Advances in the Bacteriophage-Based Precise Identification and Magnetic Relaxation Switch Sensor for Rapid Detection of Foodborne Pathogens

Abstract

The development of novel and highly specific technologies for the rapid and sensitive detection of foodborne pathogens is very important for disease prevention and control. Bacteriophages can recognize viable and unviable bacteria, replacing antibodies as the recognition element in the immune response, which are currently being widely developed in novel precise identification biosensors. Magnetic relaxation switch sensors based on the magnetic relaxation signal has been used to construct a variety of background-free novel biosensors in recent years, which can realize rapid detection of foodborne pathogens. This chapter will mainly introduce the latest developments and future prospects of bacteriophages in the field of accurate identifications for foodborne pathogens. At the same time, it will introduce the research progress and development direction of novel magnetic relaxation switch sensors for detecting foodborne pathogens.

1. Introduction

Food safety is one of the key issues that people are most concerned about. Food poisoning caused by foodborne diseases is a major problem in food safety. Pathogens are infectious agents that can cause foodborne diseases, which include fungi, protozoans, bacteria and viruses. They enter the human body through various modes of infection like food, water and air, and are responsible for deaths worldwide. The major foodborne pathogens include Salmonella, Listeria monocytogenes, Escherichia coli, Campylobacter, and Staphylococcus aureus. These pathogens are ubiquitous, which not only affect the quality of food ingredients, destroy nutritional components, but also can induce different diseases in humans by directly invasion or secreting certain toxins, leading to serious health risks and economic burdens [1]. Pathogen diagnosis requires convenient and rapid analytical methods to provide accurate identification. Highly sensitive and accurate analytical methods are also essential for timely clinical decision-making and management of epidemics for infectious diseases [2]. Meanwhile, identifying pathogens rapidly in the early stage of infection is significant to decrease high mortality caused by ingestion of contaminated foods [3].

Biosensors are a high-tech analytical device developed by the interdisciplinary integration of physics, chemistry and biology. They mainly use biomolecular recognition elements (such as antibodies, enzymes, nucleic acids, etc.) to recognize the targets analyte, and then converts to optical, electrical, magnetic or other signals that are easy to capture and recognize by a transducer for easy readout. Because of its advantages of high efficiency, easy automation and simple operation, it has been widely used in the field of food safety [4, 5].

It is anticipated that the future research direction of developing novel biosensors is to achieve the accurate identification of pathogens and eliminate the interference from food sample impurities. As emerging technologies, bacteriophage can recognize viable and unviable pathogens and can be a precise recognition element to replace traditional antibodies in the immune response, contributed to construct various precise biosensors. And the magnetic relaxation switch sensors which can realize high signal-to-noise ratio and background-free detection have also attracted extensive attention in the field of rapid detection for foodborne pathogens. In recent years, extraordinary progress has been made in terms of bacteriophage- and biosensor-based detection methods, focusing on their potential use in the field of rapid detection for foodborne pathogens, and becoming frontier research hotspot. This chapter will mainly introduce the latest developments and future prospects of bacteriophage-based precise identification technologies, which allow accurate detection of foodborne pathogens. Additionally, it will summarize the research progresses and development directions of novel magnetic relaxation switch sensors in the field of rapid detection for foodborne pathogens.

2. Introduction of bacteriophage

Bacteriophages (shortened to phage), composed of protein and nucleic acid, are viruses that can specifically infect bacteria and proliferate in host bacteria. Phage was first discovered by Frederick W. Twort in 1915 and subsequently isolated by Felixd’Herelle in 1917, who named phages according to their properties [6]. According to their basic structural forms, phages can be classified into icosahedral phages without tail structure, icosahedral phages with tail structure, and filamentous phages. Most phages are icosahedral phages with tail structure [7]. The head capsid and tail are composed of proteins. These phages consist of thehead capsids containing the genetic material of the phage (DNA or RNA), and the tails that have special receptors to recognize the cell surface of the host bacteria, which are related to phage specificity (Figure 1). Phage typically could also be categorized into lytic or temperate phages based in their life cycle.

Figure 1.

Structure of a typical bacteriophage.

Especially in the recent years, with a better understanding of the detailed knowledge phage characteristics, its application in clinical disease treatment, foodborne pathogen detection and other aspects has been gradually expanded. Phages offer several advantages, such as simple structure, easy to generate quantities in within a short time, high specificity to host bacteria, harmless to human, good stability, and the ability to distinguish viable and unviable bacteria. Their short preparation time and low cost are also the advantages over antibodies. Based on these advantages, a variety of methods have been developed to detect pathogenic bacteria that using phages as the probes. Phages have shown a good application prospect in the field of rapid detection for pathogenic bacteria.

