Background

Metal–organic frameworks (MOFs) show attractive applications in various fields including gas adsorption (Li et al. 2009) and chemical separation (Maes et al. 2010), and catalysis (Lee et al. 2009). Zeolitic imidazolate framework (ZIF) materials belong to an important class of MOF (Phan et al. 2010), which exhibit the tunable pore size, chemical functionality of classical MOFs, exceptional chemical stability, and structural diversity of zeolites (Wu et al. 2007). Because of these features, ZIFs show great promise for enzyme immobilization (Hou et al. 2017; Wu et al. 2017). Hou et al. (2015) reported the construction of mimetic multi-enzyme systems by embedding GOx in ZIF-8 and application of this system as biosensors for glucose detection, exhibiting extraordinary electro-detection performance, and lower detection limit. Lyu et al. (2014) reported one-step immobilization process for protein-embedded metal–organic frameworks with enhanced activities, in which the Cyt c immobilized by ZIF-8 carrier exhibited an enhancement of enzyme activity compared with free Cyt c.

In order to separate the MOF-based materials easily, previous studies have reported on MOF-functionalized magnetic nanoparticles which can be recycled under magnetic field and have excellent physical and chemical characters of the MOF shell (Ke et al. 2012). Further investigations on magnetic MOFs with core–shell structure are still needed because the controllable growth of the MOF crystals on the magnetic nanoparticles remains a great challenge (Nong et al. 2015). For instance, before the MOF crystal growth process, the magnetic nanoparticles need to be surface-modificated by styrene sulfonate (Zhang et al. 2013), polyacrylic acid (Jin et al. 2014), chitosan (Xia et al. 2017), and SiO2 (Wehner et al. 2016). It is worthy to note that cellulose may be acted as a promising surface material, because of its abundance of hydroxyl groups. These hydroxyl groups may promote the adsorption of metal ions for the formation of MOF crystal (Liu et al. 2012). Cellulose is the most abundant renewable polysaccharide on earth, which is sustainable, biocompatible, biodegradable, and non-toxic (Lavoine et al. 2012). Previous studies showed that cellulose can dissolve in NaOH/urea aqueous media under – 12 °C, and surface modified the Fe3O4 magnetic nanoparticles (Cai and Zhang 2005). Thus, investigating the growth of the MOF onto the cellulose-modified Fe3O4 is of interest.

The glucose oxidase (GOx) is an aerobic dehydrogenation enzyme, which has played an important role on de-oxidization, glucose removal, and gluconic acid synthesis. It is widely used in forage, medicine, and other fields (Wong et al. 2008). In recent years, in order to overcome the disadvantages of the free GOx such as poor mechanical stability, difficult separation, and non-recyclability (Cao et al. 2016), several nanoparticles, such as titanium dioxide nanotubes (Ravariu et al. 2011), Fe3O4/APTES (França 2014), Ag@Zn-TSA (Dong et al. 2016), and ZIF-8 (Wu et al. 2015) were attempted to be used as enzyme carriers for the immobilization of GOx. Among these nanoparticles, carriers containing metal–organic frameworks (MOFs) have received more and more concern because of their excellent physical and chemical properties mentioned above.

In this study, a new core–shell magnetic ZIF-8-coated magnetic regenerated cellulose-coated nanoparticle (ZIF-8@Cellu@Fe3O4) was fabricated. The as-prepared ZIF-8@Cellu@Fe3O4 was structurally characterized in detail. The glucose oxidase (GOx) was embedded in the pores of the ZIF-8@Cellu@Fe3O4 with a high relative activity recovery and protein loading.

Methods

Preparation of magnetic regenerated cellulose-coated nanoparticle (Cellu@Fe3O4)

The Fe3O4 nanoparticles were fabricated by co-precipitation method according to our previous literatures (Deng et al. 2016; Cao et al. 2017): 2.43 g FeCl3·6H2O and 0.9 g FeCl2·4H2O were dissolved in 200 mL deionized water at room temperature. The mixture was added dropwise into a 25% ammonia solution with stirring, N2 purge, and the pH at 10. The temperature was raised to 60 °C and kept for 1 h; the magnetite precipitate was collected with an external magnet and washed three times with deionized water.

