Cloning and expression of Bacillus phytase gene (phy) in Escherichia coli and recovery of active enzyme from the inclusion bodies

Authors


Vudem Dashavantha Reddy, Director, Centre for Plant Molecular Biology, Osmania University, Hyderabad-500007, India. E-mail: vdreddycpmb@yahoo.com

Abstract

Aims:  To isolate, clone and express a novel phytase gene (phy) from Bacillus sp. in Escherichia coli; to recover the active enzyme from inclusion bodies; and to characterize the recombinant phytase.

Methods and Results:  The molecular weight of phytase was estimated as 40 kDa on SDS-polyacrylamide gel electrophoresis. A requirement of Ca2+ ions was found essential both for refolding and activity of the enzyme. Bacillus phytase exhibited a specific activity of 16 U mg−1 protein; it also revealed broad pH and temperature ranges of 5·0 to 8·0 and 25 to 70°C, respectively. The Km value of phytase for hydrolysis of sodium phytate has been determined as 0·392 mmol l−1. The activity of enzyme has been inhibited by EDTA. The enzyme exhibited ample thermostability upon exposure to high temperatures from 75 to 95°C. After 9 h of cultivation of transformed E. coli in the bioreactor, the cell biomass reached 26·81 g wet weight (ww) per l accounting for 4289 U enzyme activity compared with 1·978 g ww per l producing 256 U activity in shake-flask cultures. In silico analysis revealed a β-propeller structure of phytase.

Conclusions:  This is the first report of its kind on the purification and successful in vitro refolding of Bacillus phytase from the inclusion bodies formed in the transformed E. coli.

Significance and Impact of the Study:  Efficient and reproducible protocols for cloning, expression, purification and in vitro refolding of Bacillus phytase enzyme from the transformed E. coli have been developed. The novel phytase, with broad pH and temperature range, renaturation ability and substrate specificity, appears promising as an ideal feed supplement. Identification of site between 179th amino acid leucine and 180th amino acid asparagine offers scope for insertion of small peptides/domains for production of chimeric genes without altering enzyme activity.

Introduction

Phytate, myo-inositol hexakisphosphate, is the primary source of inositol and the chief storage form of phosphorus accounting for 50–80% of the total phosphorus content of cereals, legumes and oilseed crops (Harland and Morris 1995). It acts as an antinutrient by chelating important dietary minerals, such as zinc, iron, copper, magnesium and calcium, making them unavailable for absorption in the intestines of humans and monogastric animals, such as poultry, fish and swine (Cheryan 1980). Phytate has been implicated in the precipitation of metal-binding enzymes and other proteins and also in the reduction of the digestibility of proteins, starch and lipids (Kies et al. 2006). Polyphenols and phytic acid were found to affect starch digestibility through interaction with amylase (Thompson and Yoon 1984). Action of various enzymes, viz, amylase, trypsin, acid phosphatase and tyrosinase were shown to be inhibited by phytic acid and inositol pentaphosphate (Harland and Morris 1995). Phytases, myo-inositol hexakisphosphate hydrolases, catalyse the hydrolysis of phosphomonoester bonds of phytate, thereby producing lower forms of myo-inositol phosphates and inorganic phosphates, which is deemed as an important metabolic process in various micro-organisms. Phytate phosphorus is biologically unavailable to various nonruminant species as they lack the requisite gastrointestinal enzymes capable of hydrolysing the substrate (Iqbal et al. 1994; Baruah et al. 2005). Undigested phytate remains in monogastric-animal-derived manure causing serious phosphorus pollution, contributing to eutrophication of surface waters. Supplementation of animal feeds with phosphorus to meet dietary requirements is not only expensive, but also potentially polluting and nonsustainable. Hence, introduction of phytases into animal feed is expected to resolve the problem of phosphorus pollution (Kim et al. 1999; Knowlton et al. 2004; Vohra et al. 2006).

