Highly Sensitive Amperometric α-Ketoglutarate Biosensor Based on Reduced Graphene Oxide-Gold Nanocomposites

Herein, a rapid and highly sensitive amperometric biosensor for the detection of α-ketoglutarate (α-KG) was constructed via an electrochemical approach, in which the glutamate dehydrogenase (GLUD) was modified on the surface of reduced graphene oxide-gold nanoparticle composite (rGO-Aunano composite). The rGO-Aunano composite was one-step electrodeposited onto glassy carbon electrode (GCE) surface and was characterized by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS), and electrochemical techniques. In addition, the rGO-Aunano/GCE was also found to electrocatalyze the oxidation of β-nicotinamide adenine dinucleotide (NADH) at the peak potential of 0.3 V, which was negatively shifted compared with that at bare GCE or Aunano/GCE, illustrating better catalytic performance of rGO-Aunano. After the modification of GLUD, the GLUD/rGO-Aunano/GCE led to effective amperometric detection of α-KG through monitoring the NADH consumption and displayed a linear response in the range of 66.7 and 494.5 μM, with the detection limit of 9.2 μM. Moreover, the prepared GLUD/rGO-Aunano/GCE was further evaluated to be highly selective and used to test α-KG in human serum samples. The recovery and the RSD values were calculated in the range of 97.9–102.4% and 3.8–4.5%, respectively, showing a great prospect for its real application.


Introduction
α-Ketoglutarate (α-KG) is an important metabolic marker of early diagnosis of cancer and microbial fermentation monitoring and also is a vital intermediate in the tricarboxylic acid cycle as well as a node to connect the carbonnitrogen metabolism in cells [1][2][3][4]. It has been emphasized to possess exciting angiogenesis suppressor activity [5][6][7]. Furthermore, micronutrient application of α-KG has exhibited beneficial effects on several malignant tumors with only minor negative effects on normal cells [4,5]. erefore, accurate and sensitive methods for the detection of α-KG are urgently required. Till now, several methods have been developed for the quantification of α-KG, such as electrochemical [8], gas chromatography-mass spectrometry (GC-MS) [4], and high-performance liquid chromatography (HPLC) [5,9]. Among them, electrochemical methods have become considerably meaningful due to the advantages of convenience, high speed, low cost, and easy-to-use. Currently, Wang and his coworkers have developed an electrochemical biosensor for rapid and sensitive detection of α-KG based on ruthenium-rhodium modified carbon fiber enzyme microelectrode [8]. However, a large overpotential was found at the microelectrode and the detection of α-KG was prone to suffer interference from other compounds. erefore, it is important to prepare a capable modified electrode with lower overpotential for α-KG detection and meanwhile maintain the bioactivity of the immobilized enzyme.
Graphene (Gr) has become extremely attractive in many fields since it has superiorities such as good biocompatibility [10], large surface-to-volume ratio, excellent conductivity [11], electron mobility, and flexibility [12]. In the electrochemical biosensing fields, its large surface-to-volume ratio is helpful for increasing the loading amount of the enzyme, and its excellent conductivity is favourable for transferring electrons between the electrode surface and electrolyte [13][14][15][16][17][18][19]. However, the practical applications of Gr are challenged by its irreversible agglomeration or even restack to form graphite due to the Van der Waals interactions in the drying state [20]. Recent research found that noble metal nanoparticles could be used to effectively prevent the massive agglomeration of Gr sheets. Si got a Pt nanoparticles (Pt nano )-Gr composite with the Gr partially exfoliated from its drying aqueous dispersions [21]. Tang et al. founded a rapid and efficient one-step approach to prepare Gr-Ag nano composite by simultaneous reduction of Gr oxide (GO) and Ag + with formaldehyde as the reducing agent [22]. Li et al. developed a sensor for the detection of paracetamol based on Pd nano -GO composite, which was obtained by a one-pot chemical reduction using Pd 2+ as a precursor without large aggregation [23]. Also, noble metal nanoparticles-Gr composites have been found to have excellent catalytic performance in electrochemical biosensors [24][25][26]. Leng et al. demonstrated that the Pd nano -GO composite could electrocatalyze the oxidation of rutin, with the detection limit of 0.001 μM [27]. Govindhan et al. prepared Au nano -Gr composite which could catalyze NADH with the detection limit as low as 1.13 nM, since the composite provided a large electrochemical active surface area and a favourable environment for electron transfer from NADH to the electrode [28]. erefore, in this work, we were ready to construct an α-KG biosensor based on reduced GO-Au nano (rGO-Au nano ) composite.
In this paper, the rGO-Au nano composite was one-step electrodeposited onto the glassy carbon electrode (GCE) surface, which exhibited a lower working potential for the electrochemical oxidation of NADH, and served as the platform for the immobilization of glutamate dehydrogenase (GLUD). α-KG was detected via monitoring the NADH consumption, and the mechanism was described as in the following equation [8]: Catalyzed by GLUD, α-KG was conversed to L-glutamate in the presence of NH 4 + and NADH. e more amount of α-KG contained in solution, the more NADH was consumed. en, if a certain amount of NADH was added in advance, the remaining amount of NADH would decrease along with the increase of α-KG. erefore, the concentration of α-KG was inversely proportional to the catalytic current of NADH, which provided the quantitative basis for α-KG detection.

