Dimerization of human butyrylcholinesterase expressed in bacterium for development of a thermally stable bioscavenger of organophosphorus compounds

Abstract

Human butyrylcholinesterase (BChE) is widely distributed plasma enzyme. For decades, numerous research efforts have been directed at engineering BChE as a bioscavenger of organophosphorus insecticides and chemical warfare nerve agents. However, it has been a grand challenge to cost-efficiently produce BChE in large-scale. Recently reported studies have successfully designed a truncated BChE mutant (with amino-acid substitutions on 47 residues that are far away from the catalytic site), denoted as BChE-M47 for convenience, which can be expressed in E. coli without loss of its catalytic activity. In this study, we aimed to dimerize the truncated BChE mutant protein expressed in a prokaryotic system (E. coli) in order to further improve its thermal stability by introducing a pair of cross-subunit disulfide bonds to the BChE-M47 structure. Specifically, the E377C/A516C mutations were designed and introduced to BChE-M47, and the obtained new protein entity, denoted as BChE-M48, with a pair of cross-subunit disulfide bonds indeed exists as a dimer with significantly improved thermostability and unaltered catalytic activity and reactivity compared to BChE-M47. These results provide a new strategy for optimizing protein stability for production in a cost-efficient prokaryotic system. Our enzyme, BChE-M48, has a half-life of almost one week at a 37°C, suggesting that it could be utilized as a highly stable bioscavenger of OP insecticides and chemical warfare nerve agents.

Introduction

Butyrylcholinesterase (BChE, EC 3.1.1.8) is a relatively abundant nonspecific cholinesterase enzyme in plasma capable of hydrolyzing many different choline-based esters [1]. It is involved in various interesting physiological functions and drug development efforts, including the hydrolysis of acetylcholine to maintain the steady state of the central nervous system, the metabolism of cocaine [2], as a treatment for obesity through the hydrolysis of the hunger hormone ghrelin [3, 4], a therapeutic for Alzheimer’s disease [5], and the detoxification of organophosphorus (OP) esters including various nerve agents and pesticides [6, 7]. In its active site is a nucleophilic serine residue (Ser198) that rapidly reacts with OP esters to form an irreversible covalent bond, simultaneously inhibiting BChE and destroying the OP poison [8]. BChE functions as a stoichiometric scavenger that rapidly and irreversibly sequesters OP compounds. However, only one OP compound can bind to each BChE molecule, which has a high molecular weight. So, in order for this to be an effective treatment, there must be relatively high levels of BChE in circulation to inactivate the same molar amount of the OP agent [9]. Even so, the administration of BChE as a prophylactic treatment for OP detoxification by a number of neurotoxic OP nerve agents is attractive due to the enzyme’s rapid reaction rate with a variety of OP compounds [8]. In fact, when administered as a pretreatment, injection of BChE resulted in an increased chance of survival against poisonous nerve agents in mice, rats [10, 11], guinea pigs [12], and monkeys [13, 14]. However, before prophylactic treatment with high quantities of exogenous BChE can be implemented, a highly pure enzyme must be produced in order to prevent instigating an immune reaction or other side effects. Thus, the development of a high-quality enzyme preparation protocol to produce enzymes free of endotoxins and large quantities for bioscavenging in humans is of vital practical significance.

