Cloning, Overexpression, Refolding, and Purification of the Nonspecific Phospholipase C fromBacillus cereus

PROTEIN EXPRESSION AND PURIFICATION ARTICLE NO.

10, 365–372 (1997)

PT970756

Cloning, Overexpression, Refolding, and Purification of the Nonspecific Phospholipase C from Bacillus cereus Cristina A. Tan, Michael J. Hehir, and Mary F. Roberts1 Merkert Chemistry Center, Boston College, 2609 Beacon Street, Chestnut Hill, Massachusetts 02167

Received January 21, 1997, and in revised form April 9, 1997

Bacillus cereus secretes a nonspecific phospholipase C (PLC) that catalyzes the hydrolysis of phospholipids to yield diacylglycerol and a phosphate monoester. B. cereus PLC has been overexpressed with its signal sequence in Escherichia coli using a T7 expression system. The expressed enzyme formed intracellular inclusion bodies which were solubilized in the presence of 8 M urea. Renaturation was initiated by gradual removal of urea and addition of zinc ions. The signal peptide was specifically cleaved by a protease, clostripain, added when the urea concentration was 1.5 M. Factors that led to protein reaggregation included rapid removal of urea, use of Tris instead of barbital buffer, and presence of the signal peptide when the urea concentration was below 1.5 M. The folded protein was purified by Q-Sepharose Fast Flow chromatography to yield a preparation ú99% pure. The final yield of active enzyme was 30–40 mg per liter of culture. The recombinant PLC exhibited biochemical and kinetic properties identical to those of extracellularly produced PLC from B. cereus. Site-specific mutagenesis of Asn-134 was carried out as a test of the general effectiveness of the refolding procedure. q 1997 Academic Press

Non-specific phospholipase C (PLC,2 EC 3.1.4.3) plays a crucial role in the regulatory functions related to cellular signal transduction pathways in mammals (1,2). The enzyme catalyzes the hydrolysis of phospho1 To whom correspondence should be addressed. Fax: (617)-5522705. 2 Abbreviations used: PLC, phospholipase C; DAG, diacylglycerol; SPM, sphingomyelin; SPMase, sphingomyelinase; diC6PC, 1,2-dihexanoylphosphatidylcholine; diC7PC, 1,2-diheptanoylphosphatidylcholine; diC14PC, 1,2-dimyristoylphosphatidylcholine; PC, phosphatidylcholine; CMC, critical micelle concentration; IPTG, isopropylb-D-thiogalactoside; PCR, polymerase chain reaction; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; TX-100, Triton X-100; LB, Luria broth.

diester bonds in phospholipids to generate a water-soluble phosphate monoester and diacylglycerol (DAG), a hydrophobic membrane-localized activator of protein kinase C. PLC from Bacillus cereus is a monomeric, exocellular, zinc metalloenzyme (3). It has been shown to be a useful model for the poorly characterized mammalian PLC, since it can mimic the ability of mammalian PLC to enhance prostaglandin synthesis (4). Moreover, antibodies to the enzyme purified from B. cereus cross-react with some of the mammalian activity, which suggests that the bacterial enzyme is antigenically and structurally similar to the mammalian enzyme (5). The enzyme acts stereospecifically on phospholipids with a broad range of headgroups and exhibits interfacial activation by preferentially binding and catalyzing the hydrolysis of substrates clustered as aggregates rather than monomers (6). PLC is a phosphodiesterase since phosphatidic acid is neither a good substrate nor an inhibitor (7). PLC is synthesized as a 283-residue precursor with a 24-residue prepeptide and a 14-residue propeptide (collectively termed signal peptide). The mature form of the enzyme contains 245 amino acids and has a molecular weight of 28,520 Da. The backbone carboxyl and amide groups of the N-terminal residue of the active enzyme, Trp-1, are involved in coordination to zinc which may explain why the unprocessed enzyme remains inactive (8). Details of the active site of PLC have been provided by crystal structures of the enzyme complexed with several inhibitors (9–11). A potential nucleophile for attack on the phosphorus has been proposed based on a structure of PLC complexed with a phosphonolipid inhibitor; however, this assignment is tentative because the conformation of the phosphonolipid in the structure is different from that found in natural phospholipid substrates (11). A mechanism with a slightly different arrangement of catalytic groups based on distance geometry calculations and molecular modeling has also been proposed (12). There are few mechanistic 365

