H-rGO-Pd NPs Nanozyme Enhanced Silver Deposition Strategy for Electrochemical Detection of Glypican-3

Glypican-3 (GPC3), as an emerging biomarker, has been shown to be beneficial for the early diagnosis and treatment of hepatocellular carcinoma (HCC). In this study, an ultrasensitive electrochemical biosensor for GPC3 detection has been constructed based on the hemin-reduced graphene oxide-palladium nanoparticles (H-rGO-Pd NPs) nanozyme-enhanced silver deposition signal amplification strategy. When GPC3 specifically interacted with GPC3 antibody (GPC3Ab) and GPC3 aptamer (GPC3Apt), an “H-rGO-Pd NPs-GPC3Apt/GPC3/GPC3Ab” sandwich complex was formed with peroxidase-like properties which enhanced H2O2 to reduce the silver (Ag) ions in solution to metallic Ag, resulting in the deposition of silver nanoparticles (Ag NPs) on the surface of the biosensor. The amount of deposited Ag, which was derived from the amount of GPC3, was quantified by the differential pulse voltammetry (DPV) method. Under ideal circumstances, the response value was linearly correlated with GPC3 concentration at 10.0–100.0 μg/mL with R2 of 0.9715. When the GPC3 concentration was in the range from 0.01 to 10.0 μg/mL, the response value was logarithmically linear with the GPC3 concentration with R2 of 0.9941. The limit of detection was 3.30 ng/mL at a signal-to-noise ratio of three and the sensitivity was 1.535 μAμM−1cm−2. Furthermore, the electrochemical biosensor detected the GPC3 level in actual serum samples with good recoveries (103.78–106.52%) and satisfactory relative standard deviations (RSDs) (1.89–8.81%), which confirmed the applicability of the sensor in practical applications. This study provides a new analytical method for measuring the level of GPC3 in the early diagnosis of HCC.


Introduction
Globally, hepatocellular carcinoma (HCC) has been evaluated as one of the most common malignant malignancies with high prevalence and fatality rates [1]. The survival rate of HCC patients has been suggested to increase with the earlier detection of this malignancy [2]. The existing imaging methods, including ultrasound (US), magnetic resonance imaging (MRI), computed tomography (CT), and photoacoustic imaging (PAI), find difficulty in detecting HCC, and all these detection methods require expensive instruments as well as specialized personnel. Therefore, the development of HCC diagnosis is crucial to increase the chance of effective treatment and reduce HCC-related mortality [3].
Immunoassay detection of serum biomarkers including enzyme-linked immunoassay (ELISA) and bioluminescence enzyme immunoassay (BLIEA) has been widely used in clinical practice for the diagnosis of HCC [4][5][6]. Glypican-3 (GPC3), as an emerging biomarker, has been shown to be beneficial for early clinical screening of HCC [7,8]. A study showed that the serum concentration of GPC3 (108.67 ± 230.04 ng/mL) in HCC patients was much higher than those in healthy persons (3.99 ± 7.68 ng/mL) [9]. Therefore, the use of GPC3 as one of the detection criteria for HCC will potentially improve the detection rate of HCC.
ELISA serum kits are widely used to detect antigens or antibodies in clinical practice by using double antibodies and labeled-horseradish peroxidase (HRP) to form sandwichtype structures [10]. On the one hand, antibodies have strong specificity in recognizing the target protein. On the other hand, the sandwich structure provides a more stable result for HRP color development. It is worth mentioning that the advent of nucleic acid aptamer complements the disadvantages that antibodies are not easy to store and has the advantage of antibody specificity [11,12]. Compared with natural enzymes such as HRP, artificial nanozymes have the advantages of high catalytic efficiency, high stability, and low price [13,14].
