Cloning, over-expression, purification, and characterisation of N-acetylneuraminate synthase from Streptococcus agalactiae

Protein Expression and Purification 27 (2003) 346–356 www.elsevier.com/locate/yprep

Cloning, over-expression, purification, and characterisation of N-acetylneuraminate synthase from Streptococcus agalactiae Venty Suryanti,a,1 Adam Nelson,b and Alan Berrya,* a

School of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT, UK b Department of Chemistry, University of Leeds, Leeds LS2 9JT, UK Received 5 August 2002, and in revised form 23 September 2002

Abstract N-acetylneuraminate synthase (NeuAc-synthase; E.C. 4.1.3.19) is one of the two enzymes responsible for sialic acid (N-acetylneuraminic acid) synthesis in bacteria. Potential genes encoding NeuAc synthase in Streptococcus agalactiae and Bacillus subtilis were identified from a BLAST search of the EMBL/GenBank/DDBJ database using the E. coli neuB gene sequence as a probe and the genes cloned and expressed at high level in Escherichia coli. The neuB gene of S. agalactiae was shown to encode an active NeuAc synthase, whereas the spsE gene product from B. subtilis did not have this activity. Expression of the native S. agalactiae neuB gene product enzyme in E. coli resulted in a product that was prone to proteolysis during purification so the protein was tagged with a hexa-histidine tag at its N-terminus and the enzyme was rapidly purified to homogeneity by ammonium sulphate fractionation and Ni-chelating affinity chromatography in two steps. Measurement of the subunit molecular mass by electrospray ionisation mass spectrometry (Mr ¼ 38, 987  3) and of the native molecular mass by gel filtration chromatography (Mr ¼ 78,000) clearly demonstrated that the enzyme is dimeric. The effects of EDTA, temperature, and pH on the activity of the S. agalactiae NeuAc synthase were examined. Enzyme activity was maximal at pH 7 and was dependent on the presence of metal ions such as Mg2þ , Mn2þ or Co2þ . The purified enzyme was inhibited by the reagent phenylglyoxal and the substrates N-acetyl mannosamine or phosphoenol pyruvate afforded protection against this inhibition, suggesting that one or more arginine residues are involved in substrate recognition and binding. The ease of expression and the properties of the enzyme should now permit a thorough study of the specificity of the enzyme and provide the prerequisites for attempts to alter this specificity by directed evolution for the production of novel sialic acid analogues. Ó 2002 Elsevier Science (USA). All rights reserved.

Sialic acid (N-acetylneuraminic acid, NeuAc) is an essential component of complex oligosaccharides which perform pivotal roles as recognition signals in a variety of biological processes. These events are particularly well exemplified in host–pathogen and host–parasite interactions (for a review of the roles of sialic acid see [1]), where the oligosaccharide is often required for invasion, infectivity, and survival of the invading organism in the host [2]. For example, Escherichia coli K1 and Neisseria meningitidis groups B and C, which *

Corresponding author. Fax: +44-113-233-3167. E-mail address: [email protected] (A. Berry). 1 Present address: Department of Chemistry, Faculty of Mathematics and Natural Sciences, University of Selebas Maret Surakarta, Surakarta 57126, Indonesia.

cause urinary tract infections and neonatal meningitis, respectively, produce colominic acid [3], a capsular homopolymer containing sialic acid residues a(2–8)- or a(2–9)-ketosidically linked that have been identified as pathogenic determinants. Sialic acid analogues thus represent attractive targets for novel chemotherapeutic agents against bacterial and viral infections [4], as well as other disease states where cell–cell recognition is compromised, e.g., cancer metastases. However, sialic acid analogues are very difficult to chemically synthesise de novo and although a number of syntheses have been reported using the enzyme NeuAc lyase, its use has been restricted partly because of its specificity and partly because large amounts of pyruvate are required to drive the equilibrium forward towards sialic acid synthesis.

1046-5928/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 1 0 4 6 - 5 9 2 8 ( 0 2 ) 0 0 6 3 3 - 2

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

In bacteria, two enzymes are involved in the synthesis of N-acetyl neuraminic acid. N-acetylneuraminate lyase (NeuAc lyase or NeuAc aldolase, E.C. 4.1.3.3) condenses N-acetyl-D -mannosamine (ManNAc) and pyruvate to form N-acetylneuraminate in a reversible reaction. This enzyme has been purified from many sources and its three-dimensional structure has been solved by X-ray crystallography [5]. The other enzyme N-acetylneuraminate synthase (NeuAc synthase, E.C. 4.1.3.19) is much less well studied. It condenses ManNAc and phosphoenolpyruvate (PEP) to form sialic acid in an irreversible reaction. The enzyme has been identified in E. coli [6], N. meningitidis [7], and Campylobacter jejuni [8] and the neuB gene from E. coli, encoding the enzyme [9,10], has been cloned and expressed [9]. These enzymes represent attractive targets for the protein engineer to alter their properties to open the route to the facile enzymic synthesis of a range of sialic acid analogues. In this paper, we report that the Streptococcus agalactiae neuB gene product, which has 56% identity and 72% similarity to the E. coli neuB gene, has NeuAc synthase activity, whereas the B. subtilis spsE gene product, which has 37% identity and 51% similarity with the E. coli K1 neuB gene, does not encode NeuAc synthase activity. The NeuAc synthase from S. agalactiae was purified to homogeneity as an N-terminal His6 tagged protein using Chelating Sepharose Fast Flow gels immobilised with Ni2þ and characterisation of this enzyme was also carried out.