1046-5928/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

pepa

AP: PEP

366

TAN, HEHIR, AND ROBERTS

studies of this enzyme, although there is some indirect evidence for the order of product release (13). Site-specific mutagenesis is a powerful tool for studying structure–function relationships in enzymes. Nonspecific PLC has not been a good target for mutagenesis studies, in part because earlier attempts to express B. cereus PLC in Escherichia coli were not encouraging (8). High levels of expression with leaky high-level expression vectors were usually toxic to the host cell. To approach this problem, we have recloned the plc gene into the tightly regulated pET vector (14) and developed a new method for expressing large quantities in E. coli. Moreover, we have devised a simple scheme for purification of the recombinant PLC that takes advantage of its denaturation in 8 M urea and refolding on addition of Zn2/. MATERIALS AND METHODS

Materials. The PCR kit Gene-Amp was purchased from Perkin–Elmer (Branchburg, NJ), DNA ligation kit from Novagen, Inc. (Madison, WI), a gene cleaning kit from BIO 101 Inc. (La Jolla, CA), DNA sequencing kit from USB (Cleveland, OH), and QuickChange sitedirected mutagenesis kit from Stratagene Cloning Systems (La Jolla, CA). Restriction enzymes and other enzymes for cloning were obtained from New England Biolabs (Cambridge, MA) and USB (Cleveland, OH). Primers for PCR, sequencing, and site-directed mutagenesis were obtained from Operon Technologies (Alameda, CA). Ultrapure urea was obtained from USB. Barbital buffer was purchased from Sigma (St. Louis, MO), [35S]dATP was obtained from Amersham Life Sciences (Cleveland, OH). All other materials were reagent grade. Dialysis tubing (molecular weight cutoff 12,000–14,000) was from Spectrum (Houston, TX). Clostripain (EC 3.4.22.8) was purchased from Sigma and used with no further purification. Native PLC from B. cereus was purchased from Sigma or purified according to Little and co-workers (15). Plasmids and bacterial strains. The PLC-overproducing B. cereus strain ATCC10987 was obtained from American Type Culture Collection (Rockville, MD). The pET23a expression vector, BL21(DE3) pLysS, and Novablue competent cells were obtained from Novagen. Epicurian Coli XL-1 Blue Supercompetent cells were obtained from Stratagene Cloning Systems. Bacterial culture media. The B. cereus medium contained the following (g liter01): Sigma peptone, 10; yeast extract, 10; NaCl, 5; Na2HPO4 , 0.4. The pH was adjusted to 7.4 with 1 N NaOH. LB medium for growth of E. coli contained (g liter01): Bactotryptone, 10; yeast extract, 5; NaCl, 10. The pH was adjusted to 7.5 with 1 N NaOH. If necessary, ampicillin and chloramphenicol were added to final concentrations of 100 and 34 mg/ ml, respectively.

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

Isolation of the genomic DNA of B. cereus. A single colony of B. cereus ATCC10987 was grown in 5 ml B. cereus medium overnight at 377C. The next day, 1 ml of the overnight culture was used to inoculate 50 ml fresh medium; this culture was subsequently grown for 6 h at 377C. After the cells were harvested, the genomic DNA was isolated using sodium dodecyl sulfate and proteinase K to lyse the cell membrane and cetyltrimethylammonium bromide to remove the contaminating components (16). Plasmid constructs. We have designated the B. cereus gene encoding the nonspecific PLC to be plc as previously proposed (8). Chromosomal DNA from B. cereus was digested with NotI. PCR was used to amplify the plc gene. The 5* primer contained an NdeI site and the 3* primer contained an EcoRI site. The PCR fragment was digested with NdeI and EcoRI, then ligated into the plasmid pET23a to create the plasmid pMR1 carrying the plc gene. The entire nucleotide sequence of the plc gene was verified by the dideoxy chain-termination method using a Sequenase Version 2.0 sequencing kit (USB). Expression of recombinant PLC. A single colony of E. coli BL21(DE3) pLysS cells harboring plasmid pMR1 was grown at 377C with shaking in 5 ml of LB medium containing ampicillin and chloramphenicol until the OD600 reached 0.6–0.8. A 1-ml aliquot of this culture was diluted into 50 ml of the same medium and grown at 377C with shaking. When the cells reached an OD600 of 0.6–0.8, expression from the T7 promoter was induced by adding IPTG to a concentration of 0.4 mM. Growth was continued for another 2–3 h. The cells were harvested by centrifugation at 5000 rpm in a Beckman JA 17 rotor for 5 min at 47C. Purification of inclusion bodies. The cell pellet was washed with 10 ml of ice-cold wash buffer (50 mM Tris– HCl, 2 mM EDTA, pH 8.0), and centrifuged at 5000 rpm in a Beckman JA17 rotor for 5 min at 47C. The pellet was resuspended in 5 ml wash buffer, and hen egg white lysozyme was added to a final concentration of 100 mg/ml. After incubation for 20 min at room temperature MgCl2 and DNase were added to final concentrations of 10 mM and 20 mg/ml, respectively. After incubation for an additional 20 min at room temperature with gentle shaking, the suspension was sonicated on ice with four 15-s pulses. The inclusion bodies were washed by resuspension in wash buffer at least twice, followed by centrifugation at 4000 rpm for 25 min at 47C. Denaturation and refolding of PLC. The inclusion bodies were denatured with 20 ml of buffer containing 8 M urea and 25 mM barbital buffer, pH 7.5 (adjusted by acetic acid). The solution was stirred at room temperature for at least 1 h or until the sample turned clear. The amount of urea was gradually adjusted to 4

