Abstract

Electrochemical enzymatic biosensors represent a promising, low-cost technology for point-of-care (POC) diagnostics that allows fast response and simple sample processing procedures. In this review, we summarize up-to-date literature on NAD+/NADH (β-nicotinamide adenine dinucleotide)-dependent electrochemical dehydrogenase biosensors and highlight their applications in human physiological fluids. A brief comparison of various enzyme immobilization procedures is first presented, discussing preparation processes and principal analytical performance characteristics. In the following section, we briefly discuss classes of biosensors based on redox mediators-mediated electron transfer systems (METs). Finally, the conclusion section summarizes the ongoing challenges in the fabrication of NAD+-dependent electrochemical dehydrogenase biosensors and gives an outlook on future research studies.

1. Introduction

The history of sensor design and fabrication began in 1962, when Clark and Lyons first proposed the concept of enzyme-based electrode [1]. Inspired by the seminal work of Clark and Lyons, various types of enzymatic electrochemical biosensors have been developed successively for the detection of diverse targets (lactate, ethanol, bile acid, etc.), which enabled high-throughput and onsite analysis of biological samples [24]. All such procedures need the addition of sensing elements to the electrode structure through multiple strategies, including physical adsorption, covalent bonding techniques, and mediators [57].

In recent years, enzymatic electrochemical biosensors have gained popularity owing to the high biocatalytic activity and specificity of enzymes, along with the financial accessibility to purified enzymes [8]. β-nicotinamide adenine dinucleotide (NAD+/NADH) is an important coenzyme couple involved in diverse enzymatic reactions during which electrons and protons are transferred [9]. Amperometric techniques and cofactor regeneration approaches bring new opportunities for the fabrication of biosensors and the development of new electrochemical devices. Nevertheless, the existing sensing applications are limited and hampered by drawbacks such as sophisticated immobilization and stabilization protocol for enzymes, selectivity and stability in clinical complex samples, and the need for cofactor regeneration [10]. To expand the analytical possibilities of NAD+/NADH-dependent electrochemical dehydrogenase biosensors, studies are underway to improve immobilization methods, miniaturization of sensor components, and improvement of enzyme stability [11, 12]. Diverse nanomaterials derived from carbon materials, metal materials, polymer composites (e.g., conducting polymers or molecularly imprinted polymers), and hybrid materials (e.g., hydrogel) have also been attempted.

This article reviews the recent developments in NAD+/NADH-dependent electrochemical dehydrogenase biosensors, which include different strategies for biosensor construction, various immobilization methods, analysis performance, and application of sensors in actual sample analysis. Along with this, the merits and challenges of current NAD+-dependent electrochemical dehydrogenase biosensors are highlighted and discussed. As far as we know, this is the first review on NAD+-dependent electrochemical dehydrogenase biosensors in human physiological fluids.

2. Basic Principle of Electrochemical Biosensors

Electrochemical biosensors pose an attractive solution for point-of-care (POC) diagnostics because they are readily integrated with microelectronics and they require minimal instrumentation. Enzyme-coupled biosensing electrodes are designed to immobilize biocatalysts near the electrode surface where the biocatalysts involved must catalyze specific electrochemical reactions. Typically, the biometric element is attached to the surface of the transducer, which provides substrate specificity. The consumption or production procedure is then detected by a transducer which generates a measurable signal (most often an electric current) proportional to the concentration of analytes as shown in Figure 1. Electrochemical dehydrogenase biosensors operate in the presence of NAD+/NADH, acting as a medium to shuttle electrons between the enzyme and electrode. The general equation of electrochemical dehydrogenase biosensors is given as follows:

3. Enzyme Immobilization Technology

As powerful biocatalysts, enzymes have unique substrate specificity and high catalytic activity. For the preparation of electrochemical sensors, the immobilization of enzymes is a very complicated process and has a great impact on the performance of the sensors [13]. Immobilization of enzymes commonly is performed by four methods, including physical adsorption on a support material, covalent binding to a surface (that provides stronger, more stable, and irreversible linkages compared to other methods), entrapment within polymers, and crosslinking between molecules. Detailed results are presented in Table 1.

