supplemented medium and no induction by 4-OH-PAA or benzoic acid was reported by Levitov et al. (1967). It has been stated that 4-OH-PAA is not a true intermediate of the PAA catabolic pathway in E. coli W (Ferrndez et al. 1998). The fact that 4-OH-PAA and PAA are catabolized via two different pathways can

explain the differences in utilization of PAA and 4-OHPAA by various strains of E. coli. The PGA producer, E. coli W (ATCC 11105), is able to grow on PAA and 4OH-PAA as well as our isolate. Other strains of E. coli (B, C, K12 and NCTC 5928) could use only one or other of these compounds (Daz et al. 2001). In E. coli W, proof that two unrelated pathways are involved in the degradation of benzoic acid or b-phenylpropionic acid and PAA has also been reported (Garca et al. 1999). In contrast, our isolate was not able to use b-phenylpropionic acid. Although two separate pathways for catabolism of benzoate and 4-hydroxybenzoate have been described in P. putida (Nichols and Harwood 1995), our strain utilized only benzoic acid. These differences can be explained by the existence of different pathways involved in the aerobic catabolism of these compounds in various bacterial strains. Induction of the synthesis of acylase in Achromobacter sp. CCM 4824 by 4-OH-PAA and benzoic acid suggests the existence of common regulators or regulatory mechanisms involved in the synthesis of the enzyme and in aerobic catabolism of aromatic compounds. In addition to the mechanism of induction of the biosynthesis of the enzyme, PAA has also been reported to influence membrane transport and intracellular proteolytic activity, resulting in a higher yield of PGA (Ignatova et al. 2000). In P. putida U, a PAA transport system (PATS) induced by PAA has been described (Olivera et al.1994). However, similar compounds with an odd number of carbon atoms (benzoic acid, 3-phenylpropionic acid) do not induce the system, which suggests that the true inducer molecule of PATS is phenylacetylcoenzyme A (Schleissner et al. 1994). A 4-OH-PAA transport system—performing efficiently at very low external concentrations of 4-OH-PAA—that is also induced by PAA was demonstrated in E. coli W (Prieto and Garca 1997). In Achromobacter sp. CCM 4824, no difference in the composition of fatty acids was found in cells grown with or without PAA. Therefore, this strain does not seem to use a mechanism of tolerance to PAA similar to that described for 4-hydroxybenzoic, i.e., an increased rigidity of the cell membrane resulting from a change in the profile of unsaturated fatty acids (RamosGonzles et al. 2001). The activity of acylase in the cell-free enzyme extract of Achromobacter sp. CCM 4824, indicated a dependence on the acyl moiety of the substrate. Preliminary experiments indicated a large difference between the acylase activity of Achromobacter sp. CCM 4824 and other acylases. In our case, ampicillin and amoxicillin are hydrolyzed 1.5- and 2-times faster than penicillin G, respectively. To date, the highest reported value of the RAA/P ratio (i.e., the ratio of ampicillin/penicillin G activities) is 0.83 for PGA of Alcaligenes faecalis (Baker 1992). Ampicillin acylase from Pseudomonas melanogenum has a very poor activity towards amoxicillin (Kim and Byun 1990) and the AEH from Acetobacter turbidans has the highest activity towards cefalexin (Takahashi et al. 1974; Ryu and Ryu 1988; Fernndez-Lafuente et al.

