Alternative Cefazolin Synthesis with a Cephalosporin-Acid Synthetase

Two pathways of combined chemical and biocatalytic synthesis of the antibiotic cefazolin (CEZ) from 7-amino-cephalosporanic acid (7-ACA) with the immobilized recombinant cephalosporin-acid synthetase as the biocatalyst are compared. The first pathway involved chemical substitution with 2-mercapto-5-methylthiadiazole to modify the 3-acetoxy group in 7-ACA with subsequent biocatalytic acylation of the amino group of the product, 7-amino-3-[2-methyl-1,3,4-thiadiazol-5-yl)-thiomethyl]-3-cephem-4-carboxylic acid (TDA), with the methyl ester of 1(Н)-tetrazolylacetic acid. An alternative pathway involved biocatalytic acylation of the 7-ACA amino group to form an intermediate (S-p CEZ) that was chemically transformed into CEZ at the next step without isolation from the reaction mix. Analysis and optimization of each of the biocatalytic processes showed that 7-ACA acylation had a number of important advantages over TDA acylation with respect to the process yield, final concentration of the product in the reaction mix, and the tolerance of the process conditions with respect to enzyme activity and stability. Given the obvious environmental advantages of the process of chemical S-p CEZ transformation into CEZ over the process of TDA production from 7-ACA, we conclude that the second pathway of combined chemical and biocatalytic CEZ synthesis is preferable.

Keywords: cefazolin; biocatalytic synthesis; biocatalysis; cephalosporin-acid synthetase; solubility; 7-aminocephalosporanic acid

Cefazolin ((6R,7R)-3-[(5-methyl-1,3,4-thiadiazol-2-yl)thiomethyl]-8-oxo-7-[(1H-tetrazol-1-yl)acetylamino]-5-thia-1-azabicyclo[4.2.0]oct-2-en-2-carboxylic acid, CEZ) is among the most important representatives of the class of cephalosporin-acids, which includes more than fifteen semisynthetic parenteral β‑lactam antibiotics [[1]]. This antibiotic is included in the list of basic pharmaceuticals recommended by the World Health Organization.

The current volume of CEZ production worldwide is approximately 1000 metric tons per year (calculated for free acid) [[3]]; its production bases on chemical synthesis. The precursor 3-(acetyloxomethyl)-7-amino-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-en-2-carboxylic acid (7-amino-cephalosporanic acid, 7-ACA) is transformed into 7-amino-3-[2-methyl-(1,3,4-thiadiazol-5-yl)-thiomethyl]-3-cephem-4-carboxylic acid (TDA) via substitution of the 3-acetoxy group mediated by 2-mercapto-5-methylthiadiazole (MMTD) in the presence of strong acids in a nonaqueous medium [[4]]. The TDA amino group is subsequently acylated by activated derivatives of 1(Н)-tetrazolylacetic acid (TzAA) in a nonaqueous or a water-organic solvent medium at a low temperature (–40°С) via the protection of the carboxyl group [[6]].

Transformation catalyzed by enzymes in the form of heterogenous biocatalysts is a promising alternative to the chemical synthesis of β-lactam antibiotics, including CEZ and other cephalosporin-acids. The use of biocatalytic technologies enables a decrease in the environmental load, as well as the production of chemicals of a higher purity [[2], [8]–[13]].

The combined chemical and biocatalytic synthesis of CEZ from 7-ACA can be implemented in two ways (Fig. 1). In the first case, conventional chemical synthesis is used to transform 7-ACA into TDA (Fig. 1, transformation 1), and the TDA amino group is then subjected to biocatalytic acylation to form CEZ (Fig. 1, transformation 2). The alternative pathway involves biocatalytic acylation of the amino group in 7-ACA (Fig. 1, transformation 3) with the subsequent chemical transformation of the obtained CEZ derivative (S-p СЕZ) into the target antibiotic (Fig. 1, transformation 4). An activated TzAA derivative, namely, its methyl ester (METzAA), is the acylating agent in both cases, i.e., synthesis via acyl transfer (synthesis under kinetic control) is implemented [[2], [14]], with TDA or 7-ACA as the precursor β-lactam, the key amino acid (KA), and METzAA as the acylating agent (AA).

Graph: Fig. 1. Pathways of combined chemical and biocatalytic CEZ synthesis. Transformations: 1, chemical synthesis of TDA from 7‑ACA; 2, biocatalytic synthesis of CEZ from TDA; 3, biocatalytic synthesis of S-p CEZ from 7-ACA; 4, chemical synthesis of CEZ from S-p CEZ.