150 mg of Fe3O4 was dispersed in 30 mL aqueous solution containing 7 wt% of NaOH and 12 wt% of urea and pre-cooled to – 12 °C for more than 1 h. Then, 100 mg of microcrystalline cellulose was added into the above suspension. After 1 h of freezing, the microcrystalline cellulose was dissolved completely. Then the deionized water was mixed with the above mixture and the cellulose-coated Fe3O4 (Cellu@Fe3O4) was formed. The Cellu@Fe3O4 was collected with an external magnet and washed three times with deionized water.

Preparation of magnetic ZIF-8 nanoparticles (ZIF-8@Cellu@Fe3O4)

Zinc nitrate hexahydrate was dissolved in deionized water (40 mM, 2 mL) mixed with 10 mg Cellu@Fe3O4 under stirring for 20 min. Then 2-methylimidazole (160 mM, 2 mL) was added into the mixture and stirred for 3 h (Liang et al. 2015).

Preparation of GOx-loaded ZIF-8@Cellu@Fe3O4 nanocomposite

The synthesis processes of GOx-loaded ZIF-8@Cellu@Fe3O4 nanocomposites are illustrated in Scheme 1. Zinc nitrate hexahydrate was dissolved in deionized water (40 mM, 2 mL) mixture and stirring with 10 mg of Cellu@Fe3O4 for 20 min. Then 2-methylimidazole (160 mM, 2 mL) was added into the mixture and stirred for 10 min (Lyu et al. 2014; Liang et al. 2015; Du et al. 2017). Free GOx was dissolved in buffer (200 mM, pH 4.0–8.0); 0.2 mL free enzyme (100 U/mL) was added into solutions. The reaction lasted for (0.5–3 h) and immobilized at (10–50 °C, 200 rpm). The immobilized GOx was separated through an external magnetic field. Then, the un-immobilized GOx was removed by continuous washing until no protein was detected. The washing solutions were collected to detect the amount of un-immobilized GOx. The amount of immobilized GOx loaded on the ZIF-8@Cellu@Fe3O4 was calculated as the difference between the initial and the un-immobilized GOx. The GOx-loaded ZIF-8@Cellu@Fe3O4 was named as GOx-ZIF-8@Cellu@Fe3O4.

Scheme 1
scheme 1

Preparation scheme of GOx@ZIF-8@Cell@Fe3O4

Full size image

Enzyme activity assay and protein concentration

Protein concentration was determined according to the Bradford method using bovine serum albumin as standard (Lowry et al. 1951).

The activities of free and immobilized glucose oxidase were determined by indigo carmine method (Zhou et al. 2008). Glucose oxidase was dissolved in 1 mL phosphate buffer (200 mM, pH 7.0) and then 4 mL 0.2 mol/L glucose solution was added. The solution was mixed at 37 °C for 10 min. 3 mL acetic acid–sodium acetate (0.1 M acetic acid 500 mL and 0.1 M sodium acetate 30 mL, pH 3.5) was added as buffer solution, and 1.3 mL indigo carmine (0.1 mM) was used as redox indicator. After treated at 100 °C (boiling water) for 13 min, the absorbance of solution at wavelength of 615 nm was measured.

Activity recovery (%) was calculated as follows:

$$= \,100\, \times \,\frac{{{\text{Activity}}\;{\text{of}}\;{\text{immobilized}}\;{\text{enzyme}}\, ( {\text{U)}}}}{{{\text{Activity}}\;{\text{of}}\;{\text{free}}\;{\text{enzyme}}\;{\text{used}}\;{\text{for}}\;{\text{immobilization}}\, ( {\text{U)}}}}.$$

Enzyme loading (%) was calculated as follows:

$$= \,100\, \times \,\frac{{{\text{Enzyme}}\;{\text{content}}\;{\text{of}}\;{\text{immobilized}}\;{\text{enzyme}}\;\left( {\text{mg}} \right)}}{{{\text{Content}}\;{\text{of}}\;{\text{enzyme}}\;{\text{used}}\;{\text{for}}\;{\text{immobilization}}\;\left( {\text{mg}} \right)}}.$$

