Improvement of Catalytic Properties of Escherichia coli Penicillin G Acylase Immobilized on Glyoxyl Agarose by Addition of a Six-Amino-Acid Tag

A tag of three lysines alternating with three glycines was addedto the C-terminal end of the ß chain of penicillinG acylase (PGA). This modification improved the immobilizationefficiency of PGA on glyoxyl agarose and the catalytic propertiesof the PGA derivative, although it impaired the posttranslationalsteps of overexpressed protein maturation.

INTRODUCTION

Top

Introduction

Penicillin G acylase (PGA) is used as a catalyst for the enzymaticsemisyntheses of several penicillins and cephalosporins (11,15-17, 21). An immobilized derivative of PGA is usually employedto ensure suitable enzyme stability and consequently enzymerecovery and reuse (7, 10). Depending on the immobilizationsupport and procedure, different catalytic properties can beobtained in the synthesis of ß-lactam antibioticsby kinetically controlled N acylation (20). Immobilization proceduresusually involve the free amino groups of the enzyme. Immobilizationon supports activated with aldehyde groups (glyoxyl supports)proceeds via interaction with the region of the protein surfacethat is richest in lysines or that bears the most reactive ones.Lysine residues are present near the active site of PGA (13).If interactions between the enzyme and the solid support occurthrough those lysine residues, PGA-catalytic properties maybe affected by the consequently limited accessibility of theactive site.

Here we propose a new strategy to facilitate access of the substrateto the active site of PGA by addition of a tag consisting ofthree lysines alternating with three glycines at the end ofthe ß chain. This strategy is aimed at favoring immobilizationon glyoxyl agarose at the tag side, far from the active site.A similar strategy was successful in the case of a subtilisinderivative in which the enzyme activity actually improved afterthe introduction of a cysteine on the enzyme surface, far fromthe active site (14). The rationale for adding lysine residuesto PGA is that increasing the lysine content on a lysine-richregion of PGA surface has been shown to increase the numberof bonds between protein and glyoxyl agarose and to induce apreferential orientation towards the support surface (1).

Cloning and PCR mutagenesis.

Top

Cloning and Pcr mutagenesis.

The Escherichia coli pac gene was amplified by PCR using 5'CACCAATGAAAAATAGAAATCGTATGA 3' as the forward primer and 5'AAAGTAACCAGCCCTCCAACAT 3' as the reverse primer and cloned intothe pET101/D-TOPO expression vector to obtain the pETPAC construct.The mutant 3G3Kpac was obtained by PCR, with the same forwardprimer and the reverse primer 5' TTACTACTTACCCTTGCCCTTACCTCTCTGAACGTGCAACACTT3'. The mutated XbaI-HindIII 3G3Kpac fragment was subclonedinto the pET28a vector (Novagen) to obtain the pET28-3G3KPACconstruct.

Expression and purification of PGA and 3G3KPGA.

Top

Expression and Purification of...

Wild-type PGA overexpression was performed by growing a 4-literaerated Luria-Bertani BL21(DE3) culture (0.5-bar air pressure,150-rpm stirring) at 28°C. The temperature was lowered to22°C after induction at an optical density at 600 nm of0.6 with 0.5 mM IPTG (isopropyl-ß-D-thiogalactopyranoside).Expressing cells were collected after 1 day from induction.The periplasmic extract (3) was purified on a Q-Sepharose column(Pharmacia), with a 0-to-75 mM NaCl gradient in 50 mM Tris HCl(pH 8.5) (average yield was 18 mg of purified protein per 30g of wet cells). pET28-3G3KPAC was expressed in BL21 Star(DE3)cells as described previously, except that the temperature forthe induction of expression was lowered to 20°C, due tothe increased sensitivity and lower yield of processing andmaturation (2, 4, 8, 13, 18, 19) of the modified enzyme (datanot shown). The periplasmic extract was precipitated with 75%(NH₄)₂SO₄ and the pellet, resuspended in 1 M (NH₄)₂SO₄ and 50mM potassium phosphate (pH 7), was purified on a phenyl Sepharosecolumn with a gradient of 0 to 30% ethylene glycol in 50 mMpotassium phosphate (pH 7). The average yield was 26 mg of purifiedprotein per 30 g of wet cells.