Phytases are widely distributed in plants (Laboure et al. 1993; Dionisio et al. 2007), micro-organisms (Nakamura et al. 2000; Cho et al. 2005; Huang et al. 2006) and in certain animal tissues (Iqbal et al. 1994). In recent years, phytases have been studied intensively because of the great interest evinced in using these enzymes for reducing phosphorus supplements in animal feed and food items. Phytase not only satisfies the high phosphorus requirements, but also aids in efficient assimilation of other essential dietary factors (Han et al. 1997; Camden et al. 2001; Cufadar and Bahtiyarca 2004; Liao et al. 2005). Yip et al. (2003) demonstrated improved growth rate in transgenic tobacco expressing Bacillus phytase. Introduction of microbial phytase genes by genetic engineering and their expression in the edible parts of food crops, such as rice, improved the bioavailability of various mineral nutrients (Lucca et al. 2001; Hong et al. 2004). Scope for nutritional enrichment of animal feed and the grave problem of phosphorus pollution opened up bright prospects for research on phytase.

Phytases are derived from a number of sources, including micro-organisms, plants and animals. Until recently, fungal phytases have been commonly employed for use on a commercial scale in feed industry, but now research is targeted towards the use of bacterial phytases, as Bacillus phytase exhibits substrate (phytate) specific activity. Bacillus phytases have been studied extensively because of the immense potential of these enzymes having unique characteristics, feasibility of mass production for market and applicability in animal feed (Powar and Jagannathan 1982; Shimizu 1992; Kim et al. 1999; Park et al. 1999; Choi et al. 2001; Tye et al. 2002; Gulati et al. 2007). Expression of fungal phytases in bacterial systems proved futile owing to their inability to glycosylate the expressed protein into an active form (Phillipy and Mullaney 1997). Limited studies focussing on economically competitive expression and/or secretion systems for phytase were carried out (Mayer et al. 1999; Miksch et al. 2002). Kim et al. (1998) reported the expression of Bacillus phytase in Escherichia coli that amounted to only 20% of the total soluble proteins. However, production of Bacillus phytase in active form in E. coli has met with limited success owing to the enzyme being sequestered to inclusion bodies (Kerovuo et al. 1998). Kerovuo and Tynkkynen (2000) evaluated Lactobacillus plantarum as an extracellular expression system for the phyC gene from Bacillus subtilis, but the expression and secretion levels were too low to be considered for industrial use.

The present investigation deals with cloning and expression of a novel phytase gene (phy) from Bacillus sp. in E. coli. Further, it also reports the recovery of active phytase from the inclusion bodies produced in the transformed E. coli. In pilot scale experiments, the transformed E. coli, expressing phy gene upon cultivation in the bioreactor, could scale up the production of active phytase enzyme.

Materials and methods

Cloning of phytase gene

Genomic DNA was purified from the Bacillus sp., isolated at the Centre for Plant Molecular Biology, Osmania University, Hyderabad (India), according to the method of Ausubel et al. (1993). An additional methionine residue was incorporated at the N-terminus by adding an NdeI site at the 5′ end. Two forward primers 5′-CAT ATG AAG GTT CCA AAA ACA ATG CTG C-3′; 5′-CAT ATG TAT GTG AAT GAG GAA CAT CAT-3′ and the reverse primer 5′-CTA GCC GTC AGA ACG GTC TTT CAG C-3′, were designed based on conserved sequences and used for amplification of the phytase coding sequence and the truncated phytase (sans the first 81 nucleotides). Amplifications were carried out using 5 ng of genomic DNA as template with 1·5 units of Taq DNA polymerase, 200 μmol l−1 dNTP and 1·5 mmol l−1 magnesium chloride in Peltier Thermal Cycler PTC-200 (MJ Research, Ramsey, Minnesota, USA). The conditions of PCR were: 94°C for 45 s, 52°C for 45 s and 72°C for 2 min for 35 cycles. The amplified PCR products were ligated using T4 DNA ligase into the pTNOT vector linearized with XcmI. The genes were subcloned into the bacterial expression vector pET 21a (+) downstream to the T7 promoter at NdeI and BamHI sites. The recombinant plasmids harbouring the entire and truncated phytase gene were designated as pSPHY and pPHY, respectively, and the clones containing the phytase genes were subjected to sequence analysis. The deduced amino acid sequences were compared with the NCBI protein database by Blast search. The plasmid constructs were transformed into the E. coli strain BL21 (DE3) to check for protein expression.

