Application of Voltammetric Sensors for Pathogen Bacteria Detection: A Review

Abstract

In recent years, new strategies for bacteria determination have been developed in order to achieve rapid detection and adequate limits of detection for quantification of microorganisms. This review classifies voltammetric sensors according to whether the bacteria are directly or indirectly detected. Direct methods are based on the recognition of the bacteria themselves, either in labeled or label-free mode. In contrast, indirect methods detect a metabolite produced by the bacteria. New trends in bacteria sensors involve DNA analysis, which makes it possible to improve the sensitivity and specificity of measurements. Voltammetric sensors provide good linear ranges and low limits of detection and are useful for analysis of food and clinical and environmental samples.

1. Introduction

Biological pathogens, including bacteria, viruses, fungi and parasites, are ubiquitously distributed in nature, as a result of their natural life cycles, as well as in artificial environments, mainly due to human activities. Frequently, they entail some kind of risk to human and/or animal health, causing diverse diseases (such as infectious diseases and allergies) as a result of direct exposure to the pathogen or through contact with associated allergens or toxins [1].

Exposure to biological pathogens is particularly significant in several professional areas; for example, (1) agriculture and livestock production, as well as related activities, specifically raw food processing; (2) human and veterinary healthcare services; (3) analytical laboratories; and (4) wastewater treatment plants [2,3,4]. The risks associated with exposure must be assessed and controlled; hence, recurrent analyses of diverse types of samples (water, food, soil, air and biological specimens) from different environments have to be performed.

Considering this threat to human health, the demand for simple, rapid and low-cost instruments for accurate analysis is growing, especially for in situ determination. Identification and quantification of biological pathogens in environmental, food and biological samples are carried out with diverse analytical techniques. The most common approaches are based on culturing methods (in agar plate cultures or other liquid cultivation media), enzyme-linked immunosorbent assays (ELISA) and polymerase chain reaction (PCR) technology [5,6]. Culturing methods, as well as nucleic acid amplification techniques, are time-consuming and, therefore, less suitable for rapid detection. ELISA assays also require culture enrichment and, furthermore, may be affected by cross-reactivity effects. These methods are known for their high specificity, but obtaining the results can take several hours or even days and they involve complicated procedures and high costs [7].

Other analytical techniques enable faster and more accurate pathogen identification from culture-positive specimens; namely, matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry and fluorescence in situ hybridization [8,9,10]. Nonetheless, the highly sophisticated and costly equipment and the need for trained personnel impair their extensive implementation in laboratories. Furthermore, most techniques need complex sample pretreatments before the analysis in order to isolate and concentrate bacteria.

The use of microfluidic devices for identification and quantification of pathogens has been described [11,12,13]. Microfluidic device-based methods make it possible to simplify analytical procedures and automatically couple pre-treatment procedures with various detection techniques [14]. However, in-line sample pre-treatments are still problematic, particularly for complex samples. An interesting alternative is the use of sensors, including optical and electrochemical sensors. In general, these can be considered suitable alternatives to conventional pathogen detection methods since they enable real-time analysis without the need for culture enrichment or DNA amplification and can be applied to different types of samples [12].

Diverse electrochemical sensors have been developed for the rapid and accurate identification and quantification of biological pathogens in various samples using different types of detection, including impedance, amperometry and voltammetry [15]. Their low cost, portability, fast response and adequate sensitivity are among their main advantages [16], and their limited selectivity, particularly in the analysis of samples with complex matrices, can be overcome by using specific biorecognition elements, such as antibodies, nucleic acids or peptides [17].

Voltammetry is an interesting option based on a redox reaction directly or indirectly related to the analyte (pathogen). In direct analysis, pathogens bind to the modified electrode surface, which causes a signal change correlated with the pathogen concentration [18]. Indirect detection consists of the measurement of an electroactive molecule hydrolyzed by an endogenous enzyme of the pathogen [19] or the direct measurement of an electroactive molecule produced by the pathogen and secreted to the surrounding medium [20,21].

Cyclic voltammetry (CV), differential pulse voltammetry (DPV) and square wave voltammetry (SWV) are the most commonly used voltammetry techniques for pathogen determination in diverse samples. This review describes some representative examples of voltammetric sensors for direct/indirect determination of pathogens and presents their advantages and disadvantages for real sample analysis.

2. Voltammetric Microbial Detection

2.1. Sensors for Bacterial Detection

Different methodologies involving voltammetric sensors and biosensors are used for pathogen analysis. Some methodologies include the direct use of the pathogen for qualitative and quantitative detection. In this case, voltammetric sensors can be classified in two types. In the first type, detection is performed through direct interaction between the electrode and the bacteria, without the need for biological recognition agents. These sensors tend to be simpler and more stable and robust compared to sensors employing biorecognition elements, which facilitates their implementation in the analysis of pathogens [22]. Figure 1 shows a general scheme for the construction of voltammetric sensors and biosensors that require the presence of the bacteria directly in the detection, where the crucial step is the modification of the electrode. Different materials have been used for the construction of the electrode, including nanoparticles, carbon-based materials, molecularly imprinted polymers, antibodies and bacteriophages.

In this context, one of the first sensors developed with voltammetric detection for bacteria consisted of the use of a glassy carbon electrode modified with tetracycline.

This sensor did not present specificity towards a single bacterium since it allowed the detection of Escherichia coli, Salmonella, Bacillus subtilis, Staphylococcus aureus and Lactobacillus lactis. In this case, the tetracycline acts as a promoter in the transfer of electrons between the bacteria and the surface of the glassy carbon electrode, resulting in a change in the current; this is due to the fact that it can bind to small subunits of the liposomes of the bacteria membrane. The change in the signal current is attributed to the electrooxidation of electronic mediators in the electron transport chains, such as NADH, FADH, cytochrome C, b, aa3 (cytochrome oxidase) and NADPH, resulting in anodic peak potentials in the interval from 0.837 to 0.690 V. Due to the similarity of the potentials, it is assumed that the sensor acts in a similar way on the bacteria, using the same molecules for the voltammetric response. The generated current is then associated with the concentration of bacteria present in the sample. The sensor presents limits of detection (LODs) of 2.0 × 10⁴ to 5.0 × 10⁴ CFU mL⁻¹ for the analysis of the mentioned microorganisms in food samples [23].