Materials and methods Materials Agarose was obtained from Life Technologies (Paisley, UK) and acrylamide/bisacrylamide mixture was from Severn Biotech (Kidderminster, UK). All other chemicals were from Sigma Chemicals (Poole, UK). The restriction enzyme, BamHI, calf intestinal alkaline phosphatase (CIAP), T4 DNA ligase, DNA markers, dNTPs, and Pfu DNA polymerase were from Promega (Madison, WI, USA). The restriction enzyme NdeI was from New England Biolabs (Hertfordshire, England, UK). DNA purifications were carried out using the Wizard Plus DNA Purification System Minipreps (and Maxipreps) from Promega (Madison, WI, USA). Oligonucleotides were from Genosys (Genosys Biotechnologies, Pampisford, Cambridgeshire, UK). Diethylaminoethyl-cellulose (DE-52) resin, cellulose nitrate membrane filters 0.2 lm and 90 mm filter papers were from Whatman International (Maidstone, England, UK). Chelating Sepharose Fast Flow was from Amersham Pharmacia Biotech AB (Uppsala, Sweden) and the HiLoad 16/20 Superdex 200

347

column was supplied by Pharmacia Biotech (Milton Keynes, UK). Sepharose CL-4B and chelating resin (iminodiacetic acid) were from Sigma Chemicals (Poole, UK). Bacterial strains and plasmid Escherichia coli KM3 [D(his gnd) Dlac araD fda ptsF ptsM rpsL pro NAR /F0 proAþ Bþ lacIq lacZ DM15] was described previously [11] and E. coli BL21(DE3) [E. coli B F dcm ompT hsdS(rB m B ) gal (DE3)] was from Stratagene Europe (Amsterdam, The Netherlands). Bacillus subtilis (NCIMB 3610) and S. agalactiae (NCIMB 8778) were from The National Collection of Industrial and Marine Bacteria (NCIMB) (Aberdeen, Scotland, UK). The expression plasmid pET23a was from Novagen (Madison, WI, USA). Enzyme assay NeuAc synthase activity was assayed using a modification of the method described previously [12]. Samples of enzyme were incubated with 10 mM ManNAc, 8.75 mM PEP in a final volume of 50 ll of 125 mM Tris buffer, pH 7.0, at 37 °C for 15 min. In assays to determine the substrate specificity of the enzyme, N-acetylmannosamine was replaced with various sugars at a concentration of 10 mM and the incubation time was extended to 30 min. After incubation, the reaction was stopped by adding 137 ll periodic acid solution (2.5 mg/ ml sodium periodate in 57 mM H2 SO4 ) and the tubes were shaken and incubated at 37 °C for an additional 15 min. After this time, 50 ll sodium arsenite solution (25 mg/ml sodium arsenite in 0.5 M HCl) was added and the tubes were shaken until the yellow–brown colour disappears. One-hundred ll of 2-thiobarbituric acid solution (71 mg/ml of 2-thiobarbituric acid adjusted to pH 9.0 with NaOH) was added and the tubes were then shaken. The tubes were heated in a vigorously boiling water bath for 7.5 min. The pink-red chromophore generated was extracted by vigorous shaking in 1 ml cyclohexanone and the tubes were then centrifuged at 5000g for 5 min. The absorbance of the organic phase was determined at 549 nm with a Jasco V-560 spectrometer. The assay was quantified by the construction of a standard curve of N-acetylneuraminate with every determination of enzyme activity. One unit NeuAc synthase is defined as the amount of enzyme that synthesises 1 nmol N-acetylneuraminate in 30 min at 37 °C. The effect of pH on the enzyme activity was investigated by measuring the activity at a range of pHs as follows: The purified enzyme (50 lg), 10 mM ManNAc, and 8.75 mM PEP were incubated in various buffers [125 mM sodium acetate (pH 4.0–6.0), 125 mM Tris– HCl (pH 6.0–10.0), and 125 mM glycine–NaOH (pH 10.0–12.0)] for 30 min at 37 °C and the amount of sialic

348

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

acid formed was determined using the thiobarbituric acid assay. The thermal stability of the S. agalactiae NeuAc synthase was examined by incubating the enzyme (50 lg) for 30 min at 0, 4, 25, 37, 50, 60 or 80 °C, after which 10 mM ManNAc and 8.75 mM PEP in 125 mM Tris– HCl buffer, pH 7.0, were added and incubated at 37 °C for 30 min. The amount of sialic acid formed was assayed using the thiobarbituric acid assay. Cloning Primers used for the PCR were designed based upon the known nucleotide sequence of the relevant genes [10,13–15]. The primers were 50 -AGT CAG TCC ATA TGA GTA ATA TAT ATA TCG-30 (E. coli forward) and 50 -AAC AGG ATC CTT ATT ATT CCC CCT C-30 (E. coli reverse), 50 -AGT CAG TCC ATA TGG TTT ATA TTA TTG C-30 (S. agalactiae forward) and 50 -TTG TGG ATC CTT ATT ATA ACT CGG GC-30 (S. agalactiae reverse), and 50 -GGC GAA CAT ATG GCA GCG TTT CAG ATC-30 (B. subtilis forward) and 50 -GGC GAA CAT ATG GCA GCG TTT CAG ATC-30 (B. subtilis reverse). Forward primers contained an NdeI site and reverse primers a BamHI site (underlined). The primer used for gene cloning and concomitant insertion of a hexa-histidine tag at the Nterminus of the protein was 50 -GCG GGC CAT ATG GAA CAC CAT CAC CAT CAC CAT GTT TAT ATT ATT CAG AG 30 ) (NdeI site is underlined). Total genomic DNA preparations from E. coli, S. agalactiae, and B. subtilis were prepared for use as the template in the PCR as described by Kirby et al. [16]. PCRs were carried out in a 50 ll reaction mixture consisting of 5 lg template DNA with 100 pmol each forward and reverse primers, dNTPs (0.05 mM of each), 3 U Pfu polymerase, and 5 ll of 10 Pfu DNA polymerase buffer (10 mM Tris–HCl, pH 9.0, 0.1% (w/v) Triton X-100, 50 mM KCl, 7 mM MgSO4 , and 0.3 mM MnCl2 ). Amplification involved an initial denaturation step at 95 °C for 5 min, followed by 35 cycles of 95 °C for 1 min, 55 °C for 1 min, and 72 °C for 2.5 min, and then a final extension at 72 °C for 2 min. The PCR products were purified from 0.8% agarose gels and then digested with NdeI and BamHI. The genes were ligated into NdeI/BamHI treated pET23a and transformed into the non-expression host, E. coli KM3. The plasmids containing inserts of the correct size (judged by agarose gel electrophoresis) were designated pETneuBEc, pETneuBSa, and pETspsEBs and these plasmids were transformed into the E. coli expression strain, BL21(DE3). The E. coli BL21(DE3) harbouring the correct plasmids were grown overnight at 37 °C in 10 ml of 2xTY media containing ampicillin at 50 lg/ml and the whole cell samples were run on a 15% SDS– PAGE to check for protein expression.