An attempt was made to broaden the pH range, a nine amino acid domain, NSRHGARYP, corresponding to the fungal acid phosphatase activity, was introduced into the phytase gene between 179th amino acid leucine and 180th amino acid aspargine. The corresponding oligonucleotides, 5′-AAT TCC CGT CAC GGT GCG CGT TAC CCG-3′ and 5′-AAT TCG GGT AAC GCG CAC CGT GAC GGG-3′ were used for adapter assembly and then ligated into the EcoRI linearized pPHY clone. The resulting gene construct pRPHY was transformed into E. coli BL21 (DE3) cells and checked for protein expression and activity.

Expression of recombinant phytase

Hundred millilitres of Luria Bertani (LB) medium containing 70 μg ml−1 ampicillin was inoculated with 1 ml each of overnight culture of the different clones and agitated at 220 rev min−1 at 37°C. The cultures were induced with 1 mmol l−1 isopropyl β-d-thiogalactopyranoside (IPTG) once the OD600 reached 0·8 (log phase), and were allowed to grow under the same conditions. The uninduced and 3, 6 h and overnight induction samples were collected. Cells were lysed with 2× SDS buffer (Sambrook and Russel 2001) and electrophoresed on a 12% SDS-PAGE gel along with a standard protein marker.

Purification of induced protein

Purification of the induced protein was done by harvesting cells from 1 l induced culture subjected to centrifugation at 5000 g and 4°C for 5 min. The pelleted cells were resuspended in buffer containing 50 mmol l−1 Tris-Cl (pH 8·0) and 25% sucrose. The suspension was sonicated on ice at 50% level every 5 s for 30 pulses using the Labsonic Sonicator (B. Braun Biotech International, Penang, Malaysia). To this, lysozyme was added to a final concentration of 0·1 mg ml−1 and incubated at room temperature (28°C) for 30 min. Later, lysis buffer (50 mmol l−1 Tris-Cl, pH 8·0, 0·1% Triton X-100, 1% Na deoxycholate and 100 mmol l−1 NaCl) was added and incubated at room temperature for 60 min. Subsequently, it was centrifuged at 11 000 g for 20 min at 4°C. The pellet was initially washed with wash buffer (50 mmol l−1 Tris-Cl, pH 8·0, 100 mmol l−1 NaCl) containing 0·5% Triton X-100 and then with wash buffer alone. The inclusion bodies obtained were solubilized in 5 ml 8·0 mol l−1 urea in Tris-Cl (pH 7·0) buffer. Protein concentration was determined according to Bradford’s method (Bradford 1976), using the protein assay kit (GeNei, Bangalore, USA) with bovine serum albumin as the standard.

Co-solute-assisted protein refolding

The denatured protein was diluted to a final concentration of 0·1 mg ml−1 in refolding buffer (100 mmol l−1 Tris-Cl, pH 7·0, 0·5 mmol l−1 CaCl2) containing one of the different refolding co-solutes, viz 20% sucrose, 0·4 mol l−1l-arginine, 2·0 mol l−1 proline, 20% glycerol and 50 mmol l−1 glycine. Protein aggregation was monitored by the percentage transmittance of these solutions at OD600.

The purified and denatured protein was diluted to a final concentration of 0·1 mg ml−1 in refolding buffer (100 mmol l−1 Tris-Cl, pH 7·0, 0·5 mmol l−1 CaCl2) with and without proline. Experiments were carried out at 10°C and room temperature (28°C) to optimize the appropriate protein refolding temperature. Enzyme activity was tested every 24 h for a period of 1 week.

To optimize the appropriate concentration of proline for protein refolding, the refolding experiments were carried out using 0, 0·5, 1·0, 1·5, 2·0 and 2·5 mol l−1 proline in the refolding buffer.