The enzyme used in β-lactam (CEZ or S-p CEZ) synthesis under kinetic control catalyzes three competing reactions [[2], [14]]:

− synthesis of the target product via transfer of the acyl group from the AA (METzAA) to the amino group in the С7 position of the β-lactam moiety of KA (TDA or 7-ACA) accompanied by methyl alcohol formation;

− hydrolysis of the target product to form the corresponding KA and TzAA;

− METzAA hydrolysis to form TzAA and methyl alcohol.

The target-product yield is determined by the ratio of the rates of the reactions listed above and depends on the reaction conditions and the starting concentrations of the precursor β-lactam and AA.

Optimization of the biocatalytic synthesis of a β-lactam antibiotic for the development of a cost-effective process should be directed towards a high yield of the transformation product at the highest possible concentration of the KA () and the lowest possible molar excess of AA over KA ( M/M) [[2]]. This results in a high concentration of the target product (CEZ or S-p CEZ) in the final reaction mix at the minimal residual concentrations of the substrates (TDA or 7-ACA and METzAA) and the by-product (TzAA), which is necessary for the development of an efficient isolation procedure for the antibiotic.

Cephalosporin-acid synthetase (CASA, EC 3.5.1.11), also called cephazolin synthetase [[2]], is one of the enzymes that show high specificity in the synthesis of CEZ and other cephalosporin-acids. The enzyme is synthesized by Escherichia coli strains produced by chemical mutagenesis, such as strain 1787 (Culture Collection of the State Research Center for Antibiotics, Russia [[15]–[17]]), FU-99-S [[18]], VPKM V-10182 [[19]], and CGMCC No. 3508 [[19]]. Cloning of the CASA gene isolated from the VPKM V-10182 strain yielded VPKM V-12206 [[20]] and VPKM V-12316 [[3]] E. coli strains that produced recombinant CASA with a high efficiency. The CASA gene was identified as a direct homolog of the penicillin G acylase (PGA) gene from the E. coli strain ATCC 9637, which harbors multiple mutations beneficial for the level of enzyme biosynthesis and its synthetase activity [[20]].

The CASA from the recombinant E. coli strain VPKM V-12316 immobilized via covalent binding to an epoxy-activated microporous carrier [[3]] (Immobilized Enzyme CASA, IECASA) was used as the biocatalyst in the present study.

Earlier studies of biocatalytic CEZ production catalyzed by PGA or CASA from E. coli strains [[15]–[18], [22]–[24]] were summarized in a review article dedicated to cephalosporn-acid synthesis [[2]]. A high yield of 98% was achieved when 7-ACA was transformed into the S-p СЕZ intermediate via biocatalytic acylation with METzAA, but only at a low initial 7-ACA concentration ( = 50 mM) and a threefold molar excess of METzAA relatively to 7-ACA, = 3 M/M. PGA immobilized on glyoxyl agarose served as the biocatalyst [[23]]. The CEZ yield exceeded 90% when the biocatalytic transformation of TDA into CEZ was performed in an aqueous medium and CASA from the mutant E. coli strain FU-99-S immobilized in a polyacrylamide gel [[2], [18]] or CASA from the recombinant E. coli strain VPKM V-12206 covalently bound to a macroporous carrier was used as the biocatalyst [[2], [21]], but the initial TDA concentration was relatively low ( = 60–80 mM) and the molar excess of METzAA over TDA was high ( ≥ 4 M/M).

The effects of the рН and initial TDA and METzAA concentrations on the efficiency of CEZ synthesis under kinetic control were addressed in an earlier study by the authors, with allowance for the рН dependences for the solubility of all components of the reaction mix [[3]]. The biocatalytic procedure for CEZ synthesis based on the obtained data could be implemented in an aqueous environment in a controlled stepwise pH gradient and enabled the use of TDA at a high initial concentration ( = 150–200 mM) and a CEZ yield of 92–95% at ≤ 3.5 M/M. We immobilized recombinant CASA from a VPKM V-12316 E. coli strain on a macroporous carrier to obtain the biocatalyst used for the synthesis; operational tests performed at the Sichuan Industrial Institute of Antibiotics (China) demonstrated high biocatalyst stability [[3]].

The goal of the present work was to develop a method of CEZ production that would include the biocatalytic synthesis of an intermediate from 7-ACA via acyl transfer and to compare the efficiency of two alternative methods of antibiotic production involving the IECASA biocatalyst.