Al-Fandi et al. described the determination of E. coli O157:H7 using a ZnO nanoparticle sensor on cellulose paper. Bacterial detection was performed at an approximate potential of 0.6 V, and the mechanism of detection produced an increase in the electrochemical current that was observed in the presence of the cell membrane and cytoplasm originating from the pathogen. The bacteria spread over the electrode surface, enhancing the conductivity and, thus, increasing the electrochemical current, reaching an LOD of 1.0 × 10³ CFU mL⁻¹. The sensor performance was considered adequate since it could be successfully used for the determination of E. coli O157:H7 in biological samples [24].

Panhwar et al. proposed another method for the determination of E. coli in water samples based on the use of Fe₃O₄ nanoparticles functionalized with L-cysteine. The mechanism of the interaction between the bacteria and the sensor relied on the fact that the structure of the bacterial cell membrane promotes different interactions with nanoparticles, and L-cysteine helped in the adsorption of Fe₃O₄ nanoparticles on the surface of the cellular membrane of bacteria. The peptidoglycan layer that lies between the outer and inner membranes of bacteria allowed L-cysteine to enter the cell wall, promoting a physical interaction with the nanoparticles since the positively charged nanoparticles could interact with the negatively charged cell membrane, causing the membrane to break. An anodic peak of 0.4 V was obtained and a linear increase in current correlated with the pathogen concentration. This sensor was proved to be highly selective in the concentration interval from 1.0 × 10¹ to 1.0 × 10⁵ CFU mL⁻¹, with an LOD of 10.0 CFU mL⁻¹ [25]. Graphene-supported mesoporous ZrO₂ and In₂O₃ sensors have also been proposed to detect E. coli O157:H7 in the interval from 1.0 × 10¹ to 1.0 × 10¹⁰ CFU mL⁻¹. In contrast, Fatema et al. proposed the binding of E. coli to the surface of the electrode [18]. This modification produced a decrease in the current generated by the presence of an electroactive species in the electrolyte solution. The current linearly diminished with the E. coli concentration because the E. coli blocked the electron transfer from the electrode, thus increasing the resistivity of the sensor. With this methodology, an LOD of 10.0 CFU mL⁻¹ was obtained [18].

Another type of voltammetric sensor is based on a mechanism of detection in which bacteria are introduced into the cavities of a tailored agent present at the surface of a modified electrode, causing a decrease in the peak current of the electrode. Sensors of this type enable specific recognition of bacteria and are based on the use of molecularly imprinted polymers (MIPs). MIPs allow the detection and quantification of different bacterial pathogens in a selective way, with suitable LODs in different matrices [26,27]. MIPs are materials prepared through the polymerization of monomers, crosslinkers and other components in the presence of a template. For the analysis of bacteria, the template employed is the target bacterium. After the formation of the MIP, the template is removed, and free cavities are created that retain the same shape, size and chemical functionality toward the target bacterium [28,29,30]. Finally, the MIP forms an artificial receptor that, when specifically bound to the bacteria, produces a change in the current that enables detection [28]. Direct bacterial imprinting and, thus, the generation of recognition and binding sites in polymeric materials have facilitated the development of new bacterial detection methodologies that are feasible and low-cost options when the size and morphology of the cavities formed are critical factors for the detection [26].

For example, Khalid et al. described the construction of a sensor based on MIPs, using polydopamine to modify the surface of a gold screen-printed electrode for the detection of Salmonella Typhimurium in a linear interval from 1.0 × 10² to 1.0 × 10⁹ CFU mL⁻¹ in meat samples, with an LOD of 47.0 CFU mL⁻¹ [31]. A MIP-based sensor has also been developed for the detection of Klebsiella pneumoniae, for which the electrode was modified with polypyrrole. The developed sensor was intended to be applied for the analysis of urine samples; it has a linear interval from 1.0 × 10⁻¹ to 1.0 × 10⁵ CFU mL⁻¹ and an LOD ≈ 2.0 CFU mL⁻¹ [32]. Another sensor for K. pneumoniae was developed using methacrylamide, acrylamide and N-vinylpyrrolidone with graphene, achieving an LOD of ≈2.0 CFU mL⁻¹ and a linear interval from 1.0 × 10¹ to 1.0 × 10⁵ CFU mL⁻¹ [33].

As mentioned above, sensors that do not incorporate recognition agents usually have low specificity. Therefore, biochemical and microbiological pretreatments have been coupled to obtain more specific electrochemical sensors [34]. These additional steps allow the sensor to detect the target pathogen or microorganism with increased sensitivity and specificity, although such additional steps require more time, adequate conditions for storage and proper equipment [22]. This type of sensor is described below.

2.2. Biosensor for Direct Bacteria Detection

Sensors that incorporate a biological recognition agent to detect bacteria are called biosensors. Biosensors with voltammetric detection are the most widely used in the development of methodologies for direct bacteria detection because they have high specificity [35]. The most frequently used biological recognition agents for biosensor construction are antibodies [22]. Since antibodies are compounds generated as immune responses to bacterial pathogens, these molecules have high specificity and are usually used for the modification of gold electrodes. Gold has been widely used in the development of biosensors because of its simple modification and high conductivity, although carbon-based electrodes have also been used [36,37]. In biosensors for bacteria detection, the mechanism of interaction of biosensors for bacteria detection consists of employing a redox probe. The most widely used redox probe is the pair K₃[Fe(CN)₆]/K₄[Fe(CN)₆], but other redox probes can be used, such as electroactive iron species. There are two detection strategies. The first consists of the capture of the bacteria on the electrode surface through interactions with a specific immobilized antibody, which leads to a decrease in the electron transfer rate on the surface of the biosensor with the redox probe and is considered label-free detection. The second approach consists of labeling the bacteria: in this case, the label is an electroactive molecule capable of generating a response by itself; thus, the concentration of the label detected is directly proportional to the number of bacteria present in the electrode surface [38,39].