Phytase assay

Phytase activity was measured by incubating 1 μg of the refolded enzyme with 600 μl of 2·0 mmol l−1 sodium phytate in 100 mmol l−1 Tris-Cl buffer (pH 7·0) supplemented with 2·0 mmol l−1 CaCl2.2H2O. Enzyme reaction was carried out at 37°C for 30 min, and then, the reaction was stopped by adding 750 μl of 5% trichloroacetic acid. The liberated phosphate was measured at 700 nm by following the production of phosphomolybdate with 1·5 ml of colour reagent (freshly prepared by mixing four volumes of 1·5% ammonium molybdate solution in 5·5% sulfuric acid and one volume of 2·7% ferrous sulfate solution). One unit of phytase activity was defined as the amount of enzyme hydrolysing 1 μmol of phosphate per minute under the assay conditions (Choi et al. 2001).

Enzyme assays were performed to check for the various parameters influencing the enzyme activity, and all assays were run in triplicate with an average of three independent experiments. Defined buffers used in enzyme activity assays to check for the pH optima were as follows: 100 mmol l−1 glycine (pH 3·0, 4·0), 100 mmol l−1 sodium acetate (pH 5·0, 6·0) and 100 mmol l−1 Tris-Cl (pH 7·0, 8·0, 9·0). To determine the enzyme kinetics, activity was tested at various substrate concentrations of 0·25, 0·5, 1·0, 2·0, 3·0 and 4·0 mmol l−1 sodium phytate. For determining the acid phosphatase activity, assays were carried out using 2·0 mmol l−1p-nitrophenyl phosphate in place of phytate.

Effect of divalent cations on enzyme activity

Metal requirement for the active conformation of enzyme was tested by refolding the denatured enzyme in the buffer (100 mmol l−1 Tris-HCl, pH 7·0, 2·0 mol l−1 proline) supplemented with one of the seven divalent cations, viz, Ca2+, Cu2+, Fe2+, Mg2+, Co2+, Zn2+ and Mn2+ at 0·5 mmol l−1 concentration. Enzyme assays were performed as described earlier.

Thermal stability of the enzyme

Thermal stability assays were conducted by denaturing the enzyme at temperatures ranging from 75 to 95°C for 10 min in 100 mmol l−1 Tris-HCl (pH 7·0) in the presence of 5 mmol l−1 Ca2+, followed by cooling to room temperature (28°C) for 1 h prior to enzyme assay conducted at 37°C.

Cultivation of recombinant Escherichia coli in bioreactor

Cultivation of E. coli BL21 (DE3)-pPHY was carried out in a 10 l Bioflo 2000 (New Brunswick Scientific, Edison, NJ, USA) bioreactor at 37°C using 300 rev min−1 agitation; pH 7·0; 30·2 mg l−1 dissolved oxygen; and 0·2 min−1 VVM (volumes of air per minute per volume of batch). Foaming was controlled by the addition of 1·0 ml of antifoam agent from Sigma Aldrich. An inoculum of E. coli BL21 (DE3)-pPHY was grown in 500 ml Erlenmeyer flask containing 100 ml LB seed medium with 70 μg ml−1ampicillin for about 16 h. The seed (OD600 2·28) was used to inoculate 6·0 l of terrific broth medium containing 0·2% glycerol and 100 μg ml−1 ampicillin in the bioreactor. Ampicillin was added once again after 2 h of seed inoculation to a final concentration of 100 μg ml−1 to maintain the selection pressure. Lactose (0·4%) and IPTG (1·0 mmol l−1) were added after 3 and 4 h of seed inoculation, respectively. Uninduced and 3- and 6-h induction samples were collected and checked on a 12% SDS-PAGE gel. Enzyme recovery and assays were performed as described earlier.

In silico modelling of the proteins

The phytase protein PHY was modelled using the modeller from Accelerys module to determine its structure and also the structural integrity and stability upon incorporation of the acid phosphatase domain of the protein pRPHY.