Two Phase I clinical trials (NCT00333528 and NCT00744146) were conducted to evaluate the safety of human BChE (hBChE), demonstrating that hBChE is safe for use in humans . In addition, multiple clinical trials [15, 16] were conducted to evaluate the safety of an hBChE mutant (known as CocH1 or E14-3) designed in our lab [17-19], demonstrating that the hBChE mutant (in albumin fusion formulation, known as Albu-BChE, Albu-CocH1, Albu-CocH, or TV-1380 in literature) [15, 16, 20-23] is also safe for use in humans. On the other hand, deriving hBChE from human plasma or transgenic animals as a therapeutic is expensive, making large-scale production for therapeutic uses not feasible. Thus, research efforts have been directed at expressing recombinant hBChE via cell culture, which is a more convenient and cost-efficient method [24, 25]. Native hBChE exists primarily as a tetrameric structure, consisting of four identical subunits and one polyproline-rich peptide source [26]. The C-terminus of hBChE features a tetramerization domain made up of 40 residues which non-covalently interacts with the polyproline-rich peptide to form a tetramer [26]. However, expression of recombinant full-length hBChE via cell culture produces low yield of the enzyme [27, 28] as a mixture of monomer, dimer, and tetramer. This can be rectified by deleting the tetramerization domain, which results in the production of pure monomer [29] and has been shown to induce higher levels of expression in Chinese hamster ovary (CHO-S) cells (increasing the production from ~1 mg/L to 3-5 mg/L) [27] and insect cells (increasing the production from ~1 mg/L to 4 mg/L) [28]. It has been a grand challenge to efficiently express hBChE via cell culture [30]. However, the absence of post-translational modifications results in an unstable enzyme with a short half-life and low catalytic activity and while this can be mitigated with post-purification modifications, they are expensive and make the cost of large-scale production unreasonable [30].

To decrease the cost of hBChE protein production, Escherichia coli (E. coli), which is a much cheaper protein expression system, has been tested for hBChE expression. However, initial attempts to express hBChE in E. coli were not successful, because it produced hBChE as non-foldable inclusion bodies, due to the presence of the disulfide bridges and N-glycosylation sites [31, 32]. The finding also implies that the known stability of native hBChE or hBChE expressed in mammalian cells is attributed to the post-translational modifications, particularly glycosylation associated with the well-known nine N-glycosylation sites [31, 32]. So, the absence of post-translational modifications in E. coli results in an unstable enzyme with a short half-life and low activity.

Recently, catalytically active and stable human acetylcholinesterase (hAChE) and hBChE were successfully expressed in Escherichia coli (E. coli) for the first time using an automated structure- and sequence-based algorithm Protein Repair One Stop Shop (PROSS) [32, 33] to design thermally stable mutants. The new BChE construct with the maximal catalytic activity (known as hBChE-7) [32] had an optimized codon bias for expression in E. coli and contained 47 mutations [32]. For convenience, the BChE protein entity (hBChE-7) with the 47 mutations is denoted as BChE-M47 throughout the rest of this report. BChE-M47 was truncated after residue 529, deleting 45 amino acids off the C-terminus. Despite maintaining the dimer interface (without any amino-acid residue substitution on the interface) using the PROSS process, the oligomeric state characterization revealed that BChE-M47 was actually a monomer rather than a dimer [32]. While this monomer is catalytically active, expressing BChE with a higher thermostability in a prokaryotic system could contribute to the development of a more effective bioscavenger.

In the present study, we aimed to dimerize the truncated BChE protein expressed in a prokaryotic system (E. coli) in order to further improve its thermal stability by introducing cross-subunit disulfide bonds to BChE-M47, a similar strategy which was used in our previously reported studies to dimerize a highly efficient cocaine hydrolase (CocH) [34] or a mutated cocaine esterase (CocE) [35].

Based on our previous work on the CocH dimerization [34] and further computational modeling (see below), we added the E377C/A516C mutations to BChE-M47, which are expected to introduce a pair of cross-subunit disulfide bonds (C377a-C516b and C516a-C377b) and produce a stable dimer. Note that V377 of wild-type BChE was changed to E377 in BChE-M47. So, compared to wild-type BChE (hBChE), there are a total of 48 mutations in our newly designed E377C/A516C mutant of BChE-M47, denoted as BChE-M48 for convenience. BChE-M47 and BChE-M48 were prepared for the characterization of their oligomeric forms, catalytic activity against butyrylthiocholine iodide (BTC) and acetylthiocholine iodide (ATC), thermostability at a physiological temperature (37°C), and the reactivity with paraoxon. The results of this study demonstrate that the formation of desirable cross-subunit disulfide bonds was successful, introduced a stable dimer structure to the enzyme, and significantly improved the enzyme’s thermal stability without significantly changing its functions.