In recent years, electrochemical detection has alternatively been used as a powerful technique for many point-of-care (POC) sensors due to its inherent advantages of low cost, portability, and fast response [15,16]. For instance, Zhou et al. designed a dualrecognition sandwich electrochemical biosensor for circulating tumor (CTC) detection based on PdPtCuRu nanospheres with catalytic hydrogen peroxide properties to improve the specificity, and the constructed sandwich biosensor exhibited good specificity and accuracy in spiked serum samples [17]. A study showed that sandwich structure greatly improves the stability, sensitivity, and specificity of the sensor to detect the target protein [18]. Therefore, the construction of a sandwich electrochemical biosensor could provide new opportunities to improve the sensitivity for HCC detection.
Nanozymes are simulated enzymes composed of nanomaterials, which have two characteristics of nanomaterials and enzyme-like activities. Nanozymes can solve the shortcomings of the natural enzyme such as high cost and variability, but the activity is slightly lower than that of the natural enzyme. Nanozymes and reaction substrates are accompanied by electron transfer and valence changes, showing REDOX enzyme activities. Compared with ordinary bio-signal, the electrochemical signal generated by nanozymecatalyzed amplification technique can be enhanced by the reaction of the chromogenic substrate [19,20]. Among them, peroxidase-like catalytic silver deposition has been an effective way for signal enhancement due to high catalytic activity, in situ highly efficient enzymatic catalysis of Ag NPs deposition, and increased conductivity resulting from Ag deposition [21]. In another aspect, Hemin, an important natural porphyrin iron compound, is often used as a catalyst to replace natural enzymes due to its good peroxidase properties [22]. Similarly, palladium nanoparticles (Pd NPs) can be used not only as a highly stable catalase but also as a conductivity material to improve electrical conductivity [23]. Reduced graphene oxide (rGO), a two-dimensional structure of carbon materials, is a useful carrier with strong electrical conductivity and high biocompatibility due to its large specific surface area [24].
Given the above considerations, our study aimed to generate a novel sandwichstructure electrochemical biosensor for GPC3 detection based on hemin-reduced graphene oxide-palladium nanoparticles (H-rGO-Pd NPs) nanozyme peroxidase-like catalytic silver deposition for signal amplification, combining GPC3 aptamer (GPC3 Apt ) and GPC3 antibody (GPC3 Ab ) as recognition elements. GPC3 Apt was labelled on the binding sites of H-rGO-Pd NPs nanozyme through π-π action and Pd-N coordination interaction. The H-rGO-Pd NPs-GPC3 Apt signal probe not only improves the electron transfer rate but also enhances the number of fixed biomolecules. GPC3 Ab , as a capture probe, was adsorbed on the surface of a Au NPs@rGO-modified screen-printed electrode (SPE). In the presence of GPC3, an "H-rGO-Pd NPs-GPC3 Apt -GPC3-GPC3 Ab " sandwich-structure complex with peroxide properties was formed by specific binding, and further catalyzed the reaction of H 2 O 2 with AgNO 3 to deposit Ag NPs on the sensor surface. The dissolution current of Ag NPs could be measured using differential pulse voltammetry (DPV). The analytical performance in terms of working curve, linear range, sensitivity, specificity, reproducibility, and stability of the proposed GPC3 electrochemical nanobiosensor was discussed.