E. coli detection employing a label-free biosensor and based on a carbon screen-printed electrode modified with gold nanoparticles and anti-E. coli antibody was described by Vu et al. This biosensor presented an LOD of 15.0 CFU mL⁻¹ and an interval of quantification from 1.0×10¹ to 1.0×10⁶ CFU mL⁻¹. The bacteria–biosensor interaction produced a decrease in the current related to the E. coli concentration [40]. S. aureus and E. coli were detected by Svalova et al. [38] using a biosensor based on an electrodeposited poly-3-ethynylthiophene and bovine serum albumin film. This biosensor showed LODs of 15.9 CFU mL⁻¹ for S. aureus and 7.2 CFU mL⁻¹ for E. coli, with an interval of applicability from 1.0 × 10² to 1.0 × 10⁶ CFU mL⁻¹

Recently, other recognition elements, such as bacteriophages, have been applied due to their specificity, low cost and high stability. The target bacteria can be detected and quantified with bacteriophage-based biosensors because bacteriophages specifically bind to the bacteria, since they are a type of virion that infect bacteria [41]. Bacteriophages can strongly adsorb to lipopolysaccharides contained in the cell wall of viable bacterial pathogens [42]. Using this recognition agent, a biosensor based on the immobilization of T4 bacteriophages was developed for the quantification and voltammetric detection of E. coli B. The binding bacteriophage E. coli B promotes an increase in current with respect to the E. coli concentration. An LOD of 1.4 × 10³ CFU mL⁻¹ over a wide linear interval from 1.9 × 10¹ to 1.9 × 10⁸ CFU mL⁻¹ was described [43].

However, indirect detection of bacteria using biosensors has been preferred because of its better sensitivity and specificity compared to direct testing. Labeled antibodies are generally used for electrochemical detection. Horseradish peroxidase (HRP)-labeled antibody is the most widely used indirect detection agent. The labeled antibody binds to the surface of bacteria attached to the electrode and, subsequently, a horseradish peroxidase-catalyzed reduction occurs upon the addition of hydrogen peroxide and a compound, such as thionine, is oxidized, allowing electron transfer and detection [39,44].

Gram-negative bacteria have been detected by biosensors based on HRP-labeled antibodies. A biosensor for the determination of S. gallinarum was constructed using screen-printed carbon electrodes modified with 1-butyl-3-methylimidazolium hexafluorophosphate and anti-S. gallinarum antibody. The developed biosensor presented a linear interval from 1.0 × 10⁴ to 1.0 × 10⁹ CFU mL⁻¹ and an LOD of 1.0 × 10³ CFU mL⁻¹ [44]. The determination of S. Typhimurium was carried out using a glassy carbon electrode modified with chitosan and anti-S. Typhimurium antibody. The pathogen could be detected in milk and egg samples in a concentration interval from 5.6 × 10¹ to 5.6 × 10⁸ CFU mL⁻¹ and with an LOD of 35.0 CFU mL⁻¹ [39]. A biosensor was also constructed using graphene sulfonate modified with poly-(3,4-ethylenedioxythiophene)-gold nanoparticles and anti-E. coli antibody for the detection of E. coli in water and milk, presenting an LOD of 34.0 CFU mL⁻¹ and a linear interval from 7.8 × 10⁴ to 8 × 10⁶ CFU mL⁻¹ [45].

Moreover, other pathogens, such as Bacillus cereus and E. coli O157:H7, have been detected through immunomagnetic separation using magnetic nanoparticles modified with polyaniline and with the respective anti-bacteria antibodies. The extraction was carried out through direct adsorption of the target bacteria to the sample solution, which was placed on a screen-printed carbon electrode sensor for direct detection. The binding of the bacteria with the sensor decreased the electrochemical signal in direct relation to the increase in the concentration of the bacteria. The sensors presented linear intervals of 4.0 × 10⁰ to 3.9 × 10² CFU mL⁻¹ (B. cereus) and 6.0 × 10⁰ to 5.9 × 10⁴ CFU mL⁻¹ (E. coli), obtaining LODs of 40.0 CFU mL⁻¹ (B. cereus) and 6.0 CFU mL⁻¹ (E. coli) in pure culture [46].

Table 1. Voltammetric sensors and biosensors for the detection of pathogenic bacteria.