Results

Sequence homology of phytase gene

The nucleotide and the deduced amino acid sequences of Bacillus phytase gene (phy) have been subjected to blast analysis. The gene disclosed 85% identity with phyC gene of Bacillus amyloliquefaciens strain BAP, and 84% identity with that of phyC gene of B. subtilis, phy gene of B. amyloliquefaciens strain FZB45 and phytase gene of Bacillus sp. SD01N. At the amino acid level, it showed 73% identity with PHYC of B. amyloliquefaciens, PHY of Bacillus licheniformis (strain DSM 13 / ATCC 14580), 72% identity with the phytase of Bacillus sp. SD01N and 71% identity with PHYC of B. subtilis. The complete pSPHY gene sequence has been deposited in GenBank with accession number EF536824.

Biophysical characteristics of phytase proteins

Molecular weights of the proteins encoded by the entire (pSPHY) and truncated phytase (pPHY) clone, as estimated by SDS-PAGE, are approximated to 43 and 40 kDa, respectively (Fig. 1a). In both the cases, the expressed proteins were found in the insoluble cytoplasmic fraction as inclusion bodies. These proteins were purified and solubilized in 8·0 mol l−1 urea, and the denatured proteins showed aggregation when diluted in the refolding buffer devoid of co-solutes. To promote protein refolding and to prevent protein aggregation, various co-solutes were added to the refolding buffer. Protein aggregation was tested as a function of per cent transmittance (%T600) of the ‘refolded protein’. With the exception of proline and l-arginine, the %T600 of the protein solutions refolded in the presence of various co-solutes (sucrose, glycerol and glycine) was in the range of 10–20%. At <2·0 mol l−1 proline concentration, the solutions containing ‘refolded protein’ were faintly turbid. Both 2·0 mol l−1 proline and 0·4 mol l−1l-arginine could suppress protein aggregation as evidenced by 100% transmittance at OD600. However, the active enzyme was recovered only from the truncated protein refolded in the presence of proline as a co-solute, while the nontruncated protein proved to be inactive. The purified phytase enzyme was observed as a single protein band on SDS-PAGE (Fig. 1b), which showed a specific activity of 16 U mg−1 protein. The theoretical pI of the enzyme was estimated as 4·90.

Figure 1.

 (a) Expression of Bacillus sp. phytase in Escherichia coli. Lane 1, uninduced sample; lanes 2 and 3, 3- and 6-h induction sample of phytase (∼43 kDa) with signal sequence encoded by pSPHY; lanes 4 and 5, 3- and 6-h induction sample of phytase (∼40 kDa) lacking signal sequence encoded by pPHY; lane M, protein marker. (b) Expression of Bacillus sp. phytase in E. coli. Lane 1, uninduced sample; lane 2, 6-h induction sample of phytase lacking signal sequence (pPHY) showing induced band of phytase; lane 3, purified phytase (∼43 kDa) with signal sequence encoded by pSPHY; lane 4, purified phytase (∼40 kDa) encoded by pPHY; lane M, protein marker.

Refolding, substrate-specificity and enzyme kinetics

Observations on protein refolding revealed that maximum enzyme activity (16 U mg−1) was obtained with 2·0 mol l−1 proline. A gradual decrease in the recovery of activity has been observed with decreasing concentration of proline, while only negligible activity was noted in the absence of proline (Fig. 2). Data recorded on protein refolding temperatures revealed 10°C as optimal as continuous improvement in the enzyme activity has been observed up to 4 days. Whereas, at room temperature (28°C), only 50% enzyme activity could be observed after 24 h without any further improvement with increasing time of incubation. Similar trend was observed at higher temperatures over 28°C.

Figure 2.

 Phytase activity recovered at different concentrations of proline in the refolding buffer. The values are average of three experiments with three replicates each. I, standard error.