Results and Discussion
2.1. The Analysis Principle and Feasibility of GPC3 Electrochemical Aptasensor Figure 1A illustrates the detection principle of the GPC3 electrochemical nanobiosensor based on H-rGO-Pd NPs nanozyme peroxidase-like catalytic silver deposition for signal amplification, combining the GPC3 aptamer (GPC3 Apt ) and the GPC3 antibody (GPC3 Ab ) as recognition elements. Firstly, the H-rGO-Pd NPs nanozyme with good peroxidase-like catalytic properties was prepared by a two-step reduction method, and the H-rGO-Pd NPs-GPC3 Apt detection probe was prepared through π-π action, and Pd-N coordination interaction. Then, the Au NPs@rGO was modified on the surface of pretreated SPE by electrodeposition, leading to the formation of Au NPs@rGO/SPE. After that, GPC3 Ab was immobilized on Au NPs@rGO/SPE by Au-N bonding as well as physical adsorption. When the GPC3 was added, the specific recognition reaction between the GPC3 Ab and GPC3 produced an antibody-antigen complex and arranged on the electrode surface. Next, the H-rGO-Pd NPs-GPC3 Apt detection probe was fixed on the electrode surface by π-π bond and electrostatic adsorption. Both GPC3 Ab and H-rGO-Pd NPs-GPC3 Apt specifically bonded with GPC3 to form the H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab sandwich-structure complex with a stable spatial structure as well as catalytic performance, which could induce the reduction in the Ag ions in the solution containing H 2 O 2 and AgNO 3 solution for the deposition of Ag NPs on the surface of Au NPs@rGO/SPE. The metallic Ag NPs deposited on Au NPs@rGO/SPE could produce detectable anodic stripping signals, which can be determined by DPV. Since the amount of H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab affects the Ag NPs deposition which further leads to the change of the sensor response current, the standard curve was determined by studying the relationship between the sensor response current and GPC3 concentration. Herein, H-rGO-Pd NPs revealed good conductivity, nontoxicity, and high peroxidaselike catalytic properties because of the peroxidase properties of Hemin, good catalystsupporting material of rGO, and the efficient catalytic synergies of Pt NPs.
By using the DPV method, the feasibility of the GPC3 electrochemical nanobiosensor was determined under the potential voltage range of (−0.2-0.4 V) ( Figure 1B). In the absence of Ag (i.e., no silver deposition, curve a), there was an insignificant response current, indicating that the determination of GPC3 is achieved by the anodic stripping peak current of the depositing Ag. In the absence of GPC3 protein (curve b), the response current was slightly higher than that of the curve a. This reason for this phenomenon was that H 2 O 2 was able to react with AgNO 3 slowly without catalytic substances, which left a small amount of deposited Ag on the electrode surface. Under the catalysis of H-rGO-Pd NPs peroxidase-like activity, the GPC3 nanobiosensor detected a significant current response (curve c and curve d). Moreover, the current response significantly increased when the concentration of GPC3 went up from 10.0 to 50.0 µg/mL. It was positively correlated with GPC3 concentrations. The feasibility analysis suggested that the "H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab " sandwich complex can effectively catalyze the reaction of H 2 O 2 and AgNO 3 on the surface of electrodes. Being heavily coated with Ag NPs amplified the current signal of the sensor, further indicating that the electrochemical nanobiosensor was capable of detecting GPC3.

Characterization of H-rGO-Pd NPs
The UV-vis spectra of rGO, hemin, and H-rGO-Pd NPs were shown in Figure 2A. The peak of 263 nm was the strongest absorption peak of rGO (curve a), indicating that rGO was successfully reduced [25]. The peak of 382 nm was the Soret peak of hemin (curve b). H-rGO-Pd NPs (curve c) had similar absorption peaks at 265 nm and 390 nm, which was consistent with rGO and hemin, and the absence of a characteristic absorption peak for the reduction of Pd ions to zero-valent Pd NPs [26], which indicated that H-rGO-Pd NPs may be successfully synthesized. FT-IR spectra of rGO (curve a), hemin (curve b), and H-rGO-Pd NPs (curve c) were shown in Figure 2B. All of them had the characteristic absorption peaks of O-H at 3434 cm −1 , a constriction vibration peak of C-H at 2918 cm −1 , a vibration peak of C=O at 1630 cm −1 , and a vibration peak of C-O at 1384 cm −1 . H-rGO-Pd NPs and hemin have a com-mon asymmetric stretching peak at 850 cm −1 which is from the formation of Fe-O bond in hemin [27]. Moreover, compared with the strong intensity of the peak of hemin at 1384 cm −1 (C-O), the weak intensity of H-rGO-Pd NPs can be described by the binding interaction between H-rGO and the Pd NPs, indicating the H-rGO-Pd NPs may have been successfully synthesized.
Characterization of the H-rGO-Pd NPs was performed using the Zeta particle size analyzer. As shown in Figure 2C, the Zeta potential of rGO (curve a), hemin (curve b), H-rGO (curve c), and H-rGO-Pd NPs (curve d) were 1 mV, −1 mV, 0 mV, 12 mV, respectively. The H-rGO-Pd NPs were shifted to positive potential due to the presence of a small amount of Pd 2+ in the solution. It can be seen that H-rGO-Pd NPs were successfully prepared.