515

2001; Polderman-Tijmes et al. 2002). In contrast to the acylase activity of Achromobacter sp. CCM 4824, both AEH have a reduced activity towards penicillin G and substrates with hydroxylated aromatic ring of the acyl moiety. In addition, the different RAA/P ratio for PGA activity of Achromobacter sp. CCM 4824 after induction by PAA, similar to that by benzoic acid or 4-OH-PAA, suggests the presence of several enzymes of different specificity with different regulation of their syntheses. Also, the differences in RAA/P ratio observed under the conditions of a partial or complete induction at various stages of cultivation of CCM 4824 support this assumption. The novel penicillin acylase recently purified from strain CCM 4824 (krob at al. 2003) hydrolyzed ampicillin, amoxicillin and cephalexin at the same rate, which was twice as fast as that with penicillin G. Further study of this strain will help explain those differences. The hypothesis that penicillin acylases play a role as scavengers for many different natural esters and amides that, after their hydrolysis, can be transported and mineralized by central catabolic pathways (Valle et al. 1991; Roa et al. 1995; Luengo et al. 2001), is supported, among other evidence, by the localization of PGAencoding gene pac in proximity to the gene cluster involved in utilization of 4-OH-PAA (Prieto et al. 1996). These relationships have been studied mostly in E. coli, which has become a model system for the research of biochemical, genetic, evolutionary and ecological aspects of catabolism of aromatic compounds (Daz et al. 2001). From this point of view we may suppose that a physiological study of our strain Achromobacter sp., with its high tolerance to PAA, and the study of expression of the pag gene may increase our understanding of coordinately regulated catabolic pathways and of how bacteria become highly specialized for degradation of a family of structurally related compounds. The high thermostability and pH stability of acylase activity determined in the cell-free extracts from Achromobacter sp. CCM 4824 are important characteristics for viability of the enzyme in conditions of long-term use under reaction conditions of hydrolysis or synthesis. Stability of the enzyme at acidic pH could be advantageous for the syntheses of b-lactam antibiotic derivatives. These properties, together with the different profile of activities against b-lactam substrates, make the new strain Achromobacter sp. CCM 4824 promising also from the industrial point of view. Acknowledgement This work was supported by Institutional Research Concept No. AV0Z5020903.

References Alvaro G, Fernandez-Lafuente R, Rosell CM, Blanco RM, GarcaLopez JL, Guisn JM (1992) Penicillin G acylase from Kluyvera citrophila; new choice as industrial enzyme. Biotechnol Lett 14:285–290