Activity of phytase enzyme was tested using sodium phytate as the substrate. Enzyme refolded in the presence of Ca2+ showed hydrolysis of the phytate, while it failed to hydrolyse the p-nitrophenylphosphate. Enzyme refolded in the presence of one of the divalent cations, viz, Cu2+, Fe2+, Mg2+, Co2+, Zn2+ and Mn2+, failed to show any enzyme activity even in the presence of Ca2+ in the reaction mix. Addition of 2·0 mmol l−1 EDTA to the reaction mix resulted in complete inhibition of the enzyme activity. The enzyme followed simple Michaelis–Menten kinetics, and the Lineweaver–Burk plot (data not shown) resulting in a Km value of 0·392 mmol l−1 for hydrolysis of sodium phytate.

Effect of pH and temperature

The enzyme was found active at 37°C between pH 5·0 and 8·0 with the peak in activity at pH 7·0 (Fig. 3). Although the enzyme exhibited activity at various temperatures ranging from 25 to 70°C, an optimum activity has been observed at 55°C and pH 7·0 (Fig. 4). The enzyme showed a recovery of ∼45% activity when it was preincubated at pH 3·0 and 37°C for 6 h prior to phytase assay. The enzyme upon denaturation for 10 min at 75, 85 and 95°C, followed by 1 h renaturation at the room temperature (28°C) in the presence of 5 mmol l−1 Ca2+, restored 86%, 54%, and 37% activity (Fig. 5). However, no higher activity was observed after prolonged refolding time after heat denaturation. Temperature mediated denaturation of the enzyme in the presence of reduced Ca2+ (1 mmol l−1) failed to show recovery of activity.

Figure 3.

 Effect of pH on the phytase activity. The values are average of three experiments with three replicates each. I, standard error.

Figure 4.

 Temperature optima for phytase activity. The values are average of three experiments with three replicates each. I, standard error.

Figure 5.

 Thermostability of the phytase.

Cultivation of Escherichia coli in the bioreactor

The transformant with pPHY was cultivated in the bioreactor on a pilot scale to assess its productivity. The cell mass obtained after 9 h of cultivation was found to be 26·81 g ww per litre yielding 4289 U enzyme activity when compared to 1·978 g ww per litre mass producing 256 U activity in shake-flask cultures.

Activity of modified phytase

The recombinant phytase enzyme, containing the insertion of a nine amino acid domain, encoded by clone pRPHY exhibited no acid phosphatase activity, but showed only phytase activity similar to that of unmodified, refolded recombinant phytase enzyme. Furthermore, in silico analysis of the modified protein disclosed identical deduced pI value of 4·8, and was classified as a stable conformational protein.

In silico modelling of the proteins

In silico modelling of the proteins revealed β-propeller structure of the phytase with a six-fold symmetry axis lying on the shaft of the propeller. The phytase protein pRPHY also revealed similar structure despite domain insertion in between the Leu 179 and Asp 180 with the formation of a looping out structure of the inserted stretch of amino acids (Fig. 6).

Figure 6.

 Structure of phytase proteins. (a) pPHY encoded phytase as viewed from the propeller shaft.,(b) pRPHY encoded phytase with an inserted domain looping out (indicated by arrow).

Discussion

The enormous potential for application of phytase has motivated researchers to attempt cost-effective production of the enzyme in the bacterial system. This has resulted either in the poor expression levels and/or formation of inclusion bodies. The difficulties associated with the isolation, purification and in vitro refolding of recombinant phytase from inclusion bodies at economically feasible levels constitute a major challenge. This is the first report of its kind dealing with the recovery of active Bacillus phytase enzyme from the inclusion bodies produced in the transformed E. coli.

Nonfunctionality of the whole protein encoded by the full length pSPHY and enzyme activity displayed by the truncated protein encoded by pPHY (sans 81 bp) imply that the first 27 amino acids at the N-terminus interfere with the protein refolding, thereby affecting its function. Using the Signal 3·0 Server (http://www.cbs.dtu.dk/services/SignalP/), the 27 amino acid sequence was predicted to be the signal peptide of the enzyme. Full-length protein encoded by pSPHY amply demonstrates the nonfunctionality of Bacillus signal sequence in E. coli.

