**1. Introduction**

Carbonic anhydrase (CA, E.C. 4.2.1.1), a zinc metalloenzyme, can efficiently catalyze the reversible hydration of CO2 to bicarbonate ion [1,2]. In nature, chemical fixation of CO2 usually requires high temperatures, high pressure and long reaction times, while carbonic anhydrase can fix carbon dioxide and have a high reaction rate under low pressure conditions [3]. With the intensifying greenhouse effect, the use of carbonic anhydrase is considered as a green and efficient strategy for bio-mineralization of CO2 [4–7]. However, the poor stability and reusability of carbonic anhydrase limits its application in practice [6,8]. Enzyme immobilization on solid supports has been developed to overcome these shortcomings [9–13]. Previous immobilization studies mainly focused on commercial carbonic anhydrase on a wide range of solid supports such as: polyurethane foam [14], SBA-15 [15], mesoporous silica [16], MIL-160 [9] and ZIF-8 [17]. However, high cost and protracted immobilization processes greatly limit the application of immobilized CA [4,10,12].

Fusion protein expression is a common strategy to obtain enzyme proteins. Due to the convenience of purification using Ni-chelating affinity chromatography, histidine tags are widely used for enzyme protein expression [18,19]. Nickel ions modified materials for enzyme purification and immobilization

were developed which are based on the a ffinity of nickel ions and histidine labels [20–22]. However, until now, there are few piece of research on the one-pot immobilization of CA via Ni2<sup>+</sup> modified materials. The introduction of histidine-tag in CA is more conducive to reducing costs and shortening immobilization time.

Metal-organic frameworks (MOFs) are a new kind of organic-inorganic hybrid porous material [23,24]. Since MOF materials are considered to be potent supporting matrices for enzyme immobilization, many kinds of commodity enzymes were immobilized successfully [25–28]. However, other researchers have spent their energy on the design of Zn2+, Cu2+ or Fe3<sup>+</sup>-based MOFs for enzyme immobilization [29–31]. Inspired by the specific a ffinity between Ni2<sup>+</sup> and histidine-tags [32], we focus on utilizing Ni-based MOFs to achieve one-pot immobilization of histidine-tagged enzymes. To date, a number of Ni-based MOFs have been reported, however, studies on immobilized enzymes by Ni-based MOFs are rare [33]. Therefore, it is worthwhile to focus on the adsorption characteristics of Ni-based complexes itself.

Herein, the recombinant carbonic anhydrase II from human (hCA II) with His-tag and without His-tag were successfully designed and overexpressed in *Escherichia coli* (*E. coli*) (DE3). Meanwhile, Ni-based MOF (Ni-BTC) nanorods were synthesized by employing a one-pot hydrothermal process and were used as immobilization support for CA. Compared to our previous work [34] in this work we were able to carry out the protein immobilization process at room temperature, while achieving very high binding e fficiencies. Our approach to realizing His-hCA II one-pot immobilization by Ni-BTC is shown in Figure 1. As a result, the His-tagged hCA II could be e fficiently immobilized on Ni-BTC under optimal conditions by a simple mixing step. However, when hCA II (without His-tag) was mixed with Ni-BTC under the same conditions, the hCA II (without His-tag) could not be e fficiently immobilized on Ni-BTC. This means that Ni-BTC has specific binding abilities with regards to His-hCA II. Furthermore, a specific protein purification step could be omitted in our application, which could save on production costs in practical applications. In addition, the stability of the free and immobilized His-hCA II under di fferent conditions was also investigated.

#### **2. Results and Discussion**

#### *2.1. Synthesis and Characterization of Ni-BTC*

Because of the a ffinity of Ni2<sup>+</sup> and histidine tags, the nickel-modified materials could easily realize the His-tagged enzyme's separation and immobilization in one step. However, most of the previously reported Ni2<sup>+</sup> modified materials required complex and lengthy reaction conditions [20,21]. In this work, the light green Ni-based MOF materials were easily obtained and used as support for His-tagged enzyme immobilization.