Bacteria	Electrode	Detection	Lineal Interval (CFU mL−1)	Sample	LOD (CFU mL−1)	REF
E. coli,	Glassy carbon/tetracycline	CV	2.0 × 104–5.0 × 107	Food	2.0 × 104	[23]
B. subtilis,			2.0 × 104–5.0 × 107		2.0 × 104
S. aureus,			3.0 × 104–9.0 × 107		3.0 × 1047
Salmonella and			3.0 × 104–1.0 × 108		3.0 × 104
L. lactis			5.0 × 104–1.6 × 108		5.0 × 104
E. coli O157:H7	ZnO nanorods	CV	−−−	−−−	1.0 × 103	[24]
E. coli	L-cysteine	CV	1.0 × 101–1.0 x 1010	Tap water	10.0	[25]
E. coli O157:H7	ZrO2 and In2O3/graphene	CV	1.0 × 101–1.0 × 105	−−−	10.0	[18]
S. aureus	Mercury/carbon nanotubes	CV	−−−	Human blood and serum	−−−	[47]
Salmonella	Poly(dopamine)	CV	1.0 × 102–1.0 × 109	Meat	47.0	[31]
K. pneumoniae	Pyrrol	DPV	1.0 × 100–1.0 × 105	Urine	≈2.0	[22]
K. pneumoniae	Methacrylamide/acrylamide/N-vinylpyrrolidone/graphene	CV	1.0 × 101–1.0 × 105	−−−	≈2.0	[33]
E. coli O157	Screen-printed carbon/gold nanoparticles/anti-E. coli O157	CV	1.0 × 101–1.0 × 106	−−−	15.0	[40]
S. aureus and	3-ethynylthiophene/bovine serum albumin/anti-E. coli or anti-S. aureus	CV	1.0 × 102–1.0 × 106	−−−	15.9	[38]
E. coli					7.2
B. cereus and	Screen-printed carbon/magnetic nanoparticles/polyaniline anti-B. cereus or anti-E. coli	CV	4.0 × 100–3.9 × 102 6.0 × 100–5.9 × 104	−−−	40.0	[46]
E. coli O157:H7					6.0
E. coli B	T4 bacteriophage	DPV	1.9 × 102–1.9 × 109	−−−	14.0	[43]
Shigella flexneri	Multiwalled carbon nanotubes/sodium alginate/anti-S. flexneri HRP	CV	1.0 × 104–1.0 × 1011	−−−	3.1 × 103	[48]
E. coli O157:H7	Sulfonated graphene/poly-(3,4-ethylenedioxythiophene)/gold nanoparticles/anti-E. coli HRP	DPV	7.8 × 101–7.8 × 106	Water and milk	34.0	[45]
S. gallinarum	Anti-S. gallinarum gold nanoparticles/anti-S. gallinarum HRP	CV	1.0 × 104–1.0 × 109	−−−	3.0 × 103	[44]
E. coli O157:H7	Pencil graphite/chitosan/multi-walled carbon nanotubes/gold nanoparticles/polypyrrole/anti-E. coli	CV	3.0 × 101–3.0 × 107	−−−	30.0	[49]
S. Typhimurium	Glassy carbon chitosan/S. Typhimurium/anti-S. Typhimurium HRP	CV	5.6 × 101–5.6 × 108	Milk and eggs	35.0	[39]
S. Typhimurium	Chitosan/gold nanoparticles/S. Typhimurium/anti-S. Typhimurium HRP	CV	1.0 × 101–1.0 × 105	−−−	5.0	[50]
S. aureus	Single-walled carbon nanotubes/anti-S. aureus	DPV	1.0 × 101–1.0 × 107	−−−	13.0	[51]
E. coli O157:H7	Methylene blue nanocomposites/magainin I	DPV	1.0 × 102–1.0 × 107	−−−	32.0	[52]
Salmonella	Glassy carbon/grapheme MOFs/CoFe/gold nanoparticles/anti-Salmonella	CV	2.4 × 102–2.4 × 108	Milk	120.0	[53]

CV: cyclic voltammetry, DPV: differential pulse voltammetry, MOFs: metal–organic frameworks.

compares different sensors and biosensors using voltammetric detection that have been applied for the determination of different bacterial pathogens. CV and DPV are the most commonly used detection techniques and E. coli is the most frequently studied pathogen, and LODs up to 2.0 CFU mL⁻¹ have been obtained. Moreover, for the detection of other bacteria, such as K. pneumoniae, LODs close to 2.0 CFU mL⁻¹ have been obtained, which is competitive with conventional analysis methods. Although the sensors that have been developed recently have adequate sensitivities and stability and allow the detection of multiple pathogens simultaneously, they have the disadvantage of a lack of specificity, which makes the sensors prone to false positives [22,34]. For this reason, the development of biosensors has utilized indirect methods because they allow higher specificity and sensitivity; however, antibodies require careful storage and handling because they can denature at unsuitable temperature and pH [22]. Hence, other biological recognition agents that allow high specificity towards the target bacteria have been sought.

Most of the methodologies require viable bacterial cells. Therefore, the use of microbiological treatment methods is necessary, making the methodologies laborious, expensive and time-consuming. Furthermore, the increase in the number of assay steps can lead to sample contamination [34]. Therefore, voltammetric sensors and biosensors that employ indirect detection and do not require direct interaction with bacterial pathogens have been recently described [15].

2.3. Sensors for Indirect Methods

Although recent approaches in voltammetric sensors for bacteria detection are based on the whole-cell recognition (direct detection), one of their main disadvantages is the use of bacterial cultures. These approaches require specific working conditions due to bacteriological hazards, especially when MIPs are fabricated [54]. Bacterial growth involves the use of specific media, humidity, pH and oxygen conditions and, furthermore, these techniques require specific laboratory equipment and trained personnel [21]. In this context, an alternative indirect technique to quantitatively determine the presence of bacteria has been proposed in recent years. This new strategy involves the use of bacteria-related compounds for detection. These compounds have an electroactive nature and are produced in the presence of bacteria or present in the bacteria cellular structure. The selection of related compounds for the determination of bacteria depends mainly on their cellular structure and the generated immune response (Figure 2).

In recent years, the development of sensors for the detection of toxic lipopolysaccharides (LPSs) has gained attention, and there are a great number of examples described in the literature. LPSs, also known as endotoxins, are present in the outer membrane of Gram-negative bacteria (GNBs) and, thus, are considered the optimal biomarker for bacteria detection. The importance of the determination of toxic LPSs lies in the fact that, although the bacteria die, the bacterial endotoxins are neither eliminated nor destroyed and can cause infections and fatal diseases. GNBs are one of the most widespread threats to human health and include Pseudomonas aeruginosa, Salmonella enterica and E. coli species. Different approaches have been developed to determine LPSs, including voltammetric methods that use electrodes modified with a recognition molecule to which LPSs bind. Proteins and poly-e-lysine have been used as recognition molecules, and enzymes have been employed to improve sensitivity [55]. The LPS content can be electrochemically estimated from the current generated in the oxidation or reduction of molecules, such as ferricyanide/ferrocyanide in the presence or absence of LPSs, since the current is dependent on the amount of LPSs. Another approach consists of the capture of LPSs on the surface of a modified electrode, followed by the attachment of an electroactive species to LPSs that provides an amplified response with Fe²⁺ ions. MIPs [54] and cell-based biosensors using mammalian cells [56] have also been proposed and are shown in Table 2. The indirect determination of LPSs has also been described and consists of the determination of substances that are generated by the cell as part of an immune response against the presence of an endotoxin. Nitric oxide (NO) is a compound often used as marker of immune response when a macrophage cell detects the presence of LPSs. These organisms are capable of expressing NO synthase and producing NO. Jiang et al. described the elaboration of a miniaturized electrochemical cell sensor for in situ detection of NO released from LPS-treated macrophage cells. The LOD of this approach was 3.5 × 10⁻³ ng mL⁻¹ and the sensor was applied for the determination of Salmonella in fruit juice samples [56].