Assay of crude cell lysates To investigate the activities of the cloned gene product from E. coli, S. agalactiae, and B. subtilis; the E. coli BL21(DE3)/pETneuBEc, E. coli BL21(DE3)/pETneuBSa, and E. coli BL21(DE3)/pETspsEBs strains were grown overnight at 37 °C in 10 ml of 2xTY media containing 50 lg/ml ampicillin. The cultures were then centrifuged at 13,000g for 5 min. The cells were lysed by freeze thawing after first adding 500 ll of 10 mg/ml lysozyme in 100 mM Tris–HCl, pH 7.0. The crude lysates were then assayed by the thiobarbituric acid assay using equal amounts of enzyme in each assay (75 lg). Purification of enzymes The E. coli BL21(DE3)/pETneuBSa cells were grown overnight in 2xTY media (5L) supplemented with 50 lg/ ml ampicillin and harvested by centrifugation in a Heraeus Contifuge 17RS continual action centrifuge at 15 000g. The cell pellet was re-suspended in 50 ml of 125 mM Tris–HCl buffer, pH 7.0, and the cells were disrupted in a French Press at 4 °C. The extract was clarified by centrifugation at 11,4000g for 1 h and then fractionated with ammonium sulphate (0–25%, 25–65%, and >65% ammonium sulphate saturation). The fractions were dissolved in minimal volumes of 125 mM Tris–HCl buffer, pH 7.0, and dialysed against the same buffer. The 25–65% ammonium sulphate fraction was shown to contain the bulk of the NeuAc synthase activity and applied to a DEAE-cellulose (DE-52) column (13 cm  2:6 cm diameter), equilibrated with 100 mM Tris–HCl, pH 7.0. After loading the enzyme sample, the column was washed with the same buffer until the absorbance at 280 nm of eluate produced a baseline at zero and then washed with 1 column volume of 0.1 M NaCl in the same buffer. The enzyme was then eluted with 1 column volume of 0.2 M NaCl in 100 mM Tris–HCl, pH 7.0. The peak fractions containing the enzyme activity were pooled and the enzyme was concentrated by the addition of (NH4 )2 SO4 to 80% saturation, collection of the precipitate, dissolution in a small volume of 125 mM Tris–HCl buffer, pH 7.0, and dialysis against the same buffer. The enzyme was then applied onto a Blue Sepharose column, equilibrated with the same buffer. The column was washed with 10 mM Tris–HCl, pH 6.0, and the enzyme was then eluted with 0.1 M NaCl in 10 mM Tris–HCl, pH 6.0. The peak fractions containing the enzyme activity were pooled and the enzyme was concentrated as above. The His6 -tagged protein was purified from E. coli BL21(DE3)/pETneuBSa-NHis grown overnight at 37 °C in 2xTY media supplemented with 50 lg/ml ampicillin. These cells were harvested and lysed as above. The supernatant was clarified by centrifugation at 11,400g for 1 h and the 25–65% saturated ammonium sulphate

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

fraction was collected and dialysed extensively against 125 mM Tris–HCl buffer, pH 7.0. Imidazole was added to a final concentration of 5 mM to the dialysed ammonium sulphate fractionation and the protein was then applied onto an immobilised metal ion affinity chromatography (IMAC) column charged with Ni2þ prepared from Chelating Sepharose Fast Flow gel (Phamacia) (column 10 cm  2:6 cm diameter previously equilibrated with the same buffer) charged with Ni2þ according to manufacturerÕs instructions. The column was then washed with 100 ml of 125 mM Tris–HCl, pH 7.0, containing 5 mM imidazole, followed by 100 ml of each of 25, 50, 100, and 200 mM imidazole in the same buffer. The flow rate was 3.0 ml/min and 10 ml fractions were collected. The elution profile contained 3 peaks, each of which was analysed on a 15% SDS–PAGE. NHis6 -tagged NeuAc synthase eluted in 50 mM imidazole in 125 mM Tris–HCl, pH 7.0, and the peak fractions containing the enzyme activity were pooled, concentrated, and re-dissolved as above. Effect of EDTA or metal ions A reaction mixture of enzyme (50 lg), 10 mM ManNAc, 8.75 mM PEP, and EDTA (1 mM, 10 mM or 100 mM) in 125 mM Tris–HCl buffer, pH 7.0, was incubated at 37 °C for 30 min. The amount of sialic acid formed was then measured using the thiobarbituric acid assay. To investigate the effect of metal ions on the activity, similar assays were carried out in the absence of EDTA, but with 1 mM metal ion present. Apo-NeuAc synthase was prepared by dialysis of 1 mg/ml enzyme against 100 mM Tris–HCl buffer, pH 7.0, containing 100 mM EDTA. This sample was then freed of EDTA by gel filtration on a Superdex 200 column previously equilibrated with 100 mM Tris–HCl buffer, pH 7.0, which was kept free of metal ions by the use of an in-line pre-column containing chelating resin (iminodiacetic acid immobilised on 1% cross-linked polystyrene) before the sample injector. Reactivation of the apoenzyme with metals was carried out by incubating apoenzyme (15 lg), with 10 mM ManNAc, 8.75 mM PEP, and 1 mM metal ion in a total volume of 50 ll at 37 °C for 2 h, followed by assay using the thiobarbituric acid assay. Phenylglyoxal modification The enzyme (3 mg/ml) was dialysed into 120 mM sodium bicarbonate buffer, pH 8.0, and incubated at room temperature in the presence or absence of 2–10 mM phenylglyoxal in the same buffer. During the course of the reaction, suitable aliquots (25 ll) were removed and assayed using the thiobarbituric acid assay. Individual curves were fitted with an exponential equation. The enzyme was also incubated with phenylglyoxal in the