TEM and SEM were used to characterize the morphology of Ni-BTC. As shown in Figure 1A,B respectively, the as-prepared Ni-BTC were nanorods with an average length of 3 μm and an average width was 0.5 μm. Furthermore, the Ni-BTC had a uniformly smooth long, stick-like structure and it was highly aggregated. Next, the distribution of elements in nanorods was studied by energy dispersive X-ray spectral element mapping. The elemental distribution of elements in Figure 1C was known, and we found that carbon, nitrogen, oxygen and nickel element were distributed homogeneously throughout the Ni-BTC structure. Our results with regard to the structure of synthesized Ni-BTC MOF materials (Figure 1D) appear to be in agreemen<sup>t</sup> with those reported elsewhere [35]. These results were also in agreemen<sup>t</sup> with our X-ray spectroscopy (XPS) spectrum (Figure S1). Signal peaks of Ni 2p, O 1s, N 1s, and C 1s were detected, indicating that Ni-BTC was successfully synthesized with nickel acetate and H3BTC.

The chemical structure of Ni-BTC nanorods had been characterized previously which provided us with knowledge about the linker's integration and the oriented growth of the Ni-BTC nanorods and its cubic geometry [35]. As shown in Figure S2, Ni-BTC nanorods kept stable at temperatures below 100 ◦C and showed weight loss (26.1 wt %) between 100 ◦C and 300 ◦C, the reason might be the vaporization of water and solvent. Significant weight loss (43.1 wt %) occurred between 300 ◦C and 600 ◦C, which could be assigned to the decomposition of H3BTC [34].

**Figure 1.** TEM images (**A**) and SEM images (**B**) of Ni-BTC and (**C**) corresponding carbon; nitrogen; oxygen and nickel elemental mapping of Ni-BTC; (**D**) The synthesis process and structure of the 2D Ni-MOF.

#### *2.2. Optimal Conditions for His-hCA II Immobilization*

We first prepared the His-hCA II with a histidine-tag in *E. coli* BL21 (DE3). The plasmid map of pETDuet-1-His-hCA II is shown in Figure S3. Following overexpression in *E. coli* BL21 (DE3), His-hCA II was extracted in a phosphate buffer (50 mM, pH 7.4). The purified His-hCA II appeared as a single band on the silver-stained gels (Figure S4). In order to ge<sup>t</sup> the best fixed conditions, 100 μL of His-hCA II crude cell lysate was incubated with Ni-BTC under different conditions. The influence of mass ratios on Ni-BTC to crude protein on protein loading and activity recovery of His-hCA II @Ni-BTC were investigated. As shown in Figure 2A, activity recovery of His-hCA II @Ni-BTC increased somewhat from a ratio of 2.5:1 to 5:1 and declined as the ratio increased further. The decline in activity recovery can possibly be attributed to the greater inclusion of heteroproteins from within the crude extract at the higher concentrations adsorbed on the Ni-BTC. Protein loading of His-hCA II @Ni-BTC as the ratio increased from 2.5:1 to 12.5:1 (Figure 2A) increased linearly. The elevated protein loading at a ratio of 12.5:1 also implies the increased adsorption of Ni-BTC on hybrid protein. On the basis of the maximum activity recovery of His-hCA II @Ni-BTC at the ratio of 5:1, we selected this protein loading ratio for further experiments. In addition, we also calculated the amount of immobilized His-hCA II (Table S1), the quality of bounding protein was calculated by the difference between the amount of protein added to the reaction mixture and that in the leachate and washing solutions after immobilization. As a result, the catalytic activity of an immobilized enzyme was 3.1 times that of a free enzyme and the protein yield of crude cell lysate was 31.8%.

