Advances in gene-targeted diagnostics for pathogenic Escherichia coli

Abstract

Pathogenic Escherichia coli (PEC) strains are important pathogens that causes a variety of infectious diseases in humans. Traditional bacterial culture and biochemical identification methods are time-consuming and lack specificity, making rapid and accurate diagnosis challenging. Molecular detection techniques, with their high sensitivity, specificity, and rapidity, have become increasingly important for identifying, typing, and tracing PEC. This review systematically sorts out the main target genes currently used for molecular detection of PEC, including virulence genes and marker genes of intestinal PEC and extraintestinal PEC. It also details the progress of mainstream molecular assays, such as polymerase chain reaction and its derivatives, multiple isothermal amplification technologies, biosensors, genome sequencing, integrated devices, and emerging detection methods. Despite significant advancements, challenges remain in differentiating live from dead bacteria, handling complex sample matrices, standardization, and cost control. Future developments in PEC molecular testing will focus on integrating macro-genomics, biosensors, and artificial intelligence to achieve more automated, intelligent, high-throughput, and field-deployable solutions, thereby providing robust support for disease prevention, control, and food safety assurance.

---

Yu Zhang's research group

Our research group is committed to the R&D, application, and industrialization of nanozymes, magnetic micro-/nanomaterials, molecular imaging, tumor diagnosis and treatment integration, biosensors, functional hydrogels, molecular/immunological diagnostic technologies, etc. Specifically, within the domain of molecular and immunological technologies for point-of-care testing (POCT), the research endeavors of our team center on the following areas: high-efficiency nucleic acid extraction, nucleic acid tests, high-performance fluorescent microspheres, and multi-cascade fluorescence signal amplification technologies. Additionally, we are engaged in the modification and application of functional micro-/nanomaterials, fluorescence immunochromatography, biosensors, etc. These research activities are directed towards providing theoretical and technical support for the expeditious and precise detection of foodborne pathogens, zoonotic pathogens, and disease-related molecular markers. Furthermore, the detection kits and equipment developed by the team are well-suited for a variety of applications, including scientific research, medical diagnosis, food safety and quality testing, prevention and control of animal diseases, etc.

1. Introduction

Escherichia coli (E. coli), a Gram-negative bacillus, is widely found in the intestines of warm-blooded animals and is an important part of the normal intestinal flora. However, certain specific strains of E. coli are able to cause a variety of diseases in humans and animals by acquiring virulence factors, and these strains are known as pathogenic E. coli (PEC).¹ Globally, PEC are one of the major pathogens of a wide range of infectious diseases such as foodborne diseases, intestinal infections, urinary tract infections (UTIs), neonatal meningitis, and sepsis, and poses a serious threat to public health and safety.

Conventional PEC detection methods principally depend on bacterial culture, morphological observation, biochemical identification, and serologic typing. Although these methods are well established, they suffer from several practical drawbacks: a prolonged turnaround time (often several days), labor-intensive procedures, and an inability to detect non-culturable or fastidious strains. Moreover, serotyping can yield inaccurate results because of cross-reactivity among O and H antigens.² The advent of molecular technologies has precipitated the development of nucleic-acid-based detection methods. These techniques have emerged as a pivotal means to address the limitations of conventional methods due to their expeditious, precise, and sensitive characteristics. Therefore, continual refinement of molecular methods for PEC is essential for early diagnosis, outbreak control, and ensuring food safety.³

The objective of this review is to categorize and summarize the main target genes currently used for the molecular detection of PEC, together with the principles guiding their selection. We also systematically introduce the principles of leading molecular techniques—such as nucleic acid hybridization, amplification technologies, biosensors, sequencing, and emerging assays—and their specific applications in identifying different PEC types (Scheme 1). Emphasis is placed on comparing key performance parameters, including the detection time, limit of detection, and detection range. By synthesizing existing research, we aim to provide a valuable reference for investigators and clinical laboratories. Finally, we discuss future trends and challenges in molecular PEC detection to promote the development of more efficient and accurate strategies.

Scheme 1 Current molecular methods for the detection of pathogenic E. coli.

2. Classification and pathogenic characteristics of PEC

The classification of PEC is complex. Strains are typically categorized according to their pathogenic mechanism, clinical symptoms, virulence factors, and O:H serotype characteristics. Broadly, PEC strains fall into two groups: (a) intestinal pathogenic E. coli (InPEC), which mainly cause intestinal infections—most commonly diarrheal diseases. (b) Extraintestinal pathogenic E. coli (ExPEC), capable of causing infections in tissues outside the intestinal tract.⁴ Although these groups share some typing schemes, their pathogenic spectra and virulence traits differ.

2.1. Main types of InPEC and their characteristics

2.1.1. Enterotoxigenic E. coli (ETEC). ETEC are the leading cause of diarrhea in infants and young children, and travelers’ diarrhea in developing countries. Its pathogenicity involves adhesion to small-intestinal epithelial cells via fimbriae (e.g., the colony-forming antigens CFA/I, CFA/II and CFA/IV) and the production of at least one enterotoxin: heat-labile enterotoxin (LT) or heat-stable enterotoxin (ST).

LT is structurally and functionally similar to V. cholerae toxin; it activates adenylyl cyclase, raising intracellular cyclic adenosine monophosphate (cAMP) levels. ST, conversely, activates guanylate cyclase and increases cyclic guanosine monophosphate (cGMP). Both toxins disrupt intestinal electrolyte and water balance, causing watery diarrhea.⁵

2.1.2. Enteropathogenic E. coli (EPEC). EPEC rank among the leading bacterial causes of diarrhea in infants and young children, particularly in developing countries. Their characteristic pathogenic mechanism is causing “attaching and effacing (A/E)” lesions on intestinal epithelial cells, which are encoded by genes localized to the locus of enterocyte effacement (LEE), leading to the destruction of intestinal microvilli and the formation of pedestal-like structures.⁶ EPEC achieve tight adhesion by binding to the host cell surface receptor Tir (secreted by EPEC itself and inserted into the host cell membrane) via the bacterial outer membrane protein intimin (encoded by the eae gene). Typical EPEC (tEPEC) harbor the EAF (EPEC adherence factor) plasmid, which encodes a bundle-forming pilus (BFP) that mediates bacterial aggregation and initial adhesion, whereas atypical EPEC (aEPEC) lack the EAF plasmid.⁷

2.1.3. Enterohemorrhagic E. coli (EHEC)/Shiga toxin-producing E. coli (STEC). STEC, also known as verotoxin-producing E. coli (VTEC), are defined by their ability to produce Shiga toxin (Stx), including Stx1, Stx2, and their numerous variants. Stx inhibits eukaryotic protein synthesis and induces cell death. EHEC/STEC infections range from mild diarrhea to severe hemorrhagic enteritis. While E. coli O157:H7 remains the most extensively studied EHEC strain, many non-O157 serotypes also pose significant public-health threats. Some EHEC/STEC strains (e.g., O157:H7) additionally harbor the LEE pathogenicity island and can induce attaching-and-effacing lesions.⁸

2.1.4. Enteroaggregative E. coli (EAEC). EAEC strains are characterized by a stacked-brick aggregation pattern on the surface of HEp-2 cells cultured in vitro or on the intestinal mucosa. Their pathogenicity is associated with a variety of virulence factors, including aggregative adhesion fimbriae (AAF), dispersin, plasmid-encoded toxin (Pet), and EAEC heat-stable enterotoxin 1, as well as several other toxins. These organisms are important causes of acute and persistent diarrhea in children globally and have been linked to traveler's diarrhea and infections in immunodeficient populations.⁹

2.1.5. Enteroinvasive E. coli (EIEC). EIEC strains are highly similar to Shigella spp. in genetics, biochemical properties, and pathogenesis. They can invade colonic epithelial cells, proliferate intracellularly, and spread to neighboring cells, thereby eliciting an inflammatory response and tissue destruction, which leads to dysentery-like symptoms accompanied by fever, abdominal pain, and mucopurulent bloody stools. Their invasive capacity is mainly encoded by a large invasive plasmid (pINV).¹⁰

2.1.6. Diffusely adherent E. coli (DAEC). DAEC strains are defined by their diffuse adhesion pattern in HEp-2 or HeLa cell cultures, i.e., the bacteria are uniformly distributed over the entire cell surface. Their pathogenic mechanisms and exact roles in diarrheal disease are not fully understood, but several studies have suggested that these strains are associated with diarrhea in specific pediatric age groups, particularly children aged 1–5 years. The Afa/Dr adhesin family, a well-characterized set of virulence factors in DAEC, mediates binding to host cell-surface molecules such as DAF (decay-accelerating factor, CD55) and other receptors.⁹

2.1.7. Adherent-invasive E. coli (AIEC). AIEC strains are able to adhere to the surface of intestinal epithelial cells and invade the interior of the cells, leading to cellular damage and inflammatory responses. They are also able to survive and replicate in macrophages, thereby triggering chronic inflammation.¹¹

2.2. Main types of ExPEC and their characteristics

ExPEC strains can establish infections at extraintestinal sites; their virulence-factor repertoire differs from that of IPEC strains, enabling them to adapt to the extraintestinal environment and overcome host defense mechanisms.

2.2.1. Uropathogenic E. coli (UPEC). UPEC strains are the leading cause of both community-acquired and hospital-acquired UTIs, accounting for 80–90% of all cases. They carry a diverse arsenal of virulence factors that aid in their adherence to urinary epithelial cells (e.g., via type I and P trichomes), resistance to host immunity (e.g., polycarboxylic polysaccharides), acquisition of nutrients (e.g., the iron carrier system), and damage to host tissues (e.g., the hemolysin and cytotoxic necrosis factor).¹²

2.2.2. Neonatal meningitis E. coli (NMEC). NMEC strains rank among the principal pathogens responsible for neonatal bacterial meningitis, a disease associated with high morbidity and mortality. Their hallmark virulence trait is the production of K1-type capsular polysaccharide, which confers antiphagocytic activity and facilitates traversal of the blood–brain barrier and subsequent invasion of the central nervous system. Additional virulence determinants—including the invasion of brain endothelium proteins, outer-membrane protein A, the FimH adhesin, and cytotoxic necrotizing factor 1—contribute to their pathogenic repertoire.¹³

2.2.3. Avian pathogenic E. coli (APEC). APEC strains are major avian pathogens that trigger a broad spectrum of diseases—respiratory infections, septicemia, airsacculitis, peritonitis, and pericarditis—inflicting substantial economic losses on the poultry industry.¹⁴

2.2.4. Sepsis-associated E. coli (SEPEC). SEPEC strains are E. coli lineages capable of breaching the bloodstream and causing sepsis—a life-threatening systemic inflammatory response syndrome. They commonly possess an ExPEC-linked virulence arsenal that includes serum-resistant capsules, lipopolysaccharide (LPS), iron-scavenging systems, adhesins, and selected toxins. Their virulence repertoire frequently overlaps with those of UPEC and NMEC strains.¹⁵

3. Target genes for molecular detection of PEC

3.1. Principle of target gene selection

In PEC molecular detection, the choice of appropriate target genes is critical, as it directly determines assay specificity, sensitivity, and reliability. Ideal targets should satisfy the following criteria: (a) specificity: the sequence should be present (or highly enriched and conserved) only in the intended PEC pathotypes and absent—or markedly divergent—in non-pathogenic or unrelated strains, thereby minimizing false positives. (b) Conservation: the sequence must be highly conserved across strains of the same pathotype, ensuring that primers and probes recognize the broadest possible range of target isolates and reduce false negatives arising from the genetic drift. (c) Pathogenic association: priority should be given to genes encoding key virulence determinants—such as toxins, adhesins, invasins, capsule components, iron-acquisition systems, or their regulators—whose presence strongly correlates with pathogenic potential. (d) Stability: chromosomally located genes are generally preferred over plasmid-borne ones, because the latter are more prone to insertion, deletion, or point mutations that could disrupt primer/probe binding and compromise assay performance. (e) Copy number: multi-copy targets (e.g., certain insertion sequences or ribosomal RNA genes) can enhance sensitivity, but their copy-number stability among target strains and absence from non-target bacteria must be verified. For most virulence genes, simple presence/absence is the primary consideration.

3.2. Common target genes for InPEC testing

3.2.1. Target genes for ETEC assays. The identification of ETEC strains relies primarily on detecting the enterotoxins—heat-labile (LT) and/or heat-stable (ST)—they produce.

The lt operon comprises eltA (encoding the enzymatically active A subunit of LT) and eltB (encoding the receptor-binding B subunits). Once secreted, LT ADP-ribosylates the α-subunit of host adenylyl cyclase, raising intracellular cAMP and driving the secretion of water and electrolytes into the intestinal lumen.¹⁶

The st gene (encoding heat-stable enterotoxin) consists of two main types associated with human disease. The sta (estA) gene encodes STp (porcine, also known as STIa or ST1), which acts on guanylate cyclase C in the intestinal epithelium, resulting in elevated levels of cGMP, inhibition of fluid absorption, and stimulation of chloride secretion. The stb (or est) gene encodes STh (human, also known as STII), whose mechanism of action is not fully understood. STh is primarily found in pigs, but it can also be detected in some populations. ST possesses a low molecular weight and is not immunogenic. However, it exerts a pronounced diarrhea-inducing effect.¹⁷

Fimbrial adhesins such as fimH have been evaluated as detection targets. ETEC strains adhere to the small-intestinal epithelium via an array of fimbriae—collectively termed colonization factor antigens (CFAs)—that are indispensable for colonization. These factors are encoded by multigene clusters (e.g., cfa for CFA/I, CS1–CS6, and additional variants) that usually reside on plasmids. Because the repertoire of CFA genes is highly heterogeneous, a single fimbrial marker can miss many ETEC isolates. Consequently, toxin genes remain the more reliable and broadly applicable detection targets.¹⁸

3.2.2. Target genes for EPEC assays. EPEC strains are characterized by their ability to form attaching-and-effacing (A/E) lesions on intestinal epithelia. The eae gene, located within the LEE pathogenicity island, encodes the outer-membrane adhesin intimin; its product mediates the tight adherence required for A/E lesion development.¹⁹ Although the LEE island also harbors genes for the type III secretion system and its effectors, eae remains the most widely used molecular marker for these strains.

Typical EPEC (tEPEC) strains carry the bfpA gene, which encodes the bundle-forming pilus subunit BfpA. This pilus mediates bacterial auto-aggregation and initial adhesion to intestinal epithelial cells. The bfpA locus resides on the EPEC adherence factor (EAF) plasmid. Atypical EPEC (aEPEC) strains lack the EAF plasmid and, consequently, the bfpA gene.²⁰

Thus, eae⁺ serves as a screening marker for both tEPEC and aEPEC, whereas the eae⁺/bfpA⁺ genotype specifically identifies tEPEC. The combination eae⁺/stx⁻ is required—but not sufficient—to distinguish EPEC strains from EHEC/STEC isolates.

3.2.3. Target genes for EHEC/STEC assays. EHEC/STEC strains are defined by their ability to produce Shiga toxins. The stx1 gene and its variants (stx1a, stx1c, and stx1d) encode Shiga toxin 1 (Stx1), which blocks protein synthesis and ultimately causes cell death.²¹ The stx2 gene family (stx2a–stx2g and additional alleles) encodes Shiga toxin 2 (Stx2), whose isoforms differ in toxicity. Overall, Stx2 is considered more potent than Stx1 and is linked to a greater risk of severe clinical outcomes.²²

Adhesion-associated loci are also targeted in EHEC/STEC diagnostics. Strains of serotype O157:H7 and several non-O157 serotypes—O26, O103, O111, and O145—frequently harbor the eae gene, which drives attaching-and-effacing lesions and markedly enhances virulence.²³ In addition, Z3276 serves as a specific marker for STEC O157:H7 strains and is critical for biofilm formation.²⁴ Current reference protocols therefore rely on the simultaneous detection of stx1 and/or stx2 together with eae, followed by serogroup-specific wzx or wzy genes to identify the top seven EHEC serogroups (O26, O45, O103, O111, O121, O145, and O157).²³

The O157 serotype-specific genes include rfbE_O157 and fliC_H7. rfbE_O157 encodes a perosamine synthetase, which is required for the synthesis of the O157 antigen (the O side chain of lipopolysaccharide). It is a molecular marker that is specific for the identification of the O157 serogroup.²⁵fliC_H7 encodes the H7 flagellin subunit, enabling specific identification of the H7 flagellar antigen. Together, rfbE_O157 and fliC_H7 unequivocally identify EHEC O157:H7 strains.

In addition, the ehxA gene is known to encode enterohaemolysin, a toxin widely associated with EHEC/STEC virulence. The extracytoplasmic function operon (ecf1) has also been proposed as a potential marker for STEC detection.²⁶

3.2.4. Target genes for EAEC assays. The identification of EAEC is complicated by the heterogeneity of their virulence factors. Aggregative adherence regulator (aggR) encodes the transcriptional activator AggR, which controls the expression of a variety of virulence factors including AAF hyphae and dispersin. This gene is considered to be an important marker for typical EAEC (tEAEC).²⁰

Aggregative adherence fimbriae (AAF) genes include aafA (encoding AAF/I fimbriae subunit), aggA (encoding AAF/II fimbriae subunit), aafC, agg3A, etc. AAF is the key structure that mediates the characteristic “brick-pile-like” adhesion of EAEC.²⁷

Derived from a conserved pAA-plasmid fragment, the pCVD432 probe has long served as a molecular marker for EAEC. Yet its specificity is imperfect: some non-pathogenic strains carry homologous sequences, while certain bona-fide EAEC isolates lack the fragment altogether.²⁸ Consequently, pCVD432 is routinely evaluated alongside additional targets—most notably aggR—to improve diagnostic accuracy.

The toxin genes implicated in EAEC pathogenesis include astA, pet, and pic. astA encodes the heat-stable EAST1 toxin, whose contribution to EAEC disease remains controversial; the gene is also found in several other E. coli pathotypes and unrelated bacteria. pet encodes a secreted serine-protease toxin that perturbs the cytoskeleton and disrupts the intestinal epithelial barrier. pic encodes a secreted mucinase that degrades mucus, facilitating bacterial penetration and subsequent colonisation of the intestinal mucosa.⁹

3.2.5. Target genes for EIEC assays. EIEC strains are genetically and pathogenetically closely related to Shigella spp. Their invasiveness depends primarily on the pINV invasion plasmid. The multicopy ipaH gene is present on both the chromosome and the pINV plasmid of EIEC and Shigella; it encodes an effector protein of the type III secretion system that promotes intracellular survival and immune evasion. Owing to its high copy number and sequence conservation, ipaH is the most widely targeted marker for detecting EIEC and Shigella.²⁹

In addition to ipaH, EIEC strains harbor the ial and invE loci on the pINV plasmid. The ial operon encodes invasion-associated proteins and has served as a polymerase chain reaction (PCR) target, yet its presence can be unstable, making it less reliable than ipaH. invE encodes invasion-regulated proteins, but it is rarely used alone for routine detection.³⁰ In practical applications, invE is often detected in combination with ipaH or virF to improve coverage and reliability.³¹

Given the prevalence of virulence genes (e.g., ipaH) shared by EIEC and Shigella, conventional molecular methods are often unable to differentiate between these two bacterial species, instead classifying them collectively within the EIEC/Shigella group. Further differentiation necessitates a combination of biochemical tests or the detection of other specific markers (e.g., Shigella typically cannot ferment lactose, whereas EIEC can).

3.2.6. Target genes for DAEC assays. Molecular characterization of DAEC strains remains challenging, as their virulence markers are less clearly defined than those of other pathotypes. The Afa/Dr adhesin family is their hallmark: these adhesins mediate diffuse adherence to host epithelial cells. daaE encodes the structural subunit of the F1845 (DAA) adhesin, a member of the Afa/Dr family.³²afaC is a part of the Afa/Dr operon and participates in adhesin biosynthesis or translocation, whereas draE encodes the Dr adhesin itself.³³

Additional virulence-associated loci include the secreted autotransporter toxin gene (sat), the mucinase gene (pic), and the secreted IgA-like protease gene (sigA), and the type 1 fimH. sat encodes a cytotoxic autotransporter that is frequently detected in DAEC isolates. pic is also commonly found in these strains. sigA may contribute to immune evasion. fimH, which encodes the mannose-sensitive adhesin of type 1 fimbriae, is ubiquitous among E. coli, including commensals. Certain allelic variants or atypical regulatory patterns of fimH have been linked to the diffuse-adherence phenotype, yet its value as a DAEC-specific marker is limited and must always be interpreted in combination with Afa/Dr family genes.³³

The molecular diagnosis of DAEC typically relies on the detection of genes associated with the Afa/Dr adhesin family. Due to its heterogeneity, the detection of a single target gene may not be sufficient for comprehensive coverage of all DAEC strains.

