An enhanced chemiluminescence bioplatform by confining glucose oxidase in hollow calcium carbonate particles

A chemiluminescence (CL) amplification platform based on HCC/Lucigenin&GOx (HLG) film was developed. Hollow structural calcium carbonate (HCC) particles were used as alternative materials for carrying both enzyme and CL reagent. The model enzyme (GOx), immobilized in confined space of HCC particles, exhibited an improved biocatalysis. The Michaelis constant (Km) and the enzymatic rate constant (kcat) were determined to be 0.209 μM and 2.21 s−1, respectively, which are much better than those of either free GOx in aqueous solution or the GOx immobilized on common nanomaterials. Based on the HLG platform, CL signal was effectively amplified and visualized after adding trace glucose, which could be attributed to the HCC particles’ high biocompatibility, large specific surface area, attractive interfacial properties and efficient interaction with analyses. The visual CL bioplatform showed an excellent performance with high selectivity, wide linear range and low detection limit for sensing trace glucose. Because it eliminates the need of complicated assembly procedure and enables visualization by the naked eye, the sensitive and selective CL bioplatform would provide wide potential applications in disease diagnosis and food safety.

a mixture solution of GOx and lucigenin on HCC film to get a HLG film. When glucose was incubated with the HLG film to generate H 2 O 2 , enzyme reaction kinetic parameters such as enzymatic rate constant kcat and Michaelis-Menten constant (Km) on the CL platform were studied, better than those of free GOx and the immobilized GOx on the reported grapheme oxide CL system. A high sensitivity with a LOD of 9.0 pM and good selectivity were shown. All advantages would be ascribed to the HCC platform that not only provides high ratio of surface-to-volume for better mass and energy transfer, but also gives rise of confine effect for facilitating the maintenance of bioactivity and the high affinity toward substrate of immobilized enzyme. No need of complicated assembly procedure and pretreatment of sample, the sensitive and selective CL platform would provide wide potential applications in disease diagnosis and food safety.