Materials and Methods

Materials

The BChE-M47 gene sequence was synthesized by GenScript Corporation (Piscataway, NJ). The Q5^® Hot Start High-Fidelity 2X Master Mix, KLD Enzyme Mix, and SHuffle^® T7 Express Competent E. coli were ordered from New England Biolabs (Ipswich, MA). All oligonucleotides were synthesized by Eurofins MWG Operon (Huntsville, AL). The Amicon^® Ultra-15 Centrifugal Filter Unit was from MilliporeSigma (Massachusetts, United States). ATC, BTC and Enterokinase Cleavage Capture Kit were purchased from Sigma-Aldrich (St. Louis, MO). The GeneJET Plasmid Miniprep Kit, the HisPur™ Cobalt Resin, NativePAGE 4-16% Bis-Tris Protein Gel, Novex 4-12% Tris-Glycine Mini Protein Gel, SimpleBlue Safe Stain were purchased from Thermo Fisher Scientific (Waltham, MA).

Molecular dynamics simulation

Based on the BChE crystal structure (PDB ID: 6EMI) report [32], we modeled BChE-M47 dimer with the contacting interface of E377 and A516 residues. Using the PyMol software [36], E377 and A516 residues in each subunit were changed to C377 and C516, respectively, to obtain the initial structure of BChE-M48. The general procedure for the molecular dynamics (MD) simulation of each dimer structure was similar to that used in our previously reported computational study [34]. In brief, all molecular mechanics (MM)-based energy minimization calculations and MD simulations were performed using the AMBER 16 program package [37]. The AMBER ff14sb force field was used to establish the potentials of the proteins [38]. Fifteen sodium ions were used to neutralize the system for each protein. The system was then immersed in an orthorhombic box of TIP3P water molecules with a minimum solute-wall distance of 8 Å. The whole system was then equilibrated and fully energy-minimized to obtain the dimer structure with two cross-subunit disulfide bonds (C377a-C516b and C516a-C377b). Technically, we performed the MD simulation of the system without the disulfide bonds first. When the residues were close enough, we mutated them into Cys and did the energy-minimization. Subsequently, the system was gradually heated in the isothermal-isobaric (NPT) ensemble from 10 K to 300 K over 100 ps. The MD simulation was performed under a typically used temperature (300 K) for 50 ns in the production run. The Particle Mesh Ewald (PME) method was adopted to treat the long-range electrostatic interactions [39]. The SHAKE procedure was applied to constrain the bond lengths of all covalent bonds involving hydrogen atoms, with a 2 fs time step [40]. The atomic coordinates were saved every 10 ps for the productive sampling and analysis.

Construction for E. coli expression

The BChE-M47 gene was subcloned into the bacterial expression vector, pET-32b (+), between EcoRI and XhoI sites for the expression of thioredoxin-BChE-M47-6xHis (Trx-BChE-M47), which allows the protein to be purified by HisPur™ Cobalt Resin. Using the plasmid pET-32b (+)-BChE-M47 plasmid as the template and primers (listed in Table 1) with specific base-pair alterations, mutations were induced by Q5^® Hot Start High-Fidelity 2X Master Mix DNA polymerase to produce pET-32b (+)-BChE-M48. The polymerase chain reaction (PCR) products were treated with KLD Enzyme Mix to induce phosphorylation by the kinase, ligation by the ligase, and digestion of the DNA template by the DpnI endonuclease. After treatment with KLD Enzyme Mix, the PCR products were transformed into Top-10 competent cells to amplify the plasmids, before extracting the mutant DNA with the GeneJET Plasmid Miniprep Kit. DNA sequencing confirmed that expression of pET-32b (+)-BChE-M48 was successful.

Table 1.