3.2.7. Target genes for AIEC assays. AIEC strains lack a universally accepted single molecular marker. Several genes have been proposed as adjunctive targets. Type 1 fimbriae are ubiquitous among E. coli; specific fimH variants (e.g., G66S and V27A) are enriched in AIEC and enhance adhesion/invasion, yet they are not pathotype-specific. yjaA, encoding a putative stress-response protein, is up-regulated in intracellular AIEC and may facilitate survival within macrophages. fyuA, the yersiniabactin receptor gene, promotes iron acquisition and is frequently detected in AIEC. The K polysaccharide transport group II (kpsMT II) region directs synthesis of group II capsular polysaccharides that confer resistance to complement-mediated killing.³⁴

3.3. Common target genes for ExPEC testing

3.3.1. Target genes for UPEC assays. UPEC have numerous virulence factors involving adhesion, iron acquisition, toxin production and immune escape.

Adhesin genes central to UPEC pathogenesis include fimH, pap cluster, sfa/foc, and afa/dra locus. fimH encodes the FimH tip adhesin, which initiates colonisation by binding mannosylated receptors on bladder epithelial cells. While fimH is almost ubiquitous among E. coli, only certain alleles correlate strongly with cystitis and pyelonephritis. The pap cluster—particularly papA, papC and papG—mediates attachment to renal epithelial cells via P pili and is strongly associated with upper-tract infections.³⁵ The sfa/foc operons produce S and F1C fimbriae, respectively. The sfaS gene encodes the S-fimbrial adhesin that promotes binding to renal and endothelial cells. The afa/dra locus encodes non-fimbrial Dr adhesins that recognise the decay-accelerating factor (CD55). Their expression is linked to chronic or recurrent UTIs.

Toxin genes include hlyA and cnf1. hlyA encodes α-hemolysin, a pore-forming toxin that lyses erythrocytes and injures uroepithelial cells, driving inflammation and tissue damage. cnf1 encodes cytotoxic necrotising factor 1, which activates Rho-family GTPases; this triggers cytoskeletal rearrangement, multinucleate-cell formation and apoptosis, thereby enhancing invasion and inflammation.³⁶

Iron is an essential element for bacterial growth, and with very low concentrations of free iron in urine, UPEC has evolved several efficient iron acquisition systems. Iron acquisition system genes include the aerobactin receptor (iutA), iron-regulated outer membrane protein N (iroN), ferric yersiniabactin uptake A (fyuA), coli heme uptake A (chuA), etc. iroN encodes the receptor for catecholamine iron carriers such as salmochelin. fyuA encodes a receptor for Yersinia pestis, another important iron carrier. chuA encodes an outer membrane receptor for the heme uptake system, which enables the bacteria to utilize heme from the host hemoglobin as an iron source.³⁷

In addition, the kpsMT II locus orchestrates the export and assembly of group II capsular (K2) polysaccharides that shield UPEC from phagocytosis and complement-mediated lysis. The usp gene encodes a uropathogen-specific bacteriocin with DNase activity that may confer a competitive edge in the urinary tract. Genes such as C3509 and C3686 (yrbH) have emerged from comparative genomics as candidate markers of UPEC virulence, though their precise roles remain under investigation.³⁷

Molecular assays for UPEC are often performed using combinations of multiple virulence genes to improve the accuracy and coverage of the detection.

3.3.2. Target genes for NMEC assays. NMEC strains cross the blood–brain barrier largely due to their K1 capsule. The K1 polysaccharide is synthesized by the neu and kps gene clusters. neuC encodes N-acetylneuraminate synthase, catalyzing the production of sialic acid, the capsule's dominant monomer. kpsMT II drives capsule export and surface assembly. The sialic acid structure of the K1 capsule is similar to that of host cell surface molecules, which helps the bacterium evade host immune recognition.³⁸

NMEC strains rely on a dedicated invasion cassette to breach the blood–brain barrier. Key loci include ibeA, ibeB, ibeC, and ompA. ibeA encodes brain-microvascular-endothelial-cell invasion protein A, the primary adhesin/invasin that initiates contact with and entry into cerebral microvascular endothelial cells. ibeB and ibeC encode accessory invasion proteins that act in concert with IbeA to promote transcytosis. ompA encodes outer-membrane protein A, which further enhances adherence to and penetration of the blood–brain barrier.³⁹

NMEC strains also harbor the core ExPEC virulence repertoire—fimH, hlyA, cnf1, and an array of iron-acquisition genes such as iutA and fyuA. Acting in concert, these factors augment adhesion, cytotoxicity, inflammation and nutrient scavenging, thereby potentiating the overall pathogenicity of NMEC.

3.3.3. Target genes for APEC assays. APEC molecular detection panels typically target a core set of virulence-associated genes: iron uptake-related genes (e.g., iutA, iucD, and iroC), iss (encoding protectins), outer membrane protease gene (ompT), fimbrial usher chaperone genes (e.g., fimC and papC), hlyF (haemolysin genes), and wzx (O-antigen flipping enzyme gene for identification of O78 serotype strains).^40,41 Accessory markers—tsh (temperature-sensitive haemagglutinin), cvaC (colicin V synthesis), and vat (vacuolating autotransporter toxin)—are also frequently included as supportive markers.⁴² Collectively, these loci are routinely employed to identify APEC strains and to differentiate them from non-pathogenic or other PEC.

3.3.4. Target genes for SEPEC assays. SEPEC are sepsis-causing E. coli with highly heterogeneous virulence profiles, usually a combination of multiple ExPEC virulence genes. SEPEC strains typically carry a range of virulence genes that contribute to their ability to survive and multiply in the bloodstream and resist host defense mechanisms. These genes overlap considerably with the virulence genes of UPEC and NMEC.³⁹

Iron acquisition system genes, such as iutA, iroN, and fyuA, are essential for iron acquisition in an iron-restricted blood environment. The kpsMT II gene cluster in PEC synthesizes group II capsule polysaccharides, which help bacteria resist the host's immune response.⁴³ Although adhesion in the blood may be less important than on mucosal surfaces, certain adhesins may be involved in bacterial interactions with vascular endothelial cells or thrombosis. Toxin genes include hlyA, cnf1, cnf2 (cytotoxic necrosis factor 2, found mainly in animal-derived SEPEC).³⁹ Genes associated with serum resistance, such as traT (outer membrane protein), have been identified as playing a role in counteracting complement-mediated killing.

Given the heterogeneity of SEPEC, a universal molecular marker has yet to be identified. The identification of ExPEC is typically contingent upon the detection of a core set of virulence gene profiles associated with its pathogenicity. Whole genome sequencing has demonstrated an increasing capacity for the identification and characterization of ExPEC.

4. Molecular methods

The initial step in this process is the precise selection of target sequences. The sensitivity, specificity, and field applicability of detection are ultimately contingent upon the molecular technology platform utilized. Over the past decade (2015–2025), as the demand for foodborne pathogen monitoring and the concept of point-of-care testing (POCT) have evolved in tandem, E. coli detection has rapidly transitioned from traditional culture-based biochemical identification to a diverse array of “molecular-smart-miniaturized” technologies. The following sections will systematically review the six major technological trends driving this transformation: hybridization techniques, classical amplification systems represented by PCR and its derivatives, isothermal amplification strategies (e.g., loop-mediated isothermal amplification), biosensors integrating nanomaterials or optoelectronic conversion, integrated devices, and emerging frontier technologies such as clustered regularly interspaced short palindromic repeats (CRISPR). The collective integration of diverse technical approaches, encompassing sensitivity, speed, cost, and scenario adaptability, is propelling E. coli detection into a new era of “single-cell, minute-level, and visually readable” testing.

4.1. Hybridization techniques

Nucleic acid hybridization is one of the fundamental techniques in molecular diagnostics, the core principle of which is the specific binding of a labeled nucleic acid probe to a target nucleic acid sequence (DNA or RNA) in the sample.⁴⁴ The advantages of these methods are the in situ analysis, the preservation of the spatial structure of the sample, and the ease of visualization, which is ensured by the precise design of the probe sequence. This section focuses on hybridization techniques that do not rely on or are dependent on non-polymerase amplification, demonstrating their evolutionary path from classical in situ intracellular imaging to exponential signal amplification using cascade reactions.

4.1.1. Fluorescence in situ hybridization. Fluorescence in situ hybridization (FISH) is a technique that allows for microscopic visualization, identification, and enumeration of specific microorganisms. This method utilizes fluorescently labeled oligonucleotide probes, which bind to highly conserved and abundant ribosomal RNAs (rRNAs) directly in intact, physiologically active, immobilized cells.⁴⁵ This technique does not necessitate culture and furnishes morphological and spatial distribution information at the single-cell level.⁴⁶

In recent years, a variety of enhanced FISH techniques have been developed to address the limitations of conventional FISH, such as cumbersome operation and inadequate signal strength. The advent of microfluidics has led to significant advancements in the automation and high-throughput capabilities of the FISH process.^47,48 For instance, a microfluidic one-pot FISH method involves the division of the sample into a multitude of picoliter droplets, thereby facilitating the quantification of single-cell-sensitive bacteria within a dynamic range spanning four logarithms (from 3 × 10³ to 3 × 10⁷ bacteria per mL). This approach significantly reduces the total assay time to approximately 1.5 hours.⁴⁸ Conversely, the integration of flow cytometry with FISH facilitates the expeditious evaluation of a substantial cell count, exhibiting comparable sensitivity to conventional plate counts in differentiating between E. coli and non-E. coli strains.^49,50

Innovations in probe chemistry have emerged as a pivotal factor in enhancing the performance of FISH. Peptide nucleic acid (PNA) and locked nucleic acid (LNA), as artificial nucleic acid analogs, exhibit a substantially higher binding affinity and specificity to targets in comparison with conventional DNA probes (T_m increase per monomer: PNA ≈ +1 °C, LNA +3–9.6 °C; single-base discrimination ΔT_m > 5 °C).^51,52 This enhanced performance can be attributed to the unique characteristics of their neutral backbone or locked conformation. Research has demonstrated that the optimization of PNA-FISH parameters (e.g., pH and probe concentration) is imperative for the effective detection of Gram-positive and negative bacteria. Probe concentrations greater than 300 nM have been shown to be generally beneficial for both groups.⁵³ For LNA probes, high salt concentrations (2–5 M NaCl) and optimized hybridization temperatures (62 °C for probes with 64% GC) significantly enhance the signal.⁵⁴ Furthermore, multiplex techniques, such as FISH combined with spectral imaging and nucleic acid mimics, leverage multiple fluorescent dyes to differentiate up to seven distinct bacterial species concurrently.⁵⁵

4.1.2. Hybridization chain reaction. The hybridization chain reaction (HCR) has been identified as a potent isothermal, enzyme-free signal amplification technique. The fundamental mechanism entails the utilization of the target nucleic acid (DNA or RNA) as a “trigger” for two stable “hairpin” probes to sequentially open and hybridize with each other, thereby forming a long, repeating double-stranded DNA (dsDNA) polymer.⁵⁶ Each target molecule catalyzes the formation of a long strand, thereby transforming a single binding event into an exponentially amplified signal that can be easily detected.

A salient benefit of HCR is its adaptability in signal output, encompassing colorimetric, fluorescence, and electrical signals, and lateral flow assays (LFAs), thus catering to diverse assay requirements ranging from laboratory to field settings.^57–61 Furthermore, when HCR is employed in conjunction with immunoassay techniques, such as enzyme-linked immunosorbent assay (ELISA), its signal amplification effect is further enhanced. A study demonstrates that the detection limit of a sandwich ELISA based on HCR signal amplification is 185 times lower than that of a conventional ELISA, reaching 108 CFU mL⁻¹ (Fig. 1).⁶²

Fig. 1 Schematic representation of E. coli O157:H7 detection using DNA-based HCR and biotin–streptavidin amplification via ELISA, adapted from ref. 62 with permission from Elsevier (Guo, et al., Biosens. Bioelectron., 2016, 86, 990–995), copyright 2016. The terms “Bio-H1” and “Bio-H2” refer to biotinylated DNA probes. The detection signal is generated via a colorimetric reaction involving horseradish peroxidase (HRP) and its substrate/chromogen.⁶²

Another advantage of HCR is its ability to achieve ultra-high sensitivity when utilized in conjunction with other technologies. For instance, the combination of dsDNA templated copper nanoparticles (CuNPs) reduces the detection limit to 0.003 CFU mL⁻¹, and an electrochemiluminescence-based sensor has a detection limit of 38 CFU mL⁻¹.^63,64 The fluorescence-based HCR platform has been documented as having a detection limit as low as 0.49 CFU mL⁻¹ for E. coli, with a linear range extending up to 7 orders of magnitude wider.⁶⁵ This high sensitivity renders it a promising candidate for direct detection without culture.

4.1.3. Other innovative strategies. In addition to FISH and HCR, researchers have developed a variety of other innovative nucleic acid hybridization strategies, which often revolve around novel materials and signaling mechanisms.

The popularity of visualization assays can be attributed to several factors, including their cost-effectiveness, ease of operation, and the ability to visually interpret the results. These methods are predominantly predicated on the color change of gold nanoparticles (AuNPs). Under specific conditions, the presence of target DNA has been shown to either prevent or trigger the aggregation of AuNPs, resulting in a change in solution color from red to blue. The methods boast an exceptionally short detection time, usually completing within 30 minutes after genome extraction, while achieving a detection limit as low as 10–10³ CFU mL⁻¹.^66,67 A recent study has estimated that the total material cost per assay is approximately $0.69, suggesting significant scalability.⁶⁸

In addition, a range of advanced fluorescence techniques have been developed to facilitate rapid identification of E. coli. In the realm of probe technology, optimized double-stranded PNA probes have been shown to exhibit superior fluorescence signals and analytical specificity in comparison withs conventional PNA beacons.⁶⁹ The streptavidin–luciferase fusion protein exhibits a linear detection range of 10⁻¹⁸ to 10⁻¹³ mol per well.⁷⁰ The employment of nuclease-responsive DNA probes facilitates the expeditious detection of live bacteria (10³ CFU for E. coli) within one minute.⁷¹ Innovative approaches are increasingly drawing on the unique properties of advanced nanomaterials (e.g., AuNPs,⁷² carbon quantum dots (QDs),⁷³ graphene QDs,⁷⁴ upconversion nanoparticles (UCNPs),⁷⁵ graphene oxide (GO),⁷⁶ Au@Ag nanorods,⁷⁷ and carbon nanodots⁷⁸). These technologies have been shown to enhance the sensitivity, speed, and specificity of bacterial detection through a combination of probe design, nanomaterial integration, and signal amplification strategies.

4.2. PCR-based methods

PCR represents a revolutionary technological advancement in the domain of molecular biology, enabling the sensitive detection of trace quantities of nucleic acids through the exponential amplification of specific DNA fragments within a laboratory setting. As the “gold standard” of molecular diagnostics, PCR assays have evolved from simple qualitative detection, through high-throughput multiplex analysis and precise real-time quantification, to absolute quantification achieved without standard curves—providing a powerful and versatile toolbox for E. coli detection.

4.2.1. Multiplex PCR. The primary benefit of multiplex PCR (mPCR) is its high-throughput capacity, which enables the concurrent amplification of multiple distinct target sequences in a single reaction. This is of particular value in scenarios where the simultaneous detection of multiple pathogens, the identification of strain serotypes, or the screening of multiple virulence genes is required. Such applications significantly enhance detection efficiency, leading to substantial saving of both time and cost resources.⁷⁹ For instance, the hydrolysis probe assay designed for PEC serotypes (O26, O103, O111, and O121) achieves 94.8% virulence prediction accuracy by targeting single nucleotide polymorphisms.⁸⁰ Multiplex PCR, based on species-conserved genes (e.g., the cdgR gene in E. coli and the EAKF1_ch4033 gene in E. albertii), has been shown to differentiate E. coli, E. albertii, and E. fergusonii with 100% accuracy.⁸¹

The key to mPCR is primer design and target selection. Researchers have developed gene combinations for E. coli species identification (e.g., cydA, lacY, and ydiV),⁸² specific targets for the typing of the STEC serogroup (O26, O45, O103, O104, O111, O121, O145, and O157),⁸³ and primer systems for distinguishing between pathogens with close affinity (e.g., Shigella vs. EIEC).⁸⁴ The performance of these methods has been outstanding, e.g., triple-gene-based mPCR has 99.49% specificity for the identification of E. coli, and multiplex oligonucleotide ligation-PCR has more than 90% analytical specificity.^82,83 An important technological advancement is the combination of mPCR with live bacteria detection techniques. Propidium monoazide (PMA) and its analogs selectively enter membrane-compromised dead cells; upon light exposure, they covalently cross-link DNA and thus prevent downstream amplification, allowing only viable bacteria to be detected.⁸⁵ For live bacteria with intact cell membranes, PMA is inaccessible. This PMA-mPCR technology is particularly important in food safety testing. For example, a study has achieved a detection limit as low as 1 CFU g⁻¹ in complex matrices like cantaloupe after 6 h enrichment.⁸⁶

To further enhance the throughput and speed of mPCR, a variety of innovative platforms have emerged. For instance, cassette PCR utilizes an aluminum cassette device containing multiple capillaries to detect 10 targets in parallel within 75 minutes.⁸⁷ The multiplex PCR capillary electrophoresis method can simultaneously detect 13 pathogens associated with bacterial pneumonia.⁸⁸ The microarray-based MeltArray even enables 62-fold analysis of E. coli serotypes.⁸⁹ In addition, innovations in primer design strategies have also contributed greatly to the efficiency improvement. For example, the cliffhanger PCR successfully reduces the total turnaround time of laboratory samples from 20 hours to 6 hours by optimizing the primer structure, underscoring the pivotal role of molecular engineering in assay optimization (Fig. 2).⁹⁰

Fig. 2 Schematic diagram of a cliffhanger PCR primer with a Z–X–Z–S structure, adapted from ref. 90 (Schneider, et al., 2018) under a Creative Commons CC BY license. The 5′ ortho-TINA molecule protects the oligonucleotide from 5′ to 3′ exonuclease activity, while the internally placed TINA molecule blocks DNA polymerase, thereby creating a single-stranded DNA overhang on the PCR amplicon.⁹⁰

4.2.2. Quantitative PCR. Quantitative PCR (qPCR), also known as real-time PCR, is a major leap forward in PCR technology. The core principle is to introduce a fluorescent reporter group into the conventional PCR system to quantify the copy number of the starting template by monitoring the accumulation of fluorescence signals during the reaction in real time.⁹¹ Commonly, qPCR methods include the TaqMan probe method (luminescence using the separation of the fluorescent group from the quenching group after the probe is degraded) and the SYBR Green I dye method (luminescence by non-specific embedding of double-stranded DNA). The fluorescence intensity is proportional to PCR product abundance and the starting template amount in unknown samples can be accurately quantified by recording the threshold cycle at which the signal crosses a preset threshold.

Known for its high sensitivity and specificity, qPCR is widely recognized as the “gold standard” for nucleic acid quantification. In particular, in the quantification of live bacteria with real infection risk, various qPCR techniques have become powerful methods to distinguish live bacteria from dead ones. Compared to ethidium monoazide (EMA), PMA is a nucleic acid dye that is more commonly used in the detection of live E. coli.^92,93 By fine optimization of parameters such as PMA concentration (e.g., 50 μM) and light conditions (e.g., 650 W, 15 min), the amplification of DNA from dead bacteria can be inhibited to the maximum extent.⁹⁴ The data show that the difference in the cycle of quantification values (ΔCq) between PMA-treated heat-killed cells and untreated dead cells could reach 8.9–9.99, enabling effective discrimination of live bacterial signals.⁹⁵ In addition, phage-based lysis combined with qPCR, as well as qPCR methods targeting RNA (e.g., 16S rRNA, 23S rRNA), have shown promising utility in detecting live coliforms.^96–99

In addition, qPCR is also of great value in multiplex assays. The TaqMan qPCR method for five diarrheal E. coli (EPEC, EIEC, ETEC, EHEC, and EAEC) shows an analytical sensitivity of 1.6 × 10¹–1.6 × 10² copies per μL (R² > 0.999).¹⁰⁰ The quadruplex TaqMan PCR allows simultaneous detection of four pathogens, including E. coli, with amplification efficiencies of 81.81–94.84%.¹⁰¹ The detection limit of multiplex qPCR combined with magnetic capture hybridization for E. coli O157, Salmonella and Listeria monocytogenes is down to 1–10 CFU.¹⁰² Interestingly, multiplex qPCR based on melting-curve analysis (T_m: 82–93 °C) and high-throughput qPCR capable of simultaneously detecting 68 pathogen markers (sensitivity > 80%, specificity > 99%) significantly expand the pathogen screening capacity, providing efficient and accurate support for food-safety monitoring and clinical diagnosis.^103,104 The high-resolution melting curve (HRM)-qPCR technology is able to distinguish DNA sequences with only a single base difference, thus realizing rapid typing and relative quantification of different genotypes (e.g., different fimbriae genes of ETEC).^105,106

To meet the demand for rapid on-site detection, qPCR is being miniaturized and accelerated. A microfluidic qPCR system attains ultrafast thermal cycling by maximizing the surface-to-volume ratio, cutting the reaction time from 1–2 h to just 7 min without sacrificing sensitivity.¹⁰⁷

4.2.3. Digital PCR. Digital PCR (dPCR) represents the third generation of nucleic acid quantification, which shifts the quantification paradigm from signal intensity to absolute counting. The underlying principle is partitioning followed by Poisson-based counting of positive partitions.¹⁰⁸ Prior to the initiation of the reaction, the sample containing the target nucleic acid is subjected to extensive dilution and distribution into thousands of individual reaction units of microliter or even nanoliter volume. The Poisson distribution principle guarantees that each reaction unit contains at most one or zero target molecules. Subsequently, the PCR is performed independently in each unit. At the conclusion of the reaction, the number of positive reaction units and negative reaction units is determined by the presence or absence of fluorescence signals. The absolute copy number of the target molecule in the original sample is calculated directly from the proportion of positive units and the total number of reaction units.