pepa

AP: PEP

OVEREXPRESSION AND PURIFICATION OF NONSPECIFIC PHOSPHOLIPASE C

by addition of 25 mM barbital buffer, pH 7.5, and the sample was stirred overnight at 47C. Refolding was initiated by addition of zinc to a final concentration of 0.1 mM, and the stirring was continued for another 2 h. The protein was dialyzed against 1.5 M urea in 25 mM barbital, pH 7.5, and 0.1 mM ZnSO4 (refolding buffer). Clostripain was added to the dialyzed protein in a clostripain-to-inclusion body ratio of 1:50 (w/w), with 10 mM dithiothreitol added to activate clostripain. The sample was incubated at 377C for 20 min followed by heating at 507C for 20 min. The remaining urea was gradually removed by dialysis first against 0.3 M urea in refolding buffer and finally against refolding buffer without urea, followed by dialysis against the column buffer used next in an ion-exchange column (25 mM barbital, 0.01 mM ZnSO4 , pH 8.3). Refolded PLC was loaded on a 5-ml Q-Sepharose Fast Flow (Pharmacia Biotech, Piscataway, NJ) column equilibrated with column buffer. The column was washed with 4 column vol of buffer. PLC was eluted using a linear gradient up to 0.5 M sodium acetate in the same buffer. Fractions containing PLC eluted as a broad peak from 0.08 to 0.1 M sodium acetate; the PLC was identified by SDS– PAGE. Appropriate fractions were pooled and dialyzed against refolding buffer, then concentrated using Centriprep 10 (Amicon, Inc., Beverly, MA) and stored at 47C. SDS–PAGE was used to analyze the purity of the protein. Gels were stained with Coomassie brilliant blue. Protein concentration was determined by ultraviolet spectroscopy at 280 nm using an extinction coefficient of 51,000 M01 cm01 (17) recalculated based on a molecular weight of 28.52 kDa. The Bio-Rad version of the Bradford assay was used to determine the concentrations of mutant proteins, using PLC as standard. The first three amino-terminal residues of the active PLC were sequenced by Commonwealth Biotechnologies Inc. (Richmond, VA). M

Large-scale production of recombinant PLC. For cells that were grown in 1 liter medium, solubilization of inclusion bodies was done in 80 ml of 8 M urea in 25 mM barbital, pH 7.5. The rest of the refolding procedure was similar to that described above, except the concentration of ZnSO4 was 0.01 mM throughout the refolding process. Kinetic analyses of recombinant PLC. Phospholipid substrates stored in chloroform were purchased from Avanti (Alabaster, AL) and used without further purification. For preparation of assay samples, the chloroform was removed under a steady stream of nitrogen gas and the phospholipid sample was placed under vacuum overnight to remove residual organic solvent. Small unilamellar vesicles of SPM and diC14PC were prepared by suspending the dried lipid film in buffer and sonicating for 7 min with a Branson-200 sonifier

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

367

equipped with a microtip. The vesicle solution was then centrifuged at 14,000 rpm for 1 h at 47C and the pellet discarded. PLC activity at 257C was measured by titration of the product phosphomonoester with NaOH as the titrant and an endpoint of pH 8.0. Base concentration was kept between 0.5 and 1.5 mM NaOH and the base chamber was protected from CO2 absorption with ascarite. DiC6PC samples ranged in concentration from 0.05 to 5 mM; diC7PC ranged from 1.5 to 5 mM. All rates were determined in duplicate or triplicate. At high substrate concentrations, the estimated average experimental uncertainty in the rates was 8%; at low substrate concentrations, the uncertainty was 7%. Errors in Km extracted from a Michaelis–Menten treatment of monomeric substrate were estimated to be less than 20%. 31 P NMR (202.33 MHz) was used as an alternate method to monitor PLC activity toward phospholipid substrates in various aggregation states. This technique is particularly useful where the enzyme activity is low. All spectra were obtained on a Varian Unity 500 spectrometer. Monomers, detergent-mixed micelles, as well as small unilamellar vesicles (in 50 mM Tris–acetate, pH 8.0), were used in assays for PLC activity. Phosphoric acid was used as an external reference. Typical acquisition parameters for a kinetic run included 16,380-Hz sweep width, 5.1-ms pulse width (577), 1-s acquisition delay, 1.6-s accumulation time, 1 H-WALTZ decoupling, and 16 transients per time point. Hydrolysis of phospholipid substrate was monitored by measuring the integrated intensity under the resonance corresponding to the phosphate monoester (Ç4-5 ppm) as a function of incubation time. When SPM was used as a substrate, the 31P spectra were monitored for 3 h at 307C with 80 transients per time point. All other substrates were monitored for 30–60 min. Errors in specific activities were estimated to be about 15%. Site-specific mutagenesis. With double-stranded pMR1 as the template and a forward and a reverse primer that contained the desired mutation, the mutated plasmid was generated using the Quickchange method. The entire mutant plc gene was then sequenced, digested with NdeI and EcoRI, and then subcloned into NdeI/EcoRI-digested pMR1. Mutagenesis efficiency was 100%. RESULTS AND DISCUSSION