3.1. Physical Adsorption

Physical adsorption is a simple and fast way for attaching enzymes to the biosensor surface. The application of nanomaterials can enhance the performance of biosensors in several aspects, for instance, it can increase transducer stability and lifetime, improve sensitivity, and can achieve a better time of response [31].

3.1.1. Carbon Nanotubes

Among a variety of nanomaterials, carbon nanotubes (CNTs) with superior electrical conductivity, high surface area, and good chemical properties make them a promising material for enzyme immobilization. However, their hydrophobic aromatic structure is not suitable for electrostatic bonding (unless functionalized). Instead, the physical adsorption of CNTs onto aromatic residues is found to be mainly hydrophobic interactions. To test 3-hydroxybutyrate (3-HB), Khorsand et al. [14] utilized single-walled carbon nanotubes (SWCNT) as a binder to attach 3-hydroxybutyrate dehydrogenase (HBDH) to the screen-printed carbon electrode (SPCE) surface. After the addition of CNTs, the oxidation potential of NADH decreased to −0.05 V. When the biosensor was used to analyze 3-HB in serum samples, the linearity was up to 2.00 mM with the detection limit of 80.00 μM and a good storage stability (180 days) at 4°C. In a similar research study by Khorsand et al. [15], the 3-HB biosensor was established by using SWCNT to fix the cofactor NAD+ on the surface of SPCE, and then HBDH was deposited on the modified electrode. This biosensor presented the linear range of 0.01–0.10 mM and the detection limit of 9.00 μM. The established assay correlated well with the standard β-hydroxybutyrate assay kit available on the market. It is worth mentioning that only the first drop of NAD+ would be enough, instead of the addition of NAD+ for each test. Lately, a simple and rapid procedure was reported for the construction of the androsterone biosensor, in which the strong pi-stacking interaction between cofactors and CNTs provided excellent stability [16]. The use of the Nafion® film enabled the accurate detection of androsterone in the presence of interferents (uric acid and ascorbic acid). As mentioned above, such composite electrodes integrate the capacity of CNTs to facilitate electron transport with the desirable benefits of paste electrode materials. Enzymes or other substances can be physically mixed by noncovalent approaches. Meanwhile, the prepared electrodes preserve the features of traditional carbon paste electrodes such as the feasibility of achieving background current, easy renewal, and recombination properties.

3.1.2. Reduced Graphene Oxide

Reduced graphene oxide (rGO) with large specific surface areas and abundant functional groups is an ideal substrate for enzyme immobilization [32]. Moreira et al. [17] immobilized the phenylalanine dehydrogenase enzyme (PheDH) onto the paper microzone by physical adsorption for the phenylketonuria (PKU) screening in neonatal samples (Figure 2). The electrochemical oxidation was investigated by differential pulse voltammetry (DPV) at 0.6 V. The response was linear from 1.00 μM to 600 μM with the detection limit of 0.20 μM. It has been reported that the fixation of metals and metal oxide nanoparticles on the surface of RGO prevents the aggregation of graphene sheets and promotes ion transfer [33]. More recently, metal and metal oxide nanoparticles have been integrated on GO through a single-step synthesis in which metal salts were coreduced simultaneously along with GO [13, 33]. A selective biosensor for α-ketoglutarate (α-KG) analysis was developed through the attachment of glutamate dehydrogenase (GLUD) onto the surface of the rGO-Aunano composite [13]. It exhibited a linear behavior in the 66.70–494.50 μM and the detection limit of 9.20 μM. The precision of the spiked serum (n = 3) was in the range of 3.8%–4.5%, with recoveries of 97.9%–102.4%.

3.1.3. Metal Oxide

Among nanostructured metal oxides, it is noted that CeO2 nanoparticles (isoelectric point 9.2) can immobilize biomolecules with low isoelectric points via electrostatic interactions, which helps to preserve their biological activity [34]. Nesakumar et al. [18] coated a glassy carbon electrode (GCE) with a thin layer of carbon paste in which CeO2 nanoparticles with a face-centered cubic structure were embedded, and then immobilized NADH and lactate dehydrogenase (LDH) at the interface. The amperometric response to the standard concentrations of lactate was found to be linear from 0.20 mM to 2.00 mM with the sensitivity of 571.19 μA·mM−1 and a response time of ≤4 s.

