Immobilization of IMP-1 metallo-beta-lactamase on Fe₃O₄@SiO₂ as nanobiocatalyst for degradation of beta-lactam antibiotics in wastewater

Residues of antibiotics in water resources are important environmental and public health issues (Thai et al. 2018). These pollutants are discharged into the environment through various ways such as: discharged treated and untreated municipal, industrial, hospital, and livestock farming and aquaculture wastewaters. The pharmaceutical industry is one of the main sources of the release of antibiotics into water sources. Therefore, it is very important to reduce or eliminate the antibiotics released from the effluents of the pharmaceutical industry. The main concern of these substances in the environment is the development of antibiotic resistance (Rodriguez-Mozaz et al. 2015). With an increasing trend, tens of thousands of deaths from antibiotic resistance occur worldwide annually. Beta-lactam antibiotics are the most widely used group of prescribed antibiotics, accounting for 65% of the antibiotic market in 2003 (Elander 2003). This group includes cephalosporins, penicillin derivatives, monobactams, carbacephems, and carbapenems. The main characteristic of these antibiotics is a ring called beta-lactam in their molecular structure (De Rosa et al. 2021). Residues of beta-lactam antibiotics, even in very low concentrations in the environment, can lead to antibiotic resistance, and as mentioned earlier, this can cause public health and environmental problems (Lateef 2004). The Wastewater containing cephalosporin antibiotics contains inorganic salts, active pharmaceutical ingredients (APIs), and toxic organic compounds (Yang et al. 2016).

Biodegradation and adsorption are two major mechanisms for removing beta-lactam antibiotics, and volatilization and hydrolysis are less effective in the biological wastewater treatment processes. Conventional biological treatment methods, such as activated sludge processing, are not efficient enough to remove beta-lactam antibiotics from wastewater (Zhang et al. 2021). However, further treatment is required before discharging the effluent from conventional WWTPs (Pirsaheb et al. 2014). Recently, various advanced methods for removing beta-lactam antibiotics in wastewater and aqueous solutions have been investigated, including photolysis, ozone, ozone/hydrogen peroxide, activated carbon, nanofiltration, and reverse osmosis (Riaz et al. 2020). The membrane processes, such as reverse osmosis, are a relatively high-efficiency method of removing antibiotics (Couto et al. 2018). However, these processes have limitations, such as membrane clogging due to the formation of chemical deposits or microbial growth. In addition, this method cannot decompose antibiotics and only collect them from the liquid phase (Homem & Santos 2011). In the adsorption processes, such as membrane filtration processes, the mechanism of removing the contaminant is not decomposition and degradation, but its separation from the liquid phase and transfer to the solid (adsorbent) phase, which causes the contaminant to condense. In addition, they have limitations such as high operating costs and regeneration of the adsorbent bed and inefficient at high concentrations of pollutants (Crini et al. 2019). The mentioned limitations and additional treatment of effluents and wastes from membrane filtration and adsorption processes, increased interest in studying and optimizing advanced oxidation processes. The main limitation of advanced oxidation methods, despite their high efficiency in the removal of these contaminants, is the increase in operating costs and secondary contaminants due to the production of by-products, which may sometimes be more toxic and hazardous. Studies also show that under different photolysis conditions, cephalosporin antibiotics in aqueous solutions can become more toxic and more stable products than the primary compound (Ribeiro et al. 2018). As a result of the mentioned problems and issues, it is necessary to develop new methods such as enzymatic processes to effectively remove beta-lactam antibiotics from wastewater (Al-Gheethi et al. 2015). Enzymes have been recently used as a potential biocatalyst in many biotechnological sciences (Lima et al. 2017). Enzymes with excellent catalytic properties have been well studied as catalysts because of their unique physicochemical behavior (Rueda et al. 2016). Potential advantages of enzymatic treatment compared to conventional treatment include application of biodegradable compounds, efficiency at high and low pollution concentrations, a wide range of pH, temperature and, salinity, reduction of sludge volume (without biomass production), and simplicity of process control (Pandey et al. 2017). The use of enzyme technology can be considered for resolving the problem of residuals of antibiotics and other emerging pollutants from water and wastewater, especially pharmaceutical industries and hospital wastewaters, due to its cost-effectiveness and specificity of removing target contaminants without producing by-products (Bilal et al. 2019).