349

presence of ManNAc (10 mM) or PEP (10 mM). Suitable aliquots (25 ll) were again removed and assayed.

Results and discussion Directed evolution of enzyme activities or properties offers the opportunity to tailor the activity of an enzyme to a particular function and opens the way to the design of enzyme catalysts for the facile synthesis of complex biomolecules [17]. In this respect, NeuAc synthase presents an ideal target for alteration to accept novel substrates for the synthesis of analogues of sialic acid. We have chosen to use the DNA shuffling approach to directed evolution [18,19] to produce new enzyme properties or activities however, the random nature of directed evolution necessitates the screening of large areas of sequence space and this process is aided by shuffling a number of related genes—a process termed ‘‘family shuffling’’ [20]. In E. coli K1, the neuB gene is known to encode the NeuAc synthase and its sequence, and hence, the primary sequence of the enzyme are known [10]. To identify NeuAc synthase genes in other organisms, we carried out a BLAST search [21] of the EMBL/GenBank/DDBJ database using the E. coli NeuB protein sequence and the program located at http://dove.embl.heidelberg.de/Blast2. The top 15 scoring sequences (excluding the E. coli neuB gene itself) are listed in Table 1. Some of these genes have previously been assigned general roles in polysaccharide biosynthesis and some have been attributed as the N-acetylneuraminate synthase gene on the basis of similarity to E. coli neuB alone. We chose two of these genes, namely the neuB gene of S. agalatiae [13,14] and the spsE gene of B. subtilis [15], for further study. The protein products of these genes show 56% sequence identity and 72% similarity, and 37% identity and 51% similarity, respectively, to the E. coli NeuB protein, suggesting that they may encode N-acetylneuraminate synthase activity. Both genes were amplified by PCR from genomic preparations of DNA from the relevant organisms, the DNA was fully sequenced to ensure that no spurious mutations had been introduced during cloning and the genes were inserted into the expression vector pET23a. The E. coli neuB gene cloned previously by Vann et al. [9] was also amplified by PCR and cloned into pET23a as a positive control. The resultant plasmids, designated pETneuBEc, pETneuBSa, and pETspsEBs for the E. coli, S. agalactiae, and B. subtilis genes, respectively, were transformed into the expression strain of E. coli, BL21(DE3). Expression of protein was confirmed by SDS–PAGE. Fig. 1 shows that a protein of 39 kDa was expressed from the E. coli neuB gene, whereas expression from the S. agalactiae neuB gene yielded a product of 38 kDa and the spsE gene of B. subtilis, a 41 kDa protein. In each case, these results were in

350

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

Table 1 Results of a BLAST search of the GenBank/EMBL database using the E. coli neuB gene as the search sequence Identifier

Gene

Protein

Organism

% Identity

% Similarity

AB017355

neuB

Streptococcus agalactiae

56

72

CAB73754 Q9RDX5 Q58465

neuB2 neuB MJ1065

Campylobacter jejuni Legionella pneumophila Methanococcus jannaschii

51 43 46

70 62 61

BAA91818

Unnamed

Homo sapiens

36

54

Q57265 CAB73744 Q9R9S2 AAF31772 P39625

synC siaC neuB3 neuB Unnamed spsE

Neisseria meningitidis Campylobacter jejuni Aeromonas caviae Campylobacter jejuni Bacillus subtilis

36 35 34 36 37

52 55 57 52 51

Q9R9M1 CAB73396 Q9ZMQ2 Q9VG74 Q83573

rkp Q neuB1 neuB CG5232 spsE

NeuB capsular polysaccharide synthesis locus N-acetylneuraminate synthase N-acetylneuraminate condensing enzyme Spore coat polysaccharide biosynthesis protein E Weakly similar to spore coat polysaccharide biosynthesis protein SpsE Polysialic acid capsule biosynthesis protein N-acetylneuraminate synthase NeuB protein Sialic acid synthase Spore coat polysaccharide biosynthesis protein SpsE Putative polysaccharide biosynthesis protein N-acetylneuraminate synthetase Sialic acid synthase CG5232 protein Spore coat polysaccharide biosynthesis protein

Rhizobium meliloti Campylobacter jejuni Helicobacter pylori Drosophila melanogaster Treponema pallidum

34 34 30 36 28

55 51 53 53 46

The top 15 scoring sequences found are listed.