Compared with biometric elements such as antibodies and nucleic acids, phages have obvious advantages in the detection of pathogenic bacteria [8]. The features of bacteria and phages including specific interactions between phages and target bacterial cells, their infectious ability and phage-induced cell lysis, provide a basis for the detection of pathogenic bacteria [9]. At present, the phage-based detection methods are mainly based on phages (natural phages or recombinant phages) and phage components as recognition elements. Phage-based detection methods mainly include phage amplification, phage-based biosensors which combined natural phages with biosensors (such as electrochemistry biosensors and optics biosensors) and engineered phage-based methods (such as reporter phages and phage display technology). Phage component-based assays are mainly conducted by taking advantage of phage receptor binding proteins and lysin proteins. This part will discuss the research progress of the current phage-based detection methods to provide a comprehensive theoretical basis for food safety assessment.

2.1 Progress in detection methods using phage as recognition element

2.1.1 Phage amplification

Phage amplification is a classical method for the detection of foodborne pathogens. The principle of this strategy was based on the measurement of progeny phage released from the infected target bacteria. Specifically, phages were mixed with the sample solution to infect the target bacteria. Then the viricides such as ammonium ferrous sulfate are used to kill the free phages in the culture medium. After phages were released from the lysed target bacteria, the helper bacteria cells were added to propagate the phages and determine the phage titer using the double-layer plate method, so as to evaluate the number of target bacteria. This method has been used for the detection of Mycobacterium tuberculosis, Salmonella, Listeria monocytogenes, Escherichia coli, Pseudomonas aeruginosa and Campylobacter [10].

Rajnovic et al. developed a method based on the analysis of optical density kinetics in bacterial cultures with lysed MS2 phage for bacterial infection. This method can detect as few as 10 phage particles per assay volume after a phage incubation period of 3.5 h. And it could detect as low as 10⁴ CFU/mL Escherichia coli in 2 h [11]. Garrido-Maestu et al. developed a novel method based on the amplification of the Salmonella bacteriophage vB_SenS_PVPSE2, coupled with real-time PCR (qPCR) for the rapid detection of viable Salmonella Enteritidis in chicken samples [12].

2.1.2 Phage-mediated biosensors

Phage-mediated biosensors can be divided into two categories in principle. One is to use phages as recognition and capture components of pathogenic bacteria, supplemented by other substances for signal readout, but it donot lyse bacteria. The other is to use naturally occurring lytic phages to specifically lyse host bacteria and release intracellular substances, which in turn trigger the catalysis of the substances to produce signals for readout. In this section, we will review the recent developments of these two detection methods.

Due to their advantages of high sensitivity, specificity, accuracy, fast response and low cost, sensors have become one of the most widely used methods in the detection of pathogenic bacteria. Zhou et al. developed a carbon nanotube (CNT)-based impedimetric biosensing method for rapid and selective detection of viable Escherichia coli B cells. The T2 bacteriophage (virus) served as the biorecognition element, which was immobilized on polyethylenimine (PEI)-functionalized carbon nanotube transducer on glassy carbon electrode. The detection was highly selective toward the B strain of Escherichia coli and the detection limit of the biosensor is 10 CFU/mL [13]. In a recent study, Farooq et al. isolated Staphylococcus aureus-specific phages and immobilized them on modified bacterial cellulose/carboxylated multiwalled carbon nanotubes to create an electrochemical biosensor to detect bacteria in milk samples. Results showed that 3 CFU/mL bacteria were detected in the phosphate buffer and 5 CFU/mL bacteria could be detected in the milk sample within 30 min at neutral pH [14]. Optical sensor is also one of the sensors that has been widely used in detecting pathogenic bacteria. Edgar et al. constructed the detection method of Escherichia coli by combining Quantum dot and phage for the first time in 2006. Tawil et al. developed a biosensor using whole phage and surface plasmon resonance to detect methicillin-resistant Staphylococcus aureus (MRSA) at 10³ CFU/mL [15].

Phages are used to specifically lyse host bacteria to release intracellular enzymes or other specific substances. The released enzymes act as markers to catalyze the reaction of active substances to produce specific substances and generate signals that can be measured by biosensors, so as to detect pathogenic bacteria [16]. The intracellular enzymes that can be used as markers mainly include adenylate kinase, β-D-galactosidase, β-D-glucuronidase, etc. Chen et al. immobilized T7 phage particles on magnetic beads to capture and lyse Escherichia coli BL12, and then to release intracellular β-galactosidase (Figure 2). The detection limit within 2.5 h was about 1 × 10⁴ CFU/mL, and it was reduced to 10 CFU/mL after 6 h pre-enrichment [17].

Figure 2.