Cloning Strategy The genome of B. cereus contains 11 NotI sites, with the plc gene localized on the biggest NotI fragment (1.3 Mb) (18). PCR was used to amplify the plc gene from NotI-digested chromosomal DNA of B. cereus ATCC 10987. The 5* primer was designed to introduce an

pepa

AP: PEP

368

TAN, HEHIR, AND ROBERTS

FIG. 1. Subcloning of the plc coding sequence into pET23a expression vector. An NdeI site was engineered between the presequence and the prosequence, and an EcoRI site was introduced downstream of the translational stop codon. Following PCR the gene was subcloned to NdeI/EcoRI-digested pET23a to produce pMR1. The Nterminal residue of the active protein is Trp-1 whose backbone is involved in metal coordination.

NdeI site upstream of the plc gene, within the signal peptide encoding region. The 3* primer was designed to introduce an EcoRI site downstream of the gene past the stop codon (Fig. 1). After amplification, the plc gene was cloned into pET23a, an expression vector based on the T7 expression system. The sequence of the entire cloned gene was determined to be identical to that reported earlier (8). For protein production, the recombinant plasmid pMR1 was transformed into host E. coli BL21(DE3) pLysS, which contains a chromosomal copy of the gene for T7 RNA polymerase under the lac UV5 promoter. After cells were grown to OD600 Ç 0.6, IPTG was added to the culture. This induced T7 RNA polymerase, which in turn transcribed the plc gene. Based on SDS–PAGE, PLC, with a molecular weight of 28.5 kDa, constituted more than 80% of the total cell protein under these conditions (Fig. 2, lane 1). Expression and Refolding There are several potential problems that need to be considered in cloning PLC in E. coli. With the T7

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

expression system, production of the target protein occurs so rapidly that it is often not processed properly and is frequently misfolded (19,20). Hence, methodology must be available for refolding the protein. Another potential problem with this particular system is that the plc gene was cloned with its signal sequence intact. Removal of the signal sequence is critical to protein activity. E. coli may not recognize this sequence; it may also lack the appropriate protease to convert the preprotein to the active protein, although previous reports of cloning this enzyme in E. coli (8) detected a low yield of mature protein in the cytoplasm of the cell. Introduction of the initiator Met residue in place of the N-terminal Trp of the mature enzyme can be potentially problematic since the amino group of Trp-1 is involved in coordination of one of the active site Zn2/ ions. The PLC produced in the E. coli host with the T7 expression system accumulated in inclusion bodies with its signal peptide uncleaved. After lysis of the cells, the inclusion bodies were easily isolated from the majority of the proteins in the cell by washing several times and centrifuging at low speed. This treatment sedimented the inclusion bodies but not the cell debris; however, the protein so produced was inactive and had to be refolded as well as proteolytically processed. A variety of experimental conditions were explored to obtain greater solubility, including induction at lower temperature (from 20 to 307C) and use of lower concentration of IPTG for longer induction times (21), but these led only to a much lower production of PLC inclusion bodies.

FIG. 2. Analysis of the purification of recombinant phospholipase C by SDS–PAGE: lane 1, crude cell pellet after removal of the supernate medium; lane 2, commercially available PLC (the upper band corresponds to active PLC and the lower band is a degradation product.); lane 3, inclusion body protein solubilized in 8 M urea prior to addition of clostripain; lane 4, post Q-Sepharose Fast Flow, active protein after removal of the signal peptide by clostripain.

pepa

AP: PEP

369

OVEREXPRESSION AND PURIFICATION OF NONSPECIFIC PHOSPHOLIPASE C

FIG. 3. Flow chart describing the steps for purification and refolding of recombinant phospholipase C.