In addition to this, a solid ionic lactate biosensor was designed to immobilize LDH-containing NAD+ onto a doped graphene-like membrane [19]. The biosensor showed two linear responses in the concentration range of 0.55–5.55 μM and 5.50–3330.00 μM, respectively, with a detection limit of 0.165 μM. The recoveries ranged from 96.7% to 105.8%, with a relative standard deviation of (RSD) ≤ 3.16%, indicating that the biosensor was suitable for the analysis of lactate in real samples.

3.1.4. Inorganic Mesoporous Materials

Compared to polymers, inorganic mesoporous materials have drawn significant attention as support materials for molecular catalysts owing to their excellent thermal stability as well as chemical inertness. The immunity to heat can hamper proteins from experiencing abundant conformational transforms inside the pores of solid supports [11]. Hasanzadeh et al. [11] fixed proline dehydrogenase (PRODH) onto a novel carbon paste electrode modified with mesoporous silica nanomaterials which have a large surface area (362 m2·g−1). The catalytic activity of PRODH-entrapped magnetic mesoporous silica nanomaterials remained stable at 70°C. The engineered biosensor has a linear range of 0.01–0.15 μM and a detection limit of 0.006 μM, which can be used to measure L-proline in whole blood, normal, and malignant cell lines. The immobilized PRODH exhibited greater activity over wider ranges of pH values and temperatures than the free form.

3.2. Covalent Bonding

Covalent bonding offers stronger interactions than physical adsorption because it can offer an exceptionally thin, uniform, and stable surface. Chemical conjugation via the coupling of carboxylic acid group (COOH), amino group (NH2), alcohol group (OH), or azide-alkyne cycloaddition, and sulfhydryl-maleimide coupling are usually used to covalently attach hydrophilic functional groups to the surface of the nanomaterial [35]. Accordingly, most chemical covalent modifications in electrochemical dehydrogenase biosensor studies were formed using an amide bond between amine-modified oligonucleotides and the carboxylic acid groups of the nanotube [12, 20, 21]. Rahman et al. [20] constructed an electrochemical method for lactic acid detection, in which LDH and NAD+ were successively fixed on poly-5,2′-5′,2″-terthiophene-3′-carboxylic acid (pTTCA)/multiwalled carbon nanotubes (MWCNTs) membrane, followed by the activation step of N-(3-dimethylamino-propyl)-N′-ethylcarbondiimide hydrochloride (EDC). The biosensor response was linear from 5.00 μM to 90.00 μM with a limit of detection of 1.00 μM. In another study, Teymourian et al. [21] used a simple coprecipitate procedure to in-situ load magnetic Fe3O4 nanoparticles onto the surface of MWCNTs. LDH and NAD+ were immobilized through a similar procedure where the -COOH groups present on the Fe3O4/MWCNTs film form a covalent bond with the amino group of the enzyme. DPV detection of the biosensor to lactate displayed linear responses over the concentration range of 50.00–500.00 μM with a detection limit of 5.00 μM and a sensitivity of 7.67 μA·mM−1.

The determination of targets in complex substrates employing traditional electrodes remain a major challenge under the influence of interferences. In view of this, Premaratne and his colleagues [12] constructed a biosensor employing a pyrenyl carbon nanostructure complex, with the capability of eliminating interferences. This system was conjugated with a flow injection analysis (FIA) system to determine formaldehyde (HCHO) in urine (Figure 3). For this purpose, the gold screen printed electrodes (AuSPEs) were modified with polymer films via strong π-π interactions between MWNTs and 1-pyrenebutyric acid (PBA). Thereafter, the polymer-modified AuSPEs were coated with a freshly prepared mixture of 0.35 M 1-ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDC) and 0.1 M N-hydroxysuccinimide (NHS) followed by aliquots of formaldehyde dehydrogenase (FDH) solution. The constructed electrodes were connected to an internal flow cell that was concatenated to an injection pump and a sample injector. The flow injection method for the designed bioelectrode significantly reduced the LOD to 6 ppb, which was 12-fold less than the agitation-solution method. The sensor showed improved selectivity to HCHO with a moderate cross-reactivity for acetaldehyde (CH3CHO) and negligible cross-reactivity for propionaldehyde, acetone, methanol, and ethanol. The response of the bioelectrode to HCHO in 10-fold diluted urine was found to be linear from 10 ppb to 10 ppm with the detection limit of 6 ppb.