Materials. METzAA (98% purity as demonstrated by high-performance liquid chromatography, HPLC) and MMTD (purity 97%, HPLC) samples were provided by the Sichuan Industrial Institute of Antibiotics (China). The other used reagents were 7-ACA (Anhui BBCA Pharmaceutical, China, purity 97% by HPLC); CEZ sodium salt standard (Sigma-Aldrich, United States, purity 98% by HPLC); and TzAA standard (Sigma-Aldrich, United States, 99.9% purity by HPLC). An S-p CEZ sample (purity 94.8% by HPLC) used as the standard in HPLC runs and for the solubility assays was isolated from the reaction mix after the biocatalytic synthesis via precipitation at рН 2 and subsequently purified by recrystallization.

Production of the IECASA biocatalyst. The IECASA sample used in the study was prepared from CASA isolated from the E. coli cell biomass (strain VKPM V-12316) and immobilized on a Seplite LX-1000EP macroporous epoxy-activated carrier (Sunresin New Materials, China) as described in [[3]]. The synthetase activity of IECASA in CEZ synthesis [[3]] was 420 IU/g wet biocatalyst at a dry-matter content of 37.2%.

One international unit of enzymatic (synthetase) activity of the sample in a reaction of cefazolin synthesis (1 IU) was defined as the amount of enzyme that catalyzed the formation of 1 µmol product in 1 min in a solution containing 60 mM TDA and 240 mM METzAA at a рН of 7.5 and 30°С.

HPLC analysis. HPLC analysis of the samples was performed in isocratic mode on a Gilson HPLC instrument (United States) with a UV detector and a 250 × 4 mm Spherisorb ODS column (particle size 7.5 µm) at a temperature of 30°С, a flow rate of 1.0 mL/min, and a detection wavelength of 214 or 254 nm. A mix of 50 mM ammonium phosphate buffer and methanol was used as the mobile phase. Table 1 lists the conditions for HPLC analysis of various reaction mixes and the retention times (RTs) of the components.

Conditions of HPLC analysis of the reaction mixes

Biocatalytic Synthesis of S-p CEZ. Synthesis of S‑p CEZ was performed in a glass reaction vessel with a mechanical blade stirrer and systems to control and maintain the pH and temperature; the initial volume of the reaction mix was 75 mL. The substrate (7-АСА and METzAA) solutions with preselected initial concentrations ( and ) were prepared at 30°С with constant stirring. A 7-АСА aliquot was suspended in 0.3 M sodium phosphate buffer (PB) at a рН of 8.3, the suspension was stirred until the pH no longer decreased, and 2 М NaOH was added in small portions with constant stirring until the 7-АСА was fully dissolved at a рН of 6.8–7.2. A METzAA aliquot was added, the solution was stirred until the METzAA was dissolved, and the solution volume was adjusted to 75 mL by 0.3 M PB (рН of 7.0). The synthesis process was initiated at 30°С with medium-intensity stirring via the addition of 1.8 g of wet IECASA biocatalyst to the substrate solution (enzyme concentration in the reaction mix CE = 10 IU/mL). The pH decreased spontaneously to 6.0 during the synthesis and was then maintained at this level by the addition of 2M NaOH until the completion of the process. The process was terminated via the removal of the catalyst from the reaction mix on a porous glass filter.

The content of S-p CEZ, 7-АСА, METzAA, and TzAA in reaction-mix samples collected during the synthesis was assessed by HPLC to characterize the dynamics of S-p CEZ synthesis. The synthesis process was continued until a stable plateau on the curve of the time dependence of current S-p CEZ concentration was attained. The maximal degree of 7-АСА transformation into S-p CEZ (the maximal S-p CEZ yield %) was calculated according to the formula where is the maximal concentration of the target product (average value for the plateau) and is the initial KA concentration.

Assessment of bimolecular reaction constants for CEZ and S-p CEZ hydrolysis. The processes of CEZ or S-p CEZ hydrolysis catalyzed by IECASA were performed in 0.3 M PB at a рН of 6.5 in conical flasks placed into a shaker water bath at 30°С; the initial concentrations of the β-lactams to be hydrolyzed (Сo, mM) were varied in the range of 5–30 mM. Each reaction mix had a volume of 10 mL and contained 30–40 mg of wet biocatalyst with a predefined content of dry matter. The instant concentration of the amino acid (TDA or 7-ACA) that formed in the reaction mix was determined by HPLC in aliquots drawn every 3–5 min. The initial rate of CEZ or S-p CEZ hydrolysis per 1 g dry biocatalyst (V, mM min–1 g–1) was inferred from the linear part of the product accumulation curve. The results were presented in the form of 1/V dependence on 1/Co according to the Lineweaver-Burk method [[25]], and the obtained lines were processed in the Exсel software to calculate the bimolecular constants of the hydrolytic process normalized to 1 g of dry biocatalyst weight (Vmax/KM, min–1 g–1), where KM (mM) is the Michaelis constant and Vmax (mM min–1 g–1) is the maximal rate of the enzymatic reaction.