Another molecule that can be associated with the presence of LPSs is hydrogen peroxide (H₂O₂). H₂O₂ is considered an important component of immune response, since this molecule is responsible for regulating how a cell responds to bacterial threat. Hicks et al. developed a sensor capable of detecting the H₂O₂ produced intracellularly in response to LPS structures extracted from two different serotypes of bacteria (0111:B4 and Re495) [57].

A different alternative for indirect detection of bacteria consists of evaluating the presence of virulence factors. A virulence factor is a molecule generated by a pathogen responsible for the potential of the microorganism to produce infection. The development of methodologies for the voltammetric determination of bacteria using virulence factors has been focused on the analysis of Pseudomonas aeruginosa. The pathogen P. aeruginosa is an opportunistic pathogen commonly found in water matrices, categorized as a top-priority infectious agent since it is known to present advanced antibiotic resistance [58]. Conventional methods for P. aeruginosa determination are time-consuming and require expensive instrumentation [20]. Pyocianin (PYO) is an electro-active molecule that is known as the main virulence factor of P. aeruginosa. Significant amounts of PYO are produced during bacteria growth, making this molecule a marker of bacterial infection in biological samples. The pyocyanin molecule oxidizes at a negative potential (approximately −250 mV vs. Ag/AgCl) and interferes in the host cell functions owing to its redox activity-promoting virulence [59]. Due to its electro-active nature, pyocyanin can be directly detected by using electrochemical methods and without complex sensor modification. Most of the methodologies described for the determination of P. aeruginosa refer to the concentration of pyocyanin. Ciui et al. developed an in situ methodology for determination of pyocyanin on contaminated surfaces. This device consists of a wearable platform; the authors proposed the use of gloves with an integrated voltammetric sensor in an attempt to satisfy health-care applications. This device facilitates routine monitoring of medical equipment and furniture in hospitals [60]. Other metabolites of P. aeruginosa that may be interesting for bacterial detection include pyoverdine (PYOV), Pseudomonas quinolone signal (PQS), pyochelin (PCH), 2-heptyl-4-hydroxyquinoline N-oxide (HQNO) and 2-heptyl-4-hydroxyquinoline (HHQ) [58].

When it is necessary to evaluate the presence of GNBs, another alternative is to determine bacterial quorum signaling molecules. For example, N-acyl-homoserine-lactones (AHLs) are molecules that play an important role in the regulation of bacterial virulence and are considered some of the principal signal molecules for GNBs [61]. AHLs interact with eukaryotic cells, reducing immune response and, consequently, facilitating infection [62].

Another group of bacteria that have been studied by voltammetry are mycobacteria. Related compounds described in literature for indirect detection of mycobacteria include mycolic acids (MAs). MAs are α-alkyl β-hydroxylated fatty acids that are considered some of the main components present in the lipid-rich cell walls of bacteria. Some of the species that present this composition are the genera Corynebacterium, Gordona, Tsukamurella and Mycobacterium, which include highly pathogenic microorganisms, such as M. tuberculosis and M. leprae. MA concentration can be used as an indicator of the number of CFUs of a microorganism. Most of the analytical methods for MA determination (MS, HPLC, GC) require skilled personnel and very expensive instrumentation. However, voltammetric methods offer a sensitive, simple, inexpensive and user-friendly way to determine MA. Brugnera et al. [63] described the construction of a poly-L-lysine-modified electrode for the determination of Mycobacterium smegmatis in drinking water. In this work, they observed a cathodic peak at −0.73 V (vs Ag/AgCl), probably due to reduction of the H-N-C=O group [63].

Nagar et al. proposed the determination of E. coli based on the reduction of resazurin [64]. This molecule is a blue non-fluorescent dye, which is reduced to resorufin (a red fluorescent dye) in response to chemical reduction resulting from cell growth. Nagar et al.’s work compared the reduction wave of resazurin in the absence and presence of bacterial cells using cyclic voltammetry tests with four different redox mediators: [Fe(CN)₆]^3−/4−, FcMeOH, Ru[(NH₃)₆]Cl₃ and K₂[IrCl₆]. These redox mediators can be used to detect reductive activity in living bacterial cells.

Table 2. Voltammetric methods for the indirect determination of pathogenic bacteria.