Enzymatic activity.

Top

Enzymatic activity.

Hydrolytic activity of the soluble acylases was determined spectrophotometricallyby using 1.5 mM 6-nitro-3-phenylacetamido benzoic acid (NIPAB)in 50 mM phosphate buffer, pH 8.5, at 25°C (9), as wellas by 50 mM NaOH automatic titration at pH 8 using 80 mM penicillinG potassium salt (PGK) at 37°C and 10 mM R-mandelic acidmethyl ester (MAME) at 25°C, respectively. All the enzymesshowed similar activity profiles. When PGK and NIPAB were usedas substrates, the specific activity of wild-type PGA was higherthan the activities of both the commercial (Recordati) and thetagged enzyme. When MAME was used as the substrate, the activitiesof the three enzymes were comparable. Interestingly, 3G3KPGAwas slightly more active than the commercial enzyme with everysubstrate (Table 1).

View this table:
[in this window]
[in a new window]
TABLE 1. Specific activities of the soluble enzymes for different substrates^a

Enzyme immobilization.

Top

Enzyme immobilization.

Wild-type and modified PGA were immobilized on aldehyde agarose(glyoxyl agarose) by loading 0.7 to 1 mg of protein per gramof support and subsequently compared with a commercial acylasefrom E. coli immobilized by the same procedure (5, 12). 3G3KPGAwas completely immobilized under standard conditions (pH 10and 25°C), but the resulting derivative turned out to beinactive due to low stability (Fig. 1). In order to overcomethis problem, the pH was lowered from 10 to 9.5 and the temperaturefrom 25°C to 4°C; consequently, about 50% of the initialactivity was maintained after 2 h (Fig. 1). The immobilizationefficiencies of the enzymes were compared under these conditionsby measuring the amount of protein remaining in solution atdifferent times (Fig. 2). Commercial and wild-type PGA showedthe same behavior: 30% of the total protein was immobilizedafter 3 h (Fig. 2 and Table 2), and no changes in specific activitywere observed, as reported in Table 2. Remarkably, in the caseof 3G3KPGA, the immobilization proceeded much faster, and after3 h, about 50% of the total protein was immobilized (Fig. 2).As expected considering the low stability of the modified enzyme,about half of the activity corresponding to the amount of 3G3KPGAloaded was maintained after immobilization; the specific activityof 3G3KPGA decreased from 18 to 8.5 U/mg of protein (Tables1 and 2).

View larger version (15K):
[in this window]
[in a new window]
FIG. 1. Stability of the enzymes at different pHs and temperatures. Symbols: x, wild type at pH 10 and 25°C; ${blacksquare}$ , 3G3KPGA at pH 10 and 25°C; •, 3G3KPGA at pH 10 and 4°C; ${blacktriangleup}$ , 3G3KPGA at pH 9.5 and 4°C. All tests were done in the presence of 100 mM phenyl acetic acid.

View larger version (14K):
[in this window]
[in a new window]
FIG. 2. Immobilization of wild-type PGA (•), commercial PGA ( ${blacksquare}$ ), and 3G3KPGA ( ${blacktriangleup}$ ) on agarose activated with aldehyde groups at pH 9.5, 4°C, in the presence of 100 mM phenyl acetic acid. The residual protein remaining in solution at different times was measured by the Bradford assay (6). Five independent immobilization tests gave similar results and produced the five batches of immobilized derivatives used in the assays (Table 2 and Fig. 3).

View this table:
[in this window]
[in a new window]
TABLE 2. Immobilization of acylases on agarose beads bearing aldehyde groups on the surface, at pH 9.5 and 4°C^a

Determination of S/H ratio in the acylation reaction of ß-lactam nuclei.

Top

Determination of s/h ratio...