Characterization and Electrochemical Measurement.
Surface morphologies of rGO and rGO-Au nano films were investigated by scanning electron microscopy (SEM, JEOL, JSM-6510, Japan). e elemental composition analysis was performed by energy dispersive X-ray spectroscopy (EDS, Vantage 4105, NORAN). All electrochemical measurements were carried out on a CHI 650E electrochemical workstation (Shanghai Chenhua Instrument Company, China), employing a typical three-electrode cell system. A modified or bare GCE was utilized as the working electrode (V � 3 mm), whereas a saturated calomel electrode (SCE) as a reference and a platinum foil electrode as a counter. All potentials were measured against SCE. All experiments were purged with high-purity nitrogen to remove oxygen and done at room temperature (∼25°C).

Fabrication of α-KG Electrochemical Biosensor.
e fabrication process of the electrochemical biosensor was illustrated in Scheme 1. Prior to electrodeposition, the GCE was polished using 0.3 and 0.05 μm alumina powder until a mirror-shiny surface was obtained, and followed by ultrasonication in the ethanol and double-distilled water for 2 min, respectively. e cleaned GCE was then modified by the electrochemical codeposition of the GO and Au nano . According to the method reported by Liu et al. [29], GO was exfoliated in CBS by ultrasonication for 20 min to form a homogeneous few-layer GO dispersion. Cyclic voltammetric (CV) reduction was performed in the deposition solutions containing 1.0 mg·mL −1 of GO and 200 μM of chloroauric acid with magnetic stirring. e CV was carried out between −1.4 and 0.6 V at a rate of 20 mV·s −1 , and the deposition amount was optimized as six potential cycles. en, 8 μL of 110 kU·L −1 GLUD PBS (0.1 M, pH 7.2) was dropped onto rGO-Au nano /GCE surface. After natural drying, it was dipped into a 2 wt.% of glutaraldehyde for 3 s to form a protective film and stored in PBS at 4°C before use. (Figure 1(a)). ere is one anodic peak (I) and two cathodic peaks (II and III), corresponding to the redox pair of some oxygen-containing groups [30] and the irreversible electrochemical reduction of GO [31], respectively. e ongoing increase of the peak current with successive potential sweeps indicated that the rGO from GO dispersion was successfully deposited onto GCE. Figure 1(b) shows the CV of codeposition of GO and HAuCl 4 , with a completely different appearance from Figure 1(a). e reduction current was larger than that of GO electrolysis, indicating that the deposition of rGO and Au nano onto the surface of GCE was achieved [29]. e number of electrodeposition cycles was also optimized since the thickness of the rGO-Au nano composite film would influence the conductivity and stability of the rGO-Au nano /GCE. is part would be interpreted in detail in the section of the optimization of the experimental parameters. e morphologies of rGO and rGO-Au nano were characterized by SEM. Figure 1(c) shows the SEM image of rGO film, in which a typically wrinkled texture was displayed. In Figure 1(d), you could find that Au nano uniformly scattered over rGO to make the rGO-Au nano composite. e Au nano prevented the agglomeration of rGO; meanwhile, rGO enhanced the dispersion of Au nano , both of which improve the conductivity and stability of rGO-Au nano composite film. Further, the EDS spectrum confirmed the elemental composition of rGO-Au nano , in which Si element comes from the Si substrate. e FT-IR spectra of GO and rGO are shown in Figure 2.