Beta-lactamases are produced by some bacteria that play an important role in their resistance to beta-lactam antibiotics. These enzymes are classified into four classes and inactivate beta-lactam antibiotics due to hydrolyzing the beta-lactam ring (Gnanamani et al. 2017). Metallo β-lactamases are class B lactamases with three subclasses: B1, B2 and, B3. IMP-1, subclass B1, is an enzyme encoded by the transmissible bla_imp-1 gene from Pseudomonas aeruginosa. (Mojica et al. 2006). For many industrial applications, enzymes must be immobilized through simple, cost-effective protocols and used for long cycles. Enzyme immobilization usually results in catalyst stability against chemical (organic solvent), temperature changes, and ionic strength.

Also, enzymes are partially denatured in aqueous solutions, which leads to a decrease in their activity. One of the challenges of enzyme immobilization is reduced enzyme activity and stability. Some immobilization methods can increase activity and resistance to inhibitors or enzyme denaturation (Melo et al. 2017; Monteiro et al. 2019). To immobilize the enzyme, various materials are used as a base or carrier (Cui et al. 2018). The use of silica-coated iron oxide magnetic nanoparticles has attracted much attention due to their easy separation, reuse, high dispersion and surface area, low mass transfer limitation, low toxicity, and biocompatibility for enzyme immobilization (Zhang et al. 2019). Direct immobilization without any cross-linking agents would cause attendant problems such as decreased enzymatic activity due to restriction of three-dimensional conformation (Chen et al. 2018). Glutaraldehyde is one of the most widely used materials in the design and production of biocatalysts (Barbosa et al. 2014). Therefore, its use as a spacer arm can have advantages such as being inexpensive and a very effective cross-linker in immobilizing the enzyme (Wahab et al. 2020). In most previous studies, beta-lactamase inhibitors have been investigated to reduce antibiotic resistance (Zhanel et al. 2013; Lomovskaya et al. 2020). However, only a few studies have used these enzymes to remove antibiotics from aqueous solutions (Gao et al. 2018). The IMP-1 was not immobilized and studied for antibiotic removal before this study. In the present work, the IMP-1 metallo-beta-lactamase was immobilized first time by covalent bonding through cross-linking in glutaraldehyde solution as a spacer arm on magnetic nanoparticles. Then the pH values and the optimum temperature of enzyme activity, thermal and stability storage, enzyme reusability, kinetic parameters, and the capacity of magnetic nanoparticles for enzyme storage were determined. This study aims to develop a novel nanobiocatalyst through immobilization of IMP-1 metallo-beta-lactamase on magnetic nanoparticles for the first time to improve the enzyme's function in the effective degradation of beta-lactam antibiotics from water solutions and contaminated wastewater, especially pharmaceutical and hospital wastewaters.

The recombinant IMP-1 enzyme was expressed in the Escherichia coli BL21 (DE3) harboring pET28a vector carrying the bla_IMP-1 gene from P. aeruginosa based on the protocol Arjomandi et al. (2019) in the Structural Biology Laboratory of Biological Research Center of Kermanshah University of Medical Sciences. In summary, the transfer of plasmid carrying the IMP-1 gene to BL21 (DE3) cells was performed by a heat-shock method in LB (Luria Bertani) culture medium (Froger & Hall 2007).

The cells were then cultured at 37 °C in LB broth enriched with kanamycin (50 μg/ml) as selecting an agent with shaking until the optical density (OD) 600 nm reached 0.4–0.6. Protein expression was performed in the optimum condition: expression of the protein was induced by adding isopropyl-β-d-1-thiogalactopyranoside (IPTG) to a final concentration of 0.1 mM for 8 h at 25 °C. Then, for more cultivation, the culture was incubated for 2 h and centrifuged at 5,000 × g for 10 min and stored cells in −20 °C for 24 h. The cells were suspended in lysis buffer (pH 7.5) containing 50 mM Tris, 300 mM NaCl₂, and 5 mM imidazole. A protease inhibitor without EDTA, phenylmethylsulfonyl fluoride (PMSF) was added every 30 minutes for three times. After that, cells were disrupted by a sonicator (BANDELIN SONOPULS HD 2200, Germany). The lysate was clarified by centrifugation at 12,000 rpm for 30 min to eliminate the cell debris. Enzyme purification was performed by nickel-nitriloacetic acid (Ni-NTA) agarose affinity chromatography column according to the method of Schlesinger et al. (2011). Before loading the solution, the column was equilibrated by a lysis buffer as the equilibrium buffer at pH 7.5. The protein solution was then loaded onto the column. Afterward, the protein solution was dialyzed against dialysis buffer containing potassium phosphate 50 mM, NaCl 300 mM, and ZnCl₂ 100 μM.