Solubility of the compounds involved in the biocatalytic synthesis of CEZ and S-p CEZ. The reaction mix that formed during the biocatalytic synthesis of a β‑lactam antibiotic via acyl transfer contains four components: the target product (CEZ or S-p CEZ), the KA (TDA or 7-ACA), the AA (METzAA), and the byproduct (TzAA). A knowledge of compound solubility is instrumental in the optimization of biocatalytic transformation processes and the subsequent isolation of the target product [[3], [26]]. Table 2 lists the published data on the solubility of the compounds listed above and the results of 7-ACA, S-p CEZ, and CEZ solubility analysis performed in the present study.

Acid-base properties and solubility of the compounds involved in CEZ production

* The characteristic solubility of an electrolyte is identified as the solubility of its electroneutral form: the zwitterionic form of an amino acid (S±) or a noncharged form of a monocarboxylic acid (So). ** n/d, not determined.

The solubility of the nonelectrolytic METzAA and the pH dependence of the solubility for the amino-acid TDA under conditions that modeled the CEZ synthesis process were studied earlier, and the effect of TDA solution supersaturation upon a decrease in pH was demonstrated [[3]]. The obtained data enabled the development of an efficient procedure for the biocatalytic synthesis of CEZ in a stepwise рН gradient at an initial TDA concentration ( = 150–200 mM) much higher than its solubility in the neutral pH range, which is optimal for biocatalyst functioning.

The present study involved the analysis of pH dependence of the solubility of reaction mix components in S-p CEZ synthesis, namely, KA (7-ACA), the target product, S-p CEZ, and the antibiotic CEZ, which was produced in the same reaction mix without isolation of the intermediate via MMTD-mediated substitution of the 3-acetoxy group. The effect of the pH on 7-АСА, S-p CEZ, and CEZ solubility was assessed with the saturation method [[26]] under the conditions selected earlier for biocatalytic CEZ synthesis (0.3 M PB, 30°С) [[3]]. The experimental procedure was as described in [[3], [26]].

The рН dependence of the solubility (S, mM) of monocarboxylic acids S-p CEZ, CEZ, and TzAA is described by the equation

1

Graph

The рН dependence of the solubility (S, mM) of amino acids (7-ACA and TDA) that have two ionizable groups each is described by the equation

2

Graph

The equation

2.1

Graph

can be used under the conditions of complete protonation of the carboxyl group of the amino acid, i.e., in the neutral and alkaline pH range.

The designations used in equations (1), (2), and (2.1) are as follows: [H+] is the concentration of hydrogen ions at a given рН, mM; K1 is the ionization constant of the carboxyl group of a monocarboxylic acid or an amino acid, mM; K2 is the ionization constant of the amino group of the amino acid, mM; So is the solubility of the individual noncharged form of the monocarboxylic acid (characteristic solubility of the electrolyte), mM; and S± is the solubility of an individual charge-neutral zwitterionic form of the amino acid (characteristic solubility of the electrolyte), mM.

The use of the equations presented above makes it possible to calculate the constants that determine the electrolyte solubility [[26]].

Figure 2a shows the linearization of the experimental data for the pH dependence of the solubility of the 7-ACA amino acid at pH values close to the published pK2 value (Table 2) [[28]] in the coordinates of equation (2.1) (Svs. 1/[H+]). Figures 2b and 2c present the linearized experimental data for the pH dependence of the solubility of the monocarboxylic acids S‑p CEZ and CEZ in the acidic pH range, respectively. The calculated values of the characteristic solubility (S± or So) and the ionization constants of the amino group of the amino acid (pK2) or the carboxylic group of the monocarboxylic acids (pK1) are compared to the previously published data in Table 2. The pK2 and S± values for 7‑ACA determined in 0.3 M PB at 30°С were practically identical to the constants determined in 0.1 М NaCl at 20°С [[28]]. The comparison of constants for 7-ACA and TDA determined under identical conditions (0.3 M PB, 30°С) showed that MMTD-mediated replacement of the 3‑acetoxy group led to a shift of the рK2 value, a characteristic of amino-group ionization in the C7 position of the β-lactam, by 0.5 units towards the alkaline range and to a tenfold decrease of S±. This brings about a substantial decrease in the amino-acid solubility in the рН range (6.0–8.0) used for biocatalytic synthesis. MMTD-mediated substitution of the 3‑acetoxy groups in the monocarboxylic acids S-p CEZ and CEZ also led to a tenfold decrease in the characteristic solubility. The рK1 value, a characteristic of carboxyl-group ionization in the С4 position of the β-lactam, showed a slight shift towards the alkaline range (by 0.17).