The purification method that was developed focused on optimizing PLC folding in vitro (Fig. 3). Strong denaturant was needed to solubilize the protein from the inclusion bodies during the folding process. The effect of urea on the conformation of native PLC has been studied in the past (17), and we have used the stability of PLC in urea to refold the enzyme isolated from the inclusion bodies. The enzyme purified from B. cereus is extremely resistant to high concentrations of chaotropic reagent. The zinc ions contribute to the general stability of the enzyme. In the presence of zinc ions, PLC is active even at 8 M urea, but the Zn-free enzyme decreases in activity and its fluorescence properties change. Dilution of urea to 1.5 M together with addition of zinc restored the activity and fluorescence emission of the enzyme (17). Solubilizing the recombinant PLC in 8 M urea without the zinc led to complete solubilization of the inclusion body. Refolding the recombinant PLC comprised three phases: gradual removal of urea, addition of Zn2/, and cleavage of the signal peptide. Urea was gradually removed by dialysis in steps (from 8 to 4 to 1.5 to 0.3 to 0 M urea). ZnSO4 (0.1 mM) was added when the urea

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

concentration was 4 M; the three Zn2/ in the active site of PLC crosslink widely different parts of the enzyme, and stabilize the tertiary structure of the protein. Table 1 summarizes the effect of each refolding step on the specific activity of the refolded PLC. In the presence of Zn2/, a slight PLC activity was detected prior to removal of the signal peptide in 4 M urea. Addition of Zn2/ when the urea concentration was 8 M caused massive reaggregation of the protein. To generate the active protein, the N-terminal extension was removed with a protease, clostripain, added when the urea concentration was reduced to 1.5 M. This enzyme has specificity similar to that of trypsin, but shows marked preference for Trp residues toward the carboxy terminus of the substrate enzyme immediately adjacent to the major arginyl specificity site (22). Addition of clostripain at urea concentrations higher than 1.5 M caused cleavage of PLC at other Arg sites. Presumably this occurred because the enzyme was ‘‘floppy’’ and partially unfolded in higher urea with those other sites exposed. At lower urea concentration, those secondary sites became inaccessible to clostripain. Clostripain specifically cleaved the signal peptide as confirmed by N-terminal sequencing and likewise by the activity of the folded recombinant PLC. If the signal peptide was not removed before the urea concentration was below 1.3 M, precipitation occurred. This behavior suggests that the enzyme does not tolerate the N-terminal extension. The activity of recombinant PLC after removal of the signal peptide dramatically increased even in 1.5 M urea (Table 1). After incubation with clostripain, the PLC/protease/urea mixture was heated to 507C. Clostripain is TABLE 1

Purification of Recombinant Phospholipase C

Purification step

Total proteinb (mg)

Protein recovery (%)

Specific activityc (%)

Washed inclusion body 8 M urea 4 M urea 4 M urea with zince 1.5 M urea 1.5 M urea after clostripain 0.3 M urea Q-Sepharose Fast Flow

80d 4.0 2.5 2.5 2.5 2.0 2.0 1.2

— 100 63 63 63 50 50 30

— ú1 1 10 20 90 100 100

a

a

For cells grown in 50 ml medium. Protein concentration was determined using the Bradford assay with PLC as standard unless otherwise noted. c Relative specific activity using 5 mM diC7PC as the substrate. d Weight of the wet inclusion body protein after sonication, treatment with lysozyme, and washing twice. At this point PLC constituted about 80% of the total cell protein based on SDS–PAGE. The weight of the wet crude cell pellet was about 400 mg. e 0.1 mM ZnSO4 was added starting from this point of the refolding procedure. b

pepa

AP: PEP

370

TAN, HEHIR, AND ROBERTS TABLE 2

Michaelis–Menten Parameters of Native, Wild-Type, and Mutant Phospholipase Ca DiC6PC (monomer)

Enzyme

DiC7PC (micelle)

Vmax Km Vmax Kmb (mmol min01 mg01) (mM) (mmol min01 mg01) (mM)

Native PLC Recombinant N134D N134A

740 860 630 1140

0.20 0.19 0.15 0.30

1960 1750 1050 2160

0.02 0.03 —c —c

a Measured by pH-stat assays. Monomeric diC6PC ranged in concentration from 0.05 to 5 mM. Micellar diC7PC ranged from 1.5 to 5 mM. Rates were determined in duplicate or triplicate. Errors in Km for monomeric substrate are less than 20%. Apparent Km and Vmax were derived for micelles assuming Michaelis–Menten behavior and a phase separation model for micellization (6). While this may not be strictly correct, it yields parameters that may be compared for monomeric and micellar PC. Native PLC is the enzyme purified from Bacillus cereus; the recombinant enzyme is the cloned wild-type PLC. b Errors in extrapolating the low apparent Km for micelles are relatively large; e.g., the Km for diC7PC with native PLC is 0.02 { 0.01 mM . c Apparent Km values for diC7PC micelles are õ0.05 mM but the errors are too large for an accurate estimate.