Sensor	Related Compound	Bacteria	Matrix	Detection	LOD (µM)	REF
LPS-MIP-SPCE	Lipopolysaccharides	P. aeruginosa and E. coli	Water	CV	−−−	[54]
Paper-based 3D Nafion-Ppy-GO-SPCE cell	Lipopolysaccharides	S. enterica serotype	Fruit juice	DPV	1.17 × 10−4 (NO)	[56]
Poly-e-lysine-nanocarbon film electrode	Lipopolysaccharides	E. coli	−−−	CV	2.0 ng mL−1	[65]
SPE	PQS, HHQ, PCH, HQNO, PYO	P. aeruginosa	−−−	CV and SWV	−−−	[58]
Paper-based graphene sensor	Pyocianin	P. aeruginosa	Water	SWV	0.3	[20]
SPE	Pyocianin	P. aeruginosa	Human serum, blood, saliva	SWV	0.1–25.0	[66]
Boron-doped diamond	Pyocianin	P. aeruginosa	Spiked sputum	DPV	0.05	[65]
AuSPE	Pyocianin	P. aeruginosa	Human saliva	CV	2.0	[59]
SPE	Pyocianin	P. aeruginosa	Biological fluids	SWV	0.1–1.8	[67]
PANI-AuNP-ITO	Pyocianin	P. aeruginosa	Corneal Ulcer	SWV	0.5	[68]
Screen-printed sensing glove	Pyocianin Pyoverdine	P. aeruginosa	Contaminated surfaces	SWV	3.3 × 10−3 1.6	[60]
Fe3O4@SiO2-MMIP	AHLs	P. aeruginosa, aeromonas strains 128 and 130 and Aeromonas hydrophila	Bacteria supernatant samples	DPV	8.0 × 10−4	[62]
Poly-L-lysine-GCE	Mycolic acids	Mycobacterium smegmatis	Water	SWV	59.0 CFU mL−1	[63]
Two-electrode multiplexer (Au SPE)	Resazurin	E. coli	Lysogeny broth	DPV	−−−	[64]

CV: cyclic voltammetry, DPV: differential pulse voltammetry, SWV: square wave voltammetry, LPS-MIP-SPCE: screen-printed carbon electrode modified with lypopolysaccharide molecularly imprinted polymer, SPE: screen-printed electrode, PANI-AuNP-ITO: polyaniline golden nanoparticle indium tin oxide electrode, Fe₃O₄@SiO₂-MMIP: magnetic molecularly imprinted polymers on magnetic silica surface, GCE: glassy carbon electrode.

Table 2. Voltammetric methods for the indirect determination of pathogenic bacteria.

Sensor	Related Compound	Bacteria	Matrix	Detection	LOD (µM)	REF
LPS-MIP-SPCE	Lipopolysaccharides	P. aeruginosa and E. coli	Water	CV	−−−	[54]
Paper-based 3D Nafion-Ppy-GO-SPCE cell	Lipopolysaccharides	S. enterica serotype	Fruit juice	DPV	1.17 × 10−4 (NO)	[56]
Poly-e-lysine-nanocarbon film electrode	Lipopolysaccharides	E. coli	−−−	CV	2.0 ng mL−1	[65]
SPE	PQS, HHQ, PCH, HQNO, PYO	P. aeruginosa	−−−	CV and SWV	−−−	[58]
Paper-based graphene sensor	Pyocianin	P. aeruginosa	Water	SWV	0.3	[20]
SPE	Pyocianin	P. aeruginosa	Human serum, blood, saliva	SWV	0.1–25.0	[66]
Boron-doped diamond	Pyocianin	P. aeruginosa	Spiked sputum	DPV	0.05	[65]
AuSPE	Pyocianin	P. aeruginosa	Human saliva	CV	2.0	[59]
SPE	Pyocianin	P. aeruginosa	Biological fluids	SWV	0.1–1.8	[67]
PANI-AuNP-ITO	Pyocianin	P. aeruginosa	Corneal Ulcer	SWV	0.5	[68]
Screen-printed sensing glove	Pyocianin Pyoverdine	P. aeruginosa	Contaminated surfaces	SWV	3.3 × 10−3 1.6	[60]
Fe3O4@SiO2-MMIP	AHLs	P. aeruginosa, aeromonas strains 128 and 130 and Aeromonas hydrophila	Bacteria supernatant samples	DPV	8.0 × 10−4	[62]
Poly-L-lysine-GCE	Mycolic acids	Mycobacterium smegmatis	Water	SWV	59.0 CFU mL−1	[63]
Two-electrode multiplexer (Au SPE)	Resazurin	E. coli	Lysogeny broth	DPV	−−−	[64]

CV: cyclic voltammetry, DPV: differential pulse voltammetry, SWV: square wave voltammetry, LPS-MIP-SPCE: screen-printed carbon electrode modified with lypopolysaccharide molecularly imprinted polymer, SPE: screen-printed electrode, PANI-AuNP-ITO: polyaniline golden nanoparticle indium tin oxide electrode, Fe₃O₄@SiO₂-MMIP: magnetic molecularly imprinted polymers on magnetic silica surface, GCE: glassy carbon electrode.

2.4. DNA-Based Sensors

Electrochemical DNA detection has been undertaken with direct and indirect methods. Methods that use measures such as capacitance, conductivity or impedance are considered direct methods because they measure the electroactivity of DNA itself or the changes in the interfacial properties of the DNA sensor. In contrast, indirect methods use electrochemical active DNA intercalators or labels (e.g., aminated specific sensing probes, mercaptoethanol as a blocking agent and toluidine blue as a label of the DNA probe) [69] or redox mediators or particles to amplify the analytical signal [13], such as polymethylene blue nanoparticles and dumbbell hybridization chain reactions as dual amplification [70]. DNA sensors are divided into genosensors and aptasensors (Figure 3).

Genosensors are based on DNA hybridization in which an oligonucleotide or a single-stranded DNA (ssDNA) probe complementary to the DNA sequence of interest is immobilized on the detection substrate to recognize the target DNA. DNA hybridization is detected through the redox reactions of the electroactive labels or from changes in the electrochemical properties that have influence over the analytical signal. The most critical steps are immobilization and the hybridization process [13,69].

Genosensors are focused on DNA detection and offer high specificity, sensitivity and stability, since the interactions produced between nucleotides (adenine nucleotide with thymine and cytosine with guanine) are specific, together with the potential for miniaturization. Hence, these types of sensors are attractive since they enable simpler, faster and cheaper analysis compared with traditional methods [13,69].

