Silver Doped Mesoporous Silica Nanoparticles Based Electrochemical Enzyme-Less Sensor for Determination of H2O2 Released from Live Cells

In this study, a silver doped mesoporous silica nanoparticles-based enzyme-less electrochemical sensor for the determination of hydrogen peroxide (H2O2) released from live cells was constructed for the first time. The presented electrochemical sensor exhibited fast response (2 s) towards the reduction of H2O2 concentration variation at an optimized potential of −0.5 V with high selectivity over biological interferents such as uric acid, ascorbic acid, and glucose. In addition, a wide linear range (4 μM to 10 mM) with a low detection limit (LOD) of 3 μM was obtained. Furthermore, the Ag-mSiO2 nanoparticles/glass carbon electrode (Ag-mSiO2 NPs/GCE) based enzyme-less sensor showed good electrocatalytic performance, as well as good reproducibility, and long-term stability, which provided a successful way to in situ determine H2O2 released from live cells. It may also be promising to monitor the effect of reactive oxygen species (ROS) production in bacteria against oxidants and antibiotics.


Introduction
The rapid, specific, and accurate hydrogen peroxide (H 2 O 2 ) determination is quite essential in bioanalytical fields. H 2 O 2 , as the most common representative of reactive oxygen species, which endogenously produced in a cell, is a key player in various pathological processes. High levels of H 2 O 2 results in not only tissue and DNA damage, but also aging, diabetes, cancer, traumatic brain injury, and neurodegenerative disorders [1,2]. Consequently, the determination of H 2 O 2 attracted much interest as its potential as an indicator of oxidative stress-related diseases. Several analytical methods such as spectrophotometry [3], fluorimetry [4], titrimetry [5], chromatography [6], and electrochemistry [7,8] have been applied for highly sensitive H 2 O 2 determination. Amongst the above techniques, electrochemical methods are always optimal choices, due to their high sensitivity, operational simplicity, fast response, low cost, portability, and multi-analyte detection [9,10]. However, most electrode modifiers possessed high potentials, which limited the sensitive and selective great importance to minimize the overpotential of redox reaction and improve the rate of electron transfer.
Reportedly, noble metal nanoparticles [12], noble metal nanoparticles on carbon nanostructures [13], and bimetallic nanoparticles [14] have been used for the electrode surfaces modification because of their unique catalytic and electronic properties. The constructed electrochemical sensors performed a highly electrocatalytic activity toward H2O2 reduction or oxidization. Compared with the enzyme-based sensor, enzyme-less sensors are specific to analyte and overcome the limitations from enzyme, such as high cost, susceptible conformation affected by pH, temperature, etc. [15]. Silver nanoparticles (Ag NPs) exhibit good synergetic effect on H2O2 reduction [16,17], which is widely used for enzyme-less electrochemical sensors of H2O2. However, Ag NPs are easy to aggregate in solution before making of electrodes, reducing the sensitivity of the electrochemical sensor. Use of mesoporous silica nanoparticles (mSiO2 NPs) is one of the most effective ways to prevent aggregation [18] because Ag NPs can embed uniformly in the pores of mSiO2 NPs. In addition, the mSiO2 NPs possess unique properties such as large surface area, uniform pore size, ordered mesostructure, and facile surface modification [19,20], which make it a suitable electrically conductive host of Ag NPs to construct electrochemical sensors [21,22]. For example, Khan and Bandyopadhyaya [23] presented an enzyme-less amperometric H2O2 sensor based on Ag NPs impregnated amine functionalized mSiO2 NPs with a wider linearity range (5.3-124.3 mM). Azizi et al. [24] reported a new Ag-doped SBA-16 NPs modified carbon paste electrode, which could achieve detection of H2O2 in two linear ranges (20 μM to 8 mM, and 8 to 20 mM). The biosensor constructed by Ensafi et al. [10] could achieve a lower LOD of 0.45 μM. While the significant success has been made in the detection of traces of H2O2, which is summarized in the new work by Viter and Iatsunskyi [25], biosensors based on Ag-doped MCM-41 mSiO2 NPs for real-time detection of H2O2 released from live cells has not been reported yet, to the best of our knowledge.
We presented an enzyme-less electrochemical biosensor based on Ag-doped MCM-41 mSiO2 NPs to determine H2O2 released from live cells (Figure 1). Our result showed a high electrocatalytic performance of Ag-mSiO2 nanoparticles/glass carbon electrode (NPs/GCE) towards H2O2 reduction. Additionally, a good linear range from 4 μM to 10 mM with a LOD of 3 μM was obtained. The presented sensor showed high selectivity over biological interferents, such as ascorbic acid, uric acid, and glucose. Moreover, H2O2 released from pheochromocytoma (PC-12) cells could be successfully detected using the present sensor. In addition, as reactive oxygen species (ROS) including H2O2 production in bacteria such as Escherichia coli and Vaginal Lactobacilli can increase the bacteria's susceptibility to oxidative attack from antibiotic treatment [26,27], our sensor is promising to help understanding the mechanism of H2O2 producing by bacteria against antibiotics in a clinic.