inactivated by heating at this temperature (23), whereas PLC is unaffected (24). A small amount of precipitate, which could be due to denaturation of contaminating proteins, was formed on heating. This was easily removed by centrifugation or filtration. Solubilization of recombinant PLC was optimal at room temperature. On the other hand, refolding is a gradual process and has to be done at low temperature to avoid reaggregation. It is possible that the enzyme has multiple folding pathways or several folding intermediates. Some must slowly rearrange to proceed to the native form; and some get trapped in kinetic dead ends at room temperature, unable to unfold to sample more productive pathways. During dialysis to remove the denaturant, some of the PLC (Ç50%) precipitated out of solution; however, the resulting precipitate could be recycled through the folding process, to obtain more soluble PLC. The choice of the buffer during the refolding was also important. Using Tris buffer (even at low concentrations) caused aggregation of recombinant PLC. Tris has been shown to bind to the active site of PLC (9). Apparently, when it binds to partially folded protein it alters the final conformation such that the protein forms large aggregates. Purification of Recombinant PLC After refolding of the enzyme, the only major contaminant was clostripain. The isoelectric pH of clostripain

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

is 4.8 (23), whereas PLC has a pI range of 7 to 8.5 (25). Elution of the mixture through an anion-exchange column easily separated PLC from the contaminant (Fig. 2, lane 4). A gradient of sodium acetate was used as the eluant buffer because PLC is inhibited by several anions (halides and oxyanions) but not acetate (26). The pH of the column buffer was optimized to give the best elution profile of PLC without precipitating Zn(OH)2 . The typical yield of pure recombinant PLC was 30– 40 mg active protein per liter of medium. N-terminal sequencing and subsequent HPLC analysis were used to verify that the first three residues of the active recombinant protein were Trp–Ser–Ala. Kinetic Characterization of Recombinant PLC The action of recombinant PLC toward monomeric (diC6PC) and micellar (diC7PC) short-chain phosphatidylcholine substrates was investigated. Michaelis– Menten kinetic parameters for native and recombinant PLC are summarized in Table 2. Recombinant PLC exhibited interfacial activation (increase in Vmax and decrease in Km) similar to native PLC from B. cereus. The measured Km with respect to monomeric diC6PC substrate was 0.19 mM compared with an apparent Km of 0.03 mM for micellar diC7PC. For PLC isolated from the extracellular medium of B. cereus, the Km for diC6PC is 0.2 mM and the apparent Km for diC7PC is 0.02 mM. Other assay systems, notably vesicles and detergent mixed micelles, were also monitored (via 31P NMR spectroscopy) to compare recombinant and native PLC activity (Table 3). Small unilamellar vesicles of diC14PC TABLE 3

Specific Activity of Recombinant and Native Phospholipase C toward Phospholipids in Various Aggregation Statesa

Substrate (5 mM)

Native PLC

Recombinant PLC

DiC6PCb DiC14PCc DiC14PC/TX-100d SPM (30 mM)c SPM/TX-100d

730 0.6 102 3.7 1.9

620 0.5 150 2.4 3.3

a Specific activity is reported in mmol min01 mg01 as measured by P NMR; conditions are 50 mM Tris–acetate, pH 8.0, 307C. Native PLC refers to PLC purified from Bacillus cereus. Errors in specific activities for micellar systems are about 15%; for sphingomyelin vesicles or Triton X-100 mixed micelles, the low specific activities have a 50% error. b The CMC of this lipid in the assay conditions is 11 mM so that the substrate is monomeric (6). c Small unilamellar vesicles prepared by sonication. d The lipid is solubilized with 25 mM Triton X-100 to form a mixedmicelle substrate aggregate. 31