Fig. 1. Over-expression of the E. coli neuB, S. agalactiae neuB, and B. subtilis spsE gene products. The genes were amplified by PCR, ligated into pET23a, and transformed into E. coli BL21(DE3). Samples of the cell pellet were subjected to SDS–PAGE on 15% gels and stained with Coomassie brilliant blue R-250. The E. coli neuB gene expresses a product, of 39 kDa, the S. agalactiae neuB gene a 38 kDa product and the B. subtilis spsE gene gave a 41 kDa product. The markers used consisted of proteins of sizes 66, 45, 36, 29, 24, 20, and 14.2 kDa.

agreement with the predicted size of protein from the gene sequence. The presence of NeuAc synthase activity in the expressing bacteria was next checked by assaying crude cell extracts of E. coli BL21(DE3) harbouring the expression plasmids. Fig. 2 shows that crude cell extracts of E. coli BL21(DE3) expressing the neuB genes from E. coli or S. agalactiae generated significant levels of Nacetylneuraminate from N-acetyl mannosamine and PEP as judged by the thiobarbituric acid assay. These extracts also had slightly higher levels of activity with Nacetyl mannosamine and pyruvate as substrates. This may be due to the presence of either N-acetyl neuraminate lyase or pyruvate kinase in the crude cell extracts.

In contrast, the addition of cell extract from E. coli BL21(DE3) expressing the cloned spsE gene from B. subtilis to mixtures of N-acetyl mannosamine and either pyruvate or PEP showed no NeuAc synthase enzyme activity. Previous studies [7,22,23] have suggested that NeuAc synthase activity is dependent on various metal ions (see below). We therefore reassayed the crude extract expressing the B. subtilis SpsE protein in the presence of 1 mM metal ions (Mn2þ , Mg2þ or Ca2þ ) however, no activity was detected under any condition (data not shown). These results clearly demonstrate that the S. agalactiae neuB gene encodes a functional NeuAc synthase, whereas the B. subtilis spsE gene does not encode this activity and its role

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

351

Fig. 2. Thiobarbituric acid assay of cloned gene products from E. coli, S. agalactiae, and B. subtilis. Crude cell lysates of E. coli expressing the E. coli neuB gene, the S. agalactiae neuB gene or the B. subtilis spsE gene were assayed using the thiobarbituric acid assay. Purified E. coli NeuAc lyase was also assayed as a control. The reaction mixture, containing enzyme (75 lg), ManNAc (10 mM), and PEP (8.75 mM) in 125 mM Tris–HCl buffer, pH 7.0 was incubated at 37 °C for 30 min and then assayed by thiobarbituric acid assay. Experiments were carried out in triplicate and the error bars are the standard deviation of the three measurements.

in spore coat polysaccharide biosynthesis remains unclear. We decided at this point to concentrate our efforts on the purification and characterisation of the S. agalactiae NeuAc synthase. Purification of the S. agalactiae NeuAc synthase Attempts were made to purify the wild-type NeuAc synthase by a combination of ammonium sulphate fractionation, ion-exchange chromatography on DE52, and Blue Sepharose affinity chromatography (Fig. 3). However, attempts were hampered by the appearance during purification of bands of increasing intensity at 32.5 and 32 kDa. N-terminal sequencing of the bands isolated from an SDS–PAGE gel (Fig. 3) revealed that the N-terminal methionine of the full length NeuAc synthase was removed during expression to leave valine as residue number 1. Sequencing also demonstrated that the smaller fragments were proteolytic cleavage products of the NeuAc synthase between residues Lys-50 and Ala-51, and Gln-54 and Lys-55. Despite the inclusion of a cocktail of protease inhibitors at all stages of the protein purification procedure, some degradation of the wild-type NeuAc synthase was always found (data not shown). We therefore decided to adopt a rapid purification procedure based on Histagging the protein and one-step purification by Ni-NTA chromatography to reduce the time that the protein would be exposed to proteases.

Cloning, expression, and purification of His6 tagged S. agalactiae NeuAc synthase Construction of an expression system for a His-tagged version of the enzyme was carried out by inserting DNA encoding a His6 tag at the 50 end of the S. agalactiae neuB gene using PCR in a cloning process similar to that used to clone the wild-type S. agalactiae NeuAc

Fig. 3. SDS–PAGE analysis of the various stages of S. agalactiae NeuAc synthase purification. A 15% SDS–PAGE gel showing the progress of S. agalactiae NeuAc synthase purification. Lane 1, crude lysate; lane 2, 25–65% ammonium sulphate fraction; lane 3, DE52 column eluate and lane 4, Blue Sepharose column eluate. The markers (lane M) used consisted of proteins of 66, 45, 36, 29, 24, 20, and 14.2 kDa.

352

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

showed the subunit molecular mass to be 39 kDa (Fig. 4), demonstrating that the native enzyme is a dimer of identical subunits similar to the E. coli [6] and N. meningitidis [7] enzymes. N-terminal sequencing of the purified protein showed that the first five residues were NH2 -MEHHH, demonstrating that the methionine and glutamic acid encoded alongside the hexa-histidine tag were retained in the expressed protein. The molecular mass of the enzyme subunit was accurately measured as 38,989:91  3 Da by ESI-MS. The expected molecular mass for the enzyme subunit calculated from the sequence deposited in the EMBL/GenBank database by Yamamoto et al. [13] (Accession No. AB017355) modified with a hexa-histidine tag and associated Met and Glu is 39,043.70 Da. However, another sequence for the S. agalactiae neuB gene in the database (Accession No. AF355776) [24] lacks the C-terminal two residues found in the AB017355 sequence (Glu–Leu) and has an alanine at position 57 rather than glutamic acid as found in the AB017355 sequence. The AF355776 NeuB would thus be expected to have a molecular mass of 38,743.38 Da when the hexa-histidine tag was attached. To resolve the discrepancy between the observed and expected molecular masses, we sequenced the expressed neuB gene and carried out C-terminal protein sequencing of the expressed protein. These experiments showed that position 57 in the expressed sequence was an alanine and that the C-terminal two residues of the protein were Glu–Leu– COOH—the longer form of the protein. The expected molecular mass for this protein would be 38,985.66 Da, in excellent agreement with the observed value of 38,989.91 Da. The N-acetylneuraminate synthase from S. agalactiae thus appears to be naturally able to tolerate changes at position 57 and at the C-terminus, and the protein cloned in these studies is a hybrid of the two previously reported enzymes, in that it contains alanine at position 57 and the extra two residues (–Glu–Leu– CO2 H) at the C-terminus.