Characterization of rGO-Au nano Composite. CV was recorded during the electrodeposition of GO on GCE
e FT-IR spectrum of GO (a) exhibits the characteristic absorptions from oxygen-containing functional groups. In detail, the absorption band at 3437 and 1397 cm −1 can be assigned to the stretching vibration and deformation vibration of O-H, respectively. e band at 1052 cm −1 belongs to the C-O (alkoxy), while the band at 1633 cm −1 derives from the vibration of the adsorbed water molecules and/or the contribution of the skeletal vibration of unoxidized graphitic domains. After the reduction, the peaks at 1070 cm −1 assigned to epoxy groups are decreased significantly, clearly indicating the removal of oxygen-containing groups of GO (curve b in Figure 2) and suggesting the successful reduction of GO [32].  6 ] 3−/4− ) was larger than that of bare GCE (curve A), owing to that rGO enhanced the electron transfer ability. e similarly enhanced peak current at Au nano /GCE (curve C) attributed to the good conductivity of Au nano . After the rGO-Au nano composite modified onto the GCE surface (curve D), the peak current rose further owing to the synergy from rGO and Au nano . e electrochemical properties of these modified electrodes were also characterized by EIS. Figure 3 International Journal of Analytical Chemistry 3 which meant that the direct oxidation of NADH at the bare electrode was difficult. After the modification of Au nano , the anodic peak increased and the working potential negatively shifted to 0.62 V, which might be owing to the good oxidation catalytic ability of Au nano (curve B) [32]. For the rGO/ GCE as curve C, a small oxidation peak at 0.24 V was found, showing the excellent catalytic capacity of rGO to NADH and the activation energy of NADH oxidation was declined on the surface of rGO. e reason might be that the presence of abundant defects and edge plane graphite structures in rGO was thought to facilitate the heterogeneous charge transfer at the electrode interface [25,33,34]. us, the electrons released by NADH could be transferred swiftly with the aid of rGO. After the combination of Au nano and rGO, the oxidation peak current of NADH was similar to that of Au nano /GCE, and the peak potential appeared at 0.32 V. e synergic effect of the rGO and Au nano exhibited the capability as a powerful catalyst to the oxidation of NADH [35]. In addition, we also found that no obvious peaks were observed on all these four modified electrodes in PBS without NADH ( Figure S1), indicating NADH could be catalytically oxidized with the aid of Au nano and rGO. e effect of the scan rate on the oxidation current of NADH was also investigated. e CVs of the rGO-Au nano /GCE at different scan rates were recorded as Figure 3(d). A good linear relationship of the anodic peak current with the square    International Journal of Analytical Chemistry root of the scan rate in the range of 30 and 100 mV·s −1 was presented as the inset of Figure 3(d). e result revealed the oxidation of NADH on the rGO-Au nano /GCE was a diffusioncontrolling process, owing to the fast electron transfer rate between rGO-Au nano and the electrolyte.

Optimization of the Experimental Parameters.
To improve the performance of the NADH sensor, the effect of determination conditions such as the electrodeposition cycles, the applied potential, and the solution pH have been investigated in detail. e number of the electrodeposition cycles was optimized by measuring the catalytic current to 0.5 mM NADH using rGO-Au nano /GCE with a different rGO-Au nano film. e more electrodeposition cycles exerted, the thicker the rGO-Au nano composite film would be obtained. As shown in Figure 4(a), the oxidation peak current increased sharply as increasing the electrodeposition cycle number from 2 to 6, due to the enhanced conductivity and catalytic activity along with increasing rGO-Au nano deposition. However, the peak current decreased when the number of electrodeposition cycles was more than 6, owing to the fact that the thicker rGO-Au nano film was inclined to drop off. us, 6 cycles of electrodeposition of rGO-Au nano deposition were selected in this work.
In order to achieve the best electrocatalytic effect while reducing the overpotential, the applied potential should be optimized. e effect of the applied potential was shown in Figure 4(b). As the applied potential increased from 0.1 to 0.3 V, the peak current increased gradually. Further, increasing the applied potential from 0.3 to 0.7 V led to a relatively constant current value. erefore, 0.3 V was selected as the optimal applied potential. e effect of the solution pH on the oxidation of NADH at rGO-Au nano /GCE was also studied by monitoring the peak current of NADH in 0.1 M PBS with different pH values from 6.0 to 8.0. Figure 4(c) shows that the oxidation peak current of NADH increased along with the solution pH value from 6.0 to 7.2, and then decreased. erefore, pH 7.2 was selected as the optimum for further studies.