Graph: Fig. 2. pH dependence of solubility on KA and the biocatalytic synthesis products (30°C, 0.3 M PB): a, for 7-ACA in the coordinates of the equation (2.1); b, for S-p CEZ, c, for CEZ in the coordinates of equation (1).

Comparison of the constants for CEZ determined in water at 20°С [[15]] and in 0.3 M PB at 30°С showed that an increase of the ionic strength of the medium and a temperature increase did not affect the pK1 value, whereas the characteristic solubility So decreased by almost two times.

Figure 3 shows the theoretical curves for the pH dependence of the solubility of the components of reaction mixes for biocatalytic CEZ and S-p CEZ synthesis; the constants listed in Table 2 were used to construct the curves.

Graph: Fig. 3. Theoretical curves for pH dependence of solubility based on the constants from Table 2: 1, TDA, 2, 7-ACA (30°C, 0.3 M PB), equation (2); 3, CEZ; 4, S-p CEZ (30°C, 0,3 M PB) equation (1); 5, TzAA (20°C, water), equation (1); 6, METzAA (30°C, 0.3 M PB).

The solubility of the nonelectrolyte METzAA did not depend on the рН. Earlier studies [[3]] demonstrated a solubility value of 1440 mM in 0.3 M PB at 30°С (Table 2), which indicated that METzAA can be used at high concentrations as an АА in biocatalytic synthesis. The low рK1 value and the high So value for TzAA (Table 2) enabled a high solubility S = 2560 mM already at a рН of 2.7 (Fig. 3, 5). The solubility of this byproduct of acyl transfer synthesis in the pH range (6–8) used for IECASA-catalyzed reactions [[3]] is so high that its precipitation cannot hinder the process at the attainable initial concentrations of the substrates.

The low TDA solubility in the working рН range of 6.0–8.0 (Fig. 3, curve 1) was the main obstacle to the preparation of the reagent mix and process implementation at a high initial concentration of KА () without precipitation during the biocatalytic CEZ synthesis [[3]]. The solubility of 7-ACA in the pH range of 6.0–8.0 was approximately 32 times higher than the TDA solubility (Fig. 3, curves 1 and 2), and this conferred substantial advantages over CEZ synthesis to the biocatalytic process of S-p CEZ synthesis, both at the stage of the preparation of the reagent mix and during the biocatalytic transformation.

The pH dependence curve for S-p CEZ solubility is shifted towards a lower рН relative to the respective curve for CEZ (Fig. 3, curves 3 and 4), whereas the level of S values for the intermediate is more than 130 times higher than the CEZ solubility at the same рН. For instance, the S-p CEZ and CEZ solubility at рН 4.0 was 2300 mM and 17 mM, respectively. Both products are readily soluble in the working рН range of 6.0–8.0, and this allowed their persistence in the solution during biocatalytic transformation. Product solubility in a low pH range is important for the development of a procedure for product isolation from the reaction mix via precipitation. The relatively high S-p CEZ solubility at a рН close to 2.0 was a negative factor, since it resulted in product loss during precipitation. Therefore, the development of a process for the chemical transformation of S-p CEZ into CEZ directly in the reaction mix after biocatalytic transformation without isolation of the intermediate appears to be justified.

Biocatalytic synthesis of the CEZ intermediate catalyzed with IECASA. The biocatalytic transformation of 7-ACA into S-p CEZ was performed in 0.3 M PB at 30°С, CE = 10 IU/mL, with in the range of 55–320 mM and in the range of 1.5–3.5 M/M at no higher than 830 mM, so that the METzAA solubility was not exceeded under the experimental conditions (Table 2 and Fig. 3, curve 6). The high solubility of 7-ACA at near-neutral рН (Fig. 3, curve 2) enabled preparation of the solution at a рН of ~7.0 in the entire used range of . This was a substantial advantage over the preparation procedure of KА solution for CEZ synthesis [[3]], because the low TDA solubility necessitated the preparation of solutions at a high рН (up to 8.5), which is unfavorable for β-lactam stability, and the maximal , thus achieved, did not exceed 200 mM. It was necessary to start CEZ synthesis at a рН of 8.2–8.3, which is unfavorable for IECASA activity and stability [[3]].