Fig. 4. SDS–PAGE analysis of the various stages of N-His6 S. agalactiae NeuAc synthase purification. A 15% SDS–PAGE gel showing the progress of N-His6 S. agalactiae NeuAc synthase purification; (1) crude lysate, (2) 25–65% ammonium sulphate fractionation, and (3) Ni2þ -NTA column eluate. The markers (M) used consisted of proteins of 66, 45, 36, 29, 24, 20, and 14.2 kDa.

synthase gene. After ligation of the PCR product into pET23a, the resulting plasmid was designated pETneuBSa-NHis and transformed into E. coli BL21(DE3). The E. coli BL21(DE3)/pETneuBSa-NHis was grown overnight at 37 °C in 10 ml of 2xTY media containing ampicillin at 50 g/ml to check over-expression and investigate its activity. SDS–PAGE analysis showed the over-expression of a protein of the correct size for His6 tagged S. agalactiae NeuAc synthase (39 kDa) (data not shown) and the thiobarbituric acid assay suggested that the N-His6 tagged enzyme was as active as the wildtype. The enzyme was then rapidly purified to homogeneity using ammonium sulphate fractionation and immobilised metal affinity chromatography on Ni-NTA resin. The protein at various stages through the preparation is shown in Fig. 4 and the purification is summarised in Table 2. To achieve the highest purity using only one chromatography stage only the peak fractions were pooled. The final protein had been purified 5-fold over the crude over-expressed protein in 17% yield. Characterisation of the S. agalactiae NeuAc synthase The native molecular mass of the S. agalactiae NeuAc synthase was determined by means of gel filtration on a Superdex 200 column equilibrated in 50 mM Tris–HCl buffer, pH 8.0, with a flow rate of 1.0 ml/min. The molecular mass of the enzyme was estimated to be 78 kDa (Fig. 5). SDS–PAGE analysis of the enzyme

Catalytic properties The activity of the purified enzyme over a range of pH values was measured under the standard assay conditions. The enzyme was active between pH 5 and

Table 2 The purification table of the N-His6 S. agalactiae NeuAc synthase from E. coli BL21(DE3)/pETneuBSa-NHis Purification step

Total protein (mg)

Total activity (U)

Specific activity (U/mg)

Recovery (%)

Purification factor

Crude lysates 25–65% Saturated ammonium sulphate fraction Ni2þ column eluate

5400 1440

30 20

0.005 0.014

100 67

1 2.8

200

5

0.025

17

The enzyme was prepared as described in Materials and methods. The enzyme was assayed using the thiobarbituric acid assay.

5

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

353

Fig. 5. Determination of the native molecular mass of the N-His6 S. agalactiae NeuAc synthase. The native molecular mass of the N-His6 S. agalactiae NeuAc synthase was estimated by gel filtration on Superdex 200. The standards were rabbit muscle Class I fructose-1,6-bisphosphate aldolase (160 kDa), E. coli Class II fructose-1,6-bisphosphate aldolase (78 kDa), bovine serum albumin (66 kDa), triose phosphate isomerase (53 kDa), and myoglobin (13 kDa). The elution time for the S. agalactiae NeuAc synthase was 79 min and the molecular mass was estimated to be 78 kDa.

10, with maximum activity at pH 7.0 (data not shown). The purified enzyme was thermostable up to 37 °C. Around 80–90% of the enzyme activity was lost after incubation at 50 °C for 30 min and the enzyme was completely inactive after incubation at 60 °C or 80 °C for 30 min. Effect of EDTA and metal ions It was previously reported that the NeuAc synthase from E. coli was dependent on Ca2þ ions [22], and the N. meningitidis enzyme was activated by Mn2þ and inhibited by EDTA, although Komaki et al. [6] found that the E. coli enzyme was not inhibited by EDTA nor stimulated by Ca2þ or Mn2þ ions. We therefore investigated the effect of EDTA and metal ions on the NeuAc synthase from S. agalactiae. No significant sialic acid synthesis from ManNAc and PEP by the enzyme could be detected when EDTA was present at a concentration of either 1 or 10 mM, suggesting an important role for divalent metal ions in catalysis. Furthermore, assay of the purified enzyme in the presence of 1 mM Mn2þ or Co2þ resulted in stimulation of the enzyme activity by around 2-fold. Mg2þ , Ca2þ , and Ba2þ had no effect on the activity whereas Cu2þ , Zn2þ , and Fe2þ all inhibited the enzyme activity (Table 3). Monovalent metal ions (Liþ , Naþ or Kþ ) all showed a slight (20–25%) inhibition of the enzyme activity. We further investigated the ability of metal ions to re-activate the enzyme after EDTA treatment. Apoenzyme, prepared by dialysis of the purified enzyme against EDTA followed by gel filtration to remove EDTA, was treated for 2 h at 37 °C in the presence of