The purification of the enzyme was assessed by SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) method (Gallagher & Wiley 2010). The enzyme solution was stored at −20 °C. During the purification procedure, beta-lactamase activity was assayed by spectroscopy by Perkin Elmer Lambda 25 UV-Visible Spectrometer at 25 °C (Waley 1974). In brief, the 500 μl of penicillin (2 mM) as a substrate was added to 497 μl of potassium phosphate buffer (50 mM) pH 7.5, then 2.8 μl (1.6 μg) of pure IMP-1 was added by appropriate shaking. After 10 min incubation, 100 μl of EDTA was added as an inhibitor of enzyme activity. IMP-1 activity was calculated by reducing the adsorption of penicillin G at a wavelength of 235 nm due to the opening of the b-lactam ring using the standard penicillin G curve that was previously prepared. Also, the enzyme concentration was determined by the Bradford method with bovine serum albumin (BSA) as standard (Bradford 1976).

Fe₃O₄ magnetic nanoparticles were synthesized using a co-precipitation technique with some modification (Gemeay et al. 2020). In brief, the 4.40 g anhydrous FeCl₃ (0.09 mol/L) and, 3.1 g FeSO₄·7H₂O (0.045 mol/l) with a molar ratio of Fe²⁺/Fe³⁺ (1:2) were stirred in 250 ml distilled water. The pH was increased to 11 by the addition of sodium hydroxide. The reaction was continuously overhead stirred under N₂ for 1 h at room temperature, then followed heating to 80 °C for 3 h. Afterward, the obtained magnetic nanoparticles were collected by an external magnet and washed several times with distilled water to reach a neutral pH, and finally, the nanoparticles obtained were dried in a vacuum oven for 24 hours at 60 °C.

The one-step method with some modification was used to coat Fe₃O₄ with the SiO₂ shell (Lu et al. 2015). 0.5 g of Fe₃O₄ was dispersed in 750 ml of ethanol 96% (45 minutes ultrasonic), and then 15 ml of ammonia 25% (V/V) was added to 235 ml of distilled water and again was placed for 45 minutes in an ultrasonic bath. Afterward, 3 ml of tetraethyl orthosilicate (TEOS, 99%) was added dropwise and stirred for 8 hours at room temperature. The nanoparticles were separated by a magnet, washed several times with ethanol, and dried in a vacuum oven at 45 °C.

Magnetic silica-coated nanoparticles (Fe₃O₄@SiO₂) were functionalized (aminated) by (3-aminopropyl) triethoxysilane (APTES, 98%): 0.5 g of Fe₃O₄@SiO₂ synthesized in 50 ml of ethanol was dispersed in an ultrasonic bath for 45 minutes. 1.5 ml of APTES was added and stirred at reflux at 70 °C for 6 h. Magnetic nanoparticles (Fe₃O₄@SiO₂-NH₂) were separated by a magnet, washed several times with ethanol/acetone, and dried in a vacuum oven at 45 °C (Gemeay et al. 2020).

Immobilization was carried out by the method of Ma et al. (2003) with a few modifications. Glutaraldehyde 25% was used as a coupling agent (spacer arm) to immobilize the enzyme on the magnetic nanoparticles. In summary: 100 mg Fe₃O₄ @ SiO₂-NH₂ was dispersed in 250 ml of ethanol 96% v/v, and 5% glutaraldehyde solution (V/V ratio) in 50 ml of 0.01 M sodium phosphate buffer (pH 7.4) was added to the nanoparticle/ethanol mixture. It was then incubated at 300 rpm for 12 hours at room temperature. The nanoparticles were separated by a magnet and washed several times with 50 mM potassium phosphate (pH 7.5). Then, 10 mg of glutaraldehyde activated magnetic nanoparticles were added to a 15 ml falcon tube containing 3 ml of 50 mM sodium phosphate buffer and 2 ml of the enzyme at a 1.3 mg/ml concentration in the final concentration of 2.6 mg. It was incubated at room temperature on a puddle stirrer for 2 hours. The enzyme-carrying nanoparticles were then separated by an external magnet and washed several times with 50 mM potassium phosphate buffer (pH 7.5) containing 100 μM ZnCl₂. To determine the immobilization efficiency, the supernatant was collected before washing.

The FT-IR analysis was performed by a Thermo Avatar Fourier transform infrared spectrometer. The molecular and atomic crystal structure of the magnetic nanoparticles was performed at different stages of the synthesis by X-ray diffraction (PHILIPS, PW1730, X-ray diffractometer with Cu KR (λ) 0.154 nm radiation), the FESEM images, mapping, and diameter histogram were obtained by TESCAN, MIRA III operated at 10 kV electron microscope. Vibrating-sample magnetometer (VSM) analysis was carried out using an LBKFB vibrating sample magnetometer. HRTEM images were obtained by FEI, TEC9G20 high-resolution transmission electron microscopy.