S-p CEZ synthesis was initiated at a рН of 6.8–7.2 after the sequential addition of METzAA and the biocatalyst to the 7-ACA solution and was performed at a рН that spontaneously decreased from 7.0 to 6.0, i.e., under conditions optimal for IECASA functioning and stability. A рН of 6.0 was maintained until the transformation was complete. The рН decrease during β-lactam synthesis is due to the formation of free TzAA, a byproduct of hydrolysis reactions, primarily of METzAA hydrolysis (Fig. 1). The decrease in рН can be accompanied by the precipitation of nonacylated KA if its residual concentration exceeds the solubility of the specific compound at the рН level, as observed for the transformation of poorly soluble TDA into CEZ in a spontaneous рН gradient of 8.2–8.3 to 6.0 [[3]]. The critical рН level at which KA precipitation occurred depended on the ionization constants that determine the pH dependence of amino acid solubility (Table 2) and on the operation parameters and which determine the degree of KA transformation into the target β-lactam by the time a critical pH is attained at a given temperature and CE. A stepwise pH gradient was developed for CEZ synthesis: as the pH of the reaction mix decreased spontaneously from the initial value of 8.2–8.3, it was first fixed at 6.8 and then at 6.0 to ensure TDA preservation in the dissolved state [[3]].

The transformation of 7-ACA into S-p CEZ in the 7.0–6.0 pH range was performed in a spontaneous gradient in the present study, and the process was not hindered by KA precipitation, even at a high excess of METzAA ( = 3.0 M/M) that promoted intensive TzAA release, combined with of up to 275 mM. The solubility of 7-ACA decreased gradually as the pH decreased (Fig. 3, curve 2), reaching S = 60 mM at a рН of 6.0, but it was high enough for preservation of nontransformed KA in the dissolved form in the entire pH range under the selected operating conditions. A reduction of АА excess to = 2.5 M/M enabled S-p CEZ synthesis in a spontaneous рН gradient and at a higher initial concentration of the KA (7-ACA) = 320 mM.

Figure 4 shows the time dependence of the composition of the reaction mix that formed during IECASA-catalyzed S-p CEZ synthesis under kinetic control. The figure illustrates the dynamics of KA (7‑ACA) and AA (METzAA) consumption and the accumulation dynamics for the target β-lactam S-p CEZ and the byproduct TzAA. The dynamics of the relative S-p CEZ concentration reflects the time dependence of the yield of the target synthesis product relative to the substrate with the β-lactam moiety (7-ACA). The presence of a prolonged plateau on the S-p CEZ accumulation curve, which corresponds to the maximal product yield (), is an important, distinctive feature of the process. As in IECASA-catalyzed CEZ synthesis [[3]], the presence of a plateau is explained by the existence of long-term kinetic equilibrium between the processes of synthesis and hydrolysis of the β-lactam product. The preservation of a 100% balance for both the β-lactam- and the tetrazolyl-containing compounds (Fig. 4) was evidence of the absence of uncontrollable side reactions, including those related to β-lactam degradation, in the system.

Graph: Fig. 4. Time course of changes of the reaction-mix composition (relative concentrations, %) in IECASA-catalyzed S-p CEZ synthesis (30°C, 0.3 M PB, CE = 10 IU/ml, = 320 mM, = 800 mM, = 2.5 M/M). 1, 7‑ACA; 2, S-p CEZ calculated relatively to 3, balance (%) for β-lactam, the sum of relative concentrations of 7-ACA and S-p CEZ; 4, METzAA; 5, TzAA, calculated relatively to 6, balance (%) for TzAA, the sum of relative concentrations of S-p CEZ, METzAA, and TzAA.

A series of experiments at various (55–320 mM) and (1.5–3.5 M/M) values under the standard conditions selected earlier [[3]] for CEZ synthesis (30°С, CE = 10–12 IU/mL, 0.3 M PB) was performed in order to optimize the IECASA-catalyzed S-p CEZ synthesis. A spontaneous pH gradient from a рН of 7.0 ± 0.2 to рН 6.0, with the рН subsequently maintained at 6.0 ± 0.1 via the addition of 2 М NaOH was used until process termination. The maximal S-p CEZ yield ( %) calculated as the average value for the plateau on the product accumulation curve was used as a characteristic of process efficiency. The time required to attain the maximal S-p CEZ concentration in the reaction mix ranged from 25 to 50 min and depended on the operating conditions.