Results and discussion

Characterization of Cellu@Fe3O4 and ZIF-8@Cellu@Fe3O4

The X-ray diffraction patterns of the ZIF-8@Cellu@Fe3O4, Cellu@Fe3O4, microcrystalline cellulose, and naked Fe3O4 are shown in Fig. 1. However, the microcrystalline cellulose showed three peaks at 2θ = 14.8°, 16.5°, and 22.7° assigned to the (110), (110), and (200) planes which were characteristic peaks for the cellulose crystalline (I) (Edwards et al. 2012). By comparison, the Cellu@Fe3O4 displayed three diffraction peaks at 2θ = 12.4°, 20.2°, and 22.2° assigned to the (110), (110), and (200) planes of cellulose crystalline (II) (Togawa and Kondo 1999). This illustrated that after dissolving-regeneration process, the cellulose crystalline form changed from cellulose (I) to cellulose (II) (Carrillo et al. 2004). Moreover, the Cellu@Fe3O4 showed four distinct peaks at 2θ = 30.24°, 35.60°, 43.24°, and 57.16°, ascribing to the crystal plane diffraction peaks of the (220), (400), (422), and (511) diffraction peaks for Fe3O4 (JCPDS Card No. 19–0629) (Cao et al. 2014). The ZIF-8@Cellu@Fe3O4 XRD pattern is also shown in Fig. 1. The result showed that visible diffraction peaks at about 2θ = 7.3°, 10.5°, and 18.0° were assigned to the characteristic diffraction peak of ZIF-8 (Pan et al. 2011). These results indicated the formation of the ZIF-8@Cellu@Fe3O4.

Fig. 1
figure 1

The powder X-ray diffraction patterns of ZIF-8@Cellu@Fe3O4, Cellu@Fe3O4, and microcrystalline cellulose

Full size image

Figure 2 shows the FTIR spectra of ZIF-8@Cellu@Fe3O4, Cellu@Fe3O4, and microcrystalline cellulose. For the microcrystalline cellulose (Fig. 2a), the band at 1431 cm−1 was attributed to C–O–H stretching vibration. However, for Cellu@Fe3O4 (Fig. 2b), the peak at 1431 cm−1 disappeared and peak at 1421 cm−1 was observed (Colom and Carrillo 2002). This also illustrated that after dissolving-regeneration process, the cellulose crystalline form of Cellu@Fe3O4 changed from cellulose (I) to cellulose (II) (Colom and Carrillo 2002). Moreover, there was a strong absorption peak of Cellu@Fe3O4 at around 594 cm−1 assigned to the characteristic peak of Fe3O4 (Cornell et al. 1999). Also, the bands of Cellu@Fe3O4 at 3440 cm−1, attributed to hydrogen bonding of cellulose, became broader and weaker, illustrating the strong interaction between Fe3O4 and cellulose layer which were observed (Kondo et al. 1994; Kondo and Sawatari 1996; Zhang et al. 2001). The FTIR spectrum shown in Fig. 2c displays the chemical composition of the ZIF-8@Cellu@Fe3O4. A strong peak at 421 cm−1 is ascribed to the Zn–N stretch mode (Zhang et al. 2013). The broad bands around 500–1350 and 1350–1500 cm−1 were assigned as the plane bending and stretching of imidazole ring, respectively (Lu et al. 2012). These results showed that the ZIF-8 was successfully composited on to the surface of the Cellu@Fe3O4 by co-precipitation method.

Fig. 2
figure 2

FT-IR spectra of microcrystalline cellulose (a), Cellu@Fe3O4 (b), and ZIF-8@Cell@Fe3O4 (c).

Full size image

As shown in the scanning electron microscope (SEM) graphy (Fig. 3), the Cellu@Fe3O4 has an average diameter of around 29.7 nm and displays uniform structure and morphology. The size of ZIF-8@Cellu@Fe3O4 was approximated to 170 nm.

Fig. 3
figure 3

SEM graphy of Cellu@Fe3O4 (A), ZIF-8@Cellu@Fe3O4 (B)

Full size image

The vibrating specimen magnetometer (VSM) magnetization curves of the Cellu@Fe3O4 and ZIF-8@Cellu@Fe3O4 are shown in Fig. 4. Saturation magnetization (M S) was used to measure the magnetization of samples defined as the maximum magnetic response of a material in an external magnetic field (Xiao et al. 2014). It is observed that the M S of Fe3O4 is 21.37 emu/g and of ZIF-8@Cellu@Fe3O4 (4.9 emu/g) is lower than that of Cellu@Fe3O4 nanoparticles (12.8 emu/g).