Protein aggregation is a major hurdle encountered in the production of recombinant proteins. Dilution of the purified denatured phytase during refolding resulted in the protein aggregation. Unproductive protein aggregations might originate both from nonspecific (hydrophobic) interactions of predominantly unfolded polypeptide chains and from improper interactions among partially structured folding intermediates. Hydrophobic interactions might be the primary cause of protein aggregation as there is no scope for disulfide linkages owing to the absence of cysteine residues in the phytase. Protein aggregation was suppressed by various co-solutes used; yet proline (2·0 mol l−1) proved to be both an aggregation suppressor and a folding enhancer at the same time. Artificial chaperones, such as glycerol and proline, used in this study increased the viscosity of refolding solution, which might have modulated the protein refolding dynamics so that the partially folded intermediates can successfully complete the folding process rather than being blocked by aggregation. This mechanism is corroborated by increased enzyme activity with increasing proline concentration (0–2·0 mol l−1) in the refolding buffer. Osmolytes, such as proline, may mainly affect the protein intermediates, having a partially disrupted network of internal hydrophobic interactions, leading to a decrease in the extent of solvent-exposed nonpolar surface. The osmolytes are known to initiate restoration of the internal-protein-hydrophobic core for correct folding. Kumar et al. (1998) reported the formation of ordered supramolecular assembly of proline. Based on their findings on the refolding of bovine carbonic anhydrase, they proposed that proline behaves as a protein-folding chaperone owing to the formation of an amphipathic supramolecular assembly.

The liberation of phosphate (Pi) catalysed by phytase is dependent on the concentration of substrate, implicating a typical Michaelis–Menten kinetics. Total inhibition in the activity of metal-depleted enzyme signifies the metal dependency of phytase for its structural integrity and/or conformational stability. In the present investigation, EDTA was found to inactivate the Bacillus phytase, whereas Wyss et al. (1999) reported EDTA-mediated stimulation of acid phosphatase activity of fungal phytases. Metal–substrate complex seems to be the only true substrate for Bacillus phytase as metal depletion caused by EDTA inhibited the enzyme activity. Phytase refolded in the presence of Ca2+ alone could display activity, while enzyme refolded in the presence of other divalent cations, such as Cu2+, Fe2+, Mg2+, Co2+, Zn2+ and Mn2+, failed to show any activity on the Ca2+–substrate complex, suggesting that Ca2+ ions are an essential component of the enzyme structure and are responsible for attainment of active conformation. Furthermore, the enzyme activity may be attributed to the effect of electrostatic interactions of the metal–phytate complex and the metal–enzyme complex owing to the larger ionic radius of Ca2+ compared with smaller ion radius of other divalent cations. The pH conditions prevalent in the gastro-intestinal tract lead to the formation of insoluble metal–phytate complexes hindering their absorption in the intestinal tract of animals and humans (Maga 1982; Pallauf and Rimbach 1997). In the neutral pH conditions prevalent in the small intestine, phytic acid will change from the protonated form into phytates, predominantly Ca phytate, which is the true substrate of Bacillus phytase (Oh et al. 2001). Hence, Bacillus phytase with a broad pH range, developed in this investigation, might serve as an ideal enzyme for feed supplementation.

The hydrolysis rate of phytate was doubled when the assay temperature was elevated from 37 to 55°C and the optimum temperature for the enzyme activity was established as 55°C. Phytase exposed to higher temperatures (75–95°C) showed decline in enzyme activity. However, heat-denatured enzyme, when incubated at room temperature, exhibited recovery in the activity, suggesting heat denaturation of this enzyme is a reversible process. Absence of Ca2+ ions at higher temperatures and during renaturation of the heat-treated enzyme resulted in no activity. These results amply indicate that the Ca2+ ions are involved in the thermal stability of the enzyme and also that the conformational stability of the enzyme is metal dependent. Enzyme thermal stability is indispensable in animal feed applications, where the enzyme is normally incorporated into the feed prior to pelleting, and the feed reaches processing temperatures of 80 to 85°C for a brief period of 2 min (Wyss et al. 1998; Kim et al. 1999). As such, Bacillus phytase by virtue of its thermal tolerance can find definite applicability in feed subjected to pelleting processes.