3.3. Entrapment

Physical adsorption techniques lead to problems with protein desorption due to changes in external conditions, while enzyme embedding in appropriate polymer matrix lattices offers a relatively better enzyme retention.

3.3.1. Sol-Gel

Sol-gel materials offer efficient means for fixing enzymes via inorganic oxo (M-O-M) or hydroxo (M-OH-M) bridges to formulate a continuous network containing liquid phases which can then be dried out to form solid matrices.

A dual-enzyme Clark electrode for the detection of 3-HB was established by specific dehydrogenation of HBDH and salicylate hydroxylase (SHL) coated with a poly(carbamoyl) sulfonate (PCS) hydrogel on a Teflon membrane [22]. The operation of the biosensor was based on the specific dehydrogenation of 3-HB consuming NAD+ catalyzed by HBDH, which leads to the production of NADH as shown in the following equation:

Then, SHL catalyzes the irreversible decarboxylation and the hydroxylation of salicylate in the presence of oxygen and NADH as shown in the following equation:

Clearly, the presence of interference in human body fluids was found to be minimal due to the combination of HBDH and Teflon membranes. The total reaction time was less than 5 min with the linear range of 8.00–800.00 μM and the detection limit of 3.90 μM.

Employing the similar procedure, the researchers constructed the L-lactate detecting device by combining enzymes with rGO-AuNPs in a sol-gel matrix derived from tetramethoxysilane and methyltrimethoxysilane [23]. The determination of L-lactic acid could be almost free from the interferences of urate, paracetamol, and L-ascorbate. It provided a sensitivity of 154 μA/mM·cm2, a linearity of 0.01–5.00 mM, and a coefficient of variation of 2.5%. The constructed biosensor could be stored in a desiccator at 4°C·s for over 25 days.

3.3.2. Composite Membrane

Apart from immobilizing enzymes, the membrane can also insulate the electroactive substances and reduce signal interference. Nevertheless, they also have some key weaknesses including leakage and conductivity. To solve these difficulties, Hua et al. [24] established a novel simple and valid enzyme embedding device employing a nanobiocomposite synthesized by sequentially adding graphitized mesoporous carbons (GMCs), Meldola’s blue (MDB), alcohol dehydrogenase (ADH), and NAD+ into chitosan (CS) solution. In this device, CS possessed cationic properties, good membrane forming ability and adhesiveness, and great biocompatibility, which can not only be used as a dispersant of GMCs but also as a medium for fixing ADH. The constructed disposable biosensor presented a fast amperometric response (5 s), good sensitivity (67.28 nA·mM−1), wide linear range (0.50–15.00 mM), and low detection limit (80.00 μM) towards ethanol. The recoveries ranged from 97.2% to 106.0% and the coefficient of variation within and between batches was less than 5%.

The binding of aldehyde to NAD+ resulted in the formation of NADH and aldehyde being bound to the zinc active site. ADH usually catalyzes ethanol oxidation through the following bi-bi mechanism: [36]:where E represents the enzyme, A represents NAD+, B stands for ethanol, is acetaldehyde, and Q denotes NADH.