The foremost advantage of dPCR is its ability to provide absolute quantification, eliminating errors arising from variable amplification efficiency and standard-curve uncertainty. As a result, dPCR delivers higher accuracy and reproducibility, while its single-molecule sensitivity enables detection at the attomolar level or below. Research has demonstrated that dPCR possesses the capability to detect as few as 1–2 bacteria or fungi in a reaction, exhibiting an analytical sensitivity that is more than 10 times superior to that of qPCR.^109,110 This property confers an unrivaled advantage in the detection of rare targets, such as the precise quantification of low-abundance pathogens directly from complex food matrices, including apple juice or clinical samples (e.g., whole blood), without the necessity for laborious enrichment steps.^111,112 Furthermore, multiplex dPCR techniques are becoming increasingly sophisticated. One such technique is a dPCR assay for E. coli O157:H7 that uses seven-gene markers (e.g., stx1/stx2) and achieves a wide detection range of 6.6–7900 copies per μL with a detection limit as low as 0.27 copies per μL.²⁴

The unique capabilities of dPCR have spawned cutting-edge applications. In single-cell analysis, dPCR has been used to pinpoint virulence genes within individual STEC strains, enabling same-day confirmation—crucial for pathogen traceability and risk assessment.¹¹³ Digital high-resolution melting further integrates absolute quantification with genotyping, delivering a robust platform for accurate microbial identification.¹¹⁴

4.2.4. Other novel PCR-based methods. The advent of rapid visualization assays, underpinned by lateral flow and paper-based platforms, has paved the way for the POCT of molecular diagnostics. This technological feat integrates the highly specific amplification of PCR with the portable nature of LFA, heralding a new era in medical diagnostics. The fundamental principle entails the utilization of biotin- or FAM-labeled PCR amplicons, which undergo migration by capillary action on test strips, bind to gold nanoparticle-labeled probes, and subsequently undergo specific capture and coloration (red) by test-line capture probes. For instance, the PCR-LFA can complete the test in less than 10 min.¹¹⁵ Another PCR-LFA assay achieves an ultra-fast detection (less than 20 min) through 14 min of convection PCR and 5 min of LFA.¹¹⁶ In addition, the immobilization of aminated DNA is accomplished through the functionalization of cellulose filter paper, and bacterial DNA detection is achieved on the surface of the activated paper using a 16S rDNA probe.¹¹⁷ In 2022, the team further developed a nucleic acid amplification-based pathogen detection tool for rapid identification of a wide range of DNAs in the field.¹¹⁸

The field of nanomaterial-enabled PCR has witnessed significant advancements in signal amplification and detection performance through the implementation of various innovative strategies. In the colorimetric strategy, AuNPs utilize their salt-induced aggregation color change property to maintain a dispersed state of red (a positive signal) after hybridizing with single-stranded DNA generated by asymmetric PCR. This method has been successfully applied to the detection of STEC, with results obtained within one hour (Fig. 3).¹¹⁹ A fluorescence sensing method employs the UCNPs in conjunction with the fluorescent dye Nile blue, thereby achieving signal output through dsDNA de-fluorescence quenching.¹²⁰ The catalytic activity mimicry strategy innovatively exploits the enzyme-like activity of CeO₂ NPs, whose inhibition by DNA allows for indirect quantification of nucleic acids within 5 minutes using a glucose meter.¹²¹ Polydopamine nanospheres have been shown to enhance polymerase activity, thereby enabling a 10-fold increase in the detection limit of direct PCR in complex matrices.¹²² The magnetic gene sensing method has been demonstrated to provide rapid and sensitive detection of three bacterial gene fragments by amplifying them with specific primers and combining them with silica MNPs.¹²³ The GO-based method, which utilizes the binding properties of GO to single-stranded DNA and achieves sequence-specific identification by hybridization with fluorescent probes, is both simple and highly specific.¹²⁴ It is noteworthy that platinum compounds exhibit the capacity to selectively permeate the cell membrane of dead microorganisms and bind to DNA, enabling the differentiation of live E. coli and Cronobacter sakazakii in milk at 5–10 CFU mL⁻¹ within 4 hours through the implementation of platinum-PCR.¹²⁵ The integration of these techniques has propelled the evolution of PCR assays towards a paradigm of heightened sensitivity, expedited analysis, and multifaceted dimensionality.

Fig. 3 Schematic illustration of bacterial DNA detection, adapted from ref. 119 with permission from the Royal Society of Chemistry. The process of DNA sandwich hybridization entails the binding of the AuNP probe (probes 1 and 2) to the target DNA. As the salt concentration is increased, the stability of the AuNP probe and target DNA complexes is maintained, and their original red coloration is retained. In the absence of target DNA, AuNP aggregation gives the reaction solution a purple color.¹¹⁹

The utilization of portable, high-throughput assays is instrumental in enhancing diagnostic efficiency, reducing expenses, and acquiring comprehensive information. Continuous-flow PCR employs “S”-shaped microchannels through which the sample cycles among different temperature zones, amplifying target genes from E. coli within 10 min.¹²⁶ The on-chip detection module uses microcapillary electrophoresis to separate PCR products of three food-borne pathogens and eliminate false-positive interference within only 135 seconds.¹²⁷ In the context of high-throughput array research, the hydrogel DNA array employs a “one-tube” design, facilitating the sensitive detection of 10 cells within 1.5 hours.¹²⁸ The DNA microarray possesses the capacity to concurrently identify 15 distinct pathogens through mPCR and probe hybridization, with a detection limit of 10³ copies per μL.¹²⁹ These technological breakthroughs have enabled the microfluidic platform to combine rapid response (as fast as 10 minutes), ultra-high sensitivity (single-cell level), and multi-target parallel detection, providing a revolutionary solution for on-site diagnostics.

4.3. Loop-mediated isothermal amplification (LAMP)

4.3.1. Conventional LAMP. LAMP is an exceptionally efficient nucleic acid amplification technique.¹³⁰ It is predicated on the utilization of a set of four to six specially designed primers that recognize six to eight different regions of the target DNA sequence. In the presence of a DNA polymerase with strand-substitution activity (e.g., Bst DNA polymerase), the reaction system spontaneously initiates a complex series of strand-substitution and ring structure-mediated DNA synthesis reactions at a constant temperature (typically 60–65 °C). Eventually, the target sequence is amplified exponentially, producing a large amount of DNA product containing multiple inversely repeating loop structures. This substantial quantity of product serves as the foundation for the efficacy of LAMP technology in detecting the presence of the target analyte.

The performance of LAMP is characterized by its exceptional sensitivity and remarkable speed. A multitude of studies have demonstrated that the sensitivity of the aforementioned method significantly surpasses that of conventional PCR, with the capability of detecting samples at a single-copy level.^131,132 The rapidity of the reaction is equally remarkable, typically occurring within 15 to 60 minutes. A particular salient feature of the study is the wide array of methods available for interpreting the results. The precipitation of magnesium pyrophosphate, a byproduct of amplification, has been observed to induce turbidity changes that are discernible to the naked eye. A variety of indicators have been reported for use in fluorescence or colorimetric analysis, including fluorescent dyes (e.g., calcein,¹³³ SYBR Green I,¹³⁴ and SYBR™ Safe¹³⁵), metal ion-specific indicators (e.g., hydroxy naphthol blue¹³⁶ and Eriochrome Black T¹³⁷), and pH-dependent dyes (e.g., xylenol orange¹³⁸) (Table 1). Furthermore, the utilization of visualization in conjunction with LFA has been demonstrated to enhance the efficacy of the process.¹³⁹

Table 1 Indicators used to detect E. coli in the reported LAMP methods

Indicators	Principles	Signal readouts	Ref.
Calcein	As the DNA is amplified, Mn^2+ in the reaction system releases from the calcein and binds to the pyrophosphate (PPi), and the free calcein autofluoresces	Green (+), orange (−)	133
SYBR Green I	In its free state, SYBR Green I exhibits a low level of fluorescence. However, upon binding to double-stranded DNA, there is a substantial increase in fluorescence intensity	Green (+), orange (−)	134
SYBR™ Safe	When SYBR™ Safe binds to double-stranded DNA, the fluorescence intensity is significantly increased	Fluorescence (+), non-fluorescence (−)	135
Hydroxynaphthol blue (HNB)	As the reaction progresses, Mg^2+ reacts with PPi. Consequently, the HNB loses the magnesium ions, resulting in the system's conversion to a shade of azure. In contrast, the unreacted system maintains its original violet color	Blue (+), violet (−)	136
Eriochrome Black T	The depletion of free Mg^2+ by nucleic acid amplification leads to the dissociation of the Eriochrome Black T-metal complex, resulting in a shift of the reaction solution from red to blue	Blue (+), red (−)	137
Xylenol orange	The release of H^+ during amplification lowers the pH from ∼8.5 to 6.0–6.5, changing the xylenol orange color from purple to orange/yellow	Yellow (+), purple (−)	138

---

Recent advances in LAMP have enhanced PEC detection. PMA-LAMP rapidly identifies viable EHEC, Shiga toxins, and E. coli O157:H7 by targeting the rfbE, stx1/2, and wzy genes, distinguishing viable but non-culturable organisms.^140,141 Technologically, the isotachophoresis–LAMP chip integrates nucleic-acid extraction and amplification, finishing both steps within one hour.¹⁴² Furthermore, fluorescent copper nanoclusters synthesized using LAMP amplicons enable low-background signal detection.¹⁴³ The 2-deoxyadenosine-5-(α-thio)-triphosphate (dATPαS) substitution strategy combined with UDG treatment reduced background noise, achieving a detection limit of 5 CFU mL⁻¹.¹⁴⁴ UDG-assisted LAMP also mitigated false positives by excising uracil-containing contaminants, increasing sensitivity and specificity to 95%.¹⁴⁵

4.3.2. Multiplex LAMP. Multiplex LAMP technology allows for the rapid, sensitive, and specific detection of multiple pathogens in a single reaction, greatly enhancing the diagnostic efficiency and accuracy. It is particularly suitable for rapid diagnosis, public health surveillance, and clinical applications.

The MAST ISOPLEX® VTEC kit employs a triple LAMP approach with fluorescent probes to simultaneously identify stx1, stx2, and internal controls in VTEC, achieving 100 copies per test sensitivity and 100% specificity.¹⁴⁶ An extraction-free multiplex LAMP assay detects eight common respiratory pathogens within one hour, showing 94.49% specificity and 75% sensitivity compared to next-generation sequencing (NGS).¹⁴⁷ Another protocol identifies five prosthetic joint infection-related bacteria (including E. coli) with a detection limit of 10³–10⁶ CFU mL⁻¹ within 1 hour.¹⁴⁸ A foldable microdevice has been developed that integrates LAMP with colorimetric readout. It employs 2-hydroxyethyl agarose to stabilize reagents for up to 45 days and uses a graphene heater for rapid amplification, enabling detection of E. coli O157:H7 at concentrations as low as 2.5 × 10² copies per mL.¹⁴⁹

4.3.3. Digital LAMP. Recent advances in digital LAMP (dLAMP) have substantially improved pathogen detection capabilities. A commercial membrane dLAMP system achieves absolute DNA quantification of 1.1 × 10¹–1.1 × 10⁵ copies per μL through a nanopore reactor with a novel probe, which enables up to a 100-fold difference in fluorescence between negative and positive wells at a cost of less than $0.1 for a single test.¹⁵⁰ The pumpless microfluidic dual-droplet dLAMP chip employs 64-nozzle step-emulsification to generate tens of thousands of droplets (variance < 5%) and uses fluorinated oil to suppress evaporation, enabling simultaneous dual-target detection of E. coli at 19.8 copies per μL.¹⁵¹ The microdroplet injection platform integrates a molecular beacon to achieve a detection limit of 9 copies per μL in plasmids containing the malB gene, and the whole process takes less than 1.5 hours.¹⁵² The nanoporous hydrogel-based interfacial dLAMP system can complete 100 mL water sample detection (0.09–900 cells per mL) in 30 minutes.¹⁵³ In addition, the cross-linked PEG hydrogel has multiple functions such as adsorption and release, supporting the direct detection of single-cell level pathogens in fresh food within 20 minutes, with significantly better anti-interference ability than the traditional droplet dLAMP.¹⁵⁴

4.3.4. Improved LAMP. Significant advancements have been made in the field of LAMP detection technology, particularly in the realm of probe- and molecular beacon-based methods. The isothermal transcriptional amplification technique based on an aptamer-converting probe and T7 RNA polymerase achieves a detection limit of 73.2 CFU mL⁻¹ for E. coli and allowed precise quantification of E. coli O157:H7 in milk.¹⁵⁵ The colorimetric LAMP method is able to detect E. coli O157:H7 at 10 CFU g⁻¹ in beef samples within 1 hour by optimizing molecular beacon design targeting stx1/stx2 genes.¹⁵⁶

In addition, the portable nucleic acid detection platform based on LAMP technology has also made significant breakthroughs, with multiple innovative methods demonstrating excellent on-site detection performance. Specific binding of the LAMP product to the LFA is achieved using a primer biotinylated at its 5′ end, which yields a visible blue signal.¹⁵⁷ The distance-based paper analytical device utilizes the charge reaction of HNB with polyethylenimine to semi-quantify DNA by ribbon length in 5 minutes.¹⁵⁸ The rapid LAMP-based diagnostic test uses freeze-dried reaction strips to detect ETEC in fecal samples in 50 minutes, and eliminates human bias when combined with a hand-held reader.¹⁵⁹ A paper-based platform integrates LAMP with the CRISPR-Cas12a system through the smartphone, which reads fluorescence signals to specifically detect 10³ copies per μL of E. coli in soil.¹⁶⁰ The innovative powerless paper-based device utilizes the principle of polydopamine aggregation inhibition to achieve naked-eye detection of pathogens such as E. coli O157:H7 in 25–35 minutes.¹⁶¹ Interestingly, a modification-free LFA for the specific hybridization system uses oligonucleotides to recognize the single-stranded loop region of the LAMP product, which in combination with AuNP probes can be used for the detection of E. coli O157:H7 at 10 CFU mL⁻¹ in food.¹⁶²

Furthermore, several innovative breakthroughs for E. coli detection by LAMP have also emerged in recent years. A carbon-black-PDMS (carbon black-poly(dimethylsiloxane)) paper-based photothermal platform utilizes the laser-excited photothermal effect to detect E. coli O157:H7 in 15 min (Fig. 4a).¹⁶³ A nanoporous PEG hydrogel achieves direct amplification of pathogens in complex matrices (whole blood, milk, tea, etc.) through self-cleaning properties, achieving single-cell-level detection in 20 minutes.¹⁶⁴ Coupling LAMP with catalytic hairpin assembly (CHA) and a personal glucometer enhanced the signal-to-noise ratio by 12.5-fold via a TmINV-catalyzed three-way junction process.¹⁶⁵ In another approach, target nucleic acids inhibit the interaction between cationic carbon dots and AuNPs, enabling visual detection of E. coli through AuNP dispersion (Fig. 4b).¹⁶⁶ The complementary-metal-oxide semiconductor (CMOS) image sensor is capable of detecting targets as low as 10 fg μL⁻¹ in 45 minutes through real-time monitoring of optical signal changes.¹⁶⁷ The rotary diffusion method employed Janus microbeads to measure solution viscosity changes in only 10 min using 2 μL of sample.¹⁶⁸ Furthermore, the combination of the peroxidase-mimicking G-quadruplex DNAzyme has been demonstrated to achieve excellent colorimetric detection of LAMP amplicons.^169,170

Fig. 4 Improved LAMP methods for the detection of E. coli. (a) A carbon-black-PDMS-paper hybrid device for DNA amplification in POCT, adapted from ref. 163 with permission from the Royal Society of Chemistry.¹⁶³ (b) Carbon-dot-triggered aggregation/dispersion of AuNPs for colorimetric detection of E. coli DNA, adapted from ref. 166 with permission from the Royal Society of Chemistry.¹⁶⁶

4.4. Recombinase polymerase amplification (RPA)

Recombinase polymerase amplification (RPA) is a nucleic acid amplification technique that mimics the process of DNA homologous recombination repair in living organisms.¹⁷¹ The core mechanism relies on the synergistic action of three key proteins: first, the recombinase binds to a primer to form a protein–DNA complex that searches for homologous sequences in the template DNA double strand. Subsequently, with the assistance of a single-stranded DNA binding protein (SSB), the DNA double strand is locally unwound, allowing the primer to bind to the template strand. In RPA, SSB binds and stabilizes the displaced single strand, preventing re-annealing and enabling the polymerase to extend continuously. Finally, DNA polymerase with strand displacement activity extends from the end of the primer to synthesize a new DNA strand. The whole process takes place at a low constant temperature (usually 37–42 °C, close to body temperature) without the need for a heat denaturation step.

One of the advantages of RPA is its exceptional processing speed, which is unparalleled by other methods. The majority of RPA reactions are typically completed within a time frame of 8 to 25 minutes.^172,173 A recent innovation in technology, dubbed PACRAT, has demonstrated the capability to attain single-cell detection limits in a span of less than 10 minutes.¹⁷⁴ In terms of performance, RPA is comparable to qPCR, exhibiting a detection limit of 4–5 CFU mL⁻¹ and 100% specificity against non-target strains, thus demonstrating excellent sensitivity and specificity.¹⁷⁵ Due to its “body-temperature” amplification feature and ultra-fast kinetics, the reaction can be carried out with a simple heating device (even hand-held), which makes it extremely suitable for POCT and for integration into portable devices.¹⁷⁶

The combination of RPA with LFA is a prevalent strategy in the realm of RPA for POCT, also referred to as RPA-LFA. The labeling of primers, such as FAM and biotin, enables the specific capture and coloration of amplification products on the test strips. This results in assay outcomes that are readily discernible to the naked eye and is typically completed within 25 minutes.¹⁷⁷ The model has been extensively utilized in the domains of food safety and clinical diagnostics.^178–182 For instance, in the detection of lettuce samples, the sensitivity is up to 85%, and the specificity is 100%.¹⁸³ Furthermore, the integration of RPA into centrifugal chips, paper chips, and wearable microdevices has been demonstrated to be effective for automation and high throughput, facilitating the simultaneous and automated detection of multiple pathogens.^184–186

4.5. Rolling circle amplification (RCA)

RCA utilizes a ring-shaped DNA template and DNA polymerase to continuously synthesize a long single-stranded DNA containing hundreds or thousands of repetitive sequences to achieve linear or exponential signal amplification.¹⁸⁷

In recent years, RCA-based molecular technologies have made significant progress in the rapid detection of E. coli. The researchers developed a novel visualized photothermal smartphone biosensor, a system that initiates the RCA reaction by E. coli-specific aptamer recognition and binding to magnetic beads (MBs), and the RCA product released by UV cleavage hybridizes with a near-infrared-excited Cu_xS-DNA probe to achieve a wide linear range of 5–500 000 CFU mL⁻¹ for ultra-sensitive detection (1.8 CFU mL⁻¹).¹⁸⁸ The integration of microfluidics with RCA facilitates the digital quantification of nucleic acids across a dynamic range of 1.2 aM to 190 fM.¹⁸⁹ The utilization of magnetic nanoparticle (MNP) aggregation with the hyperbranched RCA product enables visualization by the naked eye, facilitating the detection of five foodborne E. coli strains with a limit of detection as low as 100 CFU mL⁻¹.¹⁹⁰ The present study utilizes a Blu-ray optical pickup unit to detect the kinetic changes of MBs in an oscillating magnetic field, enabling high-performance and low-cost quantitative detection with a detection limit of 10 pM.¹⁹¹ The DNA–AgNCs/GO fluorescent system demonstrates a detection sensitivity of 38 CFU mL⁻¹ for both E. coli and S. aureus through the mechanism of RCA amplification and ssDNA aptamer release.¹⁹² Furthermore, the specific aggregation of MNPs to recognize the RCA amplicon can be facilitated by visual observation or photomagnetic measurements, enabling the expeditious and precise determination of detection outcomes.^190,193

These technological breakthroughs reduce testing time from days to 4 hours. Combined with smartphones, microfluidics, and optics, these technologies provide high-sensitivity, low-cost solutions for POCT, demonstrating significant application value in environmental monitoring and clinical diagnostics.