Oligonucleotides for mutagenesis

Sequence Name	Sequence
E377C-F	5’-T ACC GAT TGG TGT GAT GAA GAT CGT C-3’
E377C-R	5’-TA ATG AAA CAG GAT GCT TTC-3’
A516C-F	5’-C AAA CTG CGT TGT CAG CAT TGT CGT TTT TG-3’
A516C-R	5’-GT CAT AAT ACG GCT GCT TTC-3’

Protein expression and purification

Both pET-32b (+)-BChE-M47 and the newly mutated pET-32b (+)-BChE-M48 were transformed into SHuffle^® T7 Express Competent E. coli for expression. Clones of each construct were cultured overnight in 5 mL of Luria-Bertani (LB) medium with 100 ug/mL of ampicillin at 30°C, which acted as a starter culture. Each starter culture was then transferred into one liter of LB medium 100 ug/mL ampicillin and incubated at 30°C until the OD₆₀₀ reached 0.6. Expression was induced with 0.4mM isopropyl β-D-Thiogalactoside (IPTG) and 24 hours of incubation at 18°C. Cells were pelleted by centrifugation at 4,000 rpm and resuspended in 100 mL of 50 mM Tris-HCl pH 7.5, 300 mM NaCl, 15 mM Imidazole (Buffer A). The total soluble extracts were prepared by French Press (FA032, Thermo Fisher Scientific) cell homogenization and centrifugation at 25,000rpm for 30 min at 4°C. The entire following purification process was conducted in a cold room at 4°C. The HisPur™ Cobalt Resin was pre-equilibrated with Buffer A before the total soluble extracts were loaded into the column with a flow rate of 1-2 mL/min for the binding phase. After binding, the resin was extensively washed with Buffer A until the OD₂₈₀ of the flow-through was less than 0.01. The protein was eluted with 50 mM Tris-HCl pH 7.5, 300 mM NaCl, 300 mM Imidazole. The eluate was concentrated and dialyzed in 100mM phosphate buffer (PB) pH 7.4. Utilized the Enterokinase Cleavage Capture Kit to cleave the N-terminus fusion part and obtained the tag-free BChE-M47 and BChE-M48. The concentration of the active enzyme was determined through active site titration with paraoxon as previously reported [41].

Native gel stained for enzyme activity

Electrophoresis was conducted at 4°C. The blue native polyacrylamide gel was run at a voltage of 150 V for one hour, then increased to 250 V for additional two hours. The gel was stained at room temperature with 2 mM butyrylthiocholine iodide as the substrate to test for BChE activity per the Karnovsky and Roots staining procedure [42]. Photographs were taken and analyzed by Image lab software from Bio-Rad (Hercules, CA) to assess the relative protein band proportions.

BChE activity assays

BChE has characteristic allosteric effects due to a peripheral anionic binding site around D70[43]. When the concentration of substrates is very high, a second substrate molecule can bind to that site of BChE, resulting in either an increase or decrease in catalytic activity [43]. Here we determined the catalytic activities of BChE-M47 and BChE-M48 against high concentrations of ATC and BTC at the same time under the same experimental conditions at 25°C with the modified Ellman’s method [44] in the GENios Pro Microplate Reader with XFluor software (TECAN, Research Triangle Park, NC). 100 μL of enzyme solution (100 mM PB pH 7.4) was mixed with 50 μL of 20 mM dithiobisnitrobenzoic acid and 50 μL of either BTC or ATC in varying concentrations (0.01mM to 20mM). Reaction rates were determined by recording the time-dependent absorption at 450 nm, then performing non-linear, least-squares fitting to Eq. (1) reported by Radic [45], plotted the data using GraphPad Prism 8 software (San Diego, CA).

$$V=\frac{V_{\max }(1+bS∕K_{\text{SS}})}{(1+K_{\mathrm{M}}∕S)(1+S∕K_{\text{SS}})}$$ (1)

where S represents the concentration of the substrate, K_ss is the binding constant for the substrate at the secondary binding site, and b is a factor reflecting whether the substrate has either activating or inhibiting effects. When b = 1, the enzymatic reaction follows Michaelis-Menten kinetics, without substrate activation or inhibition. When b > 1, the substrate activates catalytic activity and when b < 1, the substrate inhibits catalytic activity. Kinetic data were analyzed with Microsoft Excel.