pepa

AP: PEP

OVEREXPRESSION AND PURIFICATION OF NONSPECIFIC PHOSPHOLIPASE C

were relatively poor substrates for both PLC enzymes. Solubilizing the long-chain PC in Triton X-100 mixed micelles dramatically enhanced PLC activity for both native and recombinant PLC. The net result of these kinetic comparisons is that the recombinant and native PLC exhibited identical kinetic properties. The recombinant PLC appears to be more stable than native PLC. Native PLC undergoes proteolytic degradation over a period of weeks (as visualized by SDS– PAGE) when stored at either 4 or 0207C (Fig. 2, lane 2). This degradation is paralleled by diminishing activity of the enzyme toward phospholipid substrates. It is possible that this is caused by a small amount of contaminating proteases in the enzyme preparation from B. cereus. Recombinant PLC, on the other hand, does not show any degradation or decrease in activity when stored at 47C for 8 months. Sphingomyelinase Activity of Recombinant PLC SPM is a phospholipid with a choline headgroup and an amide instead of an ester group in what corresponds to the sn-2 chain. Native PLC isolated from B. cereus appears to hydrolyze SPM either in vesicles or in mixed micelles at a very slow rate (1-3 mmol min01 mg01). Massing and Eibl reported that PLC hydrolyses SPM 200-fold lower than SPMase does; however, B. cereus secretes SPMase as well as PLC into the growth medium so that commercial PLC is always contaminated with SPMase (27). Interestingly, the gene encoding PLC is adjacent to the SPMase gene in the B. cereus genome. Thus, it is difficult to ascertain if PLC isolated from the growth medium of B. cereus is responsible for the low SPMase activity observed or if a contaminating SPMase is the active agent. The activity of recombinant PLC was examined using SPM vesicles and SPM solubilized in Triton X-100 mixed micelles (Table 3). The similarity in specific activity of both native and recombinant enzyme clearly indicated that, indeed, PLC is capable of hydrolyzing SPM. One of the more noteworthy features of SPM hydrolysis is that unlike PC (or other phospholipids), it is not enhanced by solubilizing the SPM in detergent micelles.

371

of the denaturation/refolding/proteolysis procedure for generating recombinant PLC. Asn-134 was mutated to Asp and Ala and the protein purified as for the wildtype recombinant material. Kinetic analyses of the mutant enzymes toward diC6PC and diC7PC are summarized in Table 2. The Michaelis–Menten parameters of the wild-type and mutant proteins are similar. Interestingly, Asn-134 r Ala seems to be a more active protein than the wild type, judging from the increase in Vmax toward monomeric and micellar substrates. Preliminary studies suggest that the refolding protocol developed for recombinant PLC will also be useful for purification of mutant PLC in which residues involved in Zn/2 coordination or residues involved in substrate binding are altered. For instance, mutations on a zinc ligand (Glu-146) and a residue that interacts with the substrate headgroup (Glu-4), produced mutant proteins whose yields of the refolded protein were 50–80% that of the wild type (C. Tan and M. Roberts, unpublished results). Glu-146 is a zinc ligand originally proposed to activate the water nucleophile based on crystal structure analysis; computer modeling later suggested it was not important for catalysis (11,12). E146Q has a Vmax about 1000-fold less and a Km 10fold higher than those of the wild type. The E4A mutant showed a twofold decrease in Vmax and an about 100fold increase in Km (C. Tan and M. Roberts, unpublished results). More detailed kinetic and structural studies of these two mutants are currently being pursued. In summary, a new method of cloning and overexpressing B. cereus PLC in E. coli has been developed. The recombinant protein exhibits kinetic parameters for PC hydrolysis identical to those for native PLC isolated from B. cereus. Since the recombinant enzyme is capable of catalyzing the hydrolysis of SPM amide bond, sphingomyelinase activity is an inherent property of the enzyme. ACKNOWLEDGMENTS We thank Professor Evan R. Kantrowitz (Department of Chemistry, Boston College) for the use of instruments and advice and Ms. Tina M. Lusignolo for doing the initial studies on cloning PLC. This work has been supported by NIH GM 26762 and NSF DMB-9023617.

Site-Specific Mutagenesis of Recombinant PLC The cloning and expression method that we developed can be used to explore PLC by site-specific mutagenesis. By use of the Quickchange method (28) sitedirected mutagenesis on PLC was performed with a mutagenesis efficiency of 100%. The initial target for mutagenesis was Asn-134, whose backbone amide was proposed to interact with the sn-2 carbonyl of the phosphonolipid inhibitor (11). Modification of the side-chain residue should not affect the enzyme activity dramatically; however, it is a test of the general effectiveness

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

REFERENCES 1. Exton, J. H. (1990) Signalling through phosphatidylcholine breakdown. J. Biol. Chem. 265, 1–4. 2. Pelech, S. L., and Vance, D. E. (1989) Signal transduction via phosphatidylcholine cycles. Trends Biochem. Sci. 14, 28–30. 3. Hough, E. (1991) The active site and catalytic mechanism in phospholipase C from Bacillus cereus, in ‘‘Site-Directed Mutagenesis and Protein Engineering’’ (El-Gewely, M., Ed.), pp. 13– 23, Elsevier Science, London. 4. Levine, L., Xiao, D., and Little, C. (1988) Increased arachidonic acid metabolites from cells in culture after treatment with the