The activity assay of free and immobilized IMP-1 was carried out by the method of Messasma et al. (2018) with few modifications. Thus, a certain amount of enzyme-carrying nanoparticles were added to the final volume of 1,000 μl, including 50 mM phosphate-potassium buffer, pH 7.5, and 500 μM penicillin as a substrate, and after 5 minutes of reaction. The magnetic nanoparticles carrying the enzyme were separated by an external magnet, and the decreased absorption of the substrate was read using a spectrophotometer (Perkin Elmer Lambda 25 UV-Visible Spectrometer) at 235 nm. The reduction of a substrate concentration was determined using a standard curve. One unit of enzyme IMP-1 is the amount of enzyme that hydrolyzes one micromole of the substrate at pH 7 at 25 °C for one minute. The free enzyme activity was also measured in the same way, with the difference that at the end of the reaction time, EDTA at a concentration of 100 μmol in the reaction volume was added to complete inhibition of the enzyme activity.

The activity of enzymes strongly depends on the pH and temperature of the reaction mixture. This study investigated and determined the effect of optimum temperature and pH on IMP-1 activity. Optimum temperature and pH were determined according to Schlesinger et al. (2011) with some modifications. 1.6 μg enzyme was incubated for 5 minutes with 1 mM penicillin G in 50 mM potassium phosphate buffer (pH 7.5 containing 100 μmol ZnCl₂) from 20 to 80 °C. After 5 minutes of the reaction, 100 μl of 1 mM EDTA solution was added to the reaction mixture to inhibit the enzyme activity. To determine the effect of pH on the IMP-1 enzyme activity, the free enzyme was incubated at pH 5.5–10 (citrate buffer 5.5–6, potassium phosphate buffer 6.5–8, and carbonate buffer pH 9 and 10) for 5 minutes with 1 mM of penicillin G at 25 °C. For the immobilized enzyme, as the same as the free enzyme, the optimum temperature and pH were determined, with the difference that at the end of the reaction time, the magnetic nanoparticles carrying the enzyme were separated by a magnet and the adsorption values were read by a spectrophotometer. The mean and standard deviation penicillin degradation were determined based on 3 measurements.

To determine the thermal stability, the free and immobilized enzymes were pre-incubated at 20, 30, 40, 50, 60, and 70 °C for various time intervals 15, 30, 45, and 60 minutes and at the end of these times, their activity was measured according to the previously mentioned method. Activity at 25 °C before incubation was considered as total activity and residual enzyme activity at different temperatures and times was considered as relative activity (Gao et al. 2018).

The reusability and storage stability of the immobilized IMP-1 on magnetic nanoparticles was determined according to the method of Gao et al. (2018). The enzyme activity was assayed after each cycle. Enzyme activity was measured in 15 consecutive cycles lasting 5 minutes at 25 °C. At the end of each hydrolysis cycle, the immobilized enzyme was separated from the reaction medium by a magnet and washed with an assay buffer. The residual activity in each cycle was expressed based on the ratio of enzyme activity to its activity in the first cycle based on the percentage (relative activity). The storage stability of the immobilized and free enzyme was determined under the same storage conditions at 4 °C for 4 weeks. At the end of each week, a certain volume of both free and immobilized enzymes was harvested and an activity assay was performed. The activity of IMP-1 at the beginning of storage was considered to be 100%.

The spectrophotometric assay was used to determine the kinetic parameters of free and immobilized enzymes for penicillin G (Pen G), cephalexin (CEX), and ceftriaxone (CRO) antibiotics (Waley 1974). The different concentrations (μM) of each antibiotic at 25 °C in 50 mM potassium phosphate buffer (pH 7.5, 100 μl ZnCl₂) and certain concentrations of free and immobilized IMP-1 were used in the 1,000 μl final volume of the reaction. Measurement of activity was performed using of spectroscopy earlier described. Changes in the values of wavelength adsorption obtained results hydrolysis of the beta-lactam ring at 235, 260, and 241 nm were read for Pen G, CEX, and CRO, respectively. Antibiotic residual values were calculated using the slope of the standard curve line which had already been prepared. Then, using GraphPad prism (version 6), the Michaelis-Menten and Lineweaver-Burk plots were drawn and the values of V_max and K_m of free and immobilized enzymes were obtained. (Rodriguez et al. 2019).