In vitro measurements of thermostability

BChE-M47 and BChE-M48 samples were assayed in 100 mM PB pH 7.4, 10mg/mL bovine serum albumin (BSA) at a concentration of 10nM, then incubated at 37°C. The remaining activity was determined at 10 minutes, 1 hour (hr), 2 hr, 4 hr, 6 hr, 8 hr, 12 hr, 24 hr, then on a daily basis. The percentage of remaining activity was plotted against time to calculate the half-life (t_1/2) with GraphPad Prism 8 software (San Diego, CA).

Inhibition kinetics by paraoxon

Proteins hBChE, which we previously expressed in suspension Chinese hamster ovary (CHO-S) cells, BChE-M47, and BChE-M48 were assayed in 100 mM PB pH 7.4, 1 mg/mL BSA, mixed with 2.6 nM paraoxon in equal volumes. The samples were incubated at room temperature and the remaining active enzyme concentrations were determined at 0 min, 30 min, 1 hr, and 2 hr using the Ellman’s method. The pseudo-first-order rate constant was determined from the slope of the semi-logarithmic plot of residual enzyme concentration against time (Figure 4). The rate constant of inhibition was calculated by dividing the pseudo-first-order rate constant by the paraoxon concentration.

Figure 4.

The in vitro thermostability of BChE-M47 and BChE-M48 at 37°C shown as a 5-day time course of BChE-M47 and a 20-day time course of BChE-M48 of their remaining enzymatic activity. The half-life for each enzyme were calculated by determining the catalytic efficiency at different time points. The BChE-M47 is expressed as a pure monomer, so the data exhibited a one-phase exponential decay, demonstrating a half-life of 23.5±1.4 hours. The BChE-M48 consists of a mixture of monomer and dimer, so the data exhibited a two-phase exponential decay (t_1/2fast 23.8±3.9 hours and t_1/2slow 156.1±8.7 hours). The half-life of the BChE-M48 monomer (t_1/2fast 23.8±3.9 hours) is consistent with the BCHE-M47 monomer, so the extended half-life (t_1/2slow 156.1±8.7 hours) is a result of protein dimerization.

Results and Discussion

Insight from Molecular modeling

Depicted in Figure 1 are some interesting results of the MD simulations. According to the MD simulations, for both BChE-M47 and BChE-M48, each subunit (i.e. each monomer of the dimer) had a dynamically stable structure, with RMSD ≤ ~2.0 Å, during the MD simulation. In comparison, the RMSD for the whole dimer was much larger, due to the structural flexibility of the dimer interface. Interestingly, as shown in Figure 1A, the BChE-M48 dimer structure shows a relatively smaller root-mean-square deviation (RMSD) during the 50-ns MD simulation compared to the corresponding RMSD for the MD simulation of the BChE-M47 dimer structure, indicating that the formation of the disulfide bonds will likely improve the thermostability of the dimer. As shown in the energy-minimized structure of the BChE-M48 dimer (Figure 1B), the pair of cross-subunit disulfide bonds can be readily formed by C377a-C516b and C516a-C377b on the dimer interface. The mutations producing C377 and C516 were not in proximity to the catalytic site, thus formation of the dimer should not interfere with the catalytic activity and reactivity of BChE.

Figure 1.

(A) The time-dependent root-mean-square deviations (RMSD) of the MD-simulated BChE-M47 and BChE-M48 dimer structures to the corresponding energy-minimized structures. The RMSD refers to all α-carbon atoms in the BChE-M47 dimer (black, average RMSD = 5.13 Å), BChE-M48 dimer (blue, average RMSD = 4.32 Å), the two BChE-M48 monomers are showed in red (average RMSD = 1.73 Å) and green (average RMSD = 1.61 Å). (B) Energy-minimized structure of the BChE-M48 dimer. The protein is shown as green cartoon, all the 48 amino-acid substitutions are shown as yellow sticks. The van der Waals surface of the catalytic site (including G116, G117, S198, S199, E325, H438, and other residues within 4 Å of these six catalytic residues) of the protein is colored by atom types (carbon in green, oxygen in red, and nitrogen in blue). A pair of cross-subunit disulfide bonds (C377a-C516b and C377b-C516a) help to stabilize the dimer and, thus, improve the thermostability of the protein.