pepa

AP: PEP

372

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

TAN, HEHIR, AND ROBERTS

phosphatidylcholine-hydrolyzing phospholipase C from Bacillus cereus. Prostaglandins 34, 633–642. Clark, M., Shorr, R., and Bomalski, J. (1986) Antibodies prepared to Bacillus cereus phospholipase C cross-react with phosphatidylcholine-preferring phospholipase C in mammalian cells. Biochem. Biophys. Res. Commun. 140, 114–119. El-Sayed, M., DeBose, C. D., Coury, L. A., and Roberts, M. F. (1985) Sensitivity of phospholipase C (Bacillus cereus) activity to phosphatidylcholine structural modifications. Biochim. Biophys. Acta 837, 325–335. Tan, C. A., and Roberts, M. F. (1996) Vanadate is a potent competitive inhibitor of phospholipase C from Bacillus cereus. Biochim. Biophys. Acta 1298, 58–68. Johansen, T., Holm, T., Guddal, P. H., Sletten, K., Haugli, F. B., and Little, C. (1988) Cloning and sequencing of the gene encoding the phosphatidylcholine-preferring phospholipase C of Bacillus cereus. Gene 65, 293–304. Hansen, S., Hansen, L. K., and Hough, E. (1993) The crystal structure of Tris-inhibited phospholipase C from Bacillus cereus ˚ resolution. J. Mol. Biol. 231, 870–876. at 1.9 A Hansen, S., Hansen, L. K., and Hough, E. (1992) Crystal structures of phosphate, iodide and iodate-inhibited phospholipase C from Bacillus cereus and structural investigations of the binding of reaction products and a substrate analogue. J. Mol. Biol. 225, 543–549. Hansen, S., Hough, E., Svenson, L. A., Wong, Y.-L., and Martin, S. F. (1993) Crystal structure of phospholipase C from Bacillus cereus complexed with a substrate analog. J. Mol. Biol. 234, 179–187. Sundell, S., Hansen, S., and Hough, E. (1994) A proposal for the catalytic mechanism in phospholipase C based on interaction energy and distance geometry calculation. Protein Eng. 7, 571– 577. El-Sayed, M. Y., and Roberts, M. F. (1985) Charged detergents enhance the activity of phospholipase C (Bacillus cereus) toward micellar short-chain phosphatidylcholine. Biochim. Biophys. Acta 831, 133–141. Studier, F. W., and Moffatt, B. (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113–130.

AID

PEP 0756

/

6q13$$$221

07-10-97 01:17:41

15. Little, C., Aurebekk, B., and Otnaess, A. (1975) Purification by affinity chromatography of phospholipase C from Bacillus cereus. FEBS Lett. 52, 175–179. 16. Ausubel, F. M. (1987) ‘‘Current Protocols in Molecular Biology,’’ Greene/Wiley–Interscience, New York. 17. Little, C. (1978) Conformational studies on phospholipase C from Bacillus cereus: The effect of urea on the enzyme. Biochem. J. 175, 977–986. 18. Kolsto, A.-B., Gronstad, A., and Oppegaard, H. (1990) Physical map of the Bacillus cereus chromosome. J. Bacteriol. 172, 3821– 3825. 19. Skaliter, R., Bergstein, M., and Livneh, Z. (1996) b*, a UV-inducible shorter form of the b subunit of DNA polymerase III of Escherichia coli. J. Biol. Chem. 271, 2491–2496. 20. Moore, J. T., Uppal, A., Maley, F., and Maley, G. F. (1993) Overcoming inclusion body formation in a high-level expression system. Protein Expression Purif. 4, 160–163. 21. Blackwell, J. R., and Horgan, R. (1991) A novel strategy for production of a highly expressed recombinant protein in an active form. FEBS Lett. 295, 10–12. 22. Mitchell, W. M. (1977) Cleavage at arginine residues by clostripain, in ‘‘Methods in Enzymology,’’ vol. 47, pp. 165–170, Academic Press, San Diego. 23. Mitchell, W. M., and Harrington, W. F. (1971) Clostripain, in ‘‘The Enzymes’’ (Boyer, P., Ed.), pp. 699–719, Academic Press, New York. 24. Otnaess, A.-B., Little, C., Sletten, K., Wallin, R., Johnsen, S., and Flengsrud, R. (1977) Some characteristics of phospholipase C from Bacillus cereus. Eur. J. Biochem. 79, 459–468. 25. Bjorklid, E., and Little, C. (1980) The isoelectric point of phospholipase C from Bacillus cereus. FEBS Lett. 113, 161–163. 26. Aakre, S., and Little, C. (1982) Inhibition of Bacillus cereus phospholipase C by univalent anions. Biochem. J. 203, 799–801. 27. Massing, U., and Eibl, H. (1994) Substrates for phospholipase C and sphingomyelinase from Bacillus cereus, in ‘‘Lipases’’ (Wooley, P., and Petersen, S., Eds.), pp. 225–241, Cambridge Univ. Press, Cambridge. 28. Papworth, C., Bauer, J. C., and Braman, J. (1996) Site-directed mutagenesis in one day with ú80% efficiency. Strategies 9, 3–4.

pepa

AP: PEP