A Review of Isothermal Amplification Methods and Food-Origin Inhibitors against Detecting Food-Borne Pathogens

The isothermal amplification method, a molecular-based diagnostic technology, such as loop-mediated isothermal amplification (LAMP) and recombinase polymerase amplification (RPA), is widely used as an alternative to the time-consuming and labor-intensive culture-based detection method. However, food matrices or other compounds can inhibit molecular-based diagnostic technologies, causing reduced detection efficiencies, and false-negative results. These inhibitors originating from food are polysaccharides and polyphenolic compounds in berries, seafood, and vegetables. Additionally, magnesium ions needed for amplification reactions can also inhibit molecular-based diagnostics. The successful removal of inhibitors originating from food and molecular amplification reaction is therefore proposed to enhance the efficiency of molecular-based diagnostics and allow accurate detection of food-borne pathogens. Among molecular-based diagnostics, PCR inhibitors have been reported. Nevertheless, reports on the mechanism and removal of isothermal amplification method inhibitors are insufficient. Therefore, this review describes inhibitors originating from food and some compounds inhibiting the detection of food-borne pathogens during isothermal amplification.


Introduction
Food-borne pathogens, such as Salmonella spp., Escherichia coli O157:H7, and Listeria monocytogenes, threaten public health, causing different food-borne diseases and deaths [1]. Since small amounts of bacteria are found in food, detection methods capable of detecting even small bacterial quantities are needed [2]. The culture-based detection method, a gold standard for pathogen detection, is time-consuming and labor-intensive [3]. Therefore, polymerase chain reaction (PCR), which was developed as a rapid detection technology based on nucleic-acid bond alignment, greatly reduces detection time, and improves the sensitivity and specificity to increase the detection efficiency of food-borne pathogens [4,5]. PCR amplifies nucleic acids through denaturation, annealing, and expansion steps during different temperature cycles, but different temperature cycles can interfere with the nucleic acid amplification process [6][7][8].
An isothermal amplification method was developed to be conducted at a single temperature, which was subsequently improved by replacing the PCR thermocycling step with Bacillus stearothermophylus (Bst) DNA polymerase, phi DNA polymerase, helicase, and RNase H enzymes with strand displacement activity [9,10]. The loop-mediated isothermal amplification (LAMP), recombinase polymerase amplification (RPA), helicase-dependent amplification (HDA), nucleic acid sequence-based amplification (NASBA), rolling circle amplification (RCA), and multiple strand displacement amplification (MDA) are examples of isothermal amplification methods [11][12][