Schematic representation of detection of Escherichia coli in drinking water using T7 bacteriophage-conjugated magnetic probe [17]. Copyright 2015 American Chemical Society.

In addition, researchers combined phages and bioluminescence reagents to develop optical-based methods, such as ATP bioluminescence, NADH bioluminescence for pathogenic bacteria detection. First of all, the target bacteria were subjected to phage specific infection, and lysed to release ATP and NADH, since the content of ATP and NADH in each cell was roughly constant. ATP is then catalyzed by luciferase to react with luciferin or NADH with substrates such as FMN and aldehydes to emit light [18]. Bacterial count could be evaluated from quantitative measurements of ATP bioluminescence. Eed et al. developed an ATP bioluminescence-sensing assay to detect microbial viability. A bioluminescent recombinant Escherichia coli strain was used with luciferase extracted from transformed bacteria. Results showed that this method were more rapid and efficient than traditional plate counting assay [19].

2.1.3 Progress in detection methods based on engineered phage

Reporter phage detection techniques are based on molecular biology methods. In this method, the reporter phage containing the reporter gene is constructed first. The reporter gene is introduced into the host chromosome and encodes the expression of a fluorescent substance or a colorimetric marker dependent substrate for pathogen identification.

Reporter genes commonly used at present include firefly luciferase gene (luc), bacterial luciferase gene (luxAB) [20], green fluorescent protein gene (gfp), bacterial ice nucleoprotein gene (inaW) [21] and β-D-galactosidase gene (lacZ) [22]. The proteins expressed by these genes in target bacteria can be detected using colorimetry, fluorescence or luminescence techniques. Reporter phage technology has been successfully used to identify a variety of pathogens, including Escherichia coli, Mycobacterium, Salmonella, Staphylococcus aureus, Listeria monocytogenes. The greatest advantage of reporter phage is its ability to distinguish between viable and unviable bacteria, since phage will not be able to infect and express reporter genes in unviable bacteria. Alcaine et al. constructed a lateral flow detection method for pathogenic bacteria by bioengineering T7 phage. This assay can detect 10³ CFU/mL of Escherichia coli in broth after 7 h [23].

Phages have the unique ability to display peptides or proteins on their surfaces and can be used for the detection of foodborne pathogens. This technique named as phage display as first discovered in 1985. The proteins or peptides displayed are capable of affiniting to a variety of targets such as carbohydrates, proteins, small molecules, or whole cells. The basic principle is to fuse the gene encoding for the target peptide or proteinto the phage surface protein encoding gene, causing the mixed protein to be expressed on the phage surface [24]. Bacteriophage (M13, F1, FD, T4 and T7, etc.) are commenly used in phage display technology. McIvor et al. panned out Listeria monocytogenes polypeptides from the phage display peptide library with the ability to distinguish Listeria monocytogenes from other Listeria spp., which may have potential utility to enhance detection of Listeria monocytogenes [25]. Karoonuthaisirit et al. screened peptides specific to 8 Salmonella mixtures from the phage display peptide library. Meanwhile, SPR was used to detect Salmonella based on the screened phage polypeptides, and was capable of detection with high specificity and accuracy. The detection limits of 8.0 × 10⁷ and 1.3 × 10⁷ CFU/mL for one-time and five-time immobilized sensors, respectively [26].

2.2 Progress in detection methods using phage components

Phage components, such as RBP and lysins, not only have specific affinity to target bacteria, but also are highly adaptable to environmental conditions. RBP, located in the tail of the virion, anchor the phage to the host cell during infection by recognizing unique protein or carbohydrate (polysaccharide) sequences on the surface of the host bacteria [27]. Lysins are phage-encoded enzymes produced in infected host bacteria at the end of the lytic cycle. These hydrolases enable the phage to lyse the host cell from within and the release of progeny phage particles.

The RBP of phage not only has unique host tail recognition specificity that can specifically recognize host bacteria, but also hashigh resistance to environmental conditions, such as pH, temperature and resistance stability. The RBP of phage can be used as a potential probing element for pathogen detection. Singh et al. reported the use of the RBP of Campylobacter bacteriophage NCTC 12673 for the specific capture of Campylobacter jejuni bacteria using RBP-derivatized capturing surfaces. The detection limit of the RBP-derivatized SPR surfaces was found to be 10² CFU/mL [28]. Poshtiban et al. attached phage RBP Gp047 of phage NCTC12673 to magnetic beads. The specificity of capture was confirmed by using Salmonella Typhimurium as negative control. Total sample preparation and analysis time were less than 3 hours [29]. The specific RBP was displayed in engineered M13 phage, which has a natural potency to target the desired bacteria. The phages were bound on gold nanoparticles due to the available thiolation potency. The interaction was monitored through SPR, which detected 100 cells of Escherichia coli in less than 60 min [30].