By using SEM, the surface morphology of H-rGO-Pd NPs was characterized. As shown in Figure 2D, many dark gray particles with relatively uniform particles and a folded film-like structure appeared on the surface of H-rGO-Pd NPs, indicating that the Pd NPs were successfully attached to H-rGO, i.e., the H-rGO-Pd NPs had been successfully prepared. In addition, the SEM characterization diagram of rGO and H-rGO is shown in Figure S1. Meanwhile, the EDS energy spectrum of H-rGO-Pd NPs was measured on silicon wafers. As shown in Figure 2E, the H-rGO-Pd NPs were rich in C and O elements and Na elements and contained a small amount of Pd elements and Cl elements. Furthermore, the TEM was also used to characterize the H-rGO-Pd NPs (see in Figure S1C).
The peroxidase-like characteristics of H-rGO-Pd NPs were verified in Figure 2F. Both the TMB and H 2 O 2 were colorless and there was no absorption peak (curves 1, 2). After mixing TMB and H 2 O 2 , a weak reaction occurred, producing the mixed solution with a light blue color (insert figure in Figure 2E) and a soft absorption peak (curve 3). The results indicated that a few hydroxyl radicals in H 2 O 2 could oxidize the TMB. The H-rGO-Pd NPs solution was added to two solutions containing TMB (curve 4) or H 2 O 2 (curve 5), respectively; however, no change was found in both solutions. When all TMB, H 2 O 2 , and H-rGO-Pd NPs were mixed, the mixture turned dark blue from being colorless ( Figure 2F). Moreover, the absorption peak (curve 6) was much higher than that of the mixed solution of TMB and H 2 O 2 without H-rGO-Pd NPs. This indicated that H-rGO-Pd NPs can effectively catalyze the reaction between H 2 O 2 and TMB, and thus have peroxidase activity.
Furthermore, to explore whether the combination of GPC3 Apt and H-rGO-Pd NPs was successful, the H-rGO-Pd NPs-GPC3 Apt was characterized and analyzed by UV-vis (see Figure S2).
The Raman microscopy of each modified GPC3 electrochemical nanosensor is shown in Figure 3B. The G-band reflected the order and integrity of materials containing carbon, and the D-band indicated the degree of defect in the C atomic lattice [29]. The I D /I G ratio was 1.05 for the Raman curve of SPE (curve a). Curve b was the Raman curve of Au NPs@rGO/SPE with an I D /I G of 1.04. The Raman I D /I G was 1.02 for GPC3 Ab /Au NPs@rGO/SPE (curve c). The Raman curves d and e were GPC3/GPC3 Ab /Au NPs@rGO/SPE and H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab /Au NPs@rGO/SPE, with I D /I G of 0.95 and 0.98, respectively. Curve f was the sensor electrode where Ag NPs were deposited, and its I D /I G was 0.84, showing that the deposition of a large number of Ag NPs had filled the defects inside the composite on the electrode surface.  The redox peak of GPC3 Ab /Au NPs@rGO/SPE (Ip: 119.8 µA, curve c) decreased slightly due to the presence of GPC3 Ab , which hindered the electron transport. In particular, the interaction between GPC3 and GPC3 Ab hindered electron transfer, and the redox peak of GPC3/GPC3 Ab /Au NPs@rGO/SPE was slightly changed (Ip: 119.5 µA, curve d). When H-rGO-Pd NPs-GPC3 Apt incubated on the electrode (Ip: 110.8 µA, curve e), the current failed to rise but decreased slightly because of the current hindered by the incubation of the aptamer, to some extent. When Ag NPs continued to be deposited on the surface of the electrode due to the catalytic action of H-rGO-Pd NPs peroxidase-like activity, the current of Ag NPs/H-rGO-Pd NPs-GPC3 Apt /GPC3 Ab /Au NPs@rGO/SPE (Ip: 183.9 µA, curve f) increased significantly.