Similarly, Adhikari and coworkers [25] reported a new enzyme embedding device coated with a special cationic polymer, poly(2-(dimethylamino)ethyl methacrylate) (MADQUAT) on a SWCNT-rGO nanohybrid thin film, which catalyzed the oxidation and reduction of electroactive substances. The chemical oxidation of NAD+ by the ethanol occurred as shown in the following equation:

When the applied voltage is +0.5 V, the electrochemical reoxidation of NADH to NAD+ results in the analytical response as expressed by the following equation:

The principle of ethanol oxidation was catalyzed by ADH, which consists of four crystallographically distinct, but structurally similar, subunits arranged as two dimers. Using the device, the reduction of NADH produced by the enzyme was accomplished at a relatively lower potential (+0.5 V vs. Ag/AgCl) and the limit of detection for ethanol was 0.16 μM, with the sensitivity of 1.84 μA·mM−1; these results showed that ADH catalyzing ethanol with MADQUAT as a redox mediator was successful. The accurate determination of ethanol in complex specimens is of great significance in clinical and forensic medicine. The biosensors for ethanol detection are based on either alcohol oxidase (AOX) or ADH. The ADH-based biosensor was superior to the AOX-based on comparisons of stability and specificity [37].

3.4. Crosslinking

Typical crosslinking occurs through the application of a chemical agent called a cross-linker, most commonly a lysine linker, to covalently concatenate two residues that are close in space within or between proteins. The most commonly used crosslinking additives are EDC/NHS [26] and glutaraldehyde (GA) [2730] for crosslinking tissue scaffolds and enhancing structural stability.

3.4.1. EDC/NHS

An amperometric urea biosensor was developed by immobilizing urease (Urs) and glutamate dehydrogenase (GLDH) on a silicon substrate precoated with N2-incorporated diamond nanowire (N-DNW) thin films which were synthesized by a microwave plasma enhanced chemical vapor deposition (MPECVD) technique [26]. The EDC–NHS-initiated -COOH group was covalently connected to the NH2 terminal of Urs and GLDH to form a covalent amide bond (CO-NH). The biosensor exhibits good performance in sensitivity (6.18 μA/mg·dL/cm2), linearity range (10.00–100.00 mg/dL), lower detection limit (3.87 mg/dL), and fast response time (>10 s). The biosensor maintained 80.0% of its original activity towards urea after being stored in the refrigerator at 4−6°C for 1 month.

3.4.2. GA

A model enzyme glucose dehydrogenase (GDH) was attached to the CNT-CS membrane and used for detecting glucose in the urine matrix without interference [27]. GDH could be covalently fixed by the reaction of amino groups in highly biocompatible and hydrophilic CS with the bifunctional crosslinking agent GA. The device permitted relatively rapid (∼60 s) determination of glucose at low potentials (0.40 V) without the presence of redox mediators. The biosensor displayed a fast response time (<5 s), a wide measuring range (5.00–300.00 μM), a high sensitivity (80 ± 4 mA·M−1·cm−2), and a low detection limit (3 μM). Koide et al. [28] reported a similar procedure for the detection of bile acids in urine by employing an electrochemical biosensor coated with Nafion®, which both reduced the presence of interfering substances and enhanced the long-term stability of the reference electrode. Then, three enzymes (bile acid sulfate sulfatase: BSS, 3α-hydroxysteroid dehydrogenase: 3α-HSD, and NADH oxidase: NHO) were fixed with GA onto the Nafion® coating sensor chip. The response time was 5 min with a linearity of 2.00–100.00 μM. It also exhibited high reproducibility (CV values: ≤10%) and continuous repeatability of measurement (CV value: 5%–11%).

Furthermore, several devices for the detection of serum ethanol were successfully manufactured by using a chemical crosslinking method with GA as the crosslinker. Luo et al. [29] produced a disposable serum ethanol biosensor that fixed ADH and NAD+ on the Nafion®-MDB-modified SPE based on the cross-linking method. When the voltage is applied to the Ag/AgCl reference electrode at −0.17 V, the response time of the biosensor is less than 30 s, and the linear range is 5.00 mM. Gao et al. [30] also produced a new electrochemiluminescence (ECL) ethanol biosensor based on Ru(bpy)32+ and ADH fixed by the rGO/bovine serum albumin (BSA) composite membrane. The ECL response of the ADH/Ru(bpy)32+/rGO/BSA electrode to standard concentrations of ethanol was found to be linear in the range of 1.00–2000.00 μM with a detection limit of 0.10 μM. The ECL strength maintained about 90% of the original value after being kept in the refrigerator at 4°C for one month.