4.6. Other emerging isothermal amplification techniques

A range of other innovative isothermal amplification techniques are emerging, each with unique principles and advantages, further enriching the toolbox for rapid molecular diagnostics. Recombinase-aided amplification (RAA),¹⁹⁴ dual-priming (self-priming and pairing-priming) isothermal amplification,¹⁹⁵ cross-priming amplification (CPA),^196–198 isothermal multiple self-matching initiated amplification (IMSA),^196–198 RNase H-based isothermal exponential amplification (RH-IEA),¹⁹⁹ thermophilic helicase-dependent amplification (tHDA),²⁰⁰ nucleic acid sequence based amplification (NASBA),²⁰¹etc., also demonstrated their potential in specific applications (Table 2). For instance, a pretreatment method involving PMAxx in conjunction with real-time RAA is able to detect viable bacteria down to a concentration of 5.4 CFU mL⁻¹.¹⁹⁴ Furthermore, pure signal amplification strategies, such as nicking endonuclease signal amplification (NESA), have been shown to release a substantial number of signaling molecules through enzymatic cycles.²⁰² When coupled with highly sensitive capillary electrophoresis, the detection limit can be pushed to an ultra-low level of 2.5 fM (equivalent to 3 CFU mL⁻¹), which demonstrates the great potential of signal amplification strategies in lowering the detection limit.

Table 2 Emerging isothermal amplification techniques for the detection of E. coli

Methods	Signal readouts	Target bacteria	Target genes	Reaction temperature (°C)	Reaction time (min)	Detection limits (CFU mL^−1)	Detection/linear ranges (CFU mL^−1)	Ref.
a Original data expressed in copies per μL. b Original data expressed in mol L^−1.
RAA	Fluorescence	E. coli O157:H7	fliC	39	20	5.4 (pure culture), 7.9 (milk)	5.4 to 5.4 × 10^6 (pure culture), 7.9 to 7.9 × 10^6 (milk)	194
DAMP	Fluorescence	E. coli	malB	60	60	100 ^a	10^2 to 10^7 ^a	195
CPA	Fluorescence	ETEC	LT-I	60	60	50	5 × 10^1 to 1 × 10^6	198
CPA	Visual	ETEC	STa	63	60	1.5 × 10^3	1.5 × 10^3 to 1.5 × 10^8	196
CPA	Visual	ETEC	LT-II	62	90	50	5 × 10^1 to 5 × 10^6	197
IMSA	Visual	ETEC	LT-I	62	45	25	2.5 × 10^1 to 1 × 10^6	198
IMSA	Visual	ETEC	STa	61	45	1.5 × 10^2	1.5 × 10^2 to 1.5 × 10^8	196
IMSA	Visual	ETEC	LT-II	60	60	25	2.5 × 10^1 to 5 × 10^6	197
RH-IEA	Visual, fluorescence	E. coli O157:H7	16S rRNA	55	60	1 × 10^2 (milk)	1 × 10^2 to 1 × 10^6 (milk)	199
tHDA	Gel electrophoresis	E. coli O157:H7	eae	68	60	1 × 10^2	1 × 10^2 to 1 × 10^7	200
NASBA	Fluorescence	E. coli	16S rRNA	41	60	5–50	5 to 5 × 10^4	201
NESA	Laser-induced fluorescence assay	E. coli	Gn	37	60	3	1 × 10^−14 to 1 × 10^−11 ^b	202

---

4.7. Biosensors

The advent of novel materials, micro- and nano-fabrication methods, and information technology has precipitated a paradigm shift in the approach to the detection of E. coli. Conventional amplification techniques are being progressively superseded by a more integrated, intelligent, and informative approach. Biosensors represent a pivotal facet of this technological evolution, ushering in a new era in the domain of detection.

4.7.1. Electrochemical biosensors. Electrochemical sensors convert biorecognition events such as nucleic acid hybridization and aptamer-target binding into measurable electrical signals (e.g., current, voltage, or impedance) through careful design.²⁰³ Their core strengths are their extremely high sensitivity and easy miniaturization. Currently, the fusion of molecular technology and biosensors has become a cutting-edge paradigm with great application potential and widespread interest. Among them, nucleic acid hybridization (e.g., DNA probes^204–208 and PNA probes²⁰⁹) and amplification (e.g., RCA,^210–212 RPA,^213,214 strand displacement amplification (SDA),²¹⁵ and CHA²¹⁶) are the two dominant technological pathways, which provide a strong support for the highly sensitive detection of biosensors. When combined with the CRISPR/Cas12a system, the electrochemical sensor enables highly sensitive detection of E. coli (5–5.28 CFU mL⁻¹).^212,217,218

In the domain of biosensor innovation research, the integration and implementation of novel nanomaterials, including gold nanostars,²¹⁹ MoS₂,^220,221 zinc oxide (ZnO) nanorods and carboxylated graphene nanoflakes,²²² TiO₂,^223,224 and MoSe₂,²²⁵ have facilitated substantial enhancement in detection performance. Furthermore, a number of studies have employed strategies such as the use of DNA walkers,^210,226 aptamers,²²⁷ HRP signal amplification,²²⁸ and nanomachines²²⁹ to realize a novel mode of detection that does not require labeling, amplification, or signal cascade amplification.

4.7.2. Optical biosensors. Optical sensors utilize changes in optical phenomena such as fluorescence, colorimetry, and surface plasmon resonance (SPR) to report detection results.

In the domain of fluorescence strategies and innovative research, researchers have developed a variety of novel biosensors for rapid detection of E. coli. A biosensor composed of a single tube has been developed for the multiplexed detection of bacterial RNA.²³⁰ This biosensor utilizes a three-way junction nucleic acid structure and a fluorescent RNA aptamer that is encoded by the T7 promoter. An all-DNA biosensing system utilizes a four-way junction structure to convert the DNAzyme response into an amplified signal. The process of target recognition is initiated by the cleavage of DNAzyme by RNA, which subsequently triggers the assembly of a CHA and the generation of a G-quadruplex fluorescence signal.²³¹ The PCR system based on upconversion-AuNPs achieves a detection limit of 14 CFU mL⁻¹ through the fluorescence resonance energy transfer principle.²³² The 3D DNA walker system is an innovative apparatus that combines AuNP trajectories with CHA reaction to achieve a low detection limit (28.1 CFU mL⁻¹) through cascade signal amplification triggered by enzyme fragments.²³³ During bacterial recognition, the released walker strand via strand displacement reaction activates the DNA walker system. By utilizing AuNP tracks for signal anchoring and amplification, the cascade CHA reaction ultimately leads to a significant enhancement in the fluorescence signal and detection sensitivity. The magnetic biosensor employs a triple amplification mechanism of DNAzyme–RCA–CuNCs, utilizing MB separation and copper nanocluster luminescence properties to lower the detection limit to 1.57 CFU mL⁻¹.²³⁴ The Internet of Things (IoT)-integrated paper-based point-of-care system achieves simultaneous detection of multiple pathogens through LAMP and real-time fluorescence imaging.²³⁵ The chemiluminescent lateral flow biosensor employs CMOS imaging to detect 16S rRNA without amplification, providing a low-cost solution for POCT.²³⁶ Collectively, these technologies offer a multifaceted solution, ranging from highly sensitive laboratory testing to rapid diagnostics in field settings.

A variety of innovative sensing technologies with high sensitivity and versatility have been reported in the application and innovative research of colorimetric strategies. A sandwich aptamer sensor-based colorimetric detection method immobilizes split aptamer fragments by AuNPs and MBs, triggers RCA using terminal deoxynucleotidyl transferase, and ultimately realizes multi-analyte detection by HRP-catalyzed TMB (3,3′,5,5′-tetramethylbenzidine) color development.²³⁷ A CRISPR/dCas9-SERS biosensor integrates LAMP and smartphone platforms to guide the self-assembly of AuNPs to form SERS hotspots through dCas9 recognition of amplicon repeat sequences, and is capable of simultaneous detection of S. aureus, P. aeruginosa, and E. coli O157:H7 in 50 min.²³⁸ The DNA-programmed, dumbbell-shaped Au–Pt nanoparticles exhibit excellent performance via polyT20-mediated anisotropic growth: they combine a 95% bactericidal efficiency (5 min) and a detection limit of 2 CFU mL⁻¹ under light.²³⁹ In addition, the platform makes innovative use of functionalized nanoparticles and filamentous sugars to achieve rapid and sensitive detection of pathogens by triggering biochemical reactions and optical signal shifts upon changes in the binding of DNA amplicons.^240,241

4.7.3. Other innovative biosensors. The DNAzyme integrated plasmonic nanosensor employs bacteria-specific RNA-cleaving DNAzyme probes as recognition elements, thereby generating signals through the localized surface plasmon resonance (LSPR) of the quadrupolar plasmons on nanoparticles.²⁴² This assay enables visual detection of as few as 50 bacteria per mL in complex samples (e.g., milk, serum, and juice). Furthermore, the implementation of lubricant-infused surfaces has been demonstrated to facilitate a 4-fold enhancement in the signal-to-noise ratio, thereby enabling the detection of 250 CFU mL⁻¹ in milk.²⁴³

Another category of breakthrough technologies includes the hydrogel sensors and multifunctional DNA probe systems based on multiple strategies. The hydrogel sensor utilizes a target-induced DNAzyme cleavage mechanism to achieve visual detection, resulting in gel dissolution and the release of encapsulated gold nanoparticles.²⁴⁴ This sensor demonstrates a detection limit of 10 CFU mL⁻¹ in lake water samples and exhibits an accuracy of 96% true positives and 100% true negatives when integrated with an artificial intelligence (AI) model. A bimodal hydrogel biosensor based on DNA-modified phage probes for rapid screening and high-precision quantitative detection of E. coli O157:H7 by mutual verification of dual signals from fluorescence detection and microfluidic electrophoresis (Fig. 5).²⁴⁵ In addition, the utilization of a label-free DNA probe strategy (e.g., prob-polyA-probe) has been demonstrated to enhance the stability of the electrode surface through the use of polyA fragments.²⁴⁶ This approach exhibits a sensitivity of up to 10 fM, enabling the distinction of single nucleotide polymorphisms with a high degree of precision. Furthermore, the DNA-functionalized colloid has been shown to identify pathogens through the analysis of genomic repetitive sequences.²⁴⁷ This method facilitates the detection of a wide range of 10–10¹⁰ copies per mL without the requirement of amplification, thereby ensuring a high degree of specificity and applicability to real samples.

Fig. 5 A dual-mode hydrogel array based on a phage-DNA probe for the detection of E. coli, adapted from ref. 245 with permission from Elsevier (Anal. Chim. Acta, 2024, 1287, 342053), copyright 2024. Target recognition and the RCA process within the hydrogel involving phage-DNA probes and target bacteria. A microfluidic chip or fluorescence signal analysis can be used for signal readouts.²⁴⁵

Furthermore, novel methodologies are emerging in the field, including the use of metamaterial-assisted terahertz sensors,²⁴⁸ shear horizontal surface acoustic wave sensors,²⁴⁹ electrical double layer-gated field-effect transistor-based biosensors,²⁵⁰ thin-film transistor sensors,²⁵¹ machine learning-assisted custom cross-response sensing array sensors,²⁵² multichannel series piezoelectric quartz crystal sensors,²⁵³ and light-addressable potentiometric sensors.²⁵⁴ These strategies utilize novel approaches to capture the weak changes in molecular interactions from different dimensions, thereby opening up a new technological path for bio-detection. This demonstrates the significant innovative potential of cross-disciplinary technological fusion.

4.8. Genome sequencing technologies

4.8.1. NGS. Whole genome sequencing (WGS) is a highly sophisticated molecular biology technique that provides the ultimate molecular resolution for accurate strain characterization at the whole genome level. Its value is irreplaceable in the fields of epidemiological traceability, foodborne disease outbreak investigations, and drug resistance/toxicity gene profiling.^255,256 Research has demonstrated a high degree of consistency between WGS and epidemiologic findings regarding the distinction between VTEC outbreak-associated and non-associated strains.²⁵⁷ The advent of reduced sequencing costs and the maturation of bioinformatics analysis processes (e.g., GeneSippr process) have culminated in the ability to complete the entire process from a single colony to obtaining identification results in less than 9 hours.²⁵⁸ This development signifies the emergence of a powerful tool for routine surveillance.

The NGS technology has been demonstrated to be a revolutionary method for the identification of all microorganisms (e.g., bacteria, fungi, and viruses) present in complex clinical samples (e.g., blood, stool, and bile) in a single step, thereby obviating the need for culture and eliminating bias.^259–261 This is of particular significance in the diagnosis of pathogens or mixed infections that are challenging to detect using conventional culture methods. A multitude of comparative clinical studies have demonstrated that metagenomics NGS (mNGS) exhibits a remarkably elevated positive detection rate and sensitivity for pathogens in bile and blood samples when contrasted with conventional culture methods.^261,262 Notably, in instances of culture-negative suspected infections, mNGS has been shown to yield pivotal diagnostic insights, underscoring its potential as a valuable diagnostic tool.

4.8.2. Nanopore sequencing. Third-generation sequencing technology, represented by Oxford Nanopore Technologies’ MinION, is characterized by its ultra-long read lengths of tens of thousands of base pairs, real-time data generation, and extreme portability.²⁶³ The long read length helps to span repetitive sequences for easy genome assembly and structural variation analysis, and is suitable for serotyping.^264,265 In addition, it can directly sequence RNA for prokaryotic transcriptome analysis or detect DNA/RNA chemical modifications by analyzing small changes in current signals, providing a new dimension for functional genomics research.^266,267 Additionally, locked nucleic acid combined with nanopore sensors significantly improves SNP identification.²⁶⁸ Despite its relatively high raw read length error rate, its application in rapid pathogen identification and surveillance in the field is promising through algorithmic corrections (e.g., AssociVar) and technology iterations.²⁶⁹

4.8.3. Emerging sequencing technologies. Researchers have developed a variety of advanced sequencing and analytical technologies for improving the sensitivity and accuracy of bacterial gene expression analysis, pathogen detection, and genome modification identification. For example, RamDA-seq combined with Cas9 for scRNA-seq enables cDNA amplification and sequencing library preparation from low-abundance bacterial RNA, significantly improving the sensitivity of gene detection at the single-cell level.²⁷⁰ The APERO algorithm detects small transcripts by analyzing sequenced fragment boundaries, improving the accuracy and robustness of sRNA detection.²⁷¹ The Seroplacer algorithm and single-marker gene sequencing provide new approaches for serotyping.²⁷² The smRandom-seq assays enable high-throughput single-microbe transcriptome analysis through droplet technology and CRISPR-mediated rRNA depletion.²⁷³ The ssDRIP-seq methods provide high-resolution and strand-specific information for genome-wide R-loop detection.²⁷⁴ NanoMod tools identify DNA modifications by comparing the raw signal distribution to identify DNA modifier bases.²⁷⁵ Electron probe diagnostic nucleic acid analysis, on the other hand, is validated as a method for rapid identification of pathogens in food matrices.²⁷⁶

4.9. Other emerging approaches

4.9.1. CRISPR-based assays. The CRISPR-Cas system, a gene editing tool derived from the bacterial immune system, has been adapted for use in the field of molecular diagnostics. This adaptation is due to the system's unparalleled sequence recognition specificity, which has led to a revolution in pathogen detection. Researchers have developed a suite of innovative CRISPR-based assays—leveraging Cas9, Cas12, Cas13, Cas14, and Cas Cascade—for the sensitive and specific detection of E. coli (Table 3).

Table 3 CRISPR-based methods for the detection of E. coli

Types of Cas	Combined technologies	Target bacteria	Detection time (min)	Analytical sensitivity (CFU mL^−1)	Ref.
a Original data expressed in ng per test. b N/A indicates not available. c PMNT indicates poly[3-(3′-N,N,N-triethylamino-1′-propyloxy)-4-methyl-2,5-thiophene hydrochloride]. d Original data expressed in copies per μL.
Cas9	SDA, RCA	E. coli O157:H7	<180	40	277
Cas9 nickase	LFA, Cas9 nickase-based amplification reaction (Cas 9 nAR)	E. coli, etc.	180	100	279
Cas9 nAR version 2	Reverse transcription, nicking-extension-displacement-based amplification	E. coli O157:H7, etc.	60	1 × 10^−8 ^a	278
Cas12a	Air-displacement enhanced evanescent wave fluorescence fiber-embedded microfluidic biochip	E. coli O157:H7	45	176	281
Cas12a	A trans-acting RNA-cleaving DNAzyme, an RNA/DNA chimeric substrate	E. coli	120	1.0 × 10^2	282
Cas12a	Split T7 promoter-based three-way junction-transcription coupled with Cas12a/blocker DNA	E. coli, etc.	90	N/A^b	288
Cas12a	RPA, LFA	E. coli O157:H7	40 (fluorescence), 45 (LFA)	2.4 (fluorescence), 2.4 × 10^2 (LFA)	283
Cas12a	RAA	E. coli O157:H7	30	5.4 × 10^2	289
Cas12a	PMA, microfluidic digital chips	Viable E. coli O157:H7	120	1.2 × 10^3	290
Cas12a	RPA, PMNT^c	E. coli O157:H7	40	N/A	284
Cas13a	Confinement effect, droplet microfluidics	E. coli, etc.	∼63	N/A	285
Cas14a1	Transcription-amplified Cas14a1-activated signal biosensor, T7 RNA polymerase	E. coli	120	1.52	286
Cas cascade	Positive feedback loop	E. coli, etc.	10	1 ^d	287
Cas12a	PMA-assisted digital CRISPR microfluidic platform	Viable E. coli O157:H7	30	3.6 × 10^1	291

---

Cas9 protein-based detection strategies exhibit diverse modes of applications. An innovative fluorescence sensing method recognizes and cleaves the virulence gene of E. coli O157:H7 by the CRISPR-Cas9 system, and this cleavage event triggers the subsequent SDA and RCA, generating a large number of DNA products. These products can hybridize with fluorescent probes quenched by the metal–organic framework platform, leading to recovery of the fluorescence signal and thus enabling quantitative detection.²⁷⁷ Another technique, called Cas9nAR-v2, utilizes an engineered Cas9 nickase (Cas9n) in combination with reverse transcription and nicking displacement amplification to achieve one-pot isothermal detection of 16S rRNA.²⁷⁸ In addition, the isothermal amplification triggered by Cas9n is successfully applied to LFAs, and instrument-free visualization of Salmonella typhi and E. coli is achieved by double-labeled amplicons.²⁷⁹

The Cas12a protein is widely employed for signal amplification due to its distinctive trans-cutting activity, which involves non-specific cleavage of the surrounding single-stranded DNA upon target recognition. For amplification-free ultrasensitive detection, researchers have combined Cas12a with a portable microfluidic biochip or graphene field effect tube, both of which enable sensitive detection of coliform bacteria.^280,281 Regarding the signaling strategy, an ingenious design employs the ssDNA/tRCD (trans-acting RNA-cleaving DNAzyme) complex to temporarily lock Cas12a activity. Upon introduction of the target bacteria, the complex is specifically dissociated, thereby unlocking Cas12a's trans-cleavage activity.²⁸² In the absence of a target, the tRCD/RCS (RNA/DNA chimeric substrate) complex silences CRISPR-Cas12a, blocking fluorescence. Upon target addition, tRCD is displaced and cleaves RCS into two short fragments, freeing Cas12a to bind its DNA activator and restoring robust trans-cleavage activity. The method demonstrated 100% sensitivity and specificity for the detection of E. coli in UTI samples. Cas12a is frequently utilized in conjunction with amplification techniques, such as RPA, to establish a dual-mode detection platform for fluorescence and lateral flow chromatography.²⁸³ A visualization method known as Cas12aVIP ingeniously integrates RPA, Cas12a cleavage, and the color change of cationic conjugated polymers. In the presence of target DNA, the solution undergoes the transformation from red to yellow, facilitating visual interpretation within 40 min.²⁸⁴

It has been demonstrated that other members of the Cas family have the capacity to exhibit distinctive potential applications. Cas13a's primary function is RNA detection, and through the integration of droplet microfluidics, the reaction system is confined to a minute spatial domain, thereby enhancing the local concentration and reaction efficiency.²⁸⁵ This approach enables the absolute quantification of single-molecule RNA without the necessity of nucleic acid pre-amplification, achieving a sensitivity enhancement of over 10 000-fold. Cas14a1 is utilized in a strategy referred to as TACAS, which activates Cas14a1 to produce fluorescence signals by continuously generating RNA activators through a ligation-transcription cascade.²⁸⁶ Furthermore, CRISPR-Cascade overcomes the inherently slow trans-cleavage of Cas enzymes through an integrated positive feedback loop, enabling amplification-free, ultra-rapid detection of pathogenic DNA that yields a visible signal within 10 min.²⁸⁷

4.9.2. Other innovative molecular assays. In addition to the advent of CRISPR technology, the domain of molecular detection has witnessed the emergence of numerous innovative methodologies based on disparate principles. These methods, in concert, have catalyzed the advancement of pathogen detection technology, propelling it towards a paradigm of enhanced velocity, sensitivity, and accessibility. This transformation is achieved by the integration of novel materials, pioneering sensing mechanisms, and avant-garde signal amplification strategies.