Arroyo M, de la Mata I, Acebal C, Castill n MP (2003) Biotechnological applications of penicillin acylases: state-ofthe-art. Appl Microbiol Biotechnol 60:507–514 Baker WL (1992) Co-existence of b-lactamase and penicillin acylase in bacteria; detection and quantitative determination of enzyme activities. J Appl Bacteriol 73:14–22 Balasingham K, Warburton D, Dunnill P, Lilly MD (1972) The isolation and kinetics of penicillin amidase from Escherichia coli. Biochim Biophys Acta 276:250–256 Chiang C, Bennet RE (1967) Purification and properties of penicillin amidase from Bacillus megaterium. J Bacteriol 93:302–306 Daz E, Ferrndez A, Prieto MA, Garca JL (2001) Biodegradation of aromatic compounds by Escherichia coli. Microbiol Mol Biol Rev 65:523–569 Felsenstein J (1993) PHYLIP (Phylogeny Inference Package), version 3.5.1. Department of Genetics, University of Washington, Seattle, Wash. Fernndez-Lafuente R, Hernnd z-Jfflstiz O, Mateo C, Terreni M, Fernndez-Lorente G, Moreno MA, Alonso J, Garca-L pez JL, Guisan JM (2001) Biotransformations catalyzed by multimeric enzymes: stabilization of tetrameric ampicillin acylase permits the optimization of ampicillin synthesis under dissociation conditions. Biomacromolecules 2:95–104 Ferrndez A, Minˇanbres B, Garca B, Olivera ER, Luengo JM, Garca JL, Diaz E (1998) Catabolism of phenylacetic acid in Escherichia coli. Characterization of new aerobic hybrid pathway. J Biol Chem 273:25974–25986 Fujii T, Matsumoto K, Watanabe T (1976) Enzymatic synthesis of cephalexin. Process Biochem 11:21–24 Garca B, Olivera ER, Minˇanbres B, Fernndez-Valverde M, Canˇedo, Prieto MSA, Garca JL, Martnez M, Luengo JM (1999) Novel biodegradative aromatic plastics from a bacterial source. Genetic and biochemical studies on a route of the phenylacetyl-CoA catabolon. J Biol Chem 274:29228–29241 Garca B, Olivera ER, Minˇambres B, Carnicero D, Muniz C, Naharro G, Luengo JM (2000) Phenylacetyl-coenzyme A is the true inducer of the phenylacetic acid catabolon pathway in Pseudomonas putida U. Appl Environ Microbiol 66:4575–4578 Ignatova Z, Taruttis S, Kasche V (2000) Role of the intracellular proteolysis in the production of the periplasmic penicillin amidase in Escherichia coli. Biotechnol Lett 22:1727–1732 Illanes A, Acevedo F, Gentina JC, Reyes I, Torres R, Cartagena O, Ruiz A, Vsquez M (1994) Production of penicillin acylase from Bacillus megaterium in complex and defined media. Process Biochem 29:263–270 Jukes TH, Cantor CR (1969) Evolution of protein molecules. In: Munro HN (ed) Mammalian protein metabolism. Academic press, New York, pp 21–132 Kawamori M, Hashimoto Y, Katsumata R, Okachi R, Takayama K (1983) Enzymatic production of amoxycillin by b-lactamasedeficient mutant of Pseudomonas melanogenum KY 3987. Agric Biol Chem 47:2503–2509 Kim DJ, Byun SM (1990) Purification and properties of ampicillin acylase from Pseudomonas melanogenum. Biochim Biophys Acta 1040:12–18 Krieg NR, Holt JG (eds) (1984) Bergey’s manual of systematic bacteriology, vol 1. Williams and Wilkins, Baltimore Kutzbach K, Rauenbusch E (1974) Preparation and general properties of crystalline penicillin acylase from E. coli ATCC 11105. Hoppe-Seyler’s Z Physiol Chem 355:45–53 Levitov MM, Klapovskaja KI, Kleiner GI (1967) Induced acylase in Escherichia coli. Mikrobiologiya 36:912–917 Luengo JM, Garca JL, Olivera ER (2001) The phenylacetyl-CoA catabolon: a complex catabolic unit with broad biotechnological applications. Mol Microbiol 39:1434–1442 Maidak BL, Cole JR, Parker CT Jr, Garrity GM, Larsen N, Li B, Lilburn TG, McCaughey MJ, Olsen GJ, Overbeek R, Pramanik S, Schmidt TM, Tiedje JM, Woese CR (1999) A new version of the RDP (Ribosomal Database Project). Nucleic Acids Res 27:171–173

516 Nara T, Misawa M, Okachi R, Yamamoto M (1971) Enzymatic synthesis of d()-a-aminobenzylpenicillin. Part I. Selection of penicillin acylase-producing bacteria. Agric Biol Chem 35:1676–1682 Nichols NN, Harwood CS (1995) Repression of 4-hydroxybenzoate transport and degradation by benzoate: a new layer of regulatory control in the Pseudomonas putida b-ketoadipate pathway. J Bacteriol 177:7033–7040 O’Connor KE, Witholt B, Duetz W (2001) p-Hydroxyphenylacetic acid metabolism in Pseudomonas putida F6. J Bacteriol 183:928–933 Ohashi H, Katsuda Z, Nagashima M, Kamei T, Yano M (1989) Expression of the Arthrobacter viscosus penicillin G acylase gene in Escherichia coli and Bacillus subtilis. Appl Environ Microbiol 55:1351–1356 Okachi R, Kato F, Miyamura Y, Nara T (1973) Selection of Pseudomonas melanogenum KY 3987 as a new ampicillinproducing bacteria. Agric Biol Chem 37:1953–1957 Olivera ER, Reglero A, Martnez-Blanco H, Fernndez-Medarde A, Moreno MA, Luengo JM (1994) Catabolisms of aromatics in Pseudomonas putida U. Formal demonstration that phenylacetic acid and 4-hydroxyphenylacetic acid are catabolized by two unrelated pathways. Eur J Biochem 221:375–381 Olivera ER, Minˇambres B, Garca B, Muniz C, Moreno MA, Ferrndez A, Daz E, Garca JL, Luengo JM (1998) Molecular characterization of the phenylacetic acid catabolic pathway in Pseudomonas putida U: the phenylacetyl-CoA catabolon. Proc Natl Acad Sci USA 95:6419–6424 Parmar A, Kumar H, Marwaha SS, Kennedy JF (2000) Advances in enzymatic transformation of penicillins to 6-aminopenicillanic acid (6-APA). Biotechnol Adv 18:289–301 Plhcˇkov K, Kyslk P, Becˇka S, Sobotkov Ir (2002) The strain of microorganism Comamonas testosteroni CCM 4824. Czech patent No. 291 154 Polderman-Tijmes JJ, Jekel PA, de Vries EJ, van Merode AEJ, Floris R, van der Laan JM, Sonke T, Janssen DB (2002) Cloning, sequence analysis, and expression in Escherichia coli of the gene encoding an a-amino acid ester hydrolase from Acetobacter turbidans. Appl Environ Microbiol 68:211–218 Prieto MA, Garca JL (1997) Identification of the 4-hydroxyphenylacetate transport gene of Escherichia coli W: construction of a highly sensitive cellular biosensor. FEBS Lett 414:293–297 Prieto MA, Daz E, Garca JL (1996) Molecular characterization of the 4-hydroxyphenylacetate catabolic pathway of Escherichia coli W: engineering a mobile aromatic degradative cluster. J Bacteriol 178:111–120