4. Mediated Electron Transfer

To help to overcome the limitations of accessibility and proximity and to reduce the susceptibility to interfering species by lowering electrode potentials, redox mediators have been used in electrochemical dehydrogenase biosensors. The medium is attached to the redox enzyme and mediates electron transfer from the enzyme’s redox center to the surface of the working electrode (Figure 4). A suitable mediator in the course of electrocatalytic reaction should be (i) long-term stability, (ii) antifouling effect, (iii) higher electron transfer rate constant (ks), (iv) formal potential (E0’) of the redox mediator should be less than the oxidation potential of NADH, and (v) reduction of considerable overpotential. Mediators used for electrochemical dehydrogenase biosensors in human physiological fluids can be divided into four types as quinine groups, metals and metal complexes, aromatic diamines, and organic dyes (Table 2).

4.1. Quinine Groups

Luo et al. [38] constructed a disposable amperometric biosensor fabricated by immobilizing ADH and NAD+ coated with Nafion® combined with AuNPs onto the surface of SPCE modified with MDB. As an electronic medium, MDB could facilitate the conversion of NADH to NAD+ at a potential of 0.0 V (vs. saturated calomel reference electrode, SCE), thereby, eliminating interferences of oxidizable substances present in real blood samples. The biosensor was linear from 2.00 mM to 8.00 mM with a correlation coefficient of 0.996. In addition to this, a reagentless biosensor has been successfully constructed to detect glutamate in food and clinical samples by mixing unpurified MWCNTs with CS to immobilizing GLDH and NAD+ and coating them layer by layer on the surface of MDB-modified SPCE [6]. The glutamate content in the sample was calculated by employing the standard addition method where the currents produced by each addition were plotted and the resulting line was extrapolated. The electrocatalysis occurred at very low potentials (from approximately +0.1 V vs. Ag/AgCl). However, analytical indexes such as sensitivity, linear range, and detection limit were not explained. It has also been proposed that MDB hindered the development of biosensor electrodes by inhibiting the NAD+-dependent enzyme. These mediators were covalently bound to important thiol groups in the enzyme. However, the inhibitory effect can be eliminated by employing a 1,10-phenanthroline quinone (1,10-PQ) medium, which can hinder 1,4-nucleophilic addition with enzyme amino acid residues [39]. These mediators incorporated the reactive quinone double bonds into heteroaromatic rings. When the SPCE bioelectrodes incorporating 1,10-PQ, NAD+, and HBDH were used to analyze 3-HB in blood, it exhibited a linear response range of 0.00 mM–6.00 mM, using a small volume of blood (5 μl) and it was stable for up to 18 months at 30°C, and shortened the total assay time to 30 s. In another research, Wang et al. [40] constructed a “pop-up” electrochemical paper-based analytical device (pop-up-EPAD) that allowed an enzymatic assay for 3-HB in blood to be read with a commercial glucometer. NAD+ and HBDH reagents were stored on the devices. The sample and 1,10-PD were added to the reaction area on the top layer of the paper device and then entered into the detection area. Once the sample passed through the electrodes, the glucose meter began testing at the potential of +0.2 V and displayed the final values. The pop-up-EPADs exhibited the linear range of 0.10–6.00 mM and the detection limit of 0.30 mM.

4.2. Metal and Metal Complexes

Liao et al. [41] built an amperometric biosensor by employing an iridium nanoparticle as a redox mediator doped into a carbon paste and utilized it for triglyceride (TG) in human serum. The catalytic and electrochemical procedures of TG can be described by the following reactions:

Followed by the electroregeneration of the Ir3+ at an applied potential of +0.15 V vs. Ag/AgCl which is expressed as

The linearity was found to be in the range of 0.00–10.00 mM with a sensitivity of 7.50 nA·mM−1. The method for the detection of TG can overcome the interference of UA and AA.