The application of nanotechnology has greatly enhanced the detection signal. One colorimetric method achieved visual detection of target DNA down to 15 aM by capturing multiple clusters of AuNPs on the surface of MBs and utilizing their plasma effect to enhance the signal, which is comparable to the performance of PCR for E. coli in a case study.²⁹² Another strategy combines exonuclease III-assisted isothermal amplification with an inverse magnetic separation strategy, and employs a Cu/Cl-co-doped nanozyme with excellent peroxidase activity as the signal reporter.²⁹³ The resulting three-mode platform provides fluorescence, colorimetric, and photothermal readouts, achieving a detection limit for E. coli as low as 6.1 CFU mL⁻¹.

Single-molecule detection techniques are designed to bypass amplification and directly analyze individual target molecules. The “kinetic locking” method, based on single-molecule force spectroscopy, enables direct in vitro visualization of protein–DNA binding and precise measurement of the hybridization energy of ultrashort nucleotides for the detection of E. coli.²⁹⁴ The single-molecule conductance technique, on the other hand, exploits the unique charge-transport properties of RNA:DNA hybrids not only to detect target RNA with single-molecule sensitivity, but also to distinguish different E. coli serotypes that differ by only a single-nucleotide polymorphism.²⁹⁵

Combining molecular recognition with electrical signal conversion is an effective way to realize rapid and portable detection. An electrochemical chip based on electroactive DNA enzymes can detect E. coli as low as 10 CFU in less than one hour by releasing an electroactive barcode through bacteria-specific DNA enzyme cleavage and transducing the signal to a nanostructured electrode.²⁹⁶ Another potentiometric sensor based on a polymeric membrane ion-selective electrode is applied to the simultaneous detection of E. coli and S. aureus nucleic acids at a single electrode with a detection limit down to the fM level by using an innovative periodic constant-current polarization technique.²⁹⁷

Novel label-free biosensors have also been developed to take advantage of changes in physical or chemical properties triggered by biomolecular interactions. A sensor based on liquid crystals utilizes the hybridization of target DNA with a probe to restore the homologous conformation of the liquid crystal. The signal change can be observed by polarized light microscopy, enabling the detection of target DNA down to 0.02 nM.²⁹⁸ The engineered zinc finger protein (ZFP)-based platform utilizes the ability of ZFP to specifically recognize dsDNA by immobilizing it on a microchip or MB to directly capture pathogen-specific DNA sequences (e.g., Shiga toxin genes), thus realizing PCR-free detection at the fmol level.^299,300

4.10. Integrated devices

The advancement of molecular technology, characterized by its transition from laboratory settings to field applications, is a pivotal catalyst in the ongoing development of contemporary technologies. This section will focus on innovative POCT devices and platforms that miniaturize, automate, and integrate complex molecular diagnostic processes. These processes are key vehicles for achieving the goal of “sample-to-answer”.

4.10.1. DNA microarrays. The field of POCT has seen significant advancements in recent years, with DNA microarrays being optimized for high-throughput and parallel analysis capabilities. The DNA microarrays, based on signaling probes, have been optimized for probe design by thermodynamic analysis. These microarrays enable direct detection of 5S rRNA from total RNA within 30 minutes without amplification and labeling.³⁰¹ Furthermore, these microarrays reduce the time required to detect tens of thousands of DNA fragments to less than 5 minutes by concentrating the samples onto microarrays by electrophoresis and scanning the samples for detection with MBs.³⁰² In the realm of innovative devices, a chemiluminescent digital microtiter array chip encapsulates individual bacteria in skin-scaled microtiter holes, thereby enabling rapid (2–4 hours) and accurate quantification of live bacteria by detecting the catalytic luminescence reaction of β-D-glucosidase.³⁰³

4.10.2. Microfluidic chips. Microfluidic chips represent the core technology for automating POCT by manipulating fluids on a micron scale and integrating multiple steps, such as sample processing, nucleic acid amplification and signal detection—in an area of one square inch. A centrifugal microfluidic utilizes centrifugal force generated by spinning to drive the liquid, allowing the entire process from bacterial thermal lysis, PCR amplification to microarray hybridization to be automated on the chip.³⁰⁴ With the combination of LAMP and colorimetric assays, the fully automated analysis of pathogens in milk samples can be completed in less than 65 minutes.³⁰⁵ Digital microfluidics (DMF) automates the entire process from MB DNA extraction to colorimetric LAMP by precisely manipulating individual droplets through an array of electrodes in about 60 minutes.³⁰⁶ A finger-driven portable biosensor integrating immunomagnetic separation, nucleic acid extraction, and RPA-CRISPR/Cas12a detection enables the quantification of 10 CFU mL⁻¹E. coli O157:H7 without the need for external pumps or specialized personnel.³⁰⁷

In terms of the diversity of application methods, the microfluidic chip integrates a variety of amplification techniques, such as continuous flow PCR and LAMP.^308,309 The detection modalities are equally varied, including label-free real-time imaging based on LSPR,³¹⁰ as well as label-free detection combined with a prefolded G4-ThT fluorescence reporter system.³¹¹ In terms of innovative devices and materials, the embedded paper-based microchip simplifies the operation by embedding a paper sheet pre-prepared with LAMP reagents into the reaction chamber and driving the sample through centrifugal force for reaction.³¹² In addition, the combination of PMAxx and LAMP in a microfluidic chip enables the specific detection of a wide range of viable bacteria.³¹³ The reagent consumption is only 1/25 of the conventional method, thereby greatly reducing the cost.

4.10.3. POCT devices. For portable and handheld diagnostics, an integrated paper-based biosensor combines nucleic acid extraction, amplification, and detection on a single piece of paper.³¹⁴ Equipped with a handheld battery-operated heater and a smartphone readout, it reduces the total detection time from over 5 hours in conventional methods to about 1 hour. Another ingenious device, the pipette-driven capillary array comb, utilizes a standard 1 mL pipette for DNA capture, LAMP, and endpoint detection via a smartphone. The entire process can be completed in less than 85 min.³¹⁵

In terms of integration and automation, researchers have developed a variety of microdevices that do not require bulky external instrumentation. For example, sequential mixing of samples and reagents is cleverly achieved through a folding or sliding structure that integrates DNA purification, LAMP, and colorimetric detection based on fuchsin without the need for pumps or centrifuges.^316,317 Furthermore, a fully integrated fluidic cartridge, which has the reagents and components required for bacterial lysis, magnetic extraction, DNA elution, and multiplex LAMP pre-packaged, enables fully automated detection of the four major pathogens involved in UTIs.³¹⁸

In terms of innovative application methods and materials, a notable platform utilizes MNPs that can generate heat under light to achieve photothermal lysis, efficient LAMP, and specific ELISA-based detection in a single microreactor.³¹⁹ This design not only greatly improves the analytical sensitivity in complex samples such as whole blood, but also improves the specificity of amplification by photonic heating. Another innovation is the modular microfluidic sensor, which is designed to separate the nucleic acid processing module (using unmixed filtration and magnetic dispensing) from the electrochemiluminescence detection module, realizing a rapid and multiplexed detection of 16S rRNA in 30 min, with a detection limit of 64.8 CFU mL⁻¹ for E. coli in blood samples.³²⁰

4.11. Commercial kits

With the maturity of the technology, a series of molecular test kits for E. coli have been commercialized to provide a standardized solution for routine testing. These commercialized kits mainly focus on the detection of key virulence genes of PEC, such as stx1/stx2, eae, etc. According to the reference materials provided by PubMed, the products available in the market are mainly based on PCR and LAMP technologies. The following is a list of commercially available PCR-based kits: STEC Gene Screen series (3M™ Molecular Detection Assay 2),³²¹ Amplidiag® Bacterial GE (Mobidiag Ltd, Finland),³²² BactoReal® Kit (E. coli Typing series) (Ingenetix, Austria),¹⁴⁶ RIDA® GENE (pathogenic E. coli diagnostics) (R-Biopharm, Darmstadt, Germany),³²³ Mastitis 4 series (DNA Diagnostic A/S, Risskov, Denmark),³²⁴ Seeplex UTI ACE detection (Seegene, Eschborn, Germany),³²⁵ Cycleave PCR EHEC Typing Kit (TAKARA, Kusatsu, Japan),³²⁶etc. Interestingly, the GeneFields® EHEC/SS kit (Kyokuto, Tokyo, Japan) offers optional products for visualizing and judging PCR results.³²⁷ Additionally, commercially available kits based on LAMP include: Loopamp Escherichia coli O157 detection kit (Eiken, Tokyo, Japan),³²⁶ MAST ISOPLEX® VTEC kit (Mast, Bootle, UK),¹⁴⁶etc. These kits provide optional support for detection tools for different application scenarios. These kits provide optional support for different application scenarios.

5. Discussion and prospects

5.1. Applicability of target genes for different PEC types

The detection of different types of PEC requires the selection of optimal methods according to specific needs. For E. coli O157:H7 and other important STEC strains, food safety surveillance recommends rapid screening by targeting the stx1, stx2, and eae genes as well as serogroup-specific genes (e.g., rfbE_O157), and whole-genome sequencing (WGS) is the gold standard for strain typing. ETEC can rapidly diagnose traveler's diarrhea by multiplex PCR/qPCR targeting the lt/st genes. EPEC requires focusing on the eae gene and combining it with bfpA to differentiate between typical and atypical subtypes. EAEC requires simultaneous detection of multiple targets such as aggR, pCVD432, etc., due to its high degree of heterogeneity. EIEC relies on the ipaH gene, but needs to be aware of cross-reactivity with shigellosis. UPEC requires multiple testing for virulence genes such as fimH, etc., in addition to urinary culture to assess pathogenicity. Multiplex testing techniques allow the assessment of pathogenic potential. In outbreak traceability studies, sequencing technologies allow precise analysis of the chain of transmission and virulence evolution. All tests require attention to the timeliness of the samples, and the detection of pathogens in the feces decreases daily after the onset of diarrhea, with remedial sampling using rectal swabs if necessary.

5.2. Comparative analysis of different molecular methods

In order to more clearly assess the applicability of various molecular assays for PEC detection, a comparative analysis is conducted in terms of several key dimensions (Table 4). The dimensions of concern encompass detection time, analytical sensitivity and specificity, quantitative ability, cost, operational complexity, major advantages and disadvantages, and typical application scenarios.

Table 4 Comparative analysis of different molecular methods for the detection of PEC

Methods	Detection time (excluding pre-treatment)	Analytical sensitivity	Analytical specificity	Quantitative ability	Cost	Operational complexity	Major advantages and disadvantages	Typical application scenarios
PCR	2–4 h	Middle	High (primer-dependent)	Qualitative	Low	Middle	Technological maturity, low cost; easy to contaminate, qualitative	Initial screening, laboratory tests
mPCR	2–4 h	Middle	High (primer-dependent)	Qualitative	Low	Middle	Time saving, sample saving, low cost; competition between primers, complicated method optimization	Initial screening, laboratory tests
qPCR	1–3 h	High	High (primer/probe-dependent)	Relative/absolute quantitative	Middle	Middle	Quantitative, fast, high sensitivity; expensive instrument, easily affected by inhibitors	Laboratory tests, precision quantification, clinical diagnostics
dPCR	2–4 h	Extremely high	Extremely high	Absolutely quantitative	High	High	Absolute quantification, high sensitivity, anti-interference; expensive instrumentation, complicated procedure	Trace detection, copy number variation, reference standard quantification
LAMP	15–60 min	High	High (primer-dependent)	Qualitative, quantitative	Middle	Low	Fast, thermostatic, visualization of results; complex primer design, false positives, difficult to achieve multiplexed detection	POCT, screening in resource-limited areas
RPA	5–20 min	High	High	Qualitative, quantitative	Middle	Low	Rapid, thermostatic; complex reaction system, sensitive to background inhibition	POCT, screening in resource-limited areas
RCA	30–90 min	High	High	Qualitative, quantitative	Middle	Middle	Single primer, multiplexable; long reaction time, ligation step required	Laboratory tests, POCT
Biosensors	1–30 min	High	High	Qualitative, quantitative	Middle-high	Middle	Sensitive, portable; cross-reactive, high matrix interference	Laboratory tests, clinical diagnostics, POCT
NGS	Several days	High	Extremely high	Relative quantification	High	Very high (with data analysis)	High information content, possibility of discovering unknowns, strong traceability; expensive, slow, complicated to analyze	Outbreak traceability, strain typing, unknown pathogen discovery, microbial community analysis
CRISPR-Cas	0.5–1.5 h (frequently coupled amplification)	High	Extremely high	Qualitative, quantitative	High	Middle	Specific, sensitive, easy to POCT; frequently dependent on pre-amplification, critical gRNA design	Laboratory tests, rapid diagnostics, POCT
Integrated devices	10–60 min	High	High	Qualitative, quantitative	High	Low (instrument automation)	Integrated, quantifiable, anti-pollution; expensive equipment, high cost per test	Rapid diagnosis, POCT, screening in resource-limited areas

---

In the selection of molecular methods, the exact choice depends on application needs, resources and requirements for test results. For example, in outbreak investigation and fine-tuned traceability, NGS (especially WGS) is preferred because of its ability to provide genome-wide information to differentiate strains and trace the chain of transmission at the highest resolution, despite its higher cost and analytical complexity.³²⁸ In clinical rapid diagnostics or on-site testing, LAMP and some biosensors/POCT technologies (e.g., LFAs) are favored for their speed, simplicity, and low equipment requirements, with LAMP giving results in less than 1 hour and LFA even in less than 15 minutes, making them well suited for emergency or resource-poor areas.³²⁹ In large-scale screening of food and environment, qPCR has become a common method due to its high sensitivity, high specificity, quantification and relatively fast speed, and LAMP and LFA also have great potential due to their rapidity and field applicability. In scientific research and in-depth study of strain characterization, genome sequencing and gene chips/microarrays can provide comprehensive genetic information for the discovery of new genes, study of disease-causing mechanisms, and so on. For quantitative monitoring, qPCR is the most commonly used quantitative technique, while dPCR provides absolute quantification without the need for standard curves, and is particularly suitable for the precise measurement of low-copy targets.¹⁰⁰ To achieve simultaneous multi-target detection, multiplex PCR—particularly multiplex qPCR—consolidates multiple assays into a single reaction, boosting efficiency.^103,104 Gene chips/microarrays provide the greatest multiplexing depth, whereas NGS delivers ultra-high-throughput, multi-target analysis.²⁵⁷

In practice, it is often necessary to combine the advantages of different methods. For instance, LFA or LAMP can be utilized for rapid preliminary screening in the field, followed by qPCR for the confirmation and quantification of positive or suspicious samples, and WGS for in-depth analysis of outbreak strains of importance or research samples. The differentiation of live/dead bacteria remains a common challenge and is currently based on the detection of RNA (e.g., by reverse transcription-qPCR or reverse transcription-LAMP) or the selective removal of the DNA of dead bacteria in combination with pretreatment with dyes such as PMA/PMAxx.^{85,96–99,313}

5.3. Implications for surveillance of emerging PEC subtypes

This review systematically documents the continuous emergence of new PEC subtypes, which raises a critical question: can detection technologies keep pace with pathogen evolution? In response, our work compiles key functional genes—such as those encoding virulence factors and specific antigens—into a target repository, as detailed in the Introduction. Furthermore, the gene-centered detection platforms emphasized here (e.g., isothermal amplification and biosensors) are modular and programmable, allowing for rapid updates when new variants emerge.

Rather than relying on the full genomic context of a particular subtype, these methods target unique “genetic fingerprints” responsible for pathogenicity. Therefore, when genomic surveillance identifies a new suspected variant, researchers can rapidly screen for novel targets via bioinformatic comparisons and validate them using the described platforms. This workflow—genomic surveillance → bioinformatic design → platform validation—constitutes a proactive and sustainable paradigm for pathogen monitoring.

5.4. Current challenges and bottlenecks

Despite the substantial advancements in molecular detection technologies for the identification of PEC, numerous challenges persist.

First, the high degree of plasticity of the E. coli genome allows it to evolve continuously through mechanisms such as horizontal gene transfer. This has led to the emergence of new serotypes, virulence factor combinations, and drug-resistant strains. As a result, it is imperative that detection methods and target databases should be continuously updated.

Secondly, conventional molecular methods (e.g., PCR and qPCR) are unable to differentiate between live and dead bacteria, which may lead to false-positive results in food and environmental monitoring. PMA-qPCR (or reverse transcription-qPCR) technologies based on DNA (or mRNA) detection are still in need of optimization and standardization, although some progress has been made.

Thirdly, the presence of inhibitors within complex sample matrices (e.g., humic acids and polyphenols) has the potential to compromise analytical sensitivity.³³⁰ Nucleic acid amplification-based methods (e.g., qPCR and RPA) often require enrichment culture or efficient nucleic acid extraction/purification to dilute or remove inhibitors, which partially sacrifices the speed advantage of being “culture-free”. In comparison, amplification-free techniques (e.g., some CRISPR-based assays) may exhibit greater tolerance to inhibitors, albeit potentially at the cost of sensitivity.

Fourthly, there is an urgent need to establish standardized operating procedures, reference substances, and quality control systems. The high cost of some advanced molecular technologies (e.g., WGS, dPCR, and microarrays) limits their application in grassroot units and resource-limited areas. The development of low-cost, high-performance detection technologies and reagents is a common global need.

Fifthly, high-throughput technologies such as NGS generate massive amounts of raw data, the effective processing, analysis, and interpretation of which require a robust bioinformatics infrastructure, specialized analytical skills, and high-quality, continuously updated reference databases. This is a high threshold for many laboratories.

Sixthly, the simultaneous detection of multiple pathogens or multi-virulence genes on POCT devices remains a technically challenging endeavor. The challenge lies in preserving the portability, ease of use, and cost-effectiveness of the device while maintaining its performance.

Notably, this review focuses on the English-language literature and may not fully reflect regional research or commercial kits available in non-English-speaking markets.

5.5. Future prospects

In the future, molecular technologies for PEC will become more rapid, sensitive, comprehensive, intelligent, economical, and user-friendly. Novel molecular markers that better reflect strain pathogenicity, host specificity, geographic origin, or evolutionary trends will continue to be identified and validated using WGS, macrogenomics, and bioinformatics. The analytical performance of these assays is being continuously improved through the development of novel nucleic acid probes (e.g., LNA and PNA), optimization of primer-design algorithms, exploration of new signal-amplification strategies (e.g., enzyme-catalyzed amplification), and enhancement of background-signal suppression techniques. Furthermore, AI and big data analytics are revolutionizing WGS by extracting critical insights from complex data. The development of AI-powered predictive models that integrate WGS with epidemiological data will significantly advance outbreak forecasting and risk assessment.

Transcending the inherent trade-off between inhibitor tolerance and analytical sensitivity represents a pivotal challenge. A promising path involves the development of integrated strategies that combine innovative sample preparation with the engineering of robust core enzymes, such as recombinases and polymerases resistant to common inhibitors. Within this framework, CRISPR-Cas systems (e.g., Cas12 and Cas13) are particularly compelling for next-generation assays due to their high specificity and programmability.³²⁹ When coupled with isothermal amplification, they enable rapid, sensitive detection. Furthermore, the compatibility of these systems with instrument-free colorimetric or turbidimetric readouts makes them exceptionally well-suited for resource-limited settings. The ultimate goal is to create inherently resilient assays that combine high sensitivity, speed, and simplicity, enabling direct analysis in complex samples and truly fulfilling the promise of point-of-need diagnostics.