Tolba et al. used the anchor region of Listeria bacteriophage produced lysin on the bacterial cell wall as the detection matrix to measure the change of electrochemical impedance during sample passage with a limit of detection limit as 10⁵ CFU /mL in milk [31]. Chibli et al. exploited the ability of specific phage proteins, the endolysins LysK and Φ11, and the bacteriocin lysostaphin, fixed on silicon wafers to bind staphylococci. Binding was quantified by clearing assays in solution and by functionalization of silicon wafers followed by light microscopy. Bacterial binding densities on functionalized surfaces were ~3 cells/100 μm² [32]. Brzozowsk et al. presented a new type of highly sensitive label-free sensor based on long-period gratings (LPG) coated with T4 bacteriophage (phage) adhesin. The adhesin (gp37) binds Escherichia coli B by recognizing its bacterial lipopolysaccharide (LPS)[33]. The application of phage for pathogenic bacteria detection have been summary in Table 1.

Target	Limit of detection	Method	References
Escherichia coli	10^4 CFU/mL	Phage amplification	[11]
Salmonella	8 CFU /25 g		[12]
Escherichia coli	10 CFU/mL	Biosensor	[13]
Staphylococcus aureus	3 CFU/mL		[14]
Escherichia coli	10^3 CFU/mL		[15]
Escherichia coli	10^4 CFU/mL		[17]
Escherichia coli	10^3 CFU/mL	Engineered phage	[23]
Salmonella	1.3 × 10^7 CFU/mL	Phage display	[26]
Campylobacter jejuni	10^2 CFU/mL	Phage components	[28]
Escherichia coli	100 cells		[30]
Listeria	10^5 CFU /mL		[31]
Staphylococci	3 cells/100 μm^2		[32]

Table 1.

Application of phage in detection of pathogenic bacteria.

In general, the mechanism of phage’s specific adsorption on host bacteria, the advantages of phage and its application in the detection of pathogenic bacteria are introduced in this part. Phage has great application potential in the detection of pathogenic bacteria due to its advantages of good stability, easy preparation, strong specificity, high safety, and ability to distinguish viable and unviable bacteria. Bacteriophage component-based assays may have some advantages over the use of full phage particles. Phage proteins exhibit greater stability to extreme pH and temperature. The smaller protein size allows for more intensive surface modifications and targeted chemical functionalization to enhance the binding activity of these surfaces compared to the whole phage. In some cases of applications, phage-derived proteins offer another advantages, including captured intact bacteria without inducing lysis and releasing toxic products. The application of phage-derived proteins, while promising to replace antibodies used to capture and enrich bacterial pathogens, is still in its infancy and its potential is largely untapped.

In addition to the wide application of phage-based accurate identification in the field of food safety, the development of effective novel biosensing technologies with low background has also shown great application prospects. Nowadays, the application of magnetic relaxation switch (MRS) sensors in the rapid detection of foodborne pathogens increasingly attracted attention. Compared to traditional optical signal, the magnetic signal owns high specificity, for which the signal is negligible especially in biological and environmental samples. The developing MRS assays do not require complex separation and purification steps and can be performed in turbid, opaque and non-uniform medium, enabling background-free detection [34, 35]. Besides, the MRS senosrs have the advantages of fast detection, simple operation, high signal-to-noise ratio, and easy to realize on-site detection, which holds great promise for food safety. Based on this, we reviewed the MRS sensors research progress in the field of rapid detection for foodborne pathogens in next part.

3. Introduction of magnetic relaxation switch sensors

Magnetic relaxation switch sensor has been an up-and-coming biosensing technology in recent years. It uses magnetic relaxation time as a signal readout to qualitatively and quantitatively detect targets. In physics, the relaxation refers to the process of returning to an equilibrium state after a certain equilibrium state is destroyed. Classical MRS sensors generally uses magnetic material as signal probe for detection. The magnetic nanoparticles (MNPs) are a kind of promising materials that can be extensively explored in various fields including clinical medicine, magnetic resonance imaging, data storage and food safety due to their unique size and physicochemical properties [36]. The basic principle of MRS sensors is the shortening of the relaxation time of water molecules mediated by MNPs, which can result in the nonuniform magnetic field. We can measure relaxation time of water protons through the process of relaxation, and relaxation time can be used as signal readout to reflect the amounts of targets. The relaxation time includes longitudinal relaxation time (T₁) and transverse relaxation time (T₂). In recent years, with the in-depth study of MNPs and relaxation mechanism, a series of magnetic sensors have been designed and worked at the molecular and cellular levels that combine with relaxation time as signal readout for target detection [37, 38].