The inactivation of Pen G, CEX, and CRO after enzymatic hydrolysis was investigated by agar well diffusion according to the Clinical and Laboratory Standards Institute (CLSI) 27th ed. For this purpose, a suspension of ATCC 25922 E. coli with a concentration of 0.5 McFarland (108 × 1.58 cfu) was prepared in Müller-Hinton broth growth medium. Bacteria were seeded by the lawn culture method in the Muller-Hinton agar plates using a sterile cotton swab. Then 50 microliters of enzymatic reaction solution sterilized by filter 0.22 were added to the wells in the plates and incubated overnight at 35 °C. The inhibition zone diameter around the wells was measured, and the inhibition effect of antibiotics before and after enzymatic hydrolysis by IMP-1 was compared.

In this study, P ≤ 0.05 was considered statistically significant. All experiments were performed independently with three replications to calculate the mean and standard deviation. All data related to immobilization are expressed as means ± standard deviation. SPSS version 16 was used for statistical analysis.

In the present study, for the first time, recombinant IMP-1 metallo-beta-lactamase was immobilized on magnetic nanoparticles for the purpose of hydrolysis of beta-lactam antibiotics. The difference between the present work and a number of similar studies is shown in Table 1. Also, in the results and discussion section, the results obtained from each phase of the study were compared with the results of similar studies.

To facilitate study and use of enzyme for the present study, the bla_IMP gene was cloned into a T7 expression vector (pE28a) and was overexpressed in the E. coli BL21 (DE3). SDS-PAGE data showed that IMP-1 was expressed successfully. The molecular weight of the IMP-1 was approximately 29 kDa as determined by SDS-PAGE (Figure 1). Crude extraction of culture was purified by a Ni-NTA column. Then, the protein was dialyzed according to the method mentioned earlier. Purification fold and % yield after this step was 2.36 ± 0.568 and 61.55 ± 6.10, respectively (Table 2). Enzymes purification is important in studying the structural and functional properties and foretelling their applications. In the present study according to the data reported in Table 3, pure enzyme yield was 31 mg per liter culture, which is more than the amount reported by Horton et al. (2012). The enzyme yields of purified IMP-1 were between 3 and 7 mg per liter of culture. Purification was carried out by affinity chromatography with a Strep-Tactin^® Superflow^® resin.

FT-IR spectra analysis was used to prove the successive steps of modification and activation of the surface of magnetic nanoparticles. For all samples due to bending vibrations of Fe-O in the Fe₃O₄, the absorption band was observed at 568 cm⁻¹. Due to the surface modification of silica, a strong absorption band was observed at 1,092 cm⁻¹ is straight direct evidence to verify the formation of the silica shell. Also, the absorption bands 2,918 and 2,835 cm⁻¹ in the Fe₃O₄@ SiO₂ nanoparticles are related to the C-H bands as a result of the amination of them by aminopropyl (Nandiyanto et al. 2019). In addition, the absorption of primary amine may overlap by the O-H band at 3,400–3,500 and 1,630, respectively (Figure 2(a)). The results show that the modification of MNPs with the amine group was performed successfully, consistent with the results of similar studies (Chen et al. 2008).

The X-ray diffraction patterns were used to study the crystal structure of all nanoparticle samples and are shown in Figure 2(b). XRD diagram for bare Fe₃O₄ nanoparticles, peaks appeared in 2θ: 30.3° (220), 35.9° (311), 43.3° (400), 57.1° (511), and 63.10° (440) except for the peak existed at 2θ = 18°, which might be attributed to the existence of amorphous SiO2, no other significant peaks can be observed.Were in accordance with the database of Fe₃O₄ in the JCPDS (JCPDS card: 19–629) file. These results indicate that Fe₃O₄ crystal is a highly crystalline cubic spinel structure. Similar peaks were observed for other nanoparticle samples, indicating the stability of the crystal structure of nanoparticles during the silica coating process, amine functionalization, and glutaraldehyde activation. These results are coordinated with other studies. (Fang et al. 2016; Feyzi & Norouzi 2016).

The results of the VSM analysis of nanoparticle samples are shown in Figure 3. According to the results, all 4 samples are super magnetic, and the highest amount of magnetic saturation is 73.5 emu/g, which belongs to Fe₃O₄. In addition, despite the decrease in the amount of magnetic saturation in other samples due to coating and functionalization, they have appropriate magnetic properties, which are consistent with the results of other similar studies (Ghasemzadeh et al. 2015; Alemi-Tameh et al. 2016). Therefore, these nanoparticles can be used for enzyme immobilization purposes due to their ease of separation from a reaction solution. Magnetic separation provides a very convenient approach for removing and recycling magnetic biocatalysts. This feature provides reusability and convenience of operational processing.