Results and Discussion
Design strategy. Figure 1 represents the schematic procedure for fabricating the HLG film and the CL amplification strategy for sensing glucose. As a proof-of-concept, HCC particles with unique hierarchical hollow structure are prepared using a reported starch template 12 . Due to porosity structure and high interfacial area of HCC particles, both GOx and lucigenin can be efficiently entrapped via physical adsorption. It was supposed that confine effect of the porous structural HCC particles might maintain even and improve biocatalysis of the immobilized enzyme. Furthermore, the porous structural HLG platform has a large geometric surface area, which might be in favor of efficient interaction between the immobilized molecules (like GOx or CL regent) and analytes. In the present of glucose, H 2 O 2 can be generated because of the GOx biocatalyst reaction happening on the HLG film. After the injection of NaOH, CL signal can be detected due to CL reaction of lucigenin with H 2 O 2 . Owing to the enhanced enzyme bioactivity caused by confine effect of HCC, CL signal can be amplified effectively, which would result in a sensitive and specific detection of glucose. Consequently, glucose concentration can be sensitively quantified owing to amplified CL response. The facile and novel CL biosensing platform based on the HLG film would have potential applications including disease diagnosis and food safety.
Characterization. SEM image shows that the HCC particles have an average diameter of about 550 ± 50 nm and exhibit a porous spherical morphology (Fig. 2b). It is ascribed to soluble starch, used as a template molecular, can couple the crystallization of CaCO 3 nanoparticles and in situ symmetry-breaking assembly of these crystallites into hollow spherical shells 12 . Consequently, plenty of defects and pores are observed on the HCC surface. The HCC particles maintained their porous structure well even after a coating process using GOx and lucigenin solution (Fig. 2c), which was consistent with XRD results.
Characteristic diffraction peaks of vaterite polymorph, such as (1 1 0), (1 1 2), (1 1 4), (3 0 0) and (1 1 8) were observed for both HCC film and HLG film (Fig. 2B,C). It is worth noting that the HLG film maintained vaterite polymorph and GOx well (Fig. 2D). This indicates that the HCC particles of the HLG film not only keep a stable hollow structure but also provide good vessels to hold GOx. N 2 adsorption-desorption isotherms were further used to measure specific surface area of the HCC particles. After the loading of GOx and lucigenin, their specific surface area effectively decreased from 64.64 to 42.23 (m 2 /g) (Fig. S1). A conclusion can be made that the hollow structural HCC can not only be a potential holder for carrying GOx and lucigenin but also act as nanoscaled-enzymatic reactors. Enhanced and visualized CL signal of HLG film. To verify the improved CL response of the HLG film for sensing glucose, the CL behavior of lucigenin&GOx (LG) film with or without HCC particles were compared. As shown in Fig. 3A, when the LG film was incubated with 100 nM glucose for 40 min, strong CL signal was observed after the addition of 0.12 M NaOH (Fig. 3A,B). This is because the immobilized GOx catalyze the oxidation of glucose to generate H 2 O 2 that further reacts with lucigenin at in alkaline condition to produce CL emission 13 . Interestingly, for the HLG platform, its CL signal was increased 3 times higher than that of the LG film on the same experimental condition ( Fig. 3A-C). It is worth noting that either these CL response or FL response for sensing low concentration glucose (100 nM) can be visualized using a simple FluorChem M system (inset of Fig. 3A). Obviously, the CL response for sensing glucose was apparently improved by the introduction of the HCC particles. There might be two reasons to explain the CL-enhancement phenomena. One is that the lucigenin-H 2 O 2 CL reaction has enhanced emission intensity because the HCC matrix might facilitate the transformation of chemical energy into light 3 . The other one might be that the HCC matrix can provide a nanoscale reactor to get an enhanced emzyme biocatalysis due to the confine effect. Therefore, the FL behavior of lucigenin-H 2 O 2 system and HLG film were investigated under the same experimental conditions. As shown in Fig. 3B, both this maximum emission wavelengths are at 480 nm, whatever HCC particles present (d) or not (c). It indicates the CL emitter of the HLG film is still excited state N-methylacridone (NMA) 13 . Meanwhile, the CL emission efficiencies of the lucigenin-H 2 O 2 system with (a) or without HCC particles (b) were the same (Fig. 3C). Obviously, the HCC particles are not capable of improving emission efficiency of the lucigenin CL system, but enhance the biocatalysis of GOx, which is also supported by previous reports 14,15 . The enzymatic reaction kinetics of immobilized GOx in the HLG reactor. The kinetics of the enzyme reaction in the HLG reactor were studied in details, by monitoring the CL response triggered by immobilized enzyme. Figure 4 shows that in the present of glucose, the CL intensity increases with the increasing reaction time up to 40 min, and then reaches a maximum value. The Km, determining the affinity of the enzyme for the substrate, was calculated using Lineweaver-Burke plot 16,17 . The Km of the GOx immobilized on HLG film was calculated to be 0.209 μM, much lower than that of immobilized GOx without HCC particles (0.759 μM). The kcat of HLG was calculated to be 2.21 s −1 , much better than those of LG (1.06 s −1 ). This is supported that the excellent biocompatibility and confine space of HCC particles can provide a favorable microenvironment for GOx to retain its bioactivity and keep high binding affinity to the substrate glucose 18 .
Analytical performance for detecting glucose. To fabricate a biosensor having an excellent analytical performance, some key parameters were optimized including reaction time, NaOH concentration, lucigenin concentration and the amount of HCC particles in details, which were 40 min, 0.12 M, 200 μM and 0.50 mg mL −1 , respectively (Fig. S2). Especially, CL intensity dramatically increases as increasing of HCC particles from 0.10 to 0.50 mg mL −1 , and gets the highest signal at 0.50 mg mL −1 . This would be attributed to the good enzyme carrier that the HCC particles act where more HCC particles can carry more enzyme molecule. However, when the concentration is so higher than 0.50 mg mL −1 , the HCC film would be too thick to have porous structure, which inhibits the interaction with analyses. Under the optimized condition, the analytical performance of the biosensor was assessed using various concentrations of glucose. As Fig. 5 shown, the CL response increased as the concentration of glucose increased. The good linear relationship between the logarithm of CL intensity and the logarithm of glucose concentration was observed. There are two wide linear ranges from 0.01 nM to 50.0 nM and 50.0 nM to 2.0 μM. Their linear equations log I = 1.153 + 0.895 log C (r = 0.9887) and log I = 2.032 + 0.351 log C (r = 0. 0.9849). I is CL intensity and C is the concentration of glucose (pM). The limit of detection (LOD) was calculated as 9.0 pM according to LOD = 3σ /S, where σ is the standard deviation and S is the slope of the calibration curve equation. It is better than that of previous reports [19][20][21] .   Selectivity against interferences and long-term stability of the biosensor. The practical utility of a sensor often depends upon its selectivity. Therefore, some potential substances were tested one by one to evaluate the selectivity of the glucose biosensor. Their concentrations were 10.0 μM. As shown in Fig. 6, when 1.0 μM glucose solution was added, a significant enhancement of the CL intensity was observed. However, when high concentration interferences (10.0 μM) were tested, any notable response from the biosensor was not observed. An excellent selectivity the CL biosensor based on HLG film for glucose was demonstrated. Such good selectivity was also visualized depending on the CL imaging (inset of Fig. 6).
With the aim of investigating its long-term stability, the HLG film was stored at 4 °C in a refrigerator, and its CL response was tested at intervals. Initially, no obvious decrease of the CL intensity was observed. After storage for 20 days and even one month, the absorbance retained 96.4% and 90.6% of its original value, respectively (Fig. S3). The bioactivity of immobilized GOx and lucigenin were well maintained, which indicated that the HCC film provided a biocompatible microenvironment for biomolecules.