Conventional MRS assay is generally based on the aggregation or dispersion of nanoparticles leading to the change of relaxation signal, as shown in Figure 3 [39]. The application of MRS sensor started in 2001 by Weissleders’s group, who use four different types of molecular interactions (DNA-DNA, protein-protein, protein-small molecule and enzyme catalysis) to show that the nanoparticles MRS technology can detect targets in vivo with highsensitivity, efficiency, and high-throughput. This platform is based the functional superparamagnetic nanoparticles (SMNPs) aggregated or dispersed via the specific affinity reaction between antibodies and antigens, and the quantity of target analyte was closely related to the degree of the state change, associated with a considerable change of the T₂ [35]. This research lay an important foundation of MRS sensor that particularly used to analyze various biochemical samples, such as nucleic acids, pathogens, biomacromolecules and micromolecules by inducing the aggregation or dispersion of MNPs and then bringing about a switch of the relaxation time to reflect the amounts of targets. The following part will summarize the research progress of MRS sensors following classification of state-dependent MRS sensors, amount-dependent MRS sensors, paramagnetic-ion mediated MRS sensors and other MRS sensors.

Figure 3.

The principle of classic magnetic relaxation switch (MRS) sensors [39]. Copyright 2012 American Chemical Society.

3.1 State-dependent MRS sensors for pathogens detection

The principle of MRS sensors based on magnetic particles changed state is to modify the donor/receptor (such as antigen/antibody, biotin/streptavidin, aptamers, etc.) on the surface of magnetic particles to construct specific magnetic probe. In the process of analysis, the specific recognition of the donor-receptor causes the state to change from dispersion to aggregation, hence affected the uniformity of the local magnetic field. When the water molecules diffuse through these uneven magnetic fields, the lateral relaxation of protons is accelerated and caused shorter lateral relaxation time [40]. The degree of magnetic probe state and T₂ signal change are positively correlated with the content of the target substance in the sample, so as to achieve the purpose of quantitative detection.

Based on this principle, Zhao et al. proposed a sensitive and rapid method for detecting Listeria monocytogenes (L. monocytogenes) in food which is based on the change of T₂ using NMR. Firstly, nanoparticles modified with silica and coupled with anti-L. monocytogenes antibodies dispersed in solution. Once the target of L. monocytogeneswas added in solution, the functioned particles self-assembled on the surface of L. monocytogenes bring about the increase of the T₂ value of water protons. This method has a lower limit of detection limit of 3 MPN (using the most-probable-number (MPN) assay) and the upper limit of 10³ CFU/mL and only requires 40 min to complete all the tests [41]. Based on a similar principle, Yu et al. developed a reliable immunoassay for the specific detection of Cronobacter sakazakii in dairy samples with silica-coated magnetic particles to ensure the safety of infant formula powder. This method is able to detect Cronobacter sakazakii in milk powder and cheese samples at 1.1–11 MPN, but does not fit for the detection of bacteria at higher concentrations (>1100 MPN) [42].

Except through the change of T₂ to judge result, Wang et al. developed a MRS sensor based on SMNPs which uses T₂ for signal readout for the rapid detection of the foodborne pathogen Salmonella in milk samples directly. The SMNPs can switch their dispersion and aggregation states based on the presence or absence of the target, which can adjust the T₂ of adjacent water molecules. Before the immune response, the state of SMNPs was dispersed, and the value of T₂ was high. After recognition by the immune response, SMNPs changed from dispersed to clustered, and the T₂ decreased rapidly. The value was positively correlated with the concentration of Salmonella, thus realizing the detection of Salmonella in milk samples. Compared with traditional ELISA, the sensitivity of this immunosensor is increased by 20 times (MRS, 10³ CFU/mL; ELISA, 2 × 10⁴ CFU/mL), and the required detection time is drastically reduced from 2 to 4 h to 30 min [43]. Apart from these pathogens, Mycobacterium avium spp. Paratuberculosis (MAP) is the known pathogen of Johne’s disease in cattle, which is an economically devastating disease. This bacteria is difficult to grow in culture and its identification with current methods are difficult. Except causing severe intestinal inflammation in cattle, this microorganism has been isolated from blood, breast milk, and intestinal lesions of human patients. Kaittanis et al. realized the detection of MAP in milk using MRS sensor. The scheme is based on disperse nanoparticles in solution that can capture to the surface of a bacterial molecular and induce significant changes in T₂. As the amounts of bacteria increases, the available nanoparticles in solution were in a more disperse-like state, causing minimal changes in T₂. The detection limit is 1.55 × 10³ CFU/mL, and the detection process was only 30 min without interference from other bacteria [44]. Subsequently, the same group reported the use of hybridizing magnetic nanosensors (hMRS) for the detection of MAP within less than an hour. The hMRS are designed to bind to a unique genomic sequence found in the MAP genome, causing significant changes in the sample’s magnetic resonance signal. Hence they realized the detection of pathogens which can evade recognition by the immune system. This platform can detect a single MAP genome copy within 30 min [45].