Fig. 4
figure 4

Hysteresis loops of Fe3O4, Cellu@Fe3O4, ZIF-8@Cellu@Fe3O4, and GOx-ZIF-8@Cellu@Fe3O4

Full size image

Immobilization of GOx by ZIF-8@Cellu@Fe3O4 nanocomposite

Figure 5 shows the effect of buffer pH on enzyme activity recovery and enzyme loading. The results show that the highest activity recovery of immobilized glucose oxidase (GOx-ZIF-8@Cellu@Fe3O4) displayed at pH 7.0 (123.7%), with the protein loading 91.3 mg/g. The GOx-ZIF-8@Cellu@Fe3O4 exhibited relative high activity at faintly acid and neutral conditions (pH 6.5–7.0), and became deactivated in the both the acid and alkaline conditions (Cao et al. 2008). Also, ZIF-8 is very stable in neutral and basic conditions (Jian et al. 2015) and exhibit best enzyme encapsulation capacity at pH 7.5–8.0. Thus, the protein loading capacity changed depending on different pH values and the optimal pH value for GOx immobilization was 7.0.

Fig. 5
figure 5

Effects of buffer pH on activity recovery and enzyme loading of immobilized GOx

Full size image

Figure 6 shows the effect of immobilization temperature on activity recovery and enzyme loading. The results showed that the GOx-ZIF-8@Cellu@Fe3O4 with highest activity recovery was obtained at 20 °C. When the temperature was higher than 30 °C, enzyme activity recovery decreased significantly. Thus, 20 °C was selected as the immobilization temperature in the following experiment.

Fig. 6
figure 6

Effects of immobilization temperature on activity recovery and enzyme loading of immobilized GOx

Full size image

Figure 7 shows the effect of immobilization time on enzyme activity recovery and enzyme loading. The result shows that when the immobilization time was 1 h, the highest relative activity was obtained. After 1 h, the relative activity decreased gradually, which was possibly due to the partial inactivation of the enzyme with immobilized time prolonged. In conclusion, the optimum immobilization time was 1 h at which the activity recovery of GOx attained 124.2% and the enzyme loading was 94.26 mg/g.

Fig. 7
figure 7

Effects of immobilization time on activity recovery and enzyme loading of immobilized GOx

Full size image

Generally speaking, the optimal condition of the immobilization process was that buffer pH 7, temperature 20 °C, and immobilization time 1 h. At this condition, the activity recovery was 124.2%, and the enzyme loading was 94.26 mg/g. The specific enzyme activity values of immobilized and free GOx were 12.42 and 10.00 U/mg, respectively. The thermal stability of free and immobilized GOx at 65 °C was performed. The activities of both free and immobilized GOx decreased gradually with increasing of the incubation time. The immobilized GOx exhibited more than 40% of its initial activity after 4 h of incubation, while that of free GOx was about 13%. This result showed that the thermal stability of immobilized GOx was enhanced after immobilization (Zhou et al. 2012a, b). As a comparison, GOx immobilized onto a 3-(aminopropyl)triethoxysilane (APTES)-coated Fe3O4 nanocarrier which retained less than 50% of the native GOx activity (Park et al. 2011). Thus the immobilization process in the present study is promising for enzyme immobilization.

A comparative study shown in Fig. 8 explores the mechanism of the enhanced enzyme relatively activity after immobilization. The presence of Cellu@Fe3O4, ZIF-8 and 2-methylimidazole could not enhance the activity of GOx, while that of Zn2+ increased the activity of GOx by about 9.3%, which was similar with the previous literature (Lyu et al. 2014). Comparison shows that the immobilization of GOx in ZIF-8@Cellu@Fe3O4 enhanced the activity of GOx by 24.2%. These results showed that the GOx-ZIF-8@Cellu@Fe3O4 exhibited an increased relative activity compared to the free GOx. This may be attributed to the following reason: the enzyme immobilization process changed the enzyme conformation and increased the substrate affinity toward the glucose; the interaction between the GOx and Zn2+ in ZIF-8 enhanced the catalytic activity (Lyu et al. 2014).

Fig. 8
figure 8

The relative peroxidase activity of GOx, GOx-ZIF-8@Cellu@Fe3O4 composite, GOx/zinc ion mixture, GOx/2-methylimidazole mixture, GOx/ZIF-8 mixture, and GOx/Cellu@Fe3O4 mixture

Full size image

Conclusions

In conclusion, Glucose oxidase (GOx) was successfully immobilized onto the biocompatible ZIF-8@Cellu@Fe3O4 via co-precipitation process. Its morphology, structure, and magnetic properties were determined. The GOx immobilized in ZIF-8@Cellu@Fe3O4 had high protein loading (94.26 mg/g) and enhanced relative activity recovery (124.2%). The results of the present work provide an efficient enzyme immobilization process and promote the application and development of immobilized enzyme catalysis.