The effects of immobilization temperature (Figure 2B) and time (Figure 2C) of Ni-BTC on the loading and activity recovery of His-hCA II @Ni-BTC and crude protein were also investigated. Similar to previous reports, [4,10] although the amount of protein binding increased, the activity recovery reached its maximum value at 30 ◦C and then decreased with an increase of the temperature from 30 ◦C to 50 ◦C. Human carbonic anhydrase has an optimum temperature close to human body temperature to retain its activity. An increase in the immobilized time showed a near linear increase in the activity recovery of immobilized His-hCA II, increasing from 61.5% to 88.2% from 15 to 60 min, and a further increase in activity recovery from 60 to 120 min to 95% (Figure 2C). Consequently, an immobilization time of 120 min was employed for further experiments. Above all, 3 mg Ni-BTC and 100 μL cell lysate were mixed with 900 μL of phosphates buffer (50 mM, pH 8.0) for half an hour at 30 ◦C, then collected by centrifuge for 10 minutes at 12,000 rpm.

We also investigated whether the different monomers in His-hCA II @Ni-BTC could accelerate the relative catalytic activity at equal protein concentrations of His-hCA II @Ni-BTC and His-hCA II. We found that His-hCA II @Ni-BTC and His-hCA II had similar catalytic activity (*p* > 0.5) with an activity recovery of 74.5% (Figure 2D). Whereas Ni-BTC and Ni2<sup>+</sup> had very low activity recovery (<10%) and H3BTC had almost no catalytic activity of its own. The reason for the low recovery activity observed for Ni-BTC and Ni2<sup>+</sup> might be that the constituent Ni2<sup>+</sup> absorbed some of the substrate and presented a minor false positive result. Hence, we conclude that the activity recovery of His-hCA II @Ni-BTC was mainly derived from the His-hCA II enzyme. In addition, we studied the specific binding ability of His-hCA II to Ni-BTC compared to hCA II (not His-tagged). In this experiment, equal quantities of hCA II and His-hCA II were incubated with different amounts of Ni-BTC at 25 ◦C for 30 min. The results indicate that the Ni-BTC has a much greater specific binding ability to His-hCA II compared to the non-His-tagged hCA II (Figure 2E).

The activity recovery of immobilized His-hCA II was 98.99% in optimal conditions, which proved that Ni-BTC could achieve high activity recovery of His-hCA II by a simple one-pot mixing procedure.

**Figure 2.** Optimizing immobilization conditions: (**A**) ratio of the carrier to enzyme; (**B**) immobilization temperature; (**C**) immobilization time; (**D**) the relative catalytic activity of His-hCA II @Ni-BTC, purified His-hCA II, Ni-BTC, Ni2+, H3BTC; (**E**) the protein binding rate of Ni-BTC to His-hCA II and hCA II.

#### *2.3. Enzyme (His-hCA II) Stability*

The pH and thermal stability of the environmental conditions are important for practical application. The stability of free His-hCA II and His-hCA II @Ni-BTC was tested at different pH conditions (phosphate buffer, 50mM, at 30 ◦C, from pH 6 to 10). After 4 h of incubation at pH 6, the reaction was performed for 5 min at room temperature. Free His-hCA II maintained 30.8% of the original activity (pH 10 and 30 ◦C), whereas the His-hCA II @Ni-BTC retained 73.7% of its original activity (pH 10 and 30 ◦C) (Figure 3A). Furthermore, the free and His-hCA II @Ni-BTC maintained 59.5% and 80.7% of its original activity after incubation at pH 7 for 4 h. These results indicated that His-hCA II @Ni-BTC exhibits better stability than free hCA II under acidic condition, which is in agreemen<sup>t</sup> with results shown elsewhere [36]. We also tested the effects of different pH on the shedding of His-hCA II proteins from the Ni-BTC support. The results (Figure S5A) indicate that His-hCA II @Ni-BTC had different degrees of dissociation at different pH solutions. However, the degree of dissociation (protein shedding) was relatively low and had no effect on the overall catalytic activity, further demonstrating the overall pH stability of His-hCA II @Ni-BTC.