In this formula, Ip represents the current peak (µA), A represents the sensor's effective surface area (cm 2 ), D is the medium proliferation parameter [Fe(CN) 6 Figure 3D, the impedance value of the SPE was 991 Ω (curve a). After the electrodeposition of Au NPs@rGO, its impedance also dropped sharply (320 Ω, curve b). Moreover, the presence of GPC3 Ab hindered electron transfer, and the impedance became larger (414 Ω, curve c). After GPC3 was incubated, the binding of GPC3 and GPC3 Ab hindered electron transfer (423 Ω, curve d), H-rGO-Pd NPs-GPC3 Apt was dropped on the electrode, the impedance increased (504 Ω, curve e) because of the presence of the aptamer hindering the electron transfer. Finally, after silver particles were deposited on the surface of the sensor, the electrode impedance decreased sharply (132 Ω, curve f), indicating that H-rGO-Pd NPs effectively catalyzed the reaction of H 2 O 2 and AgNO 3 . It was concluded that the electrode impedance was significantly reduced by depositing a large amount of Ag NPs on its surface.

SEM Spectroscopy Characterization of the Modified Electrodes
The GPC3 electrochemical nanosensor was characterized by using SEM. The surface of the bare SPE ( Figure 4A) was relatively flat. When Au NPs@rGO were deposited on SPE, there were two different uniform particles presented on the surface of Au NPs@rGO/SPE ( Figure 4B), one was darker in color and the other was brighter. After modifying GPC3 Ab , the GPC3 Ab /Au NPs@rGO/SPE ( Figure 4C) showed a white film. When GPC3 was present, the formation of GPC3/GPC3 Ab /Au NPs@rGO/SPE ( Figure 4D) found white spheroids on the surface of the electrode. After the H-rGO-Pd NPs-GPC3 Apt was fixed, the surface of the H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab /Au NPs@rGO/SPE ( Figure 4E) became flat, and a folded film was seen. On the surface of the Ag NPs/H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab /Au NPs@rGO/SPE ( Figure 4F), Ag NPs were deposited, and shiny silver particles were found. The above results suggested that Au NPs@rGO, GPC3 Ab , GPC3, and H-rGO-Pd NPs-GPC3 Apt electrode surfaces were modified.

Optimization of Conditions for GPC3 Electrochemical Aptasensor
To improve the performance of the aptasensor, some experimental conditions including the concentration of the GPC3 Apt , incubation temperature, incubation time, and the amount of H-rGO-Pd NPs nanozymes were optimized. As shown in Figure S3, 5 µmol/L of GPC3 Apt , 25 • C of incubation temperature, 60 min of incubation time, and 4 µL of H-rGO-Pd NPs nanozymes were chosen as the optimized detection conditions for subsequent experiments.

Analytical Performance of GPC3 Electrochemical Aptasensor
Herein, the GPC3 Ab -GPC3-H-rGO-Pd NPs-GPC3 Apt sandwich nanobiosensor was constructed using H-rGO-Pd NPs-GPC3 Apt as the signal probe and GPC3 Ab as the capture probe. In the presence of H-rGO-Pd NPs with peroxidase-like catalytic properties, H 2 O 2 could reduce the Ag ions in the solution to metallic Ag, which was deposited on the surface of the electrode. The ultrasensitive detection of GPC3 was achieved by the DPV method for the anodic stripping signal of the deposited Ag NPs. Under optimal conditions, the DPV curves of the electrochemical nanobiosensor with different GPC3 concentrations were obtained and shown in Figure 5A. When the GPC3 concentration increased from 0.01 to 100.0 µg/mL, the response current of the nanobiosensor increased. This was due to H-rGO-Pd NPs-GPC3 Apt specifically recognizing more GPC3 to form more sandwich complexes, resulting in increasing metallic Ag NPs deposition and enhancing electrochemical signal responsiveness. The electrochemical nanobiosensor working curve was shown in Figure 5B. The electrochemical nanobiosensor responded logarithmically linearly to the GPC3 concentration from 0.01 to 10.0 µg/mL. The linear regression equation was Y = 5.2923 × lgX + 22.44 (Y represented the current response, X represented the concentration of GPC3) with R 2 of 0.9941. When the GPC3 concentration was from 10.0 to 100.0 µg/mL, the electrochemical nanobiosensor responded linearly to GPC3 concentrations. This equation corresponded to Y = 0.1088 × X + 25.3816 in linear regression with R 2 of 0.9715. The response time of the nanobiosensor was 30 min, and the sensitivity of the nanobiosensor was calculated to be 1.535 µA µM −1 cm −2 , which was calculated by dividing the calibration curve gradient by the electrode's valid area [33]. The detection limit of detection (LOD) for the nanobiosensor was evaluated using the signal-to-noise ratio (S/N = 3), and the LOD of 3.30 ng/mL was obtained [34].