Notably, nanomaterials (e.g., MNPs, AuNPs, GO, and QDs) play an increasingly important role in biosensors and molecular detection due to their unique physicochemical properties. They can be used for signal enhancement (e.g., SERS and fluorescence enhancement), improving the immobilization efficiency of biomolecules, accelerating reaction kinetics, enabling target enrichment, and implementing novel signal conversion mechanisms. In addition, the advent of single-cell detection technology and multi-omics integration has engendered novel instruments for the in-depth study of pathogenic mechanisms. Single-cell level analysis has been demonstrated to reveal the heterogeneity of the flora. The integration and application of genomics, proteomics, and other multi-omics technologies will provide a systematic scientific basis for the development of novel diagnostic targets.

Furthermore, the molecular detection of E. coli is advancing toward a paradigm defined by high integration and intelligent analysis. Microfluidic technology enables full “sample-to-answer” automation on a lab-on-a-chip platform, greatly improving detection efficiency. The smartphone serves as a central hub, transforming detection devices into portable platforms for control, imaging, and data transmission. Meanwhile, AI-assisted analysis extracts deeper insights from complex signals, enhancing detection sensitivity and specificity. The convergence of these technologies—namely, the automation offered by microfluidics, the portability of smartphones, and the intelligence of AI—is catalyzing next-generation field-deployable diagnostics with single-cell sensitivity.³³¹

Consequently, the establishment of a surveillance network grounded in the “one health” concept is poised to serve as a pivotal support system for the prevention and control of diseases. Integration of the monitoring data from human, animal, and environmental samples, in conjunction with the establishment of a standardized and coordinated response mechanism, will effectively enhance the prevention and control of PEC. These technological breakthroughs will collectively contribute to the advancement of the molecular testing field, propelling it towards greater precision, intelligence, and systematicity.

6. Conclusion

This comprehensive review systematically examines the research landscape of molecular detection methods for PEC. The paper begins by providing an in-depth analysis of diverse PEC strains, including ETEC, EPEC, EHEC/STEC, EAEC, EIEC, DAEC, AIEC, UPEC, NMEC, APEC, and SEPEC, with a critical exploration of their molecular identification targets. By delineating both universal identification markers and strain-specific virulence genes, the review offers a nuanced understanding of the genetic characteristics that distinguish these pathogenic variants. The analysis then proceeds to explore the strategic principles underpinning target gene selection, elucidating the optimal strategies for diverse detection objects and application scenarios. Subsequently, the review conducts a thorough evaluation of contemporary molecular technologies, ranging from PCR-based methods to emerging approaches like the use of isothermal amplification, biosensors, genome sequencing, and integrated systems. Each technology is critically assessed through key performance parameters, such as detection time, analytical sensitivity and specificity, quantitative ability, and operational complexity, providing readers with a comprehensive comparative framework. The review culminates in a forward-looking discussion of the current challenges and emerging trends in E. coli molecular detection, highlighting potential future directions for research and technological innovation in this critical field of microbiological diagnostics.

Author contributions

LLZ: conceptualization, methodology, formal analysis, writing – original draft, and writing – review & editing. JSG: conceptualization, methodology, formal analysis, and writing – review & editing. MLZ: conceptualization and formal analysis. YZ: methodology and formal analysis. XL: methodology and formal analysis. CZ: conceptualization. XC: formal analysis and resources. YZ: conceptualization, methodology, formal analysis, and writing – review & editing. QPS: conceptualization, methodology, formal analysis, and writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

Data availability is not applicable to this article as no new data were created or analyzed in this study.

Acknowledgements

We would like to express our gratitude to the researchers who conducted the studies and applied molecular assays for the detection of pathogenic E. coli, as well as to all those who provided the references used in this review. We also acknowledge the support of the Natural Science Fund Project of Jiangsu Vocational College of Agriculture and Forestry (No. 2022kj32), the Natural Science Fund Project of Colleges in Jiangsu Province (No. 22KJB180001), and the Teaching Innovation Team of Animal Husbandry and Veterinary in Jiangsu Province for this work.