Determination of glucose in human serum sample.
To illustrate the feasibility of the HLG film in biologically relevant matrix, it was employed to detect glucose in human blood serum. In this work, human blood serum was simply diluted with an appropriate dilution ratio to yield testing sample solutions. The recovery test was conducted. All the data are summarized in Table 1. These satisfied recovery results demonstrated that the HLG film offer an excellent, accurate, and precise method for the determination of glucose in a biologically relevant matrix. Importantly, these two different concentrations of glucose can be visualized in human blood serum (Fig. S4).

Application of visualization detection based on both CL and FL imaging.
Visual glucose detection has become increasingly important for the laboratory and point-of-care diagnostic applications because it eliminates the need for sophisticated instrumentation and enables visualization by the naked eye. Such detection methods are economical to implement and feature simplicity and portability. In this work, dual-signals including CL and FL imaging were used to visualize glucose at a low concentration. As shown in Fig. 7, when the HLG film was incubated with different concentrations of glucose, CL imaging was captured using a FluorChem M technique after the addition of 0.12 M NaOH. The corresponding FL imaging of the HLG film after the CL reaction was recorded using a digital camera under a UV irradiation. As the concentration of glucose increased from 0.1 nM to 2.0 μM, a visualized CL response increased while luminescence signal gradually fade (Fig. 7). To amplify the visualization detection, "ΔI CL + |ΔI FL |" is used as the response signal to quantitatively determine the concentration of glucose. It can be observed that the "ΔI CL + |ΔI FL |" also increases linearly with the increase of the glucose corresponding range from 0.1 to 1.0 μM and the corresponding linear function is (ΔI CL + |ΔI FL |) = 21.99 log C (pM) -10.36 (R 2 = 0.9743). The LOD is 1.0 nM (S/N = 3) (Fig. S5), comparable to or even better than most colorimetric glucose sensors 22,23 .

Conclusion
In this paper, we present a facile strategy to develop a novel CL platform based on HLG film for visualizing low concentration of glucose. CL data demonstrated that the entrapped enzyme exhibited its improved biocatalysis that resulted in enhanced CL signal, which might be ascribed to the confine effect of hollow structural HCC particles. Two wide detection ranges of 0.01-50 nM and 50 nM-2.0 μM were obtained. The CL platform based on HLG film exhibited a good sensitively with LOD of 9.0 pM (S/N = 3). It is worth noting that a dual-signaling amplification strategy based on "ΔI CL + |ΔI FL |" can be developed to visualize low concentration of glucose at nanomolar level. No need of complicated assembly procedure and pretreatment of sample, the high sensitive and high selective CL platform would provide wide potential applications in disease diagnosis and food safety.

Experimental
Materials. Lucigenin was purchased from Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan). Glucose oxidase (GOx) and Nafion were obtained from Sigma-Aldrich (St. Louis, USA). Sodium carbonate, calcium chloride, soluble starch, glucose, lactose, sucrose, fructose, citric acid, hydrogen peroxide (H 2 O 2 ) and sodium hydroxide were purchased from Chengdu Chemicals (Sichuan, China). All reagents were of analytical grade and used as received without further purification. All aqueous solutions were prepared using deionized distilled water.
Measurements and apparatus. CL signal was recorded using a model IFFS-A multifunctional chemiluminescence system (Ruimai Electronic Science Co, China). CL spectra and fluorescence (FL) spectra were measured by a model F-7000 Fluorescence Spectrophotometer (Hitachi, Japan). The morphology of HCC particles were charactered by SEM (Hitachi S-4800, Japan). The crystal structure of HCC particles were analyzed by a Tongda TD-3500 XRD with Cu-Ka radiation (λ = 0.15 nm) operating at 30.0 kV and 20.0 mA (Liaoning, China). The Brunauer-Emmett-Teller (BET) specific surface area was obtained from the nitrogen (N 2 ) adsorption method using a Quadrasorb SI Automated Surface Area (Quantachrome, USA). The CL images were charactered by a FluorChem M (ProteinSimple, USA). The FL images were recorded using a digital camera under irradiation a model ZF-20 D UV spectrophotometer (Henan, China).
Preparation of HCC particles. The HCC particles were synthesized according to a reported method 12 . In brief, two very dilute solutions were vigorously mixed at 30 °C for ten minutes. One is Na 2 CO 3 solution (2 mM, 110 mL). The other one is 120 mL of 1.67 mM CaCl 2 solution containing soluble starch (0.042 wt%) used as an additive.
Fabrication of HCC/Lucigenin&GOx film. HLG film was prepared as follows: The HCC colloidal suspension (0.5 mg mL −1 ) was prepared by dispersing HCC particles in deionized water and keeping ultrasonically treatment for 10 min. GOx was also dissolved in deionized water with a concentration of 10 mg mL −1 . 100 μL of the HCC colloidal suspension (0.5 mg mL −1 ) was added to a well of a polystyrene 96-well plate. Dried for 30 min at 50 °C, a porous HCC film was formed. 40 μL lucigenin/GOx mixture solution with a defined concentration was added on the HCC film, and then further dried for 30 min at 30 °C. Finally, 10 μL Nafion (0.5 wt%) was coated on the film and dried for 20 min at 30 °C to obtain the HLG film.