The results of IECASA-catalyzed S-p CEZ synthesis were summarized in the form of dependences of the maximal product yield on (Fig. 5, curves 1 and 1а) and (Fig. 6, curve 1) and compared to previous results on the biocatalytic synthesis of CEZ [[3]] (Fig. 5, curves 2 and 3; Fig. 6, curves 2 and 3). CEZ synthesis in a spontaneous рН gradient (Fig. 5, curve 3; Fig. 6, curve 3) ensured a yield of = 92 ± 2% but only at = 100–140 mM and = 3.1–3.5 M/M (Fig. 5, curve 3). An increase of the initial TDA concentration to ≥ 145 mM led to a drop of the CEZ yield to 85% or even lower (Fig. 6, curve 3) due to TDA precipitation upon a рН decrease during the process. TDA precipitation could be prevented when CEZ synthesis was performed in a stepwise рН gradient (Fig. 5, curve 2; Fig. 6, curve 2). The yield = 93.5 ± 1.5% was achieved in this case at TDA concentrations of = 150–200 mM and = 3.1–3.5 M/M. These conditions can be considered as optimal for IECASA-catalyzed CEZ synthesis.

Graph: Fig. 5. Dependence of the maximal yield of the β-lactam product ( %) on the initial KA concentration ( mM) in IECASA-catalyzed acyl transfer synthesis (30°C, 0.3 M PB, CE = 10–12 IU/ml): 1, S-p CEZ synthesis from 7-ACA and METzAA; spontaneous рН gradient from 7.0 ± 0.2 to 6.0 with the pH maintained at 6.0 ± 0.1 afterwards; = 2.9–3.1 M/M; 1а, the same at = 2.5 M/M; 2, CEZ synthesis from TDA and METzAA at = 3.1–3.5 M/M, stepwise pH gradient from 8.2–8.3 with the pH maintained at рН 6.8 ± 0.1 and рН 6.0 ± 0.1 afterwards; 3, CEZ synthesis from TDA and METzAA at = 3.1–3.5 M/M, spontaneous рН gradient from 8.2–8.3 to 6.0 with the рН maintained 6.0 ± 0.1 afterwards.

Graph: Fig. 6. Dependence of the maximal yield of the β-lactam product ( %) on the initial molar excess of AA over KA ( M/M) in IECASA-catalyzed acyl transfer synthesis at 30°C in 0.3 M PB, CE = 10 IU/ml: 1, S-p CEZ synthesis from 7-ACA and METzAA at = 180–330 mM; spontaneous рН gradient from 7.0 ± 0.2 to 6.0 with the pH maintained at 6.0 ± 0.1 afterwards; 2, CEZ synthesis from TDA and METzAA; = 150–200 mM; stepwise pH gradient from 8.2–8.3 with pH maintained at рН 6.8 ± 0.1 and рН 6.0 ± 0.1 afterwards; 3, same at = 145–170 mM; spontaneous рН gradient from 8.2–8.3 to 6.0 with the рН maintained at 6.0 ± 0.1 afterwards.

A higher S-p CEZ yield = 96.5 ± 1.5% was attained when the synthesis was performed in a spontaneous pH gradient in the 7-ACA concentration range = 175–275 mM at = 3.0 M/M (Fig. 5, curve 1; Fig. 6, curve 1). It is important that the dependence on (Fig. 6, curve 1) reached a plateau at = 3.0 M/M at high 7-ACA concentrations, whereas the reduction of the molar excess to = 2.5 M/M was accompanied by a slight decrease of the yield to = 95.0 ± 1.5% (Fig. 5, curve 1а).

The increased yield upon the replacement of TDA by 7-ACA in IECASA-catalyzed synthesis under kinetic control was due, for instance, to the lower specificity of the enzyme for the side reaction of the target β-lactam hydrolysis (S-p CEZ as compared to CEZ). IECASA specificity for the hydrolysis of CEZ and S-p CEZ was inferred from the bimolecular constants of the respective processes, which were determined according to the Lineweaver-Burk method (Fig. 7). The bimolecular constants of the hydrolytic processes (Vmax/Km, min–1 g–1) calculated from the lines shown were 0.24 ± 0.01 min–1 g–1 and 0.18 ± 0.01 min–1 g–1 for CEZ and S-p CEZ, respectively, i.e., hydrolysis of the target product of S-p CEZ synthesis by IECASA was 30% less efficient than the hydrolysis of the CEZ synthesis product.

Graph: Fig. 7. Assessment of the bimolecular constants of IECASA-catalyzed CEZ (1) and S-p CEZ (2) hydrolysis at 30°С in 0.3 M PB according to the Lineweaver-Burk method.