Moreover, Li et al. [42] fixed HBDH, NAD+, and the electron mediator of potassium ferricyanide (K3 [Fe(CN)6]) onto the SPCE precoated with a layer of hydrophilic gel sodium carboxymethyl cellulose (CMC). A comprehensive evaluation of the immobilized HBDH/NAD+/Fe(CN)63−/CMC/SPCE system showed that the procedure was suitable for the detection of 3-HB in whole blood or serum with very small sample volumes (2.0 μL) and a response time of 50 s. The amperometric response was found to be linear from 1.50 mg/L to 500.00 mg/L with a sensitivity of 0.011 μA·(mg/L)−1. At storage conditions of around 23°C, the design increased shelf-life to more than 60 days.

Moreover, Zhang et al. [43] developed a 3α-HSD-based indirect electrochemical method to accurately determinate bile acids concentrations in serum with unmodified SPCE. The amperometric response of the optimized bile acids detector on exposure to standard concentrations of bile acids was linear from 5.00 μM to 400.00 μM. Known concentrations of bile acids were spiked into serum samples and the recovery values ranged from 75.10% to 113.1%. Using the similar biosensor design, Tian et al. [4] utilized the electron mediator of tris(2,2′-bipyridine)ruthenium(III) (Ru(bpy)33+) to react with NADH, thereby indirectly detecting bile acids. It exhibited a linear behavior from 5.00 pmol/L to 150.00 pmol/L with a detection limit of 0.40 pmol/L (based on 3× the baseline noise) in a 106-fold dilution serum. The use of Ru(bpy)33+ greatly enhanced the electrical conductivity and the sensitivity of the device (Figure 5).

4.3. Aromatic Diamines

An enzyme carbon paste electrode (CPE) based on tyrosinase, salicylate hydroxylase (SH), and L-phenylalanine dehydrogenase (PADH) was established for L-phenylalanine detection [44]. With the catalysis of PADH, L-phenylalanine acids reacted specifically with NAD+ to produce NADH as shown in the following equation:

Then, the irreversible decarboxylation and the hydroxylation of salicylate catalyzed by SH were described by the following equation:

The oxidation of catechol to o-quinone by tyrosinase could be described by equation (12). Also, the reaction current was measured by the chronoamperometry assay at Eappl = −50 mV vs. Ag/AgCl.

This relies on the circulation of catechol and o-quinone between the electrode and the tyrosinase, thereby amplifying the electrocatalytic signal. The reaction time for L-phenylalanine was <60 s with a linearity range from 20.00 μM to 150.00 μM. The biosensor also lost >50% of its original response to galactose within 11 days.

In addition, Manna and Retna Raj [45] described an electrochemical device for the detection of lactic acid in human serum employing rGO-PhNO2 composites covalently functionalized with a p-nitrophenyl moiety of p-nitroaniline. The rGO-PhNHOH was formed by the potential cycling of the rGO-PhNO2-modified GC electrode. The current signal was detected by the oxidation of NADH mediated by a redox couple (-NHOH/-NO) covalently functionalized with rGO. The sensor can determinate lactate as low as 2.5 μM without interference, with a sensitivity of 10.57 ± 0.38 nA·μM−1·cm−2, a linear range up to 90.00 μM, and a response time of 23 s.

4.4. Organic Dyes

Martínez-García et al. [46] established an enzymatic electrochemical device for the detection of 3-HB in serum by fixing HBDH onto a rGO and thionine (THI)-modified SPCE. The THI’s redox mediator combined with an rGO-modified electrode could reduce the detection potential to 0.0 V. The optimized device showed a linear response range of 0.010–0.40 mM, a limit of detection of 1.00 μM, and a response time of 7 s.