Preparation of Ag-mSiO 2 Nanoparticles (NPs)
Ag-mSiO 2 NPs was prepared following slight modification of the Tian reaction [28]. A quantity of 0.5 mL 2M NaOH was added in 55 mL 7.5 mM CTAB aqueous solution to make solution A; 1 mL TEOS was added to 5 mL methanol to make solution B; 0.25 mL of TSD was added to 5 mL 35 mM AgNO 3 solution to make solution C. Under vigorous stirring, solution A was heated to 75 • C and added with 4 mL of solution B dropwise to react for 15 min. Then, solution C and 2 mL of solution B were added dropwise and reacted for 2 h. A total of 2.5 mL formalin solution was finally added and reacted for 1 h. The resulted NPs were removed from CTAB in NH 4 NO 3 ethanol solution (50 mL) at 60 • C for 10 h under stirring. The resulted Ag-mSiO 2 NPs were centrifuged, washed with distilled water thrice, and re-suspended into 50 mL deionized water for further use.

Preparation of H 2 O 2 Biosensors
The surface of a glassy carbon electrode (GCE, 3 mm in diameter) was firstly polished and cleaned to a mirror-like status. Different volumes of Ag-mSiO 2 NPs suspension (10,8,5, and 3 µL) were dropped respectively on the pretreated electrode surface and dried in the air for 1 h. Then GCE modified with a stable Ag-mSiO 2 NPs film (Ag-mSiO 2 NPs/GCE) was used for the following electrochemical experiments. The pheochromocytoma (PC-12) cells were grown in 75 cm 2 flasks containing RPMI-1640 medium including fetal bovine serum (10%), as well as penicillin and streptomycin (100 µg·mL −1 ) in 5% CO 2 atmosphere at 37 • C. When the PC-12 cells reached a 90% confluence growth, they were centrifuged and responded into 4 mL of PBS (0.2 M pH 6.8). The cell-packed pellet of 1.0-2.0 × 10 5 cells·cm −2 was obtained for amperometric analysis. After achieving a steady background noise, 60 µg·mL −1 lipopolysaccharide (LPS) was added into the cell mixture to stimulate the release of H 2 O 2 from cells. Then 500 unit·mL −1 catalase was added to decomposing H 2 O 2 . As a control group, LPS and catalase were added to PBS solution without cells. An optimized potential of −0.5 V was applied to the Ag-mSiO 2 NPs-modified GC electrode for amperometric analysis.