Table 3 The effect of metal ions on the S. agalactiae NeuAc synthase Addition

Relative activity (%) Holoenzyme

None EDTA 1 mM MgSO4 1 mM CaCl2 1 mM Ba acetate 1 mM MnCl2 1 mM CoCl2 1 mM CuCl2 1 mM ZnCl2 1 mM FeCl2 1 mM LiCl 1 mM NaCl 1 mM KCl

100 3 99 98 97 180 225 10 20 60 70 80 82

Apoenzyme 10 —

140 5 8 400 600 10 2 20 5 10 8

EDTA (10 mM) or various metal ions were added to the purified holo-NeuAc synthase and its activity was measured using the thiobarbituric acid assay. The apoenzyme was prepared by dialysis of the holoenzyme against 10 mM EDTA and the latter was removed by gel filtration (see Materials and methods). The activity of the apoenzyme and of apoenzyme reconstituted with various metal ions (1 mM) was measured.

1 mM metal ion and the activity of the enzyme was measured (Table 3). These experiments demonstrated that Mn2þ and Co2þ were able to restore activity to the apoenzyme to levels 4–6 times that of the starting enzyme. In contrast to the treatment of the native holoenzyme, Mg2þ was also able to restore activity to the apoenzyme (to approximately the level of the starting enzyme). Taken together, these results suggest that the S. agalactiae NeuAc synthase is dependent on metal ions such as Mg2þ , Mn2þ or Co2þ in a manner similar to the

354

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

N. meningitidis enzyme [22] but the nature of the metal in the native protein and its exact role remain to be determined. Substrate specificity of the S. agalactiae NeuAc synthase The substrate specificity of S. agalactiae NeuAc synthase was examined using PEP and the various sugars listed in Table 4. Of these, only N-acetylman-

nosamine (ManNAc), N-acetylgalactosamine (GalNAc), and N-acetylglucosamine (GlcNAc) gave positive responses to the thiobarbituric acid assay. The S. agalactiae NeuAc synthase showed the highest activity with ManNAc (100%), but the enzyme could use GalNAc and GlcNAc albeit at lower rates, 15% and 5%, respectively (Table 4). These results suggest that an Nacetyl group at C-2 of the six-carbon sugar is important for catalytic activity with the S. agalactiae enzyme and

Table 4 Substrate specificity of the S. agalactiae NeuAc synthase Substrate

Relative activity (%)

N-acetyl-D -mannosamine

100

D -mannose

0

N-acetyl-D -glucosamine

5

D -glucosamine

0

D -glucose

0

N-acetyl-D -galactosamine

15

L -arabinose

0

D -xylose

0

L -xylose

0

L -erthyrose

0

D L -glyceraldehyde

0

D -glyceraldehyde

0

The purified enzyme (50 lg), 10 mM sugar, and 8.75 mM PEP in 125 mM Tris–HCl buffer, pH 7.0, was incubated at 37 °C for 30 min and the reaction product was assayed using the thiobarbituric acid assay. Results are described as an activity relative to the activity with 10 mM ManNAc as substrate.

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

355

Fig. 6. Treatment of NeuAc synthase with phenylglyoxal. Panel A. The S. agalactiae NeuAc synthase (3 mg/ml) was dialysed into 120 mM sodium bicarbonate buffer, pH 8.0, and treated with phenylglyoxal, samples were removed at various times and assayed using the thiobarbituric acid assay. Results are reported as the percentage of activity at time zero. Curves illustrate single exponential fits to the data. (d), Enzyme incubated in the absence of phenylglyoxal, or in the presence of 5 mM (s), 7.5 mM (.) or 10 mM (r) phenylglyoxal. Panel B. Samples of NeuAc synthase treated as above in the absence of phenylglyoxal (d), or in the presence of 10 mM phenylglyoxal (s), 10 mM phenylglyoxal, and 10 mM ManNAc (.), or 10 mM phenylglyoxal and 10 mM PEP (r).

contrasts with the E. coli enzyme where N-acetylglucosamine was found not to be a substrate and where the enzyme is tolerant of changes in stereochemistry at C4 or C2 [9]. Residues important in catalysis To explore the possibility that arginine residues were involved in substrate recognition and binding by the NeuAc synthase, the enzyme was incubated with the arginine-specific reagent phenylglyoxal. This reagent was shown to rapidly inhibit the enzyme, a solution of the synthase (3 mg/ml) losing approximately 95% of its activity within 60 min when incubated with 10 mM phenylglyoxal (Fig. 6A). The experiments were repeated in the presence of the individual substrates, N-acetylmannosamine or PEP. The rate of inhibition of the synthase by the arginine-directed reagent was decreased in the presence of either substrate (Fig. 6B), suggesting that arginine residues are present in the active site and are involved in substrate recognition and binding. Sequence alignment of the primary amino acid sequences of the known NeuAc synthases revealed that only one arginine residue, Arg 305, of the eight arginines in the S. agalactiae NeuAc synthase, is totally conserved in all four sequences. Future experiments will be required to determine if this residue is responsible for substrate binding, and to alter the enzymeÕs specificity to accept other, new, substrates for the synthesis of novel sialic acid mimetics.

Acknowledgments We thank Dr. Alison Ashcroft for the ESI-MS data and Dr. Jeff Keen for the C-terminal protein sequencing data. Thanks are also due to the DUE Project Universitas Sebelas Maret Surakarta, Indonesia, that sponsored VS as a research student. This work was supported by The Wellcome Trust and the BBSRC and forms a contribution from The Astbury Centre for Structural Molecular Biology at The University of Leeds and the BBSRC supported North of England Structural Biology Centre (NESBiC).