**Figure 3.** Stability of the immobilized and free His-hCA II (**A**) at different pHs and (**B**) temperatures.

The thermal stability of His-hCA II @Ni-BTC compared to with free His-hCA II was investigated, too. The various temperature exposures were carried out at the temperatures indicated, while the actual enzyme assays were performed at 25 ◦C in a phosphate buffer (50 mM, pH 8.0). The results showed that the relative activities of free enzyme and immobilized enzyme decreased significantly at temperatures over 30 ◦C (Figure 3B), with the immobilized enzymes performing worse than the free enzymes. Previous studies also reported a marked loss of immobilized CA activity at these temperature [4,10]. Additional experiments were conducted to detect the protein content in His-hCA II @Ni-BTC supernatant following exposure to elevated temperatures (40 and 50 ◦C), which showed a 23.7% and 44.8% loss of proteins, respectively (Figure S5B). According to the protein shedding rate of His-hCA II @Ni-BTC at different temperatures, Figure 3B shows that the relative activity of His-hCA II @Ni-BTC could reach 82.3% at 50 ◦C while the protein concentration of His-hCA II @Ni-BTC was consistent with free enzyme. It was proved that Ni-BTC has certain supporting and protective effects on enzymes. Allowing for these protein losses as lost immobilized His-hCA II @Ni-BTC, the adjusted residual His-hCA II activity outperformed the free His-hCA II. While this might not explain the exact underpinning reasons as to why the His-hAC II proteins dissociate from the Ni-BTC support, it does explain the lower activity of His-hCA II @Ni-BTC at elevated temperatures might be because the high temperature promoted His-hCA II shedding from His-hCA II @Ni-BTC. However, at the same protein concentration, as shown in Figure S6, the catalytic activity of His-hCA II @Ni-BTC was 2.4 times that of free enzymes after being maintained at 50 ◦C for 4 h. Thus, the protective and supporting effect of Ni-BTC on the protein under thermal conditions significantly improved the stability of the enzyme.

The storage stability of free and immobilized His-hCA II was also investigated (Figure 4A). At the end of two days of storage in a phosphate buffer (50 mM, pH 8.0) at 25 ◦C, the free enzyme kept 41.8% of its original activity (i.e., a 58.2% loss), whereas the immobilized enzyme retained 66.4% (i.e., a 33.6% loss) of its original activity. These losses extended further and by the end of 10 days of storage the free enzyme kept 6.4% of its original activity (i.e., a 93.6% loss), whereas the immobilized enzyme retained 41.8% (i.e., a 58.2% loss) of its original activity after 10 days of storage. During prolonged storage, denaturation and degradation of His-hCA II proteins could significantly reduce enzyme activity. However, the immobilization of His-hCA II on Ni-BTC provided physicochemical stability and mechanical protection, which enabled the His-hCA II to maintain greater stability for long-term reactions compared to free His-hCA II. Finally, the immobilized His-hCA II was tested for recycling. The immobilized His-hCA II was separated by centrifugation after running an experimental standard reaction, after which it was washed three times in water and redispersed in the buffer by vortex mixing before repeating another experimental standard reaction. There was a gradual decrease in the remaining relative enzyme activity at each recycling, however, more than 65% of original activity remained after eight cycles (Figure 4B). This might be due to the gradual shedding of some

enzymes from Ni-BTC, and some denaturation during the recovery process. The notion of a gradual release of the enzyme from the Ni-BTC support was confirmed in a subsequent experiment (Figure S7). By determining the protein content in the supernatant from each cycle, we found that each cycle's process was accompanied with protein shedding.

**Figure 4.** (**A**) Stability of the immobilized and free His-hCA II of storage stability at 25 ◦C; (**B**) reusability of His-hCA II @Ni-BTC.

### *2.4. Kinetic Parameters*

To further evaluate the catalytic capacity of immobilized enzymes, we further compared the kinetics of free enzymes and His-hCA II @Ni-BTC. As presented in Table S2, the Km and Vmax values for free and His-hCA II @Ni-BTC were determined to be 1.82 mmol/L, 1.96 mmol/L and 0.037 mmol/min, 0.035 mmol/min for p-NPA, respectively. The substrates and possible conformational changes of the mass transfer limiting structure play an important role in increasing the Km value of the His-hCA II @Ni-BTC after immobilization.

#### **3. Materials and Methods**