Characterization
High-resolution transmission electron microscopy (HRTEM, CM200UT, Philips, FEI Co., Hillsboro, OR, USA) was used to examine the morphology and distribution of mSiO 2 NPs and Ag-mSiO 2 NPs. X-ray diffraction (XRD) pattern was analyzed by an X-ray powder diffractometer (D8 Advance, Bruker, Berlin, Germany). Ultraviolet-visible spectroscopy (UV-vis) absorbance spectra were recorded by UV-vis spectrophotometer (Nanjing Feile Instrument Company, Nanjing, China) with a wavelength range from 300 to 800 nm.
Cyclic voltammetric and amperometric analysis were performed on a CHI660E electrochemical analyzer (Chenhua Co., Shanghai, China) at room temperature. A bare Glassy Carbon (GC) electrode or Ag-mSiO 2 NPs modified GC electrode, a Platinum foil electrode and saturated calomel electrode (SCE), were used as the working, the counter and the reference electrode, respectively. Before the detection of H 2 O 2 , the solutions were purged by high purity N 2 for 30 min to remove oxygen. Figure 2A shows the UV-vis absorption spectra of mSiO 2 NPs and Ag-mSiO 2 NPs. The UV-vis absorbance spectrum of mSiO 2 NPs at 414 nm was enhanced after Ag NPs decorated. In the wide-angle XRD pattern of the Ag-mSiO 2 NPs ( Figure 2B), four well-resolved diffraction peaks at 2θ values in the range of 30 • -90 • can be indexed to face-centered cubic (fcc) Ag 111, 200, 220, and 311 reflections [29]. The typical morphology and structure of mSiO 2 and Ag-mSiO 2 NPs was observed clearly through HRTEM. Small Ag NPs were doped in the ordered porous framework of spherical Ag-mSiO 2 NPs ( Figure 2C,D), which protect themselves from aggregation. (D8 Advance, Bruker, Berlin, Germany). Ultraviolet-visible spectroscopy (UV-vis) absorbance spectra were recorded by UV-vis spectrophotometer (Nanjing Feile Instrument Company, Nanjing, China) with a wavelength range from 300 to 800 nm. Cyclic voltammetric and amperometric analysis were performed on a CHI660E electrochemical analyzer (Chenhua Co., Shanghai, China) at room temperature. A bare Glassy Carbon (GC) electrode or Ag-mSiO2 NPs modified GC electrode, a Platinum foil electrode and saturated calomel electrode (SCE), were used as the working, the counter and the reference electrode, respectively. Before the detection of H2O2, the solutions were purged by high purity N2 for 30 min to remove oxygen. Figure 2A shows the UV-vis absorption spectra of mSiO2 NPs and Ag-mSiO2 NPs. The UV-vis absorbance spectrum of mSiO2 NPs at 414 nm was enhanced after Ag NPs decorated. In the wide-angle XRD pattern of the Ag-mSiO2 NPs ( Figure 2B), four well-resolved diffraction peaks at 2θ values in the range of 30°-90° can be indexed to face-centered cubic (fcc) Ag 111, 200, 220, and 311 reflections [29]. The typical morphology and structure of mSiO2 and Ag-mSiO2 NPs was observed clearly through HRTEM. Small Ag NPs were doped in the ordered porous framework of spherical Ag-mSiO2 NPs (Figure 2 C,D), which protect themselves from aggregation.

Electrocatalytic Reduction of H2O2
The electrocatalytic activity of the Ag-mSiO2 NPs modified GCE toward the H2O2 reduction was evaluated in PBS (0.2 M, pH 6.8) with a scan rate of 50 mV·s -1 . Different volumes of Ag-mSiO2 NPs (3 μL, 5 μL, 8 μL, and 10 μL) modified on the electrode for better cyclic voltammetric (CV) responses were investigated. The modifier of 8 μL on the surface showed the best response, which was applied for the following experiment ( Figure S1). The mechanism of H2O2 reduced by Ag-doped mesoporous NPs is as follows [24]:

Electrocatalytic Reduction of H 2 O 2
The electrocatalytic activity of the Ag-mSiO 2 NPs modified GCE toward the H 2 O 2 reduction was evaluated in PBS (0.2 M, pH 6.8) with a scan rate of 50 mV·s -1 . Different volumes of Ag-mSiO 2 NPs (3 µL, 5 µL, 8 µL, and 10 µL) modified on the electrode for better cyclic voltammetric (CV) responses were investigated. The modifier of 8 µL on the surface showed the best response, which was applied for the following experiment ( Figure S1). The mechanism of H 2 O 2 reduced by Ag-doped mesoporous NPs is as follows [24]: Figure 2A shows the cyclic voltammogram of bare GC electrode (curve a and c) and Ag-mSiO 2 modified GC electrode (curve b and d) in PBS (0.2 M, pH 6.8) with the absence or presence of 0.1 mM H 2 O 2 . In the absence of H 2 O 2 , no obvious cathodic current was observed for bare GCE or Ag-mSiO 2 modified GCE (Figure 3Aa,b). However, after the addition of H 2 O 2 , the reduction current of Ag-mSiO 2 modified GC electrode (Figure 3Ad) greatly increased, while the response signal of the bare GC electrode was very weak (Figure 3Ac). Cyclic voltammogram of Ag-mSiO 2 NPs modified GC electrode for H 2 O 2 reduction at various concentrations of H 2 O 2 was shown in Figure 3B. The result revealed that the current responses increased with an increased H 2 O 2 concentration, indicating a good electrocatalytic behavior of Ag-mSiO 2 /GCE towards H 2 O 2 reduction. H2O2 + e − ↔ OHads + OH − OHads + OH − ↔ OH − 2OH − + 2H + ↔ 2H2O Figure 2A shows the cyclic voltammogram of bare GC electrode (curve a and c) and Ag-mSiO2 modified GC electrode (curve b and d) in PBS (0.2 M, pH 6.8) with the absence or presence of 0.1 mM H2O2. In the absence of H2O2, no obvious cathodic current was observed for bare GCE or Ag-mSiO2 modified GCE (Figure 3A a,b). However, after the addition of H2O2, the reduction current of Ag-mSiO2 modified GC electrode (Figure 3Ad) greatly increased, while the response signal of the bare GC electrode was very weak (Figure 3Ac). Cyclic voltammogram of Ag-mSiO2 NPs modified GC electrode for H2O2 reduction at various concentrations of H2O2 was shown in Figure 3B. The result revealed that the current responses increased with an increased H2O2 concentration, indicating a good electrocatalytic behavior of Ag-mSiO2/GCE towards H2O2 reduction.

Amperometric Determination of H2O2
The amperometric analysis was used to evaluate the electrocatalytic activity of Ag-mSiO2/GC electrode for different concentrations of H2O2. First of all, the effect of applied potential on amperometric current-time (i−t) curve was studied. Several applied potentials of −0.40, −0.45, −0.50, and −0.55 V was employed to investigate the according amperometric response of Ag-mSiO2/GC electrode in PBS (0.2 M, pH 6.8) with the successive addition of 1.0 mM H2O2. The amperometric response reached a maximum value at a potential of −0.50 V ( Figure S2), which was selected for the following amperometric analysis. Figure 4A showed the amperometric i-t curves of Ag-mSiO2 NPs/GC electrode with the successive additions of H2O2 into deoxygenated PBS (0.2 M, pH 6.8) under continuous stirring. Before adding H2O2, the amperometric current of Ag-mSiO2 NPs/GC electrode was recorded in PBS solution for 200 s to obtain a steady blank control. The Ag-mSiO2 NPs/GCE-based sensor could reach a steady-state current within 2 s after a small amount addition of H2O2, which indicated its fast response for H2O2 reduction. Additionally, a wide linear range from 4 μM to 10 mM was obtained. The LOD is calculated to be 3 μM (S/N = 3) ( Figure 4B). The present H2O2 biosensor revealed a wide linear range, low LOD, and fast response.