Detection of α-KG Using GLUD/rGO-Au nano /GCE.
Based on the catalytic mechanism described in the introduction part, the concentration of α-KG could be determined by the depleted amount of NADH at a GLUD/rGO-Au nano /GCE. at is to say, the content of α-KG was inversely proportional to the catalytic current of NADH. To obtain the optimal performance of the biosensor, the concentration of the immobilized GLUD needed to be investigated further. As seen from Figure 5(a), the current response of NADH decreased quickly along with the GLUD concentration increased from 27.5 to 110 kU·L −1 , showing a fast transform from α-KG to L-glutamate. en, the current tended to become steady with a further increase in GLUD concentration, owing to the saturation of GLUD loading capacity. From an economic perspective, 110 kU·L −1 of GLUD was selected in our experiments. Figure 5(b) shows a typical amperometric current-time curve recorded at GLUD/rGO-Au nano /GCE for the successive additions of 66.7 μM α-KG in a stirred pH 7.2 PBS containing 1 mM NADH and 1 mM NH 4 Cl. e response was very fast and the steady-state current response was attained in less than 5 s. e linear response range at GLUD/ rGO-Au nano /GCE was from 66.7 to 494.5 μM, with the sensitivity of 454 μA M −1 and a correlation coefficient of 0.9992. e calculated limit of detection was found to be 9.2 μM (S/N � 3). After the GLUD immobilization, the biosensor sensitivity was decreased due to the enzyme layer act as a barrier that hindered NADH transport. To our knowledge, there are few reports on electrochemical biosensors for the measurement of α-KG, and nearly a few references could be compared. As listed in Table 1, our sensing system exhibited a comparable linear range with a bienzymatic flow injection system [37] and a similar detection limit with the other electrochemical biosensors, suggesting the proposed sensor had a good performance in α-KG detection.

Reproducibility, Stability, and Selectivity of the Biosensor.
e reproducibility of GLUD/rGO-Au nano /GCE was investigated in 0.1 M PBS containing 1 mM NADH and 1 mM NH 4 Cl via recoding the current response during the successive addition of 66.7 μM of α-KG. Six sensors were prepared in different batches, and 10 successive measurements were implemented for each electrode, and the RSD of the current response was 4.8 and 4.2%, respectively. e stability of GLUD/rGO-Au nano /GCE was also evaluated. e current response for 312.5 μM α-KG was measured, and there was nearly no apparent loss during ∼1000 s' operation, as shown in Figure 6(a). Such good stability of GLUD/rGO-Au nano /GCE was attributed to the protection of the glutaraldehyde membrane and the biocompatibility of rGO-Au nano for GLUD immobilization.
Selectivity is important for the practical application of biosensors. An assessment of the interference on the amperometric response to NADH was examined at the GLUD/rGO-Au nano /GCE in the presence of other oxidizable substances, such as dopamine (DA), ascorbic acid (AA), and uric acid (UA). Figure 6(b) shows the amperometric responses at GLUD/rGO-Au nano /GCE after the addition of various interferents. In the presence of 0.25 mM NADH, 0.1 mM DA or UA could not induce apparent interference, while 0.1 mM AA generated an obvious interference, since AA could be oxidized at lower potential compared with DA and UA [38,39]. us, AA should be removed when the biosensor was used for practical application. One possible method for removing AA was the immobilization of ascorbate oxidase onto the electrode surface [40].

Real Sample Analysis in Human Serum.
e designed GLUD/rGO-Au nano /GCE was further evaluated in real human serum using the standard addition method. e human serum was diluted 10 times with 0.1 M pH 7.2 PBS. en, the solution was transferred into the electrochemical cell for analysis. e amperometric current-time measurement was employed for the recovery test to determine α-KG in the human serum samples. As listed in Table 2, the recoveries of α-KG between 97.9 and 102.4% with RSD values in the range of 3.8 to 4.5% were obtained, indicating the strong potential application prospect in clinical diagnosis.

Conclusion
In this work, an α-KG biosensor based on the rGO-Au nano film was successfully prepared. e rGO-Au nano /GCE platform resulted in the improvement of electrocatalytic activity towards the oxidation of NADH with lower working potential, higher sensitivity, stability, and reproducibility. After the modification of GLUD, the electrode exhibited a sensitive response to α-KG in PBS via the consumption of NADH, implying good stability of GLUD on biocompatible rGO-Au nano platform and rapid electron transfer ability between electrode and α-KG. Moreover, the biosensor was applied to determine α-KG in human serum with satisfactory recoveries, illustrating its good selectivity and antiinterference. To summarize, the α-KG sensing strategy reported in this work could be expected for real applications in the future.
Data Availability e data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.