Quax WJ (1991) Penicillin G acylase, a gene encoding the same and a method for production of this enzyme. European Patent Application EP 0 453 047 A1 Ramos-Gonzlez MI, Godoy P, Alaminos M, Ben-Bassat A, Ramos JL (2001) Physiological characterization of Pseudomonas putida DOT-T1E tolerance to p-hydroxybenzoate. Appl Environ Microbiol 67:4338–4341 Roa A, Castill n MP, Goble ML, Virden R, Garca JL (1995) New insights on the specificity of penicillin acylase. Biochem Biophys Res Commun 206:629–636 Robak M, Szewczuk A (1981) Penicillin amidase from Proteus rettgeri. Acta Biochim Pol 28:275–284 Ryu YW, Ryu DDY (1988) Semisynthetic b-lactam antibiotic synthetizing enzyme from Acetobacter turbidans: catalytic properties. Enzyme Microb Technol 10:239–245 Savidge TA, Cole M (1975) Penicillin acylase (bacterial). In: Mosbach K (ed) Methods in enzymology, vol 43. Academic Press, New York, pp 705–721 Schleissner C, Olivera ER, Fernndez-Valverde M, Luengo JM (1994) Aerobic catabolism of phenylacetic acid in Pseudomonas putida U: biochemical characterization of a specific phenylacetic acid transport system and formal demonstration that phenylacetyl-coenzyme A is a catabolic intermediate. J Bacteriol 176:7667–7676 krob F, Becˇka S, Plhcˇkov K, Fotopulosov V, Kyslk P (2003) Novel penicillin G acylase from Achromobacter sp. CCM 4824. Enzyme Microb Technol 32:738–744 Szentirmai A (1964) Production of penicillin acylase. Appl Microbiol 12:185–187 Szentirmai A (1965) Properties of penicillin acylase isolated from Escherichia coli. Acta Microbiol Acad Sci Hung 12:395–405 Takahashi T, Yamazaki Y, Kato K (1974) Substrate specificity of an a-amino ester acid hydrolase produced by Acetobacter turbidans A.T.C.C 9325. Biochem J 137:497–503 Valle F, Balbs P, Merino E, Bolivar F (1991) The role of penicillin amidases in nature and industry. Trends Biochem 16:36–40 Yabuuchi E, Kawamura Y, Kosako Y, Ezaki T (1989) Emendation of genus Achromobacter and Achromobacter xylosoxidans (Yabuuchi and Yano) and proposal of Achromobacter ruhlandii (Packer and Vishniac) Comb. Nov., Achromobacter piechaudii (Kiredjian et al.) Comb. Nov., and Achromobacter xylosoxidans subsp. denitrificans (R ger and Tan) Comb. Nov. Microbiol Immunol 42:429–438