Amperometric Determination of H 2 O 2
The amperometric analysis was used to evaluate the electrocatalytic activity of Ag-mSiO 2 /GC electrode for different concentrations of H 2 O 2 . First of all, the effect of applied potential on amperometric current-time (i−t) curve was studied. Several applied potentials of −0.40, −0.45, −0.50, and −0.55 V was employed to investigate the according amperometric response of Ag-mSiO 2 /GC electrode in PBS (0.2 M, pH 6.8) with the successive addition of 1.0 mM H 2 O 2 . The amperometric response reached a maximum value at a potential of −0.50 V ( Figure S2), which was selected for the following amperometric analysis. Figure 4A

Interferences Study
Glucose, ascorbic acid (AA), and uric acid (UA) are three commonly interferents in a physiological system. The anti-interference effect of the Ag-mSiO2 NPs /GCE based sensor toward glucose, AA, and UA was investigated. The amperometric responses of the present sensor with sequential addition of 0.05 mM H2O2, 1 mM glucose, 0.15 mM AA, and 0.5 mM UA are shown in Figure 5. It can be seen that a large amperometric response was achieved with a low concentration of 0.05 mM H2O2, while negligible current was observed with high concentrations of the other interfering species such as glucose, AA or UA. This showed the present sensor was selective to H2O2.

Reproducibility and Stability
The reproducibility of Ag-mSiO2 NPs modified single electrode was tested for five successive measurements with the relative standard derivation (RSD) in about 2.6%. Moreover, the RSD of current responses of five electrodes was investigated and found to be less than 5%, suggesting the high reproducibility and good precision of our present sensor. Furthermore, the current responses of the sensor decreased to 90.2% of its original value after being stored in the refrigerator at 4 °C for 20 days. The above results indicated the reliability and long-term stability of our fabricated H2O2

Interferences Study
Glucose, ascorbic acid (AA), and uric acid (UA) are three commonly interferents in a physiological system. The anti-interference effect of the Ag-mSiO 2 NPs /GCE based sensor toward glucose, AA, and UA was investigated. The amperometric responses of the present sensor with sequential addition of 0.05 mM H 2 O 2 , 1 mM glucose, 0.15 mM AA, and 0.5 mM UA are shown in Figure 5. It can be seen that a large amperometric response was achieved with a low concentration of 0.05 mM H 2 O 2 , while negligible current was observed with high concentrations of the other interfering species such as glucose, AA or UA. This showed the present sensor was selective to H 2 O 2 .

Interferences Study
Glucose, ascorbic acid (AA), and uric acid (UA) are three commonly interferents in a physiological system. The anti-interference effect of the Ag-mSiO2 NPs /GCE based sensor toward glucose, AA, and UA was investigated. The amperometric responses of the present sensor with sequential addition of 0.05 mM H2O2, 1 mM glucose, 0.15 mM AA, and 0.5 mM UA are shown in Figure 5. It can be seen that a large amperometric response was achieved with a low concentration of 0.05 mM H2O2, while negligible current was observed with high concentrations of the other interfering species such as glucose, AA or UA. This showed the present sensor was selective to H2O2.

Reproducibility and Stability
The reproducibility of Ag-mSiO2 NPs modified single electrode was tested for five successive measurements with the relative standard derivation (RSD) in about 2.6%. Moreover, the RSD of current responses of five electrodes was investigated and found to be less than 5%, suggesting the high reproducibility and good precision of our present sensor. Furthermore, the current responses of the sensor decreased to 90.2% of its original value after being stored in the refrigerator at 4 °C for 20 days. The above results indicated the reliability and long-term stability of our fabricated H2O2