In this context, some strategies of immobilization include the following: (1) amino oligonucleotide modified with amide compounds in Salmonella detection [71]; (2) thiol oligonucleotide probes for the analysis of Salmonella, Lysteria monocytogenes, S. aureus, E. coli O157:H7 [72], Bacillus anthracis [73] and Pneumoniae bacteria [74]; (3) a 19-mer single0stranded DNA probe modified with 5′-NH₂ for the determination of Neisseria meningitidis [75]; (4) a DNA probe of a hepatitis C-supporting amino group for the analysis of Mycobacterium tuberculosis [76]; (5) gold nanoparticles supported through histidine for the detection of Brucella [77]; (6) adsorption of a specific complementary DNA for detection of Alicyclobacillus acidoterrestris [78]; and (7) a branched-chain DNA hybrid structure immobilized on magnetic microspheres for the determination of Vibrio parahaemolyticus and S. Typhimurium [79].

In the hybridization process, complementary sequences are captured by the DNA-immobilized probe at the sensor surface. Numerous methods have been proposed for the hybridization step; for example, biotin-streptavidin for the detection of Salmonella in sandwich hybridization [71], probe sequences linked to aminohexyl for the detection of Bacillus anthracis [73], 5′-biotinylated target solutions for the analysis of Pneumoniae bacteria [74] and ssG-DNA for the determination of N. meningitidis [75]. Some of the substances used for the labeling step include alpha-naphthyl phosphate, which shows an oxidation signal in a potential scan from 0.0 to +0.6 V related to the enzymatic product (alpha-naphthol) [71]; Meldola’s blue, with a reduction signal at −0.2 V; and guanine (showing an oxidation signal in a potential scan from +0.5 V to +1.4 V), used as a hybridization indicator between a thiol-linked probe and a complementary sequence [73]. Peak currents of the oxidation signals of methylene blue and ferrocene have also been employed—aimed at an electrode surface linked to a DNA target, thereby resulting in a large detection current signal—in potential scans from −0.9 V to +0.6 V and 0.0 to −0.5 V [79,80].

Aptasensors are based on aptamers, which are short nucleic acids (12 to 80 nucleotides long) of single-stranded DNA or RNA [81]. Aptamers are selected using an in vitro selection method named Systematic Evolution of Ligands by Exponential Enrichment (SELEX) by means of an oligonucleotide library [82,83,84,85]. The selection process is based on binding, separation or partitioning and amplification [81]. Aptamers provide advantages such as stability at different temperatures, pH and ionic environments; non-complex synthesis; elevated resistance against denaturation; and long half-life. Aptamers can generate double-strand DNA by binding to their complementary strand [82].

Depending on their immobilization on the sensor, aptasensors can be classified as covalent or non-covalent. Covalent immobilization has high efficiency in binding and hybridizing DNA molecules. Examples of covalent aptasensors include gold, carbon and glassy carbon electrodes, to which one of the ends of the nucleic acid chain is linked [84]. In contrast, in non-covalent immobilization, the aptamers are immobilized via physical adsorption, electrostatic attraction or supramolecular interaction. This type of immobilization has the drawback of low stability caused by desorption on the electrode surface [83,84]. In recent years, DNA walkers, classified as nanomachines, have been applied in bacteria detection. These nanomachines are based on the natural movement of kinesins in the cargoes transported along microtubules, allowing autonomous mechanical movement along nanoscale pathways in one-dimensional, two-dimensional and three-dimensional nanoparticles [85]. A DNA walker is activated by the recognition of the analyte by the aptamer, producing several indicators that improve the specificity and affinity towards the analyte through the formation of an aptamer-target complex. Hence, DNA walkers are used as an amplification method [86]. This amplification strategy has been applied in the analysis of E. coli O157:H7 [87], P. aeruginosa [88], V. parahaemolyticus [89] and S. aureus [86,90].

Nanomaterials, such as carbon nanotubes, graphene oxide, reduced graphene oxide, carbon nanowires, silica, polymeric, silver, platinum, iron oxide, molybdenum and special gold nanoparticles, are utilized in DNA sensors to increase specificity and sensitivity [91]. Gold nanoparticles have good biocompatibility, conductivity, chemical properties and electrocatalytic activity [82]. Furthermore, nano-porous gold (NPG) possesses high conductivity and signal transducing abilities, as well as enhanced catalytic properties [83].

Table 3. DNA-based voltammetric sensors for the detection of pathogenic bacteria.