The morphology and size of the nanoparticles by FESEM are shown in Figure 4. The results showed that the nanoparticles were spherical in shape, and the size of Fe₃O₄@SiO₂ and Fe₃O₄@SiO₂-NH₂ as a result of coating with TEOS and amination by APTES were larger than bare Fe₃O₄ nanoparticles. HRTM images also showed that a clear thin layer of silica formed around the nanoparticles. (Figure 5). This showed that SiO₂ was successfully coated on Fe₃O₄ nanoparticles. Similar results have been reported in the study of Wang et al. (2012).

EDS analysis was performed for all samples of magnetic nanoparticles at different stages of synthesis, coating, amination, and activation by glutaraldehyde. The elementary composition of Fe₃O₄@SiO₂-NH₂ nanoparticles including iron, oxygen, silicon, carbon, and nitrogen proved by the EDS spectrum (Figure 6). These results show no other elements in the nanoparticle composition and the main elements are evenly distributed in the nanoparticle composition. It also further proves the synthesis of Fe₃O₄@SiO₂-NH₂. The results obtained from the EDS analysis of the present study are consistent with the results of the study of Alemi Tameh et al. (2016).

The IMP-1 metallo-beta-lactamase was immobilized by covalent bonding through cross-linking in glutaraldehyde solution as a spacer arm on magnetic nanoparticles. Thus, a covalent bond was established between the aldehyde agent glutaraldehyde and the enzyme amino enzyme. To determine the optimum condition for the immobilization process, different times were examined and optimum temperature was selected from literature. In addition to determining the optimum concentration of the enzyme, three concentrations were checked. Also, to reduce diffusion limitation, the strategy of amine-functionalized magnetic nanoparticles was used.

The IMP-1 immobilization data are shown in Table 3. The enzyme was added to the nanoparticles in three 0.17, 0.338, and 0.6775 mg/ml concentrations. At concentrations of 0.17 and 0.338 mg/ml of enzyme, activity recovery was 26.94% and 26.71%, respectively. The immobilization yield was 100%, however, at the concentration of 0.6775 mg/ml, the activity recovery was 25.96% and the immobilization yield was 81.61 ± 0.14%. The results showed that the immobilization yield and recovery activity decrease with increasing enzyme concentration. For the reason that after saturation of the nanoparticles capacity to bind the enzyme, the residual enzyme concentration in the supernatant of the immobilization solution increases and reduces immobilization yield and activity recovery. The lower-activity immobilized 1IMP-1 may be due to conformational changes in the enzyme's active site as a result of strong covalent bonds between the enzyme and the carrier (Zhang et al. 2013). Also, the free enzyme may experience protein aggregation (mainly near the isoelectric point). This may be because of undesired enzyme interactions where inactivation can stabilize incorrect enzyme structures (Siar et al. 2018). This manner was also reported in a 2016 study by Xiao et al., in this study, activity of k-carrageenase immobilized onto magnetic iron oxide nanoparticles was lower than free enzyme (Xiao et al. 2016). The capacity of nanoparticles for the IMP-1 was 221 mg/g nanoparticles, which is more than the amount reported by Lee et al. (2011). The capacity of Ni magnetic nanoparticles for immobilization of six histidine-tagged enzyme was obtained at 36 mg enzyme/1 g nanoparticle . The improved binding capacity of the MNPs is probably caused by the high surface area of their nanostructure (Wang & Lin 2012).

The data for determining the optimum pH of free IMP-1 showed that the optimum pH was 7.5 and changed to 6.5 after immobilization (Figure 7(a)). The relative activity of the immobilized enzyme at optimum pH (7.5) of the free enzyme was equal to 85% and for the free enzyme at optimum pH of immobilized enzyme was 90%. However, according to the optimum pH data, the immobilized enzyme had a suitable activity in the range of 6.5–7.5. The pH is one of the most important parameters that significantly affect the activity of enzymes. Normally, the enzyme activity changes due to the structural conversion of the immobilized enzyme in the optimum pH immobilization process. This change in the optimum pH value can be due to a change in the ionization of the basic and acidic amino acid side chains in the perimeter around the active site of the enzyme (Mazlan & Hanifah 2017). According to the results, the optimum temperatures of free and immobilized enzymes were 70 °C and 60 °C, respectively (Figure 7(b)). However, no significant difference was observed in the immobilized enzyme activity at temperatures of 50, 60, and 70 °C. Differences in the optimum pH and temperature after immobilization has been reported by other study (Bussamara et al. 2012). Various factors may be responsible for these changes, such as the three-dimensional enzyme structural changes that possibly occur during the immobilization procedure. Although the optimum temperature for immobilized enzyme activity was lower than for free enzyme, the data suggest that the immobilized form of IMP-1 may be employed for environmental applications that do not need high temperatures.