The comparisons between the developed electrochemical nanobiosensor and other GPC3 detection methods are shown in Table 1. Compared with other methods, the GPC3 nanobiosensor had a wider detection range and higher sensitivity. This may be due to the signal amplification strategy of silver deposition catalyzed by the enzyme being used to strengthen the response current signal to a certain extent, resulting in higher sensitivity. However, the LOD value obtained by this project (3.30 ng/mL) was close to that of the previous research group (2.86 ng/mL) [35], but was slightly worse than those of ELISA, BLEIA, or TRFIA [4,10,36]. On the one hand, the impedance of the "antibodyantigen-aptamer" sandwich complex was larger, resulting in slightly poor LOD. On the other hand, the "antibody-antigen-aptamer" sandwich complex exhibited spatial stereoscopic properties, which would obstruct the catalytic deposition of Ag NPs by H-rGO-Pd NPs nanozymes, to some extent, resulting in a less current change. Although the LOD value of the designed sensor was slightly higher, the antibody-antigen-aptamer sandwich structure and the enzyme-catalyzed silver deposition signal amplification strategy improved the stability and sensitivity of the sensor. Moreover, due to its easy synthesis and lower cost, the GPC3 nanobiosensor can be an ideal solution for designing high-sensitivity clinical tests.  2.8. Specificity, Stability, and Reproducibility of GPC3 Electrochemical Nanobiosensor AFP, HSA, IgE, and IgG were chosen as interfering substances to examine the specificity of the electrochemical nanobiosensor. Briefly, 1.0 µg/mL GPC3, one of the above four interfering chemicals (10.0 µg/mL), or a mixture (all interfering substances) was mixed with GPC3 in a 1:1 ratio with the same concentration (1.0 µg/mL), for which the results are shown in Figure 5C. According to Figure 5C, the response current value of GPC3 was 24.30 µA, which was about three times that of the current value of other interfering substances (AFP, 8.31 µA; HSA, 8.01 µA; IgE, 8.38 µA; IgG, 7.68 µA). In addition, the response of the mixture (21.90 µA) was very close to that of GPC3. Therefore, the presence of other interfering substances had little influence on the corresponding current value of the nanobiosensor, indicating that the specificity of the nanobiosensor would meet the detection requirements.
To detect the GPC3 protein (1.0 µg/mL) and analyze its stability by using the DPV technique, the GPC3 electrochemical nanobiosensor was kept at 4 • C in a refrigerator, and would be taken out at various intervals (1, 3, 6, and 8 days) ( Figure 5D). The results showed that the response current of nanobiosensor decreased to 99.5%, 96.0%, and 93.1% of the original current value after 3, 6, and 8 days, respectively, indicating the GPC3 electrochemical nanobiosensor had good short-term stability. Furthermore, the reproducibility of the GPC3 electrochemical nanobiosensor was also studied (see Figure S4). The relative standard deviations (RSDs) of the response currents of five sensors was 2.56%, suggesting that the GPC3 electrochemical nanobiosensor can be reproducible.