Sara et al. proposed NMR-based detection system to detect pathogenic levels of Vibrio parahaemolyticus isolated from seafood with molecular mirroring using iron nanoparticles coated with target-specific biomarkers capable of binding to DNA of the target microorganism. The detection limit was 10 ⁵ CFU/mL, and was used to prevent the pathogen spread into humans, via contaminated, raw, or undercooked seafood [46]. A high affinity and specificity of the aptamer-recognition system was established by Jia’s group that the anti-Pseudomonas aeruginosa (P. aeruginosa) aptamer was immobilized onto the surface of superparamagnetic iron oxide. Then the nanoparticles acted as switches of T₂ measurement between aggregated and dispersed states, while with and without target bacteria. This MRS sensor can sensitively detect foodborne P. aeruginosa in the real food and drinking water samples with a detection limit of 50 CFU/mL [47].

The above protocols are all single-mode detection based on magnetic, so some scholars try to develop a dual-mode detection scheme. A protocol is proposed by Tyler et al. through the unique combination of magnetic and fluorescent parameters in a nanoparticle-based platform to construct a simple Enterohemorrhagic Escherichia coli O157:H7 diagnostic technique. This nano sensor was uniquely pair together magnetic relaxation and fluorescent modalities, allowing for a dual-detection platform. In the case of bacterial contamination, the binding between the magnetic nanoparticles and bacteria influence the transfer of surrounding water protons which causes a change in the T₂ relaxation times. As the concentration of bacteria in solution rises, the magnetic nanoparticles disperse and resulting in lower T₂ values. Conversely, fluorescence emission will increase in proportion with the concentration of bacteria, due to the increased number of magnetic nanoparticles directly bound to pathogen. This dual mode scheme can detect contamination with as low as 1 CFU present in solution within less than 1 h. Furthermore, the potential ability of them to be used in commercial packaged foods such as milk has been proved [48]. Subsequently, this group based on the similar dual-mode detection scheme to realize the detection of Staphylococcus epidermidis and E. coli which were contaminants in blood. This dual-mode determination improves the accuracy of actual detection [49].

However, the disadvantage of the MRS sensors depending on the state of magnetic particles is that the magnetic signal only positively correlated with the concentration of the target within a certain range, hence the linear range is narrow. And the state change is also susceptible to interference fcaused by various factors such as the sample matrix, which is easily suffered from the nonspecific adsorption and aggregation of magnetic particles that cause the inaccuracy of detection.

3.2 Amount-dependent MRS sensors for pathogens detection

The amount-dependent MRS sensors has proposed to solve the limitations of state-dependent MRS sensors. The basic scheme of MRS sensors based on the change of magnetic particles amounts depended on the difference in the separation speed of magnetic particles of different sizes in the same magnetic field. The magnetic particles of large diameter are used as the carrier of immunomagnetic separation, and the magnetic particles of small diameter are used as the magnetic signal probe. The donor/receptor specific recognition function molecular is modified on the both carrier and probe magnetic beads. The probe specifically recognizes the magnetic particles modified with the acceptor/donor through the carrier modified with the donor/receptor, and changes the number of magnetic probes after magnetic separation and other operations, thereby realizing biosensing. When the target appears and is recognized, the amount of magnetic probes is changed after magnetic separation, thereby realizing quantitative biosensing. This mode does not need to induce the aggregation of magnetic particles, which effectively improves the stability of MRS sensors. In addition, the T₂ signal is more sensitive to the change of magnetic probe concentration, which effectively improves the sensitivity of MRS sensors.

Chen et al. firstly proposed amount-dependent MRS sensor with more convenient operation, enhanced sensitivity and better reproducibility. Magnetic beads of large size (250 nm, MB250) can be separated more quickly than those of small size (30 nm, MB30) under an external magnetic field. Based on this phenomenon, a MRS sensor combined with magnetic separation that enables one-step, sensitive detection of pathogens. The MB250 and MB30 can selectively capture and enrich the targets to form the “MB250-target-MB30” conjugate. After magnetic separation, unreacted MB30 can be used as signal readout probe and corresponds to the concentration of targets (Figure 4a). The entire immunoassay can be completed within 30 min and the detection limit is 10² CFU/mL. Compared with conventional MRS sensor, this kind of sensors could avoid the unstable state of aggregation and ensure the accuracy of the signal, which is capable for the detection of Salmonella in milk [2]. Based on the similar principles, then this group proposed a highly sensitive magnetic DNA sensor based on nucleic acid hybridization reaction and magnetic signal readout. The scheme is to design the L. monocytogenes specific probe 1 and probe 2, and label them on the 30 and 250 nm magnetic nanoparticles, respectively. After hybridization reaction to form a sandwich nanocomplex and magnetic separation, the unbound 30 nm magnetic particles can act as the T₂ signal readout probe (Figure 4b). This assay allows the one-step detection of L. monocytogenes as low as 50 CFU/mL within 2 h without DNA amplification, providing a promising detection platform for pathogenic nucleic acid [50].