Oligomeric forms of the proteins

Figure 2 is the 4-16% native polyacrylamide gel stained for the BChE activity. BChE-M48 existed as a mixture of a dimer (85%) and a monomer (15%), whereas BChE-M47 and hBChE existed exclusively as a monomer. This suggests that the pair of cross-subunit disulfide bonds (C377a-C516b and C516a-C377b) in BChE-M48 were formed successfully and the presence of this pair of disulfide bonds remarkably increased the proportion of dimer to monomer, as compared to BChE-M47 and hBChE (wild-type BChE).

Figure 2.

Blue native polyacrylamide gel stained for the BChE activity of BChE-M47, BChE-M48, and human BChE (hBChE) expressed in CHO-S.

Catalytic activity against ATC and BTC

Figure 3 demonstrates the substrate activation of BChE-M47 and BChE-M48 with ATC and BTC, Table 2 summarizes the kinetic parameters. As shown in Table 2, when reacted with BTC, both BChE-M47 and BChE-M48 had larger k_cat (35580 min⁻¹ and 33440 min⁻¹, respectively) and larger K_M values (24.8 μM and 28.5 μM, respectively) than hBChE, but slightly smaller relative catalytic efficiency (k_cat/K_M) than hBChE. Surprisingly, when BChE-M47 and BChE-M48 were reacted with ATC, both enzymes had smaller k_cat values (19960 min⁻¹ and 17270 min⁻¹, respectively), smaller K_M value (38.5 μM and 41.9 μM, respectively), and larger relative catalytic efficiency than hBChE.

Figure 3.

In vitro (100 mM phosphate buffer pH 7.4, 25°C) kinetic data obtained for BTC (A) and ATC (B) hydrolysis by BChE-M47 and BChE-M48. The reaction rates are shown in μM min⁻¹ per nM of enzymes.

Table 2.

Kinetic parameters determined for ATC and BTC

Enzyme	Substrate	kcat (min−1)	KM (μM)	kcat/KM (min−1 M−1)	RCEc	Kss (μM)	b
hBChE-7a	BTC	46715	30.0 ± 2.5	1.56 × 109	0.74	1291	2.80
hBChE b	BTC	29500 ± 1100	14.0 ± 1.8	2.11 × 109	1.00	3010d	3.36
BChE-M47	BTC	355 ± 1300	24.8 ± 3.3	1.43 × 109	0.68	4983	2.99
BChE-M48	BTC	334 ± 1030	28.5 ± 3.1	1.17 × 109	0.55	6760	3.11
hBChE b	ATC	20200 ± 910	57.0 ± 6.4	3.54 × 108	1.00	2890	2.47
BChE-M47	ATC	19960 ± 830	38.5 ± 4.5	5.18 × 108	1.46	5000	3.03
BChE-M48	ATC	17270 ± 640	41.9 ±5.3	4.12 × 108	1.16	4310	2.62

^aThe k_cat, K_M, K_ss, and b for hBChE-7 against BTC came from reference [32].

^bThe k_cat and K_M for BChE expressed in CHO cells (hBChE) against BTC and ATC, K_ss and b for hBChE against ATC came from in reference [49].

^cRCE refers to the relative catalytic efficiency (k_cat/K_M), i.e. the ratio of (k_cat/K_M) of the different BChE mutants against the same substrate.

^dThe K_ss and b for hBChE against BTC came from reference[50].

In this study, all the determined b values were > 1, indicating that an additional ATC or BTC molecule binding at the peripheral anionic binding site of both BChE-M47 and BChE-M48 will increase the catalytic activity, which is consistent with the previous report about the substrate activation feature of BChE [46]. So the BChE expressed in bacterium retained the substrate activation seen in hBChE.