Analysis of GPC3 in Actual Human Serum Samples
To verify that the nanobiosensor would be applied in actual serum detection, GPC3 in human serum samples was detected by standard addition methods under optimal conditions. The approval was first received from the Ethics Committee of the Guangxi Key Laboratory of Metabolic Disease Research of the 924th Hospital of the Chinese People's Liberation Army Joint Logistic Support Force (Guilin, China). Three kinds of serum samples were prepared and detected with the same process instead of GPC3 standard solution and the results are shown in Table 2. As seen in Table 2, the recovery rate ranged from 103.78 to 106.52%. There was a range of 1.89% to 8.81% in the RSDs value (n = 3). The developed sensor showed a promising ability for the determination of GPC3 in actual serum samples.

Electrochemical Measurements and Apparatus
All electrochemical experiments were conducted on the electrochemical workstation (CHI660E, Shanghai Chenhua Instrument Co., Ltd., Shanghai, China) at room temperature. Electrochemical measurements were performed with conventional screen-printed electrodes (SPE, PK215, Nanjing Yunyou Biotechnology Co., Ltd., Nanjing, China), one of the carbon paste electrodes served as an auxiliary electrode, the other as a working electrode (surface area = 0.07 cm 2 , Φ = 3 mm), and the reference electrode was Ag/AgCl inert.

Preparation of the H-rGO-Pd NPs Nanozyme and H-rGO-Pd NPs-GPC3 Apt Detection Probe
H-rGO-Pd NPs nanozymes were prepared by a two-step reduction method. First, welldispersed hemin-reduced graphene oxide (H-rGO) solution was prepared as described in our earlier study [35]. Briefly, 10.0 mg GO powder was dispersed in 20.0 mL pure water, and crushed by cell crusher ultrasonic for 2 h to obtain GO solution (0.5 mg/mL), then Hemin aqueous solution (0.5 mg/mL) was added to mix evenly, and 30 µL ammonia solution (NH 3 ·H 2 O) and 100.0 µL hydrazine hydrate (NH 2 -NH 2 ) were added. After mixing, the solution was centrifugally cleaned in a water bath at 60 • C for 4 h to obtain the H-rGO solution. Second, 2.0 mL of PDDA (ω = 0.2%) and 5.0 mL of NaCl (0.2 M) were added into the 10.0 mL of H-rGO solution (0.5 mg/mL) and stirred for 12 h to form PDDA-modified H-rGO solution. Then, 2.0 mL of Na 2 PdCl 4 (20 mM) and 10.0 mL of EG were added into the PDDA-modified H-rGO solution. After being stirred overnight, the pH of the solution was adjusted to 12 with 1.0 M of NaOH. Furthermore, the above solution was refluxed for 4 h at 140 • C. Finally, the H-rGO-Pd NPs were obtained by centrifuging and drying.
H-rGO-Pd NPs-GPC3 Apt detection probe was prepared through π-π action and Pd-N coordination interaction. A total of 100.0 µL of GPC3 Apt solution (5 µM) and 200.0 µL of H-rGO-Pd NPs (0.5 mg/mL) were sonically mixed overnight at room temperature. Then, the solution was centrifuged at 12,000 rpm for 20 min to remove free aptamers. Thereafter, the H-rGO-Pd NPs-GPC3 Apt detection probe (1.0 mg/mL) was obtained after the residue was dispersed in Tris-EDTA buffer.

Construction of the GPC3 Electrochemical Nanobiosensor
Firstly, the SPE was immersed in 5 mL of H 2 SO 4 solution (0.5 M) and activated by an electrochemical cyclic scanning method with a scanning speed of 0.5 V/s and a scanning voltage between 0.4 and 1.2 V for 20 cycles. Secondly, the activated SPE was placed in a 5 mL mixed aqueous solution of HAuCl 4 (2.5 mL, ω = 0.05%) solution and GO solution (2.5 mL, 1.0 mg/mL), and electrodeposited under cyclic voltammetry (CV) strategy for 120 s under magnetic stirring in the voltage range of −0.5-1.0 V. The scanning rate was 0.4 V/s and the scanning period was 10 cycles. After electrodeposition, the SPE was rinsed with water several times and dried to get Au NPs@rGO/SPE [38][39][40][41]. Thirdly, 1.0 µL GPC3 Ab was dropped on the surface of Au NPs@rGO/SPE electrode and incubated for 30 min at 25 • C. Lastly, 6.0 µL of 1% BSA solution was added dropwise to GPC3 Ab /AuNPs@rGO/SPE and incubated for 30 min at 25 • C to block non-specific active sites [42], After each step, ultrapure water was used to clean the electrodes and to dry them.