The enzyme derivatives were tested as catalysts in the kineticallycontrolled N acylation of 7-aminocephalosporanic acid (7-ACA)with MAME (substrates concentration, 5 mM), by evaluating theratio of initial rates of synthesis (S) and ester hydrolysis(H) (20). S/H is strictly related to the affinity of the enzymefor the ß-lactam nucleus and S/H is strictly relatedto the affinity of the enzyme for the ß-lactam nucleusand depends on the ability of the catalyst to adsorb the substrateinto the active center. In the case of soluble proteins, thetagged enzyme showed a higher S/H ratio ( $~$ 3) than both the wild-typeand the commercial acylases (S/H ratio, $~$ 2). This result, togetherwith the lower stability of the tagged protein than the nativeprotein, suggests that the tag induces a structural change ofPGA that involves the active site. After immobilization, theS/H ratio of the tagged enzyme was almost unchanged ( $~$ 3.5), whilethose of the wild-type and the commercial PGA decreased (S/Hratio, $~$ 1) (Fig. 3). These results suggest that the active siteof the immobilized tagged enzyme may be more accessible to theß-lactam substrate. Most likely, the orientation of3G3KPGA towards the support surface is initially influencedby the high reactivity of the tag, even though further interactionscan happen with other free amino groups on the enzyme surface.We then followed the reaction between 7-ACA and MAME (both 5mM) by high-pressure liquid chromatography, until the maximumconversion was achieved (usually in 2 to 3 h); we observed thatthe immobilized 3G3KPGA yielded a higher conversion of 7-ACAinto the acylation products than the wild-type immobilized PGA(17% of maximum conversion versus 12%). The smaller differenceof the conversion values observed with the two immobilized enzymesat the end of the reaction, compared to the differences of theS/H ratios calculated at the beginning of the reaction, maybe attributed to the hydrolysis of the acylation products thatbecomes more and more significant as the concentration of theacylations product increases during the reaction course.

View larger version (15K):
[in this window]
[in a new window]
FIG. 3. S/H ratio in the N-acylation reaction of 7-ACA. The S/H ratio in the PGA-catalyzed N acylation of 7-ACA was evaluated by measuring the initial rates of synthesis (S) and MAME hydrolysis (H). The S/H ratios of wild-type, commercial, and tagged PGA were evaluated using the soluble enzymes and the glyoxyl agarose derivatives. 7-ACA and MAME concentrations were both 5 mM (pH 6.5, 4°C).

In summary, we have demonstrated that the introduction of atag of alternating glycines and lysines to the C terminus ofthe ß chain of PGA improves the efficiency of immobilizationon glyoxyl agarose. Although the stability of the soluble taggedenzyme decreases, the flexibility and the high number of lysinesin the tag may ensure efficient immobilization under conditionsthat are unfavorable for the wild-type enzyme. The derivativeconserves the excellent catalytic properties of the solubleenzyme in the kinetically controlled N acylation of 7-ACA. Theseresults suggest that the presence of the tag influences theenzyme orientation during immobilization, allowing a betteraccessibility of the substrate to the active site. However,even if the S/H ratio of the tagged enzyme can be declared superiorto the wild-type one, the low stability of 3G3KPGA limits itsindustrial application. Different immobilization proceduresare being evaluated to improve the stability of 3G3KPGA. Further-modifiedenzymes obtained by the introduction of additional lysines indifferent areas of the PGA surface are in development for thepurpose of industrial applicability.

ACKNOWLEDGMENTS

This work was financially supported by the University of Pavia(FAR 2000 and 2001 to A.M.A and Progetto di Ateneo 2002 and2003 to M.T.).

We thank Andrea Mattevi for helpful suggestions, Auro Taglianifor stimulating discussions and for providing the clavulanicacid, and Elisabetta Andreoli for expert technical assistance.Furthermore, we thank Recordati S.p.a. (Opera) for providingthe commercial PGA.

FOOTNOTES

* Corresponding author. Mailing address: Dipartimento di Genetica e Microbiologia, Università degli Studi di Pavia, Via Ferrata 1, 27100 Pavia, Italy. Phone: (39) 0382-985549. Fax: (39) 0382-528496. E-mail: albert{at}ipvgen.unipv.it.

Present address: Department of Molecular, Cellular and DevelopmentalBiology, University of California, Santa Barbara, Calif.

REFERENCES

Top