The results showed the thermal stability of the IMP-1 decreased after immobilization. The immobilized enzyme retains 44.2% of its initial activity after 1 hour at 70 °C, while this amount is about 57.14% for the free enzyme (Figure 7(c) and 7(d)). In addition, at 50 °C and at different times, no significant difference was observed in the relative activity of free and immobilized enzymes. At 60 and 70 °C, the relative activity of free and immobilized enzymes decreased with increasing incubation time, but the slope of reduced activity for the immobilized enzyme was higher than the free enzyme. The results of this study showed that after immobilization of the enzyme, its thermal stability is reduced, but there is no significant difference in the thermal stability of the free and immobilized enzymes at 50 °C. At 60 °C the stability of the free enzyme was slightly higher than that of the immobilized enzyme, and this difference was evident at 70 °C. With regards to the application temperature of the biocatalyst for antibiotic removal is not high and is at ambient temperature, this decrease stability at 70 °C can be neglected. The decrease in thermal stability of the enzyme after immobilization may be due to the fact that after high-temperature incubation, the activity was measured at 25 °C and the immobilized enzyme could not recover due to binding to the carrier. In the study of Kanbar & Ozdemir (2010), a decrease in the thermal stability of the carbonic anhydrase enzyme was reported after immobilization (Kanbar & Ozdemir 2010).

The immobilization of enzymes has provided several advantages compared to free enzymes: stable biocatalysts, production efficiency, the possibility of reuse, and easy purification of the products. The reusability of enzyme particles is very important when considering enzymatic reactions. Enzyme reusability was considered for continuous enzyme use. Reusability studies showed that some of its activity was reduced after each cycle of immobilized enzyme use. After 15 reuse cycles, 80% of the initial activity was maintained. Various factors are effective in reducing the relative activity of immobilized enzymes after use cycles. In this study, a part of the nanobiocatalyst may have been reduced due to the use of an external magnet to separate the nanobiocatalyst from the reaction solution after each cycle and wash for reuse for the next cycle. This problem can reduce the relative activity of the immobilized enzyme after successive cycles of reuse and washing (Figure 8(b)). Storage stability data of free and immobilized enzyme showed that in the first week of storage time, there is a significant difference in reducing the activity of immobilized enzyme compared to the free enzyme, but in the following weeks, no significant reduction was observed.ed enzyme. After 4 weeks of storage at 4 °C, immobilized IMP-1 retained more than 74% of its initial activity, while the free enzyme value was 80%. However, in a similar study, a decrease in the relative activity of the immobilized TEM-1 enzyme on magnetic nanoparticles was reported to be 70% after 16 days of storage at 4 °C (Gao et al. 2018) which is higher than the amount obtained in the present study. This study observed that free IMP-1 also has appropriate storage stability without adding preservatives at 4 °C (Figure 8(a)). Factors such as organic solvents, temperature changes, and ionic strength can reduce the durability of biocatalysts. Also, enzymes are partially denatured in aqueous solutions, which leads to a decrease in their activity (Dias Gomes & Woodley 2019).

The velocity of reaction at the time the enzyme is saturated with the substrate, is considered the maximum velocity of the reaction (Vmax). The Michaelis constant (k_m) is the concentration of the substrate, when the reaction rate is half the maximum velocity. The K_m and V_max of free and immobilized IMP-1 were determined by Michaelis–Menten and Lineweaver-Burk plots (Figure 9). For free and immobilized enzymes, the highest catalytic activity was observed for Pen G and CEX, whereas Vmax/Km value for CRO was 3-fold (free enzyme) to 10-fold (immobilized enzyme) lower than for Pen G (Table 4). Also, the catalytic activity of the free enzyme (relative Vmax/Km) for CEX substrate is higher than Pen G and they are almost equal after immobility. When the enzyme is immobilized, issues of diffusion and transport arise. If there is an unstirred boundary layer near the surface, the substrate concentration at the surface may differ from the substrate concentration in the bulk fluid. In addition, the product will accumulate on the surface, potentially inhibiting the reaction. This can be a problem in a microfluidic situation, or with carriers of enzymes if they are not kept in motion. The higher the molecular weight of the substrate and product, the bigger the problem, since diffusion rates through the unstirred layer slow down as the molecules get larger. In addition, the free enzyme may experience protein aggregation (mainly near the isoelectric point). This may be caused by undesired enzyme interactions where inactivation can stabilize incorrect enzyme structures (Sanchez et al. 2016).