GPC3 Detection Based on H-rGO-Pd NPs Nanozymes-Catalyzed Silver Deposition
Firstly, 1.0 µL GPC3 standard solution (different concentration) was added dropwise onto the GPC3 Ab /AuNPs@rGO/SPE surface and incubated at 25 • C for 30 min. Secondly, 4.0 µL of H-rGO-Pd NPs-GPC3 Apt (1.0 mg/mL) solution was added dropwise onto the surface of GPC3/GPC3 Ab /AuNPs@rGO/SPE and incubated at 25 • C for 60 min. Thirdly, 6.0 µL of H 2 O 2 (100 mmol/L) and 3.0 µL of AgNO 3 (6.0 mmol/L) solution were added dropwise onto the surface of H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab /Au NPs@rGO/SPE and kept in the dark at 25 • C for 30 min. The electrode was rinsed three times with water. Lastly, the electrode was inserted into a 4.0 mL glycine-NaOH buffer solution (0.05 M, pH 8.5) containing 0.1 M HNO 3 and 0.6 M KNO 3 solution and recorded the electrochemical responses with differential pulse voltammetry (DPV) method with scanning range from −0.4 to 1.0 V with a 0.1 V/s scanning rate. Each sample was detected three times, and the results were calculated as mean ± RSD.

Detection of GPC3 Level in Human Serum Samples
To verify that the developed electrochemical nanobiosensor would be applied in serum detection, GPC3 in human serum samples was detected by standard addition methods under optimal conditions. Firstly, the human serum samples were obtained after approval from the Ethics Committee of Guangxi Key Laboratory of Metabolic Disease Research, 924th Hospital of the People's Liberation Army Joint Logistics Support Force (Guilin, China). Three kinds of serum samples were prepared for determination by mixing 1.5 µL of normal serum with 1.5 µL of GPC3 solution (50.0 µg/mL, 80.0 µg/mL, and 100.0 µg/mL). The serum samples were detected as the above process instead of the GPC3 standard solution. The DPV of the electrochemical workstation was used for the determination. The measured concentration of GPC3 in human serum samples was calculated by the calibration line. Each sample was detected three times, and the results were calculated as mean ± RSD.

Conclusions
In this study, one novel electrochemical nanobiosensor was constructed for the quantitative analysis of GPC3 based on H-rGO-Pd NPs nanozyme for signal amplification, combining GPC3 Apt and GPC3 Ab as recognition elements. In the presence of GPC3, both GPC3 Ab and H-rGO-Pd NPs-GPC3 Apt specifically bonded with GPC3 to form the H-rGO-Pd NPs-GPC3 Apt /GPC3/GPC3 Ab sandwich-structure complex with a stable spatial structure as well as catalytic performance, which could enhance H 2 O 2 to reduce the Ag ions in solution to metallic Ag, resulting in the deposition of Ag NPs on the surface of the biosensor. The amount of deposited Ag, which was derived from the amount of GPC3, was quantified by the DPV method. The developed nanobiosensor was able to determine GPC3 with the LOD of 3.30 ng/mL and showed good specificity, short-term stability, and recovery rates. Although the LOD value of the designed sensor was slightly higher, the GPC3 nanobiosensor can be an ideal solution for designing high-sensitivity clinical tests. We believe that this method may be an effective strategy for the determination of GPC3 with potential clinical applications and can be used to build accurate and simple sensors for other biomarkers.  Figure S4. Reproducibility of GPC3 electrochemical nanobiosensor.