Figure 4.

(a) A MRS sensor based on the amounts of antibody-modified MNP for Salmonella detection [2]; copyright 2015 American Chemical Society. (b) A MRS sensor based on the amounts of DNA-modified MNP for L. monocytogenes detection [50]. Copyright 2021 Elsevier.

To integrate the amount-dependent technology and realize operate on 96-well plates, Zou et al. described a novel MRS sensor for Salmonella detection. In this assay, functionalized Fe₃O₄ nanoparticle clusters (Fe₃O₄ NPC-SA) were used as probe to capture biotinylated antibody. The nanoparticles are cross-linked interaction and biotinylated antibody specifically recognize Salmonella at different sites. Then the Fe₃O₄ NPC-SA was eluted from 96-well microplates, which led to the change of transverse T₂. The strategy not only detected Salmonella at 10⁵ CFU/mL showing high sensitivity, but also addressed a common phenomenon high-dose “hook effect” in which high concentration of analyte saturation prevents the effective aggregation of nanoparticles [51]. Similarly, Ling et al. combined membrane filtration with MRS sensor to establish a new time domain nuclear magnetic resonance (TD-NMR) biosensor to monitor and control Salmonella in milk for ensuring food safety, which also can overcomes the “hook effect” in MRS sensor. And the detection limits of pure culture solution and pasteurized milk were 2.3 × 10³ CFU/mL [52]. Ting et al. developed dendritic superparamagnetic iron oxide nanoparticles(dendritic-SPIONs) combined with MRS sensor for Salmonella detection in milk. Bacterial capture antibody and biotinylated detection antibody (BT-mAb) were firstly used in the form of a sandwich in a 96-well plate to immobilize target pathogen. Then streptavidin modified polyamidoamine (SA-PAMAM) was used to connect BT-mAb and functionally modified SPIONs. PAMAM-mediated amplified functionalized SPIONs via streptavidin-biotin amplification system, aggregated nanoparticles to form dendritic SPIONs to achieve dual signal amplification. Finally, the dendritic SPIONs capture complex is eluted to test T₂ signal as the output of the target bacteria capture signal. This biosensor has detection limit of Salmonella in milk of 2.6 × 10 ⁴ CFU/mL, and showed good specificity to anti interference. Therefore, this innovative detection platform provides a novel signal amplification method [53]. These amount-dependent MRS sensors mentioned above can greatly expand the detection linear range, and avoid the interference from the sample matrix, which have the advantage of being more sensitive and stable. However, these methods still require the use of magnetic particles to convert the target signal into a relaxation signal.

3.3 Paramagnetic-ion mediated MRS sensors for pathogens detection

Conventional MRS assays employ monodispersed MNPs as the magnetic probe and modulate their states or amounts to result in the changes of transverse relaxation time of water protons. Nevertheless, the stability of MNPs when conjugating with the ligands remains an issue. The conjugation of MNPs may affect their stability, and the nonspecific interaction between MNPs and the sample matrix can result in the instability that affect the accuracy. The state-dependent and amount-dependent MRS sensors still need to be mediated by MNPs. The coupling procedure of the acceptor/donor on the surface of MNPs may be quite different for different operators, hence still insufficient for stability. Therefore, some researchers proposed novel paramagnetic ion-mediated MRS assays which have greatly improved the capability. It is much easier to prepare the aqueous solution of paramagnetic ions than that of MNPs. And its solution generally has a longer shelf life. Furthermore, paramagnetic ions have different valence states that can be interconverted by redox reactions, providing a versatile magnetic sensing platform. Wang et al. described a magnetic immunosensor relying on Mn(VII)/Mn(II) interconversion to trigger the corresponding change in the low-field nuclear magnetic resonance of the T₂. The signal of the water protons detected in Mn(II) aqueous solution is much stronger than Mn(VII) aqueous solution, hence enable to develop a background signal-free magnetic immunosensor with a high signal-to-background ratio through employ immunomagnetic separation and enzyme-catalyzed reaction (Figure 5a). The detection limit of this method for Salmonella is 20 CFU/mL, which has greatly improved the sensitivity of conventional paramagnetic ion-mediated magnetic sensors, offering a promising platform for bioanalysis [54]. On the basis of this mechanism, Li et al. presented an alkaline phosphatase (ALP)-mediated magnetic relaxation DNA biosensor enabling the rapid and sensitive analysis for L. monocytogenes in ham samples. The DNA probes were initially designed to specifically hybridize with targeted region of bacterial genomic DNA. And the amounts of ALP is related to the concentration of pathogen by DNA hybridization. Then ALP can induce Mn(VII) to convert into Mn(II) resulting in a significant change of T₂ signal through enzyme-catalyzed reaction. This MRS sensor could exhibit high sensitivity for L. monocytogenes detection with a low detection limit of 10² CFU/mL. The constructed simple and reliable DNA biosensor could offer an ideal candidate for the detection of foodborne pathogens [56].