This requirement may be easily met in the flow-through system with immobilized enzymes because the substrate is continually replenished (Cooney 2011). In the current study, the decrease in the value of V_max for Pen G and CEX after immobilization of IMP-1 may be due to the substrate diffusion limit (Nematian et al. 2020). This manner has been reported in other studies for enzymes that are immobilized in various carriers (Kharel et al. 2004). Moreover, after immobilization, IMP-1 on the carrier was significantly reduced the K_m value, indicating a greater enzyme-substrate affinity which may be because of conformational changes of the enzyme resulting in a higher affinity of these substrates for the IMP-1. However, this decrease was also observed in K_m in another study for lipase immobilization (Aghababaie et al. 2018).

On the contrary of the Pen G and CEX, K_m and V_max of IMP-1 for CRO were increased after immobilization. The free and immobilized IMP-1 for CRO is significantly lower than that of Pen G and CEX, indicating that the IMP-1 is more affinity to CRO. However, K_m and V_max of the IMP-1 increased after immobilization (decreased affinity). This may be due to changes in the three-dimensional conformation of the active site at the immobilized enzyme which has increased access to this substrate (Guzik et al. 2014).

The beta-lactamase activity differs from that found in Enterobacteriaceae and Pseudomonas. In the present study, the agar well diffusion method was used to evaluate the beta-lactamase effect of IMP-1 on the studied antibiotics. The comparison of the diameter of the zone of inhibition of growth of the E. coli between antibiotics and their enzymatic hydrolysis products showed that the antibiotics were inactivated due to the enzymatic hydrolysis. Enzyme function is carried out by the opening of the beta-lactam ring. This ring plays a key role in inhibiting cell wall synthesis and leads to bacterial death (Fuda et al. 2005).

The present study was performed to develop a novel nanobiocatalyst by immobilizing IMP-1 metallo-beta-lactamase on the magnetic nanoparticles to remove beta-lactam antibiotics from aqueous solutions. For this purpose, the recombinant IMP-1 was produced by overexpression of the bla_IMP gene cloned into the pET28a vector in E. coli and purified.

Purified IMP-1 was immobilized on the Synthesized and functionalized Fe₃O₄@SiO₂ nanoparticles by covalent cross-linking process and used to degrade three beta-lactam antibiotics. The HRTEM, FTIR, and EDS studies of the synthesized magnetic nanoparticles show that the synthesis, functionalization, activation of the nanoparticles as enzyme carriers were performed successfully. The optimum pH of free and immobilized enzyme activity was 7.5 and 6.5, respectively, and the optimum temperature of free and immobilized enzyme activity was 70 and 60 °C, respectively. Also, no significant difference was observed in the thermal stability of the enzyme after immobilization at 50 and 60 °C. The immobilized enzyme retained 44.2% of its original activity after one hour at 70 °C, compared with about 57.14% for the free enzyme. No significant difference was observed in the storage stability of immobilized and free enzymes. After 4 weeks of storage at 4 °C, the relative activity of the immobilized enzyme was 73%, while that for the free enzyme was 80%. The immobilized enzyme was reusable and after 15 cycles of reuse had an 80% relative activity. The kinetic parameters (K_m and V_max) for the three studied antibiotics showed that the enzymatic hydrolysis rate of Pen G, CEX, and CRO is reduced by the enzyme, respectively. The results also showed that these antibiotics were completely inactivated by enzymatic hydrolysis. The results of this study also showed that the enzymatic process using IMP-1 enzyme immobilized on silica-modified magnetic nanoparticles can be used as a promising approach with high efficiency and inexpensive to remove beta-lactam antibiotics from aqueous solutions. Thus, this nanocatalyst could be an appropriate choice for removing beta-lactam antibiotics from pharmaceutical industry wastewater.

R.K., A.A., and M.S. designed the study. M.S. and F.K. carried out the experiments. M.S. wrote the manuscript with support from R.K. and F.K., and S.M. All authors discussed the results and contributed to the final manuscript.

The authors gratefully acknowledge Kermanshah University of Medical Science (Grant Number: 990080) for the financial support. This work was performed in partial fulfillment of the requirement for the environmental health engineering Ph.D. of Mohammad javad Shokoohizadeh, in Faculty of Health, Kermanshah University of Medical Sciences, Kermanshah, Iran.

All relevant data are included in the paper or its Supplementary Information.