Furthermore, the TG detection device was established by fixing lipase (LIP) and GDH on the indium-tin-oxide (ITO) glass electrode coated with an electrochemically reduced GO (ERGO) membrane doped by toluidine blue (TB) [47]. Glycerol was oxidized in the presence of GDH, and TB then extracted the H+ ions released from the glycerol, leaving the hydrogen to ERGO for performing the subsequent redox reaction. At a potential of +0.34 V, ERGO(Red) was reoxidized to ERGO(Oxid), resulting in an electrochemical response. With a rapid response within 12 seconds, the device is linear in the range of 50.00 mg/dl to 400.00 mg/dl and has a sensitivity of 29 Pa·mg−1·dl.

5. Challenges and Critical Issues toward Clinical Applications

Although considerable resources have been invested in building electrochemical biosensing devices based on dehydrogenase over the past two decades, there has been little apparent success in the productization of these devices for use in human physiological fluids such as serum and urine. The critical problem of such devices is the effective exchange of electrons between the active center and the supporting composite or electrode. Direct electron transfer (DET) through the bare electrode is extremely hard because the redox-active sites of NAD+/NADH-dependent dehydrogenases are deeply embedded in the well-electrically insulated shells [13]. The most common is that distances greater than 20a prevent DET between the active site and the electrode. This problem is often addressed through the use of mediators, which can reach the active site of the enzyme and act as electron shuttles by mediating the exchange of electrons between the enzyme and the electrode. Due to its excellent electrical conductivity and high sensitivity, MET is proven to be promising in the construction of various biosensors and biofuel cells. It is also important to address the issues of mediator-based biosensors, such as leakage of mediators, the high cost of complexes, or the harmfulness of shuttling molecules must be considered in future studies.

Other challenges associated with electrochemical dehydrogenase biosensors are that NADH biosensing on bare electrodes typically leads to the accumulation of the oxidation products on the electrode surface which causes high overpotential and unstable detection signals [48]. The surface characteristics of the sensor and its selectivity to targets must be improved, otherwise many components in biological fluids such as blood and urine samples could quickly contaminate the electrode. Furthermore, the bulk of the research effort has been on small initial sample sizes, portable devices, greater sensitivity or selectivity, and wide linear ranges with low detection limits. In contrast, the number of publications dealing with thermal stability, storage stability, shelf life, and the reuse of electrochemical dehydrogenase biosensors is very small. However, the key issues that must be taken into consideration when developing a viable commercial product are its storage conditions and long-term stability (shelf life). Thus, how to use nanomaterials and mediators to develop electrochemical dehydrogenase detection devices with higher sensitivity, stability, and anti-interference ability in a complex biological environment remains a great challenge.

6. Conclusion and Future Prospects

We provided an extensive overview of the recent technological progress made toward NAD+-dependent electrochemical dehydrogenase biosensors operating in human physiological fluids. It was observed that biosensors functionalized with various nanocomposites (carbon nanotubes, graphene, metal oxide, and related) can promote the adsorption or fixation of probe biomolecules with sufficient active sites, enhanced conductivity, and higher carrier density. The immobilization procedure of redox-active enzyme on the electrode has a great influence on the indexes of the detection device. It is therefore a key issue, especially if large-scale applications are envisaged, to ensure that the attenuation of the catalytic activity of the immobilized biocatalyst is as reduced as possible. Another direction being explored is the use of mediators such as metal and metal complexes, aromatic diamines, and organic dyes as the electron shuttles between the enzyme and the transducer. Biosensors, which use electron mediators, often show high current densities and a good sensitivity. However, they are usually less stable due to the immobilization of the medium and the mediators. Thus, there still are challenges yet to be overcome to achieve better sensitivity, selectivity, stability, and lifetime, before NAD+-dependent electrochemical dehydrogenase can be used in practical applications.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Xinrui Jin, Jinglan Guo, Baolin Li, and Jinbo Liu conceptualised and designed the study. Zixin Zhu, Yusen Liu, and Min Zhong constructed figures and tables. Xinrui Jin, Jingling Xie, and Jinkang Feng drafted the manuscript. All authors contributed to writing the final drafts of the manuscript.

Acknowledgments

The work was supported by the grants from Southwest Medical University (Grant nos. 2021ZKQN063, 2020500, and S202210632152) and the Science and Technology Department of Sichuan Province (Grant nos. 2021YFS0171and 2021YFH001).