Analyte	Sample	Method	Working Electrode	LOD	Linear Range	REF
Salmonella	Salmonella amplicons	DPV	SPCE	3.0 × 10−10 M of Salmonella DNA target sequence	1.0 × 10−8–1.5 × 10−7 M of Salmonella DNA target sequence	[71]
B. anthracis	−−−	DPV	Gold and graphite electrode	2.0 × 10−11 M of B. anthracis target sequence	6.6 × 10−6–3.3 × 10−4 M of B. anthracis target sequence	[73]
Helicobacter pylori	Human gastric tissues	CV SWV	BCNE	6.0 × 10−8 g mL−1 of H. pylori DNA	7.0 × 10−7–7.9 × 10−6 g mL−1 of H. pylori DNA	[92]
N. meningitidis	Cerebrospinal fluid	CV DPV	COO−MWCNT-SPCE	2.0 × 10−6 g of N. meningitidis ssG-DNA	2.5 × 10−6–1.0 × 10−5 g 6 μL−1 of N. meningitidis ssG-DNA	[75]
E. coli	Beef meat	CV DPV	MWCNT-Chi-GCE-Bi	2.0 × 10−14 M of E. coli tDNA	2.0×10−14–1.9×10−13 M of E. coli tDNA	[93]
Streptococcus pneumoniae Lyt-1 gene sequence	Clinical sample	CV ACV	DNA probe-modified gold disk electrode	~5.0 × 10−16 M of S. pneumoniae Lyt-1 gene sequence	1.0 × 10−14–1.0 × 10−10 M of S. pneumoniae Lyt-1 gene sequence	[80]
M. tuberculosis	Clinical sample	CV	Fc-acid-OMPA deposited on gold electrode	2.0 × 10−16 M of M. tuberculosis DNA	1.0 × 10−15–1.0 × 10−10 M of M. tuberculosis DNA	[76]
S. typhimurium	Raw chicken meat	CV DPV	Glassy carbon electrode	1.0 × 101 CFU mL−1	1.0 × 101–1 × 106 CFU mL−1	[94]
S. typhimurium	Chicken meat	DPV	ssDNA-rGO-AP-GCE	1.0 × 101 CFU mL−1	1.0 × 101–1 × 108 CFU mL−1	[95]
Brucella	Cultured and human samples	CV SWV	Gold electrode	−−−	1.0 × 10−16–1.0 ×10−7 M	[77]
Salmonella	−−−	DPV	PPy-rGO-GCE-AuNP-HRP-SA	4.7×10−17 M/ 8.1 CFU mL−1	1.0 × 10−16–1.0 × 10−10 M/ 9.6–9.6×104 CFU mL−1	[96]
E. coli 055:B5	Serum	CV SWV	RGO-AuNP-GCE	3.0 × 10−14 g mL−1 of lipopolysaccharides from E. coli	1.0 × 10−13–9.0 × 10−13 g mL−1 of lipopolysaccharides from E. coli	[97]
A. acidoterrestris	Orange juice	DPV	GE-ERGO-poly(3-HBA)	1.2 × 10−8 g mL−1 of A. acidoterrestris	1.2 × 10−8–1.2 × 10−4 g mL−1 of A. acidoterrestris	[78]
V. parahaemolyticus S. typhimurium	Shrimp and fish	SWV	SPCE	4.0 CFU mL−1 7.0 CFU mL−1	1.0 × 101–1.0 × 108 CFU mL−1	[79]
Helicobacter pylori	−−−	DPV	RNA cleaving DNAzyme-G-quadruplex DNAzyme gold electrode	3.4 × 10−17 M of H. pylori tDNA/1.3 × 10−12 g of H. pylori genomic DNA	1.0 × 10−16–1.0 ×10−11 M of H. pylori tDNA/2.1 × 10−12–6.7 × 10−11 g of H. pylori genomic DNA	[98]
E. coli O157:H7	Water	CV SWV	Au-GNS	1.0 × 10−23 M of E. coli target DNA	1.0 × 10−17–7.3 × 10−17 M of E. coli target DNA	[69]
S. aureus alpha-toxin	Human serum	SWV	Au-MWCNT-BMIM-PF6-CPE	1.0 × 10−9 M of alpha-toxin	3.0 × 10−9–2.5 × 10−7 M of alpha-toxin	[82]
S. aureus	Human serum, milk and pear juice	DPV	DNA-Au NC-CS-GCE	1.0 CFU mL−1	1.0 × 101–1.0 × 108 CFU mL−1	[70]
S. aureus	Raw milk, beer and apple juice	DPV	T-H-MB-MWCNT-CS-GCE	1.0 CFU mL−1	1.0 × 101–1.0 × 107 CFU mL−1	[86]
S. aureus	Sputum	DPV	Ferrocene-planar Au electrode	20.0 CFU mL−1	1.0 × 102–1.0 × 108 CFU mL−1	[90]

CV: cyclic voltammetry, SWV: square wave voltammetry, DPV: differential pulse voltammetry, ACV: alternating current voltammetry, BCNE: bismuth-immobilized carbon nanotube electrode, MWCNT-Chi-GCE-Bi: glassy carbon electrode modified with multi-walled carbon nanotubes and chitosan followed by a deposit of bismuth, Fc-acid-OMPA: redox oligo-methoxy-phenyl-acetonitrile, ferrocene groups and carboxylic acids, ssDNA-rGO-AP-GCE: glassy carbon electrode modified with reduced graphene oxide and single-stranded DNA, PPy-rGO-GCE-AuNP-HRP-SA: glassy carbon electrode modified with polypyrrole, reduced graphene oxide nanocomposite and signal amplification with horseradish peroxidase–streptavidin-biofunctionalized gold nanoparticles, RGO-AuNP-GCE: glassy carbon electrode modified with reduced graphene oxide and gold nanoparticles, GE-ERGO-poly(3-HBA): graphite electrode modified with electrochemically reduced graphene oxide and a polymer derived from 3-hydroxybenzoic acid, SPCE: screen-printed carbon electrode, Au/GNS: gold electrode functionalized with gold nanoparticles, Au-MWCNT-BMIM-PF6-CPE: carbon paste electrode modified with gold nanoparticles, multi-walled carbon nanotubes and 1-butyl-3-methylimidazolium hexafluorophosphate as a binder, DNA-Au NC-CS-GCE: glassy carbon electrode modified with DNA, gold nanocages and chitosan, T-H-MB-MWCNT-CS-GCE: methylene blue-tagged hairpin DNA adapted to glassy carbon electrode modified with multi-walled carbon nanotubes and chitosan.

shows a summary of applications involving DNA-based voltammetric sensors in different types of samples. For every study, the linear range and the LOD are reported in different units.

3. Conclusions

Voltammetry sensors offer a brilliant future in the development of new devices for the detection of microorganisms. The studies discussed in this review demonstrate a starting point for the development of selective sensors and biosensors able to detect the presence of pathogenic bacteria in aqueous media, biological fluids and different surfaces. The reported voltammetric sensors allow rapid, non-destructive and in situ determination of microorganisms, enabling their application in the analysis of food samples and the monitoring of healthcare environments for prevention purposes. Furthermore, the use of electrodes modified with nanomaterials has gained special attention because of their increased signal intensity. Carbon-based nanosized materials—for example, as carbon nanotubes, graphene and metal nanoparticles, such as AuNPs and MIPs—have been applied for their remarkable electrochemical properties. On the other hand, current research and recent publications have been focused mainly on pathogens, but this technology could also offer great support in the detection of spoilage microorganisms in liquid foods. However, one of the main challenges is the development of specific sensors, especially when dealing with samples where mixed microflora could coexist. The research on new sensors and their commercial application will accelerate the field of microorganism detection, providing feasible tools to quantify the presence of microorganisms in liquid samples.