Optimization of the IECASA-catalyzed β-lactam synthesis process was directed towards the attainment of the maximal yield of the target product at the maximal acceptable value and a minimal acceptable value to ensure a high concentration of the target product in the reaction mix after the transformation and the minimization of the levels of contaminants (residual KА and TzAA formed in a side reaction of hydrolysis). Table 3 presents the process-efficiency parameters, such as the maximal product yield relatively to the β-lactam substrate ( %) and the final reaction-mix composition after transformation for a range of CEZ and S-p CEZ synthesis experiments for comparison. The target product (CEZ or S-p CEZ) concentration Сprod (mM), the KA (TDA or 7-ACA) concentration СКА (mM), and the total concentration of the byproducts (METzAA and TzAA) СME + TzAA (mM) in the final reaction mix were determined. The last two components were combined for the analysis, because the residual METzAA can be easily hydrolyzed to form TzAA prior to isolation of the target product.

Efficiency parameters of CEZ and S-p CEZ synthesis processes catalyzed by IECASA

CEZ synthesis under optimal conditions with a stepwise pH gradient enabled transformation with a yield of = 93.9% at = 3.4 M/M and the highest possible = 200 mM (Table 3); the antibiotic concentration Сprod in the reaction mix was 190 mM at СКА = 12 mM and СME + TzAA = 490 mM. S-p CEZ synthesis at practically the same = 204 mM was implemented in a more simple, spontaneous pH gradient with a substantially higher yield = 97.7%, notwithstanding the use of a smaller excess of the acylating agent = 3.0 M/M (Table 3). This made it possible to increase the target β-lactam concentration in the reaction mix to Сprod = 200 mM at a corresponding decrease of СКА and СME + TzAA to 5 mM and 415 mM, respectively. The use of higher initial concentrations of 7-ACA = 276 mM and = 322 mM (Table 3) allowed a high yield of S-p CEZ (approximately 95%) and a 1.3- to 1.5-fold increase of the target product concentration in the reaction mix relative to the optimal CEZ synthesis process practically without increasing in the residual KA level. СM-E + TzAA increased by 15% at = 3.0 M/M and by only 2% at = 2.5 M/M as compared to the optimal process of CEZ synthesis.

S-p CEZ can be transformed into CEZ directly in the reaction mix obtained in biocatalytic synthesis after the removal of the biocatalyst. Incubation with MMTD applied in a threefold molar excess relatively to S-p CEZ allowed a 96.5% transformation of the intermediate into CEZ at 65°С and a рН of 6.0–6.5. Optimization of the transformation process will be a part of further studies in order to develop a process for CEZ isolation and purification.

Comparison of the biocatalytic synthesis of CEZ from TDA and S-p CEZ from 7-ACA (Fig. 1, stages 2 and 3) revealed a range of important advantages of the latter process:

—preparation of the starting 7-ACA solution at a neutral рН (instead of a рН of 8.2–8.5 for the TDA) to achieve a higher

—S-p CEZ synthesis in a simple, spontaneous рН 7.0–6.0 gradient favorable for enzyme activity and stability instead of the stepwise pH gradient from 8.3 to 6.0 used in CEZ synthesis;

—an increase in the target β-lactam yield from 93.5 ± 1.5 to 96.5 ± 1.5% under conditions optimized for each process;

—a 1.3- to 1.5-fold increase in the target β-lactam concentration in the final reaction mix;

—the ability to use the product of biocatalytic 7-ACA acylation by the TzAA methyl ester as the intermediate for the production of the cephalosporin antibiotic ceftezole, rather than for CEZ production only.

CEZ production based on transformations 3 and 4 (Fig. 1) can be performed as a continuous process without S-p CEZ isolation from the reaction mix. It is necessary to note that the chemical process of S-p CEZ transformation into CEZ (Fig. 1, transformation 4) that takes place in an aqueous solution at 30°С is more environmentally friendly than TDA production from 7-ACA (Fig. 1, transformation 1) performed at a low temperature (–40°С) in the presence of toxic reagents (BF3 and CH3CN). It can be concluded that combined chemical and biocatalytic CEZ synthesis via direct biocatalytic acylation of 7-ACA appears to be a promising alternative to the conventional procedure with the biocatalytic acylation of TDA.

This work was supported by State Project no. 595-00003-19 PR.

The authors declare that they have no conflict of interest. This article does not contain any studies involving animals or human participants performed by any of the authors.

Translated by S. Semenova