Thermostability of enzymes

We determined the half-lives in vitro at 37°C (physiological temperature). In order to minimize potential variability, we prepared the enzyme samples from stocks of each enzyme, which had both been expressed and purified at the same time and under the same experimental conditions. Based on the results from native polyacrylamide gel (shown in Figure 2), BChE-M47 consisted of pure monomer, whereas BChE-M48 mainly existed in dimer (85%), but also with some monomer (15%). Thus, BChE-M47 exhibits a single-phase exponential decay (t_1/2 = 23.5 ± 1.4 hours, Figure 4) and BChE-M48, as a mixture of monomer and dimer, exhibits a two-phase exponential decay (t_1/2fast = 23.8 ± 3.9 hours and t_1/2slow = 156.1 ± 8.7 hours, Figure 4). From this data, the half-life of the BChE-M48 monomer, or the fast phase, is consistent with the half-life seen for the BChE-M47 monomer. The dimer has a significantly improved half-life (156.1 ± 8.7 hours). Thus, the improved thermostability of BChE-M48 is due to the mutation-induced dimerization.

Reactivity with paraoxon

The rate constants obtained for the inhibition by paraoxon of hBChE, BChE-M47, and BChE-M48 were (1.8 ± 0.1) × 10⁶, (2.9 ± 0.2) × 10⁶, and (2.8 ± 0.1) × 10⁶ M⁻¹ min⁻¹, respectively. The reaction rate constant we obtained for hBChE with paraoxon was consistent with previously reported results for recombinant hBChE, which indicated an inhibition rate constant of (1.6 ± 0.4) × 10⁶ M⁻¹ min⁻¹ [47]. Both BChE-M47 and BChE-M48 had a ~1.6-fold higher inhibition reaction rate constant with paraoxon than that for recombinant hBChE (rhBChE) [47], indicating that BChE expressed in bacterium could be more effective in the detoxification of paraoxon. BChE-M47 and BChE-M48 have comparable inhibition rate constant values with paraoxon, indicating that the mutation-induced dimerization of BChE-M48 did not alter the reactivity of the enzyme. This was the first time the inhibition reaction rate constants of bacterial-expressed BChE mutants with paraoxon were determined. Thus, BChE-M48 was more thermally stable than BChE-M47, as well as reacted more rapidly with paraoxon than hBChE, resulting in a more rapid scavenging of paraoxon and making it a potentially better potential therapeutic.

Due to the poor thermostability of BChE-M47, we determined the rate constants at room temperature to get more reliable results, though determining rate constants at 37°C (physiological temperature) in whole blood or plasma samples would provide more clinically relevant information for OP poisoning. Worek et al [48] reported an optimized Ellman’s method, which could be applied in the further development of stoichiometric scavengers to obtain more medically relevant data.

Conclusion

In our present study, we have demonstrated that when the double mutations, E377C/A516C, are introduced to BChE-M47, the obtained new protein entity BChE-M48 has a pair of cross-subunit disulfide bonds formed between the two BChE subunits, producing a dimer with significantly improved thermostability and unaltered catalytic activity and reactivity compared to the monomer. These results provide a new strategy for optimizing protein stability for production in cost-efficient prokaryotic systems. Based on our success with the bacterial expression of a BChE dimer with an improved thermal stability and preserved catalytic activity and reactivity, one may also utilize this strategy in the future to rationally design other therapeutic proteins with improved thermostability, while maintaining their favorable functions. Our enzyme, BChE-M48, has a half-life of almost one week at 37°C, which suggests that it could be utilized as a highly stable bioscavenger of OP insecticides and chemical warfare nerve agents.

Figure 5.

Pseudo-first-order reactions of hBChE, BChE-M47, and BChE-M48 with paraoxon. The initial concentrations of hBChE, BChE-M47, BChE-M48, and paraoxon were 1.3, 0.76, 0.79, and 2.6 nM, respectively. The pseudo-first-order rate constant was determined from the slope of the semi-logarithmic plot of residual enzyme concentration versus time.

Highlights.

• We introduced a pair of cross-subunit disulfide bonds to a truncated BChE.

• The new BChE mutant with cross-subunit disulfide bonds exists as a dimer.

• The dimer has a significantly prolonged half-life (almost a week) at 37 °C.

• The dimerization does not change the catalytic activity and reactivity.