The Bacillus phytase, exhibiting a broad pH range (5·0–8·0) with optimal pH 7·0, is of crucial importance as the pH of the animal gastrointestinal tract ranges from 2·6 to 7·0, with a variation from 4·0 to 5·0 in the stomach in the presence of food. However, this is not so with regard to the Aspergillus phytases where the enzyme activity is observed only at a low pH range of 2·0–5·0 owing to their acid phosphatase activity. Recovery of enzyme activity after incubation at low pH for 6 h indicates that Bacillus phytase has the ability to regain its native conformation and exhibit activity even after exposure to acidic conditions. Sandberg and Andlid (2002) reported that ∼35% and ∼66% of phytate was hydrolysed in the stomach and small intestine, respectively. Therefore, Bacillus phytase having a broad pH range of 5·0–8·0 and its ability to renature at near-neutral pH, after denaturation at low pH, is advantageous as a feed additive. After passing through the stomach with low pH, the enzyme can renature and function in the small intestine (with near-neutral pH), wherein phosphate absorption takes place. Bacillus phytases with broad pH range and ability to recover activity after incubation at low pH for several hours indicate that it can serve as an ideal feed additive for poultry. Bacillus phytases are on par with fungal phytases with regard to their efficacy as feed additives (Elkhalil et al. 2007). Failure of Bacillus phytase to hydrolyse p-nitrophenyl phosphate indicates the lack of acid phosphatase activity. Highly substrate specific Bacillus phytase, free from acid phosphatase activity, has great potential as an ideal candidate for genetic transformation of plants, as fungal phytases with acid phosphatase activity and broad substrate range might disturb the metabolic pathways of plant.

Phytase having additional nine amino acid domain encoded by the insertional clone (pRPHY), although showed no acid phosphatase activity, exhibited enzyme activity on par with that of enzyme encoded by pPHY suggesting that the incorporation of amino acids in the region between the 179th amino acid leucine and the 180th amino acid asparagine, neither altered the protein refolding nor the stability of the enzyme. Therefore, this region has the potential to serve as the favourable target site for incorporation of domains for production of chimeric enzymes by incorporating short peptides into the protein without affecting the enzyme activity.

When the transformed E. coli expressing Bacillus phytase was cultured in the bioreactor, it could produce a 16-fold increase in the biomass as compared with the shake-flask cultures. Thus, an enormous increase in the productivity of phytase was achieved without affecting the level of enzyme activity. Furthermore, the use of optimized culture conditions in the bioreactor can be employed to scale up the phytase production for industrial application.

In silico analyses of the phytase structure revealed that it adopts β-propeller conformation with six stranded blades as in the case of B. amyloliquefaciens thermostable phytase (Ha et al. 2000). The structural integrity of the protein pRPHY with the looping out of the inserted amino acid stretch, as determined by the in silico studies, confirm the enzyme stability as evidenced by the unaltered enzyme activity in in vitro studies.

In the present investigation, reproducible protocols for cloning, expression, purification and efficient in vitro refolding of Bacillus phytase enzyme from the transformed E. coli have been successfully developed. The novel phytase with broad pH and temperature ranges, high renaturation capability and substrate specificity, makes it an ideal feed supplement. The present results also provide impetus to enhance the activity of phytase enzyme by altering its physicochemical properties through site-directed mutagenesis. Furthermore, protein-engineering methods may be adopted to design next-generation phytase endowed with improved attributes of a befitting feed additive.

Acknowledgements

We are thankful to Prof T.P. Reddy, former Head, Department of Genetics, Osmania University, for valuable suggestions in the preparation of the manuscript. The financial assistance provided to Rao, D.E.C.S., by Council of Scientific and Industrial Research, New Delhi, India, is duly acknowledged.

Ancillary