References [1] A. Rosenberg (Ed.), Biology of Sialic acids, Plenum Press, New York/London, 1995. [2] E.J. McGuire, in: A. Rosenberg, C.L. Schengrund (Eds.), The Biology of Sialic Acids, Plenum Press, New York, 1976. [3] G.T. Barry, V. Abbot, T. Tsai, Relationship of colomnic acid (poly acetyl neuraminic acid) to bacteria which contain neuraminic acid, J. Gen. Microbiol. 29 (1962) 335–352. [4] M. Koketsu, T. Nitoda, H. Sugino, L.R. Juneja, M. Kim, T. Yamamoto, N. Abe, T. Kajimoto, C.H. Wong, Synthesis of a novel sialic acid derivative (sialylphospholipid) as an antirotaviral agent, J. Med. Chem. 40 (1997) 3332–3335. [5] T. Izard, M.C. Lawrence, R.L. Malby, G.G. Lilley, P.M. Colman, The three-dimensional structure of N-acetylneuraminate lyase from E. coli, Structure 2 (1994) 361–369. [6] E. Komaki, Y. Ohta, Y. Tsukada, Purification and characterisation of N-acetylneuraminate synthase from Escherichia coli K1– M12, Biosci. Biotech. Biochem 61 (1997) 2046–2050.

356

V. Suryanti et al. / Protein Expression and Purification 27 (2003) 346–356

[7] R.S. Blacklow, L. Warren, Biosynthesis of sialic acids by Neisseria meningitidis, J. Biol. Chem. 237 (1962) 3520–3526. [8] D. Linton, A.V. Karlyshev, P.G. Hitchen, H.R. Morris, A. Dell, N.A. Gregson, B.W. Wren, Multiple N-acetylneuraminic acid synthetase (neuB) genes in Campylobacter jejuni: identification and characterisation of the gene involved in sialylation of lipooligosaccharide, Molec. Micro. 35 (2000) 1120–1134. [9] W.F. Vann, J.J. Tavarez, J. Crowley, E. Vimr, P.R. Silver, Purification and characterisation of the Escherichia coli K1 neuB gene product N-acetylneuraminic acid synthetase, Glycobiology 7 (1997) 697–701. [10] P.W. Annunziato, L.F. Wright, W.F. Van, R.P. Silver, Nucleotide sequence and genetic analysis of the neuD and neuB genes in region 2 of the polysialic acid gene cluster of Escherichia coli K1, J. Bacteriol. 177 (1995) 312–319. [11] A. Berry, K.E. Marshall, Identification of zinc-binding ligands in the Class II fructose-1,6-bisphosphate aldolase of E. coli, FEBS Lett. 318 (1993) 11–16. [12] L.B. Rodriguez-Aparicio, M.A. Ferrero, A. Reglero, N-acetyl-D neuraminic acid synthesis in Escherichia coli K1 occurs through condensation of N-acetyl-D -mannosamine and pyruvate, Biochem. J. 308 (1995) 505–510. [13] S. Yamamoto, K. Miyake, S. Iijima, Identification and characterisation of cps (capsular) po genes from S. agalactiae type Ia. Direct submission to the EMBL/GenBank/DDBJ databases, 1998. [14] K. Miyake, S. Yamamoto, Y. Koike, M. Watanabe, S. Iijima, Molecular characterisation of type specific capsular po biosynthesis genes of S. agalactiae type Ia. Direct submission to the EMBL/GenBank/DDBJ databases, 1999. [15] P. Glaser, F. Kunst, M. Arnaud, M.P. Coudart, W. Gonzales, I.M. Hu, B. Lubochinsky, L. Marcelino, I. Moszer, S.M.

[16]

[17] [18] [19]

[20]

[21]

[22]

[23]

[24]

Presecan, E. Schneider, J. Schweizer, A. Vertes, G. Rapoport, A. Danchin, Bacillus subtilis genome project: cloning and sequencing of region from 325 to 333 deg, Mol. Microbiol. 10 (1993) 317–384. K.S. Kirby, E. Fox-Carter, M. Guest, Isolation of deoxyribonucleic acid and ribonucelic acid from bacteria, Biochem. J. 104 (1967) 258–262. F.H. Arnold, A.A. Volkov, Directed evolution of biocatalysts, Curr. Opin. Chem. Biol. 3 (1999) 54–59. W.P.C. Stemmer, Rapid evolution of a protein in vitro by DNA shuffling, Nature 370 (1994) 389–391. W.P.C. Stemmer, A. Crameri, J. Minshull, Molecular evolution of genes and pathways by DNA shuffling, FASEB J. 11 (1997) 1559. A. Crameri, S.A. Raillard, E. Bermudez, W.P. Stemmer, DNA shuffling of a family of genes from diverse species accelerates directed evolution, Nature 391 (1998) 288–291. S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25 (1997) 3389–3402. E.R. Vimr, W. Aaronson, R.P. Silver, Genetic analysis of chromosomal mutations in the polysialic acid gene cluster of E. coli K1, J. Bacteriology 171 (1989) 1106–1117. R. Brossmer, U. Rose, D. Kasper, T.L. Smith, H. Grasmuk, F.M. Unger, Enzymatic synthesis of 5-acetamido-9-azido-3,5,9-trideoxy-D -glycero-D -galacto-2-nonulosonic acid, a 9-azido-9-deoxy derivative of N-acetylneuraminic acid, Biochem. Biophys. Res. Commun. 96 (1980) 1282–1289. K. McKinnon, D.O. Chaffin, C.E. Rubens, Direct submission to the EMBL/GenBank/DDBJ databases, 2001.