Reproducibility and Stability
The reproducibility of Ag-mSiO 2 NPs modified single electrode was tested for five successive measurements with the relative standard derivation (RSD) in about 2.6%. Moreover, the RSD of current responses of five electrodes was investigated and found to be less than 5%, suggesting the high reproducibility and good precision of our present sensor. Furthermore, the current responses of the sensor decreased to 90.2% of its original value after being stored in the refrigerator at 4 • C for 20 days. The above results indicated the reliability and long-term stability of our fabricated H 2 O 2 sensor.
In addition, we compared the hydrogen peroxided sensing by AgNP impregnated silica NPs in Table 1. H 2 O 2 sensor reported here has a wide linear range and low detection limit and can detect the H 2 O 2 released from living cells. Moreover, we performed a study of the real-time determination of H 2 O 2 released from live cells. For the experiment, about 2 × 10 5 pheochromocytoma (PC-12) cells in 4 mL PBS (pH = 6.8) with 100 mM glucose were used. LPS (60 µg mL −1 ) was employed to stimulate PC-12 cells to generate H 2 O 2 . After treating with LPS, an obvious current change occurred in solution with cells, while no signal could be observed in that without cells. We also can see from Figure 6 that the current decreased rapidly with an injection of 500 unit·mL −1 catalase. As catalase is a selective scavenger of H 2 O 2 , it can be concluded that the increased current in the curve "with cells" was assigned to the reduction of H 2 O 2 .
Micromachines 2019, 10, 268 7 of 10 sensor. In addition, we compared the hydrogen peroxided sensing by AgNP impregnated silica NPs in Table 1. H2O2 sensor reported here has a wide linear range and low detection limit and can detect the H2O2 released from living cells.

Determination of H2O2 Released from Pheochromocytoma Cells
Moreover, we performed a study of the real-time determination of H2O2 released from live cells. For the experiment, about 2 × 10 5 pheochromocytoma (PC-12) cells in 4 mL PBS (pH = 6.8) with 100 mM glucose were used. LPS (60 μg mL −1 ) was employed to stimulate PC-12 cells to generate H2O2. After treating with LPS, an obvious current change occurred in solution with cells, while no signal could be observed in that without cells. We also can see from Figure 6 that the current decreased rapidly with an injection of 500 unit·mL −1 catalase. As catalase is a selective scavenger of H2O2, it can be concluded that the increased current in the curve "with cells" was assigned to the reduction of H2O2.

Conclusion
In summary, a highly sensitive and selective enzyme-less sensor of H2O2 was successfully constructed with Ag-mSiO2 NPs/GCE, enabling good electrocatalytic performance toward H2O2 reduction. The modified electrode exhibited a fast response to H2O2 concentration variation at an optimized working potential of −0.5 V. A wide linear range from 4 μM to 10 mM with a low LOD of 3 μM was obtained. In addition, the presented sensor showed high selectivity over commonly

Conclusions
In summary, a highly sensitive and selective enzyme-less sensor of H 2 O 2 was successfully constructed with Ag-mSiO 2 NPs/GCE, enabling good electrocatalytic performance toward H 2 O 2 reduction. The modified electrode exhibited a fast response to H 2 O 2 concentration variation at an optimized working potential of −0.5 V. A wide linear range from 4 µM to 10 mM with a low LOD of 3 µM was obtained. In addition, the presented sensor showed high selectivity over commonly interfering substances, such as uric acid (UA), ascorbic acid (AA), and glucose. Furthermore, the Ag-mSiO 2 NPs/GCE based enzyme-less sensor exhibits good reproducibility and long-term stability. Accordingly, the constructed biosensor could be successfully used for the determination of H 2 O 2 released from live cells. As ROS or H 2 O 2 has been associated with cell cycle arrest, apoptosis, migration, and inflammation, or the antibacterial activity of certain bacteria such as Lactobacilli, our sensor has potential to help understand the role of H 2 O 2 that plays in the biological and pathological systems. However, for monitoring the ROS or H 2 O 2 from real samples, the stability of the constructed sensor still needs to be improved with employing more advanced nanomaterials because there is a lot of interferent in real clinical samples.