References

1. E. Denamur, O. Clermont, S. Bonacorsi and D. Gordon, Nat. Rev. Microbiol., 2021, 19, 37–54.
2. K. G. Joensen, A. M. M. Tetzschner, A. Iguchi, F. M. Aarestrup and F. Scheutz, J. Clin. Microbiol., 2015, 53, 2410–2426.
3. P. M. Fratamico, C. DebRoy and D. S. Needleman, Front. Microbiol., 2016, 7, 2089.
4. J. Geurtsen, M. de Been, E. Weerdenburg, A. Zomer, A. McNally and J. Poolman, FEMS Microbiol. Rev., 2022, 46, fuac031.
5. Y. Zhang, P. Tan, Y. Zhao and X. Ma, Gut Microbes, 2022, 14, 2055943.
6. M. Cepeda-Molero, C. N. Berger, A. D. S. Walsham, S. J. Ellis, S. Wemyss-Holden, S. Schüller, G. Frankel and L. Á. Fernández, PLoS Pathog., 2017, 13, e1006706.
7. Z. Saldaña, A. L. Erdem, S. Schüller, I. N. Okeke, M. Lucas, A. Sivananthan, A. D. Phillips, J. B. Kaper, J. L. Puente and J. A. Girón, J. Bacteriol., 2009, 191, 3451–3461.
8. J. M. Hunt, Clin. Lab. Med., 2010, 30, 21–45.
9. M. A. Croxen and B. B. Finlay, Nat. Rev. Microbiol., 2010, 8, 26–38.
10. A. Ud-Din and S. Wahid, Braz. J. Microbiol., 2015, 45, 1131–1138.
11. L. Zheng, S. Duan, Y. Dai and S. Wu, World J. Clin. Cases, 2022, 10, 11671–11689.
12. M. E. Terlizzi, G. Gribaudo and M. E. Maffei, Front. Microbiol., 2017, 8, 1566.
13. H. Sun, X. Li, X. Yang, J. Qin, Y. Liu, Y. Zheng, Q. Wang, R. Liu, H. Sun, X. Chen, Q. Zhang, T. Jia, X. Wu, L. Feng, L. Wang and B. Liu, Nat. Commun., 2025, 16, 2466.
14. A. Watts and P. Wigley, Antibiotics, 2024, 13, 809.
15. R. A. Conceição, M. S. Ludovico, C. G. T. J. Andrade and T. Yano, Braz. J. Med. Biol. Res., 2012, 45, 417–424.
16. J. Salvador-Erro, Y. Pastor and C. Gamazo, Toxins, 2025, 17, 71.
17. H. Wang, E. Cox and B. Devriendt, Int. J. Mol. Sci., 2022, 23, 6589.
18. A. von Mentzer, J. Tobias, G. Wiklund, S. Nordqvist, M. Aslett, G. Dougan, Å. Sjöling and A.-M. Svennerholm, Sci. Rep., 2017, 7, 12514.
19. J. B. Kaper, J. P. Nataro and H. L. T. Mobley, Nat. Rev. Microbiol., 2004, 2, 123–140.
20. M. Ohmura-Hoshino, Y. Miyaki and S. Yashima, Heliyon, 2022, 8, e10231.
21. S. Shan, Y. Huang, Z. Huang, Z. Long, C. Liu, X. Zhao, K. Xing, X. Xiao, J. Liu, Y. Huang, W. Lai and D. Liu, Food Control, 2021, 126, 108051.
22. X. Wang, D. Yu, L. Chui, T. Zhou, Y. Feng, Y. Cao and S. Zhi, Microorganisms, 2024, 12, 687.
23. A. Nemati, M. Askari Badouei, G. Hashemi Tabar, S. Morabito and A. Dadvar, BMC Vet. Res., 2024, 20, 413.
24. C. P. Tere-Peña, M. N. Calderon-Ozuna and J. E. Leguizamón Guerrero, Methods Protoc., 2024, 7, 94.
25. A. Iguchi, H. Shirai, K. Seto, T. Ooka, Y. Ogura, T. Hayashi, K. Osawa and R. Osawa, PLoS One, 2011, 6, e23250.
26. K. W. Livezey, B. Groschel and M. M. Becker, J. Food Prot., 2015, 78, 675–684.
27. L. A. Gonyar, R. M. Smith, J. A. Giron, N. C. Zachos, F. Ruiz-Perez and J. P. Nataro, Infect. Immun., 2020, 88, e00176–e00120.
28. M. M. Aslani, M. Y. Alikhani, A. Zavari, R. Yousefi and A. R. Zamani, Int. J. Infect. Dis., 2011, 15, e136–e139.
29. M. J. C. van den Beld, R. F. de Boer, F. A. G. Reubsaet, J. W. A. Rossen, K. Zhou, S. Kuiling, A. W. Friedrich and M. A. M. D. Kooistra-Smid, J. Clin. Microbiol., 2018, 56, e00510–e00518.
30. N. Haidar-Ahmad, F. O. Manigat, N. Silué, S. M. Pontier and F.-X. Campbell-Valois, Microorganisms, 2023, 11, 1709.
31. A. Farajzadeh-Sheikh, M. Savari, K. Ahmadi, H. Hosseini Nave, M. Shahin and M. Afzali, Infect. Drug Resist., 2020, 13, 119–127.
32. M. A. Belete, T. B. Demlie, W. S. Chekole and T. S. Tessema, PLoS One, 2022, 17, e0275229.
33. K. Javadi, S. Mohebi, M. Motamedifar and N. Hadi, New Microbes New Infect., 2020, 38, 100780.
34. Q. Bonet-Rossinyol, C. Camprubí-Font, M. López-Siles and M. Martinez-Medina, Front. Cell. Infect. Microbiol., 2023, 13, 1228159.
35. N. Jahandeh, R. Ranjbar, P. Behzadi and E. Behzadi, Cent. Eur. J. Urol., 2015, 68, 452–458.
36. C. Wang, Q. Li, J. Lv, X. Sun, Y. Cao, K. Yu, C. Miao, Z.-S. Zhang, Z. Yao and Q. Wang, Mucosal Immunol., 2020, 13, 22–33.
37. J. K. Brons, S. N. Vink, M. G. J. de Vos, S. Reuter, U. Dobrindt and J. D. van Elsas, J. Microbiol. Methods, 2020, 169, 105799.
38. S. Arredondo-Alonso, G. Blundell-Hunter, Z. Fu, R. A. Gladstone, A. Fillol-Salom, J. Loraine, E. Cloutman-Green, P. J. Johnsen, Ø. Samuelsen, A. K. Pöntinen, F. Cléon, S. Chavez-Bueno, M. A. De la Cruz, M. A. Ares, M. Vongsouvath, A. Chmielarczyk, C. Horner, N. Klein, A. McNally, J. N. Reis, J. R. Penadés, N. R. Thomson, J. Corander, P. W. Taylor and A. J. McCarthy, Nat. Commun., 2023, 14, 3294.
39. V. M. Sora, G. Meroni, P. A. Martino, A. Soggiu, L. Bonizzi and A. Zecconi, Pathogens, 2021, 10, 1355.
40. S. Ievy, Md. S. Islam, Md. A. Sobur, M. Talukder, Md. B. Rahman, M. F. R. Khan and Md. T. Rahman, Microorganisms, 2020, 8, 1021.
41. G. Ranabhat, D. Subedi, J. Karki, R. Paudel, H. Luitel and R. K. Bhattarai, Heliyon, 2024, 10, e35661.
42. G. Adwan, S. Abuseir, G. Omar and M. Albzour, Vet. World, 2024, 17, 2865–2879.
43. Y. Guo, R. Xiao, J. Feng, X. Wang, J. Lai, W. Kang, Y. Li, X. Zhu, T. Ji, X. Huang, D. Pang, Y. An, L. Meng and Y. Wang, Heliyon, 2024, 10, e35991.
44. J.-T. Kuo, L.-L. Chang, C.-Y. Yen, T.-H. Tsai, Y.-C. Chang, Y.-T. Huang and Y.-C. Chung, Biosensors, 2020, 11, 8.
45. J. Barrero-Canosa, C. Moraru, L. Zeugner, B. M. Fuchs and R. Amann, Environ. Microbiol., 2017, 19, 70–82.
46. K. Fujimoto, K. Toyosato, S. Nakamura and T. Sakamoto, Bioorg. Med. Chem. Lett., 2016, 26, 5312–5314.
47. M. S. Lee, H. Hyun, I. Park, S. Kim, D.-H. Jang, S. Kim, J.-K. Im, H. Kim, J. H. Lee, T. Kwon and J. H. Kang, Small Methods, 2022, 6, e2101239.
48. Y.-T. Kao, S. Calabrese, N. Borst, M. Lehnert, Y.-K. Lai, F. Schlenker, P. Juelg, R. Zengerle, P. Garstecki and F. von Stetten, Biosensors, 2022, 12, 237.
49. Y. Xue, J. G. Wilkes, T. J. Moskal, A. J. Williams, W. M. Cooper, R. Nayak, F. Rafii and D. A. Buzatu, PLoS One, 2016, 11, e0150038.
50. X. X. Huang, N. Urosevic and T. J. J. Inglis, PLoS One, 2019, 14, e0201332.
51. C. J. Yang, L. Wang, Y. Wu, Y. Kim, C. D. Medley, H. Lin and W. Tan, Nucleic Acids Res., 2007, 35, 4030–4041.
52. D. Hur, M. S. Kim, M. Song, J. Jung and H. Park, Biol. Proced. Online, 2015, 17, 14.
53. R. Rocha, R. S. Santos, P. Madureira, C. Almeida and N. F. Azevedo, J. Biotechnol., 2016, 226, 1–7.
54. A. S. Azevedo, I. M. Sousa, R. M. Fernandes, N. F. Azevedo and C. Almeida, PLoS One, 2019, 14, e0217689.
55. A. S. Azevedo, R. M. Fernandes, A. R. Faria, O. F. Silvestre, J. B. Nieder, C. Lou, J. Wengel, C. Almeida and N. F. Azevedo, Front. Microbiol., 2022, 13, 976639.
56. M. Liu, L. Geng, F. Zhang, S. Dou, F. Li, Z. Liu, Y. Guo and X. Sun, J. Agric. Food Chem., 2022, 70, 15990–15998.
57. Y. Liu, Y. Shi, S. Wang, S. Liu, M. Shang, B. Zhao, H. Liu, C. Yang, F. Wang, C. K. Kwok and H. Wang, Anal. Chem., 2024, 96, 11951–11958.
58. X. Liu, S. Bu, H. Wei, Z. Wang, S. Yu, Z. Li, Z. Hao, X. He and J. Wan, Anal. Methods, 2021, 13, 3379–3385.
59. H. Ravan, Biosens. Bioelectron., 2016, 80, 67–73.
60. H. Yan, L. Wu, J. Wang, Y. Zheng, F. Zhao, Q. Bai, H. Hu, H. Liang and X. Niu, Anal. Methods, 2024, 16, 496–502.
61. H. Liang, D. Li, S. Jia, X. Zhang, Q. Bai, Z. Wang, Y. Cai, J. Liu and L. Chen, Mikrochim. Acta, 2025, 192, 385.
62. Q. Guo, J. Han, S. Shan, D. Liu, S. Wu, Y. Xiong and W. Lai, Biosens. Bioelectron., 2016, 86, 990–995.
63. Y. Liang and C. Zhang, Anal. Chim. Acta, 2023, 1239, 340630.
64. Y. Xiong, X. Zeng, L. Yan, Y. Wang, Y. Lin, K. Ao, P. Feng, Y. Xie and P. Chen, ACS Appl. Mater. Interfaces, 2023, 15, 27687–27695.
65. J. Liu, S. Zeng, Y. Wan, T. Liu, F. Chen, A. Wang, W. Tang, J. Wang, H. Yuan, M. Negahdary, Y. Lin, Y. Li, L. Wang and Z. Wu, Talanta, 2023, 265, 124829.
66. J. Du, Z. Huang, Z. Li, L. Wang, Q. Zhang, Y. Dai and Y. Bai, J. Microbiol. Methods, 2025, 235, 107151.
67. E. Dester, K. Kao and E. C. Alocilja, Biosensors, 2022, 12, 274.
68. S. Wei, Q. Tang, X. Hu, W. Ouyang, H. Shao, J. Li, H. Yan, Y. Chen and L. Liu, ACS Sens., 2024, 9, 325–336.
69. K. E. Mach, A. M. Kaushik, K. Hsieh, P. K. Wong, T.-H. Wang and J. C. Liao, Analyst, 2019, 144, 1565–1574.
70. D. V. Smirnova, M. Y. Rubtsova, V. G. Grigorenko and N. N. Ugarova, Photochem. Photobiol., 2017, 93, 541–547.
71. K. J. Lee, W. S. Lee, A. Hwang, J. Moon, T. Kang, K. Park and J. Jeong, Analyst, 2017, 143, 332–338.
72. C. Wang, Y. Zhang, W. Gong and S. Wang, Luminescence, 2024, 39, e4716.
73. X. Bai, L. Ga and J. Ai, Analyst, 2023, 148, 3892–3898.
74. S. M. Saad, J. Abdullah, S. A. Rashid, Y. W. Fen, F. Salam and L. H. Yih, Mikrochim. Acta, 2019, 186, 804.
75. S. Doughan, U. Uddayasankar, A. Peri and U. J. Krull, Anal. Chim. Acta, 2017, 962, 88–96.
76. F. Karimi, N. Balazadeh and B. Eftekhari-Sis, J. Appl. Genet., 2019, 60, 417–426.
77. J. Sun, J. Ji, Y. Sun, M. H. Abdalhai, Y. Zhang and X. Sun, Biosens. Bioelectron., 2015, 70, 239–245.
78. E. Martínez-Periñán, Á. Martínez-Sobrino, I. Bravo, T. García-Mendiola, E. Mateo-Martí, F. Pariente and E. Lorenzo, Anal. Bioanal. Chem., 2022, 414, 5537–5548.
79. Y. Xue, S. He, M. Li and Y. Qiu, Foodborne Pathog. Dis., 2025, 22, 193–201.
80. F. J. Velez, J. M. Bosilevac, G. Salazar, H. Kaur Kapoor, A. Mishra, E. Madoroba, K. Stanford, P. Fach, S. Delannoy, R. Stephan and P. Singh, Food Res. Int., 2024, 196, 115105.
81. R. L. Lindsey, L. Garcia-Toledo, D. Fasulo, L. M. Gladney and N. Strockbine, J. Microbiol. Methods, 2017, 140, 1–4.
82. B. Zimoń, M. Psujek, J. Matczak, A. Guziński, E. Wójcik and J. Dastych, Microbiol. Spectrum, 2024, 12, e0377323.
83. T. A. Woods, H. M. Mendez, S. Ortega, X. Shi, D. Marx, J. Bai, R. A. Moxley, T. G. Nagaraja, S. W. Graves and A. Deshpande, Front. Cell. Infect. Microbiol., 2016, 6, 92.
84. R. Dhakal, Q. Wang, R. Lan, P. Howard and V. Sintchenko, J. Med. Microbiol., 2018, 67, 1257–1264.
85. G. Xie, S. Yu, W. Li, D. Mu, Z. P. Aguilar and H. Xu, J. Microbiol., 2020, 58, 668–674.
86. K. Feng, W. Hu, A. Jiang, Sarengaowa, Y. Xu, Y. Zou, L. Yang and X. Wang, PLoS One, 2016, 11, e0166874.
87. D. P. Manage, J. Lauzon, L. M. McMullen and L. M. Pilarski, BMC Microbiol., 2019, 19, 93.
88. J. Wang, P. Zhao, M. Zhao, D. Zhang, S. Chen, Y. Liu, Y. Gao, Y. Tie and Z. Feng, Microbiol. Spectrum, 2024, 12, e0120224.
89. Q. Huang, D. Chen, C. Du, Q. Liu, S. Lin, L. Liang, Y. Xu, Y. Liao and Q. Li, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, e2110672119.
90. U. V. Schneider, N. D. Mikkelsen, F. Scheutz, A. Friis-Møller and G. Lisby, PLoS One, 2018, 13, e0199766.
91. J.-H. Kim and S.-W. Oh, Food Chem., 2020, 327, 127036.
92. L. Dong, H. Liu, L. Meng, M. Xing, T. Lan, M. Gu, N. Zheng, C. Wang, H. Chen and J. Wang, J. Dairy Sci., 2019, 102, 6914–6919.
93. F. Alexandrino, J. S. Malgarin, M. A. Krieger and L. G. Morello, Braz. J. Biol., 2021, 81, 692–700.
94. M. Miotto, C. Barretta, S. O. Ossai, H. S. da Silva, A. Kist, C. R. W. Vieira and S. Parveen, Int. J. Food Microbiol., 2020, 318, 108467.
95. R. Deshmukh, S. Bhand and U. Roy, Braz. J. Microbiol., 2020, 51, 773–778.
96. X. Fang, W. Li, C. Zhang, Z. Huang, H. Zeng, Z. Dong and W. Zhang, Orthop. Surg., 2018, 10, 40–46.
97. M. Calimci, T. Tezcan, E. K. Tayyarcan, K. Guven, I. H. Boyaci and U. Tamer, Talanta, 2025, 286, 127569.
98. L. Heijnen, H. J. de Vries, G. van Pelt, E. Stroobach, A. Atsma, J. Vranken, K. De Maeyer, L. Vissers and G. Medema, Water Res., 2024, 259, 121843.
99. D. Li, Q. Zhang, G. Liu, L. Zhang, Z. Gu, Y. Pan, X. Cui, P. He, X. Li, J. Liu, G. Liu, M. Yang and X. Tian, J. Pharm. Biomed. Anal., 2021, 198, 114000.
100. B. Xue, H. Zhang, X. Yan, X. Su, Y. Zhou, J. Xie, S. Li, Y. Zhou and L. Li, Sci. Rep., 2025, 15, 16728.
101. H. Wang, Y. Sun, J. Chen, W. Wang, H. Yu, C. Gao, T. An, Y. Wang, H. Chen, L. Zhu, Z. Jin, C. Yu, C. Xia and H. Zhang, Front. Cell. Infect. Microbiol., 2024, 14, 1468783.
102. E. Carloni, L. Rotundo, G. Brandi and G. Amagliani, Folia Microbiol., 2018, 63, 735–742.
103. X. An, J. Wang, Q. Pu, H. Li, T. Pan, H. Li, F. Pan and J. Su, Environ. Res., 2020, 190, 109982.
104. M. Hasanpour and A. Najafi, J. Microbiol. Methods, 2017, 137, 25–29.
105. F. J. Velez, J. M. Bosilevac, A. Mishra and P. Singh, J. Microbiol. Methods, 2023, 204, 106632.
106. W. Wang, R. T. Zijlstra and M. G. Gänzle, BMC Microbiol., 2017, 17, 114.
107. S. Furutani, N. Naruishi, Y. Hagihara and H. Nagai, Anal. Bioanal. Chem., 2016, 408, 5641–5649.
108. M. Moinet, R. M. Collis, L. Rogers, M. L. Devane, P. J. Biggs, R. Stott, J. Marshall, R. Muirhead and A. L. Cookson, J. Microbiol. Methods, 2024, 220, 106909.
109. L.-J. Tak, M.-K. Shin, J.-I. Yoo, M.-C. Cho and W. Kim, Front. Cell. Infect. Microbiol., 2023, 13, 1301446.
110. Y. Wouters, D. Dalloyaux, A. Christenhusz, H. M. J. Roelofs, H. F. Wertheim, C. P. Bleeker-Rovers, R. H. Te Morsche and G. J. A. Wanten, Microb. Biotechnol., 2020, 13, 657–668.
111. I. Ziegler, S. Lindström, M. Källgren, K. Strålin and P. Mölling, PLoS One, 2019, 14, e0224656.
112. L. He, D. J. Simpson and M. G. Gänzle, Food Microbiol., 2020, 90, 103466.
113. X. Liu, L. Noll, X. Shi, E. Porter, Y. Wang, C. Stoy, N. Lu, T. G. Nagaraja, G. Anderson and J. Bai, J. Clin. Microbiol., 2020, 58, e01684–e01619.
114. D. E. Gaddes, P.-W. Lee, A. Y. Trick, P. Athamanolap, C. M. O'Keefe, C. Puleo, K. Hsieh and T.-H. Wang, Anal. Chem., 2020, 92, 13254–13261.
115. J. M. Kim, J. S. Park, T. H. Yoon, J. Park and K. S. Park, Anal. Bioanal. Chem., 2021, 413, 5003–5011.
116. T.-H. Kim, H. J. Hwang and J. H. Kim, Foodborne Pathog. Dis., 2019, 16, 144–151.
117. Y. Song and P. Gyarmati, Mikrochim. Acta, 2019, 186, 642.
118. Y. Song and P. Gyarmati, Sci. Rep., 2022, 12, 4515.
119. I. A. Quintela, B. G. de los Reyes, C.-S. Lin and V. C. H. Wu, Nanoscale, 2015, 7, 2417–2426.
120. Y. Song, M. Chen, L. Han, Z. Yan, L. Pan and K. Tu, Anal. Chim. Acta, 2023, 1239, 340751.
121. H. Y. Kim, K. S. Park and H. G. Park, Theranostics, 2020, 10, 4507–4514.
122. J. Xia, T. Bu, P. Jia, K. He, X. Wang, X. Sun and L. Wang, Anal. Biochem., 2022, 654, 114797.
123. S. Liébana, D. Brandão, P. Cortés, S. Campoy, S. Alegret and M. I. Pividori, Anal. Chim. Acta, 2016, 904, 1–9.
124. A. M. Giuliodori, A. Brandi, S. Kotla, F. Perrozzi, R. Gunnella, L. Ottaviano, R. Spurio and A. Fabbretti, PLoS One, 2017, 12, e0183952.
125. T. Soejima, J.-I. Minami, J.-Z. Xiao and F. Abe, Biotechnol. Bioeng., 2016, 113, 301–310.
126. W. Dong, C. Tao, B. Yang, E. Miyake, Z. Li, D. Zhang and Y. Yamaguchi, J. Visualized Exp., 2023, 201, e63523.
127. Y. Zhang, L. Zhu, Y. Zhang, P. He and Q. Wang, J. Chromatogr. A, 2018, 1555, 100–105.
128. A. Beyer, S. Pollok, A. Rudloff, D. Cialla-May, K. Weber and J. Popp, Macromol. Biosci., 2016, 16, 1325–1333.
129. X. Ma, Y. Li, Y. Liang, Y. Liu, L. Yu, C. Li, Q. Liu and L. Chen, BMC Microbiol., 2020, 20, 177.
130. T. Notomi, H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N. Amino and T. Hase, Nucleic Acids Res., 2000, 28, E63.
131. Z. Alfahl, S. Biggins, O. Higgins, A. Chueiri, T. J. Smith, D. Morris, J. O'Dwyer, P. D. Hynds, L. P. Burke and L. O'Connor, Microbiology, 2024, 170, 001485.
132. G. B. Priya, R. K. Agrawal, A. A. P. Milton, M. Mishra, S. K. Mendiratta, B. R. Singh, D. Kumar, R. K. Gandham, Z. B. Dubal, S. Rajkhowa, A. Luke and G. Patil, Braz. J. Microbiol., 2024, 55, 1723–1733.
133. Z. Wang, X. Cui, A. Hu, Z. Lu, F. Meng, L. Zhou and X. Bie, Lett. Appl. Microbiol., 2023, 76, ovad122.
134. L. Y. Zhao, J. H. Niu, X. L. Gao, C. N. Liu, S. M. Liu, N. Jiang, X. P. Lv and S. M. Zheng, Pol. J. Vet. Sci., 2020, 23, 143–152.
135. N. Saengsawang, T. Ruang-Areerate, P. Kesakomol, T. Thita, M. Mungthin and W. Dungchai, Analyst, 2021, 145, 8077–8086.
136. M. Varona and J. L. Anderson, Anal. Chem., 2019, 91, 6991–6995.
137. S. J. Oh, B. H. Park, J. H. Jung, G. Choi, D. C. Lee, D. H. Kim and T. S. Seo, Biosens. Bioelectron., 2016, 75, 293–300.
138. W. Jaroenram, P. Cecere and P. P. Pompa, J. Microbiol. Methods, 2019, 156, 9–14.
139. N. Saengsawang, P. Ruang-Areerate, N. Kaeothaisong, S. Leelayoova, M. Mungthin, P. Juntanawiwat, P. Hanyanunt, P. Potisuwan, P. Kesakomol, P. Butsararattanagomen, P. Wichaiwong, W. Dungchai and T. Ruang-Areerate, Sci. Rep., 2023, 13, 18781.
140. L. Han, K. Wang, L. Ma, P. Delaquis, S. Bach, J. Feng and X. Lu, Appl. Environ. Microbiol., 2020, 86, e02566–e02519.
141. M. Yan, L. Xu, H. Jiang, Z. Zhou, S. Zhou and L. Zhang, Microb. Pathog., 2017, 105, 245–250.
142. M. D. Borysiak, K. W. Kimura and J. D. Posner, Lab Chip, 2015, 15, 1697–1707.
143. W. Du, S. Wang, Y. Luo, X. Tong, Y. Chen, Z. Fang, J. Shan, D. Liu, H. Ni, W. Lai and S. Shan, Anal. Chim. Acta, 2025, 1364, 344222.
144. Z. Fei, D. Zhou, W. Dai and P. Xiao, Electrophoresis, 2020, 41, 1793–1803.
145. Y. Zeng, M. Liu, Y. Xia and X. Jiang, Analyst, 2020, 145, 7048–7055.
146. M. I. Suwara, M. Bennett, I. A. P. Voto, C. A. Brownlie and E. A. Gillies, Int. J. Mol. Sci., 2024, 25, 10067.
147. Y. Si, T. Zhang, N. Chen, Y. Cheng, L. Wang, J. Yuan, G. Li, M. Zong, G. Sui and L. Fan, J. Microbiol. Methods, 2021, 190, 106339.
148. W. S. Jang, S. Park, J. H. Bae, S. Y. Yoon, C. S. Lim and M.-C. Cho, PLoS One, 2024, 19, e0302783.
149. T. N. D. Trinh, H. C. La and N. Y. Lee, ACS Sens., 2019, 4, 2754–2762.
150. X. Lin, X. Huang, K. Urmann, X. Xie and M. R. Hoffmann, ACS Sens., 2019, 4, 242–249.
151. C. Wu, L. Liu, Z. Ye, J. Gong, P. Hao, J. Ping and Y. Ying, Anal. Chim. Acta, 2022, 1233, 340513.
152. Z. Ma, M. Ma, X. Cao, Y. Jiang and D. Gao, Mikrochim. Acta, 2024, 191, 430.
153. Y. Yan, T. Yang, Z. Luo, D. Li, L. Li and X. Lin, Food Chem., 2023, 408, 135226.
154. X. Lin, M. Fang, C. Yi, Y. Jiang, C. Zhang, X. Pan and Z. Luo, Biomaterials, 2022, 280, 121278.
155. X. Long, Z. Gong, Y. Gan, P. Yuan, Y. Tang, Y. Yang and S. Zhong, Anal. Chem., 2024, 96, 15608–15613.
156. J.-E. Lee, S. H. Toushik, H.-J. Park, S.-A. Kim and W.-B. Shim, Anal. Bioanal. Chem., 2023, 415, 4973–4984.
157. Y. Sugita-Konishi, N. Kobayashi, K. Takasaki, T. Kanno, M. Itoh, Riztyan, S. Futo, H. Asakura, K. Taira and Y. Kawakami, J. Vet. Med. Sci., 2019, 81, 586–592.
158. I. Hongwarittorrn, N. Chaichanawongsaroj and W. Laiwattanapaisal, Talanta, 2017, 175, 135–142.
159. S. Chakraborty, S. Connor and M. Velagic, PLoS Neglected Trop. Dis., 2022, 16, e0010180.
160. A. Sen, M. Masetty, S. Weerakoon, C. Morris, J. S. Yadav, S. Apewokin, J. Trannguyen, M. Broom and A. Priye, Biosens. Bioelectron., 2024, 257, 116292.
161. H. A. Nguyen and N. Y. Lee, Biosens. Bioelectron., 2021, 189, 113353.
162. S. H. Reo, M. C. Cho, S. Lee, S. Kim, E. S. Lee, B. S. Cha and K. S. Park, Anal. Chem., 2025, 97, 14239–14245.
163. Y. L. Kim, D. Kim, J. Park, M. Kwak and J. H. Shin, Anal. Methods, 2022, 14, 2569–2577.
164. C. Yi, Z. Luo, Y. Lu, T. Belwal, X. Pan and X. Lin, Biosens. Bioelectron., 2021, 184, 113199.
165. L. Guo, B. Lu, Q. Dong, Y. Tang, Y. Du and B. Li, Anal. Chim. Acta, 2020, 1106, 191–198.