Figure 5.

(a) A paramagnetic ion Mn(VII)/Mn(II) mediated MRS sensor for Salmonella detection [54]; copyright 2019 American Chemical Society. (b) A MRS sensor based on phosphatase-mediated transition of hydrogels for Salmonella detection [55]. Copyright 2021 American Chemical Society.

In addition to catalyzing the redox of paramagnetic ion, the ALP can also participate in catalyzing the formation of hydrogels to cause the signal changes of T₂. Wei et al. developed a sol-gel transition of hydrogels to change T₂ signal for assaying foodborne pathogens. The ALP can catalyze the reaction to generate an acidic environment that could transform the sol-state alginate solution to hydrogel, and this process can directly regulate the diffusion rate of water protons resulting in the change of T₂ signal (Figure 5b). This biosensing strategy directly modulates the water molecules rather than conventional magnetic probes, hence displaying high sensitivity for detecting 50 CFU/mL Salmonella within 2 h and offering a straightforward and sensitive platform for pathogen detection [55]. The summary of different modes MRS sensors applied to foodborne pathogens detection are listed in Table 2.

Target	Limit of detection	Mode	References
Listeria monocytogenes	3 MPN	State-dependent	[41]
Cronobacter sakazakii	1.1 MPN	State-dependent	[42]
Salmonella	10^3 CFU/mL	State-dependent	[43]
Mycobacterium avium spp. Paratuberculosis	1.55 × 10^3 CFU/mL	State-dependent	[44]
Mycobacterium avium spp. Paratuberculosis	A single genome copy	State-dependent	[45]
Vibrio parahaemolyticus	10^5 CFU/mL	State-dependent	[46]
Pseudomonas aeruginosa	50 CFU/mL	State-dependent	[47]
Escherichia coli	1 CFU	State-dependent	[48]
Staphylococcus epidermidis and E. coli	2 CFU/mL	State-dependent	[49]
Salmonella	10^2 CFU/mL	Amount-dependent	[2]
Listeria monocytogenes	50 CFU/mL	Amount-dependent	[50]
Salmonella	10^5 CFU/mL	Amount-dependent	[51]
Salmonella	2.3 × 10^3 CFU/mL	Amount-dependent	[52]
Salmonella	2.6 × 10^4 CFU/mL	Amount-dependent	[53]
Salmonella	20 CFU/mL	Paramagnetic-ion	[54]
Listeria monocytogenes	10^2 CFU/mL	Paramagnetic-ion	[56]

Table 2.

Applications of MRS sensors in foodborne pathogens detection.

Note: MPN—most probable number; CFU—colony forming unit.

4. Summary

This section mainly introduces the research progress of phage-based precise identification methods and magnetic relaxation switch sensors in the field of rapid detection for foodborne pathogens. Bacteriophage has strong specificity to pathogenic bacteria, easy to prepare, harmless to human body, can be used as a novel identification element to detect pathogenic bacteria. It mainly realized the rapid detection of foodborne pathogens by phage amplification, genetic engineering or detect the components. Besides, as a novel multidisciplinary analysis technology, the MRS sensors have the advantages of efficient analysis, high signal-to-noise ratio and simple operation, which is based on different scheme such as state, amounts and paramagnetic ion. The development of traditional MRS sensors are mature, but mainly relies on the state, mobility and distribution of hydrogen protons in the detection system, which have the disadvantage of insufficient sensitivity or targeting. Therefore, the targeting can be enhanced by developing novel functionalized nanoparticles while increasing sensitivity. With the deepening of research, the MRS sensors will play a more important role in the rapid detection of foodborne pathogens. The future research direction of MRS sensors can focus on multiple and high-throughput detection, achieve intelligent and portable on-site rapid detection, and exploring a revolutionary magnetic sensing mechanism. In the future, these emerging sensors based on specificity of phage and the efficient readout of MRS mentioned above will be developd rapidly for foodborne pathogens detection and contribute powerful methodological guarantee food safety, as well as human health.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.