166. S. S. Arumugam, A. W. Varghese, S. Suresh Nair and N. Y. Lee, Anal. Methods, 2023, 15, 5793–5802.
167. T. Wang, S. Kim and J. H. An, J. Microbiol. Methods, 2017, 133, 1–7.
168. D. Das, H.-C. Hsieh, C.-S. Chen, W.-L. Chen and H.-S. Chuang, Small Sci., 2022, 2, 2200010.
169. A. H. Sewid, J. H. Ramos, H. C. Dylewski, G. I. Castro, D. H. D'Souza and S. Eda, PLoS One, 2025, 20, e0320393.
170. A. H. Sewid, H. C. Dylewski, J. H. Ramos, B. M. Morgan, B. D. Gelalcha, D. H. D'Souza, J. J. Wu, O. K. Dego and S. Eda, Sci. Rep., 2024, 14, 28942.
171. M. Tan, C. Liao, L. Liang, X. Yi, Z. Zhou and G. Wei, Front. Cell. Infect. Microbiol., 2022, 12, 1019071.
172. B. Raja, H. J. Goux, A. Marapadaga, S. Rajagopalan, K. Kourentzi and R. C. Willson, J. Appl. Microbiol., 2017, 123, 544–555.
173. A. Rani, V. B. Ravindran, A. Surapaneni, E. Shahsavari, N. Haleyur, N. Mantri and A. S. Ball, Sci. Rep., 2021, 11, 1881.
174. P. Khan, L. M. Aufdembrink, K. P. Adamala and A. E. Engelhart, RNA, 2024, 30, 891–900.
175. J. S. McQuillan and M. W. Wilson, Lett. Appl. Microbiol., 2021, 72, 382–389.
176. K. T. L. Trinh and N. Y. Lee, Biosensors, 2022, 12, 72.
177. Z. Bian, W. Liu, J. Jin, Y. Hao, L. Jiang, Y. Xie and H. Zhang, PLoS One, 2022, 17, e0278869.
178. B. Jin, B. Ma, J. Li, Y. Hong and M. Zhang, Microorganisms, 2022, 10, 1352.
179. M. L. Powell, F. R. Bowler, A. J. Martinez, C. J. Greenwood, N. Armes and O. Piepenburg, Anal. Biochem., 2018, 543, 108–115.
180. W. Bai, J. Chen, D. Chen, Y. Zhu, K. Hu, X. Lin, J. Chen and D. Song, Int. J. Food Microbiol., 2024, 422, 110822.
181. S. Petrucci, C. Costa, D. Broyles, A. Kaur, E. Dikici, S. Daunert and S. K. Deo, Anal. Chem., 2022, 94, 2485–2492.
182. J. Hu, Y. Wang, H. Su, H. Ding, X. Sun, H. Gao, Y. Geng and Z. Wang, Mol. Cell. Probes, 2020, 50, 101501.
183. T. Zheng, X. Li, Y.-N. Xie, B. Yang and P. Wu, Anal. Chem., 2023, 95, 6053–6060.
184. H. Ahn, B. S. Batule, Y. Seok and M.-G. Kim, Anal. Chem., 2018, 90, 10211–10216.
185. G. Choi, J. H. Jung, B. H. Park, S. J. Oh, J. H. Seo, J. S. Choi, D. H. Kim and T. S. Seo, Lab Chip, 2016, 16, 2309–2316.
186. J. Chen, Y. Xu, H. Yan, Y. Zhu, L. Wang, Y. Zhang, Y. Lu and W. Xing, Lab Chip, 2018, 18, 2441–2452.
187. H. Takahashi, M. Ohkawachi, K. Horio, T. Kobori, T. Aki, Y. Matsumura, Y. Nakashimada and Y. Okamura, Sci. Rep., 2018, 8, 7770.
188. J. J. Zhang, C. Nie, W. L. Fu, F. L. Cheng, P. Chen, Z. F. Gao, Y. Wu and Y. Shen, Anal. Chem., 2022, 94, 16796–16802.
189. M. Kühnemund, I. Hernández-Neuta, M. I. Sharif, M. Cornaglia, M. A. M. Gijs and M. Nilsson, Nucleic Acids Res., 2017, 45, e59.
190. C. Tang, H. Liu, W. Pan, M. Wang, J. Ren, Z. Chen, H. Chen, Y. Deng and S. Li, Biosensors, 2022, 12, 1075.
191. M. Donolato, P. Antunes, T. Z. G. de la Torre, E.-T. Hwu, C.-H. Chen, R. Burger, G. Rizzi, F. G. Bosco, M. Strømme, A. Boisen and M. F. Hansen, Biosens. Bioelectron., 2015, 67, 649–655.
192. Y. Zeng, P. Qi, Y. Zhou, Y. Wang, Y. Xin, Y. Sun and D. Zhang, Mikrochim. Acta, 2022, 189, 403.
193. A. Mezger, J. Fock, P. Antunes, F. W. Østerberg, A. Boisen, M. Nilsson, M. F. Hansen, A. Ahlford and M. Donolato, ACS Nano, 2015, 9, 7374–7382.
194. D. Mu, D. Zhou, G. Xie, J. Liu, Z. Wang, Q. Xiong and H. Xu, J. Dairy Sci., 2022, 105, 1028–1038.
195. X. Ding, Z. Xu, K. Yin, M. Sfeir and C. Liu, Anal. Chem., 2019, 91, 12852–12858.
196. J. Zhai, Z. Yan, F. Ping, Q. Lei, X. Chen and Y. Qi, PLoS One, 2020, 15, e0230881.
197. W. Liu, C. Yuan, L. Zhang and Y. Feng, PLoS One, 2019, 14, e0216272.
198. W. Liu, C. Yuan, L. Zhang and Y. Feng, J. Microbiol. Methods, 2019, 161, 47–55.
199. Y. Li, X. Zhang, Y. Liao, C. Shi, Y. Wang, X. Mu, Y. Xie and C. Ma, Anal. Chem., 2023, 95, 18249–18257.
200. S. F. Cheung, M. F. Yee, N. K. Le, B. M. Wu and D. T. Kamei, Anal. Bioanal. Chem., 2018, 410, 5255–5263.
201. Y. Ma, X. Dai, T. Hong, G. B. Munk and M. Libera, Analyst, 2016, 142, 147–155.
202. Z. Chu, J. Chen, J. Zhang, Q. Xie, F. Zhang and Q. Wang, J. Chromatogr. A, 2023, 1706, 464275.
203. E. Yuhana Ariffin, L. Y. Heng, L. L. Tan, N. H. Abd Karim and S. A. Hasbullah, Sensors, 2020, 20, 1279.
204. A. Rabti, R. Zayani, M. Meftah, I. Salhi and N. Raouafi, Mikrochim. Acta, 2020, 187, 635.
205. N. Shoaie, M. Forouzandeh and K. Omidfar, Mikrochim. Acta, 2018, 185, 217.
206. R. Deshmukh, A. K. Prusty, U. Roy and S. Bhand, Analyst, 2020, 145, 2267–2278.
207. F. Shi, B. Wang, L. Yan, B. Wang, Y. Niu, L. Wang and W. Sun, Bioelectrochemistry, 2022, 148, 108226.
208. R. B. Jamal, T. Vitasovic, U. Gosewinkel and E. E. Ferapontova, Biosens. Bioelectron., 2023, 228, 115214.
209. J. Zhang, J. Wang, X. Zhang and F. He, Biosens. Bioelectron., 2018, 118, 9–15.
210. Y. Li, H. Liu, H. Huang, J. Deng, L. Fang, J. Luo, S. Zhang, J. Huang, W. Liang and J. Zheng, Biosens. Bioelectron., 2020, 147, 111752.
211. A. Ben Aissa, N. Madaboosi, M. Nilsson and M. I. Pividori, Sensors, 2021, 21, 1749.
212. C. Li, Y. Liang and Q. Feng, Analyst, 2025, 150, 1158–1166.
213. R. Ge, H. Dai, S. Zhang, J. Wei, T. Jiao, Q. Chen, Q. Chen and X. Chen, Anal. Chem., 2023, 95, 7379–7386.
214. H. Lee, S. Y. Yi, J. S. Kwon, J. M. Choi, D. S. Lee, S. H. Lee and Y.-B. Shin, Biosens. Bioelectron., 2021, 179, 113042.
215. S. Wang, Y. Liu, R. Liu, L. Xie, H. Yang, S. Ge and J. Yu, Anal. Chim. Acta, 2024, 1291, 342213.
216. Y. Fan, Y. Liao, Z. Gao, H. Wang, Y. Li, C. Shi and C. Ma, J. Agric. Food Chem., 2025, 73, 4342–4352.
217. L. Huang, W. Zhang, M. Liu, Y. Gong, Q. Tang, K. Wang, X. Liao, K. Zhang and J. Wei, Bioelectrochemistry, 2025, 161, 108815.
218. J. Cui, Q. Luo, C. Wei, X. Deng, H. Liang, J. Wei, Y. Gong, Q. Tang, K. Zhang and X. Liao, Anal. Chim. Acta, 2024, 1285, 342028.
219. N. Razmi, M. Hasanzadeh, M. Willander and O. Nur, Anal. Methods, 2022, 14, 1562–1570.
220. L. Yuwen, X. Li, L. Wu, Y. Luo and S. Su, Analyst, 2023, 148, 6292–6296.
221. T. N. P. Nguyen, S. H. Nguyen and M. T. Tran, PLoS One, 2024, 19, e0299272.
222. N. Jaiswal, C. M. Pandey, S. Solanki, I. Tiwari and B. D. Malhotra, Mikrochim. Acta, 2019, 187, 1.
223. J. Yoo, H. Jeong, S. K. Park, S. Park and J. S. Lee, Biosensors, 2021, 11, 212.
224. S. Nadzirah, U. Hashim, S. C. B. Gopinath, N. A. Parmin, A. A. Hamzah, H. W. Yu and C. F. Dee, Mikrochim. Acta, 2020, 187, 235.
225. J. Narang, A. Mishra, R. Pilloton, A. Vv, S. Wadhwa, C. S. Pundir and M. Khanuja, Biosensors, 2018, 8, 77.
226. J. Zhang, W. Liu, J. Li, K. Lu, H. Wen and J. Ren, Talanta, 2022, 249, 123646.
227. T. Wu, A. K. Yagati and J. Min, Biosensors, 2023, 13, 641.
228. H. Gong, Y. Li, Y. Liu, Y. Shan, Y. Wang, C. Ma and C. Shi, Talanta, 2025, 294, 128236.
229. F. Huber, H. P. Lang, S. Heller, J. A. Bielicki, C. Gerber, E. Meyer and A. Egli, Biosensors, 2022, 12, 994.
230. M. A. Reed and Y. V. Gerasimova, Front. Chem., 2022, 10, 951279.
231. Z. Zhou, J. D. Brennan and Y. Li, Angew. Chem., Int. Ed., 2020, 59, 10401–10405.
232. L. Han, Y. Song, M. Chen, L. Pan, K. Tu and J. Su, Mikrochim. Acta, 2025, 192, 212.
233. D. Li, E. Yang, Z. Luo, Q. Xie and Y. Duan, Nanoscale, 2021, 13, 2492–2501.
234. Z. Zhou, Y. Zhang, M. Guo, K. Huang and W. Xu, Biosens. Bioelectron., 2020, 167, 112475.
235. M. Liu, Y. Zhao, H. Monshat, Z. Tang, Z. Wu, Q. Zhang and M. Lu, Biosens. Bioelectron., 2020, 169, 112651.
236. F. Jahanpeyma, M. Forouzandeh, M. J. Rasaee and N. Shoaie, Front. Biosci., 2019, 11, 122–135.
237. Y. Du, Y. Zhou, Y. Wen, X. Bian, Y. Xie, W. Zhang, G. Liu and J. Yan, Mikrochim. Acta, 2019, 186, 840.
238. H. Jiang, X. Zhu, J. Jiao, C. Yan, K. Liu, W. Chen and P. Qin, Biosens. Bioelectron., 2025, 270, 116974.
239. C. Lu, L. Tang, F. Gao, Y. Li, J. Liu and J. Zheng, Biosens. Bioelectron., 2021, 187, 113327.
240. Y. M. Park, C. H. Kim, Y. Song, S.-H. Hwang, S. J. Lee and M.-K. Lee, Talanta, 2019, 195, 97–102.
241. Y. M. Park, C. H. Kim, S. J. Lee and M.-K. Lee, Biosens. Bioelectron., 2019, 141, 111415.
242. F. Yu, Y. Li, M. Li, L. Tang and J.-J. He, Biosens. Bioelectron., 2017, 89, 880–885.
243. H. Yousefi, S. E. Samani, S. Khan, A. Prasad, A. Shakeri, Y. Li, C. D. M. Filipe and T. F. Didar, ACS Nano, 2022, 16, 29–37.
244. H. Mann, S. Khan, A. Prasad, F. Bayat, J. Gu, K. Jackson, Y. Li, Z. Hosseinidoust, T. F. Didar and C. D. M. Filipe, Adv. Mater., 2025, 37, e2411173.
245. J. Xu, J. Yu, W. Liu, Q. Jiang, Z. Yu and N. Gan, Anal. Chim. Acta, 2024, 1287, 342053.
246. L. Wang, Y. Wen, X. Yang, L. Xu, W. Liang, Y. Zhu, L. Wang, Y. Li, Y. Li, M. Ding, S. Ren, Z. Yang, M. Lv, J. Zhang, K. Ma and G. Liu, Anal. Chem., 2019, 91, 16002–16009.
247. P. Xu, T. Cao, Q. Fan, X. Wang, F. Ye and E. Eiser, Proc. Natl. Acad. Sci. U. S. A., 2023, 120, e2305995120.
248. R. Zhou, C. Wang, Y. Huang, K. Huang, Y. Wang, W. Xu, L. Xie and Y. Ying, Biosens. Bioelectron., 2021, 188, 113336.
249. S. T. Ten, U. Hashim, S. C. B. Gopinath, W. W. Liu, K. L. Foo, S. T. Sam, S. F. A. Rahman, C. H. Voon and A. N. Nordin, Biosens. Bioelectron., 2017, 93, 146–154.
250. A. K. Paulose, Y.-J. Hou, Y.-S. Huang, N. Chakkalaparambil Dileep, C.-L. Chiu, A. Pal, V. M. Kalaimani, Z.-H. Lin, C.-R. Chang, C.-P. Chen, Y.-C. Lin, C.-Y. Cheng, S.-H. Cheng, C.-M. Cheng and Y.-L. Wang, Anal. Chem., 2023, 95, 6871–6878.
251. R. A. Salinas, A. Orduña-Díaz, O. Obregon-Hinostroza and M. A. Dominguez, Talanta, 2022, 237, 122970.
252. Z. Li, Y. Jiang, S. Tang, H. Zou, W. Wang, G. Qi, H. Zhang, K. Jin, Y. Wang, H. Chen, L. Zhang and X. Qu, Mikrochim. Acta, 2022, 189, 273.
253. Y. Feng, X. Zhang, L. Su, Y. Zhang and F. He, Biosens. Bioelectron., 2019, 138, 111302.
254. Y. Tian, T. Liang, P. Zhu, Y. Chen, W. Chen, L. Du, C. Wu and P. Wang, Sensors, 2019, 19, 5473.
255. A. Holmes, T. J. Dallman, S. Shabaan, M. Hanson and L. Allison, J. Clin. Microbiol., 2018, 56, e01388–e01317.
256. B. Bogaerts, S. Nouws, B. Verhaegen, S. Denayer, J. Van Braekel, R. Winand, Q. Fu, F. Crombé, D. Piérard, K. Marchal, N. H. C. Roosens, S. C. J. De Keersmaecker and K. Vanneste, Microb. Genom., 2021, 7, mgen000531.
257. J. Rumore, L. Tschetter, A. Kearney, R. Kandar, R. McCormick, M. Walker, C.-L. Peterson, A. Reimer and C. Nadon, BMC Genomics, 2018, 19, 870.
258. D. Lambert, C. D. Carrillo, A. G. Koziol, P. Manninger and B. W. Blais, PLoS One, 2015, 10, e0122928.
259. N. Singh, P. Lapierre, T. M. Quinlan, T. A. Halse, S. Wirth, M. C. Dickinson, E. Lasek-Nesselquist and K. A. Musser, J. Clin. Microbiol., 2019, 57, e00307–e00319.
260. L. Chen, Y. Zhao, J. Wei, W. Huang, Y. Ma, X. Yang, Y. Liu, J. Wang, H. Xia and Z. Lou, Microbiol. Spectrum, 2022, 10, e0119522.
261. C. Zhang, H. Cheng, Y. Zhao, J. Chen, M. Li, Z. Yu, X. Sun, P. Li, Y. Shang, J. Ma and J. Zheng, Pol. J. Microbiol., 2022, 71, 499–507.
262. J. Wang, Y. Wang, J. Cao, Y. Chen, J. Yu and N. Sun, Front. Cell. Infect. Microbiol., 2025, 15, 1575824.
263. S. Kai, Y. Matsuo, S. Nakagawa, K. Kryukov, S. Matsukawa, H. Tanaka, T. Iwai, T. Imanishi and K. Hirota, FEBS Open Bio, 2019, 9, 548–557.
264. A. Lamas, A. Costa-Ribeiro and A. Garrido-Maestu, Methods Mol. Biol., 2025, 2852, 65–81.
265. A. Peritz, G. C. Paoli, C.-Y. Chen and A. G. Gehring, Anal. Bioanal. Chem., 2018, 410, 5439–5444.
266. Q. Liu, L. Fang, G. Yu, D. Wang, C.-L. Xiao and K. Wang, Nat. Commun., 2019, 10, 2449.
267. F. Grünberger, S. Ferreira-Cerca and D. Grohmann, RNA, 2022, 28, 400–417.
268. K. Tian, X. Chen, B. Luan, P. Singh, Z. Yang, K. S. Gates, M. Lin, A. Mustapha and L.-Q. Gu, ACS Nano, 2018, 12, 4194–4205.
269. N. Harel, M. Meir, U. Gophna and A. Stern, Nucleic Acids Res., 2019, 47, e148.
270. M. Nishimura, H. Takeyama and M. Hosokawa, J. Biosci. Bioeng., 2023, 136, 152–158.
271. S. Leonard, S. Meyer, S. Lacour, W. Nasser, F. Hommais and S. Reverchon, Nucleic Acids Res., 2019, 47, e88.
272. D. Grinevich, L. Harden, S. Thakur and B. Callahan, mSystems, 2024, 9, e0075723.
273. Z. Xu, Y. Wang, K. Sheng, R. Rosenthal, N. Liu, X. Hua, T. Zhang, J. Chen, M. Song, Y. Lv, S. Zhang, Y. Huang, Z. Wang, T. Cao, Y. Shen, Y. Jiang, Y. Yu, Y. Chen, G. Guo, P. Yin, D. A. Weitz and Y. Wang, Nat. Commun., 2023, 14, 5130.
274. W. Xu, K. Li, Q. Li, S. Li, J. Zhou and Q. Sun, Methods Mol. Biol., 2022, 2528, 445–464.
275. Q. Liu, D. C. Georgieva, D. Egli and K. Wang, BMC Genomics, 2019, 20, 78.
276. T. Blagden, W. Schneider, U. Melcher, J. Daniels and J. Fletcher, J. Food Prot., 2016, 79, 574–581.
277. X. Sun, Y. Wang, L. Zhang, S. Liu, M. Zhang, J. Wang, B. Ning, Y. Peng, J. He, Y. Hu and Z. Gao, Anal. Chem., 2020, 92, 3032–3041.
278. T. Wang, Y. Wang, P. Chen, B. Yin and B. Ye, Anal. Chem., 2022, 94, 12461–12471.
279. L. Wang, X. Shen, T. Wang, P. Chen, N. Qi, B.-C. Yin and B.-C. Ye, Biosens. Bioelectron., 2020, 165, 112364.
280. Z. Weng, Z. You, H. Li, G. Wu, Y. Song, H. Sun, A. Fradlin, C. Neal-Harris, M. Lin, X. Gao and Y. Zhang, ACS Sens., 2023, 8, 1489–1499.
281. X. Han, D. Song, W. Xu, L. Lu, A. Zhu and F. Long, J. Hazard. Mater., 2024, 469, 134037.
282. Y. Liu, Y. Wu, Y. Liu, Q. Zhang, H. Yuan, S. Li, Z. Li, B. Wang, Y. Chang and M. Liu, Anal. Chem., 2025, 97, 9310–9315.
283. J. Luo, D. Xu, J. Wang, H. Liu, Y. Li, Y. Zhang, H. Zeng, B. Deng and X. Liu, Anal. Bioanal. Chem., 2024, 416, 3509–3518.
284. W. Jiang, C. He, L. Bai, Y. Chen, J. Jia, A. Pan, B. Lv, X. Tang and X. Wu, Foods, 2023, 12, 236.
285. T. Tian, B. Shu, Y. Jiang, M. Ye, L. Liu, Z. Guo, Z. Han, Z. Wang and X. Zhou, ACS Nano, 2021, 15, 1167–1178.
286. X. Tan, X. Yang, Y. Qiao, Y. Wei, W. Shang, H. Cai, X. Luo, H. Hou, B. B. Dzantiev, Y. Wan, F. Song and J. Li, Anal. Chim. Acta, 2023, 1271, 341470.
287. J. Lim, A. B. Van, K. Koprowski, M. Wester, E. Valera and R. Bashir, Proc. Natl. Acad. Sci. U. S. A., 2025, 122, e2420166122.
288. J. Kim, G. H. Lee, D. Nam and K. S. Park, Talanta, 2025, 281, 126864.
289. T. Fang, J. Shen, J. Xue, Y. Jiang, D. Guo, J. Yang, X. Kong, X. Xu and X. Wang, J. AOAC Int., 2022, 106, 156–164.
290. W. Yin, K. Hu, Y. Yang, J. Zhuang, Z. Zou, Y. Suo, L. Xia, J. Li, Y. Gui, H. Mei, J. Yin, T. Zhang and Y. Mu, Anal. Chem., 2024, 96, 17941–17949.
291. W. Yin, K. Hu, B. Yu, T. Zhang, H. Mei, B. Zhang, Z. Zou, L. Xia, Y. Gui, J. Yin, W. Jin and Y. Mu, Lab Chip, 2024, 24, 4659–4668.
292. G. Tatulli and P. P. Pompa, Nanoscale, 2020, 12, 15604–15610.
293. Y. Ren, L. Cao, Z. Li, X. Zhang, H. Li, R. Jiao, X. Yu, Z. Liang, G. Li, X. Xia, N. Ling, D. Zhang, J. Cheng and Y. Ye, Food Chem., 2025, 471, 142776.
294. M. Rieu, J. Valle-Orero, B. Ducos, J.-F. Allemand and V. Croquette, Commun. Biol., 2021, 4, 1083.
295. Y. Li, J. M. Artés, B. Demir, S. Gokce, H. M. Mohammad, M. Alangari, M. P. Anantram, E. E. Oren and J. Hihath, Nat. Nanotechnol., 2018, 13, 1167–1173.
296. R. Pandey, D. Chang, M. Smieja, T. Hoare, Y. Li and L. Soleymani, Nat. Chem., 2021, 13, 895–901.
297. J. Ding, N. Yu, X. Wang and W. Qin, Anal. Chem., 2018, 90, 1734–1739.
298. S. Das, P. Barman, R. Chakraborty, A. Upadhyay, A. Sagdeo, P. Kula, M. K. Das and S. S. Roy, J. Colloid Interface Sci., 2025, 683, 79–89.
299. J. Shim, L. Williams, D. Kim, K. Ko and M.-S. Kim, J. Microbiol. Biotechnol., 2021, 31, 1323–1329.
300. D. T. Ha, S. Ghosh, C. H. Ahn, D. J. Segal and M.-S. Kim, Analyst, 2018, 143, 4009–4016.
301. H. Uno, H. Takeuchi, T. Yoshino, T. Tadenuma, Y. Hirakawa, Y. Maeda, T. Taguchi, T. Matsunaga and T. Tanaka, J. Biosci. Bioeng., 2022, 133, 133–139.
302. Y. M. Shlyapnikov, E. A. Malakhova and E. A. Shlyapnikova, Anal. Chem., 2019, 91, 11209–11214.
303. W. Wu, B. T. T. Nguyen, P. Y. Liu, G. Cai, S. Feng, Y. Shi, B. Zhang, Y. Hong, R. Yu, X. Zhou, A. Q. Liu, Y. Zhang, E. P. H. Yap and L. K. Chin, Biosens. Bioelectron., 2022, 215, 114594.
304. M. Geissler, D. Brassard, L. Clime, A. V. C. Pilar, L. Malic, J. Daoud, V. Barrère, C. Luebbert, B. W. Blais, N. Corneau and T. Veres, Analyst, 2020, 145, 6831–6845.
305. S. J. Oh, B. H. Park, G. Choi, J. H. Seo, J. H. Jung, J. S. Choi, D. H. Kim and T. S. Seo, Lab Chip, 2016, 16, 1917–1926.
306. S. Hu, Y. Jie, K. Jin, Y. Zhang, T. Guo, Q. Huang, Q. Mei, F. Ma and H. Ma, Biosensors, 2022, 12, 324.
307. Y. Shang, G. Xing, X. Liu, H. Lin and J. Lin, Anal. Chem., 2022, 94, 16787–16795.
308. J. Lv, W. Liu, F. Wang, L. Xiong, H. Yang, X. Yao, H. Liu and J. Zheng, Anal. Chim. Acta, 2025, 1346, 343751.
309. B. Yang, P. Wang, Z. Li, Q. You, S. Sekine, J. Ma, S. Zhuang, D. Zhang and Y. Yamaguchi, Lab Chip, 2023, 23, 2633–2639.
310. J. M. Haber, P. R. C. Gascoyne and K. Sokolov, Lab Chip, 2017, 17, 2821–2830.
311. A. Chao, Q. Hu and K. Yin, Biosensors, 2025, 15, 230.
312. M. Zhang, J. Liu, Z. Shen, Y. Liu, Y. Song, Y. Liang, Z. Li, L. Nie, Y. Fang and Y. Zhao, BMC Microbiol., 2021, 21, 197.
313. X. Lin, X. Jin, W. Du, X. Shan, Q. Huang, R. Fu, W. Lv, H. Yang, Y. Su and G. Huang, Anal. Methods, 2021, 13, 3569–3576.
314. J. R. Choi, J. Hu, R. Tang, Y. Gong, S. Feng, H. Ren, T. Wen, X. Li, W. A. B. Wan Abas, B. Pingguan-Murphy and F. Xu, Lab Chip, 2016, 16, 611–621.
315. J. Hui, Y. Gu, Y. Zhu, Y. Chen, S. Guo, S. Tao, Y. Zhang and P. Liu, Lab Chip, 2018, 18, 2854–2864.
316. K. T. L. Trinh, T. N. D. Trinh and N. Y. Lee, Biosens. Bioelectron., 2019, 135, 120–128.
317. P. T. Trieu and N. Y. Lee, Anal. Chem., 2019, 91, 11013–11022.
318. N. Li, Y. Lu, J. Cheng and Y. Xu, Lab Chip, 2020, 20, 384–393.
319. C.-Y. Chen, H.-W. Yang, P.-H. Hsieh, C.-H. Hsieh and M.-H. Wu, Biosens. Bioelectron., 2022, 215, 114574.
320. X. Mao, Y. Lu, Z. Gao, J. Zhong, A. Xiao, J. Lin, J. Hu, B. Shu and C. Zhang, Anal. Chem., 2025, 97, 5085–5092.
321. M. L. Rosauer, C. Barnes, K. M. Silbernagel and K. M. Koch, J. AOAC Int., 2022, 105, 1136–1145.
322. G. A. M. Tarr, C. Y. Lin, B. Vandermeer, D. L. Lorenzetti, P. I. Tarr, L. Chui, L. Hartling and S. B. Freedman, Clin. Chem., 2020, 66, 302–315.
323. A. Hahn, M. Luetgehetmann, O. Landt, N. G. Schwarz and H. Frickmann, Trop. Med. Int. Health, 2017, 22, 1371–1376.
324. S. Magro, E. Visentin, A. Costa, M. Penasa, F. Cendron, P. Moroni, E. Chiarin, M. Cassandro, M. Santinello and M. De Marchi, BMC Vet. Res., 2024, 20, 249.
325. A. van der Zee, L. Roorda, G. Bosman and J. M. Ossewaarde, PLoS One, 2016, 11, e0150755.
326. A. Yamazaki, M. Honda, N. Kobayashi, N. Ishizaki, H. Asakura and Y. Sugita-Konishi, J. Vet. Med. Sci., 2018, 80, 706–709.
327. H. Maruoka, A. Hinenoya, N. Yasuda, A. Takeda, S. Inoue, T. Sumi, K. Koitabashi, H. Yasue, K. Kogou and S. Yamasaki, J. Microbiol. Methods, 2019, 161, 111–117.
328. N. Gupta and V. K. Verma, Microbial Technology for the Welfare of Society, 2019, vol. 17, pp. 313–341.
329. A. Rani, V. B. Ravindran, A. Surapaneni, N. Mantri and A. S. Ball, Int. J. Food Microbiol., 2021, 349, 109233.
330. M. Sidstedt, P. Rådström and J. Hedman, Anal. Bioanal. Chem., 2020, 412, 2009–2023.
331. L. Xiao, Z. Li, X. Dou, Y. Long, Z. Li, Y. Liu and Q. Dong, npj Sci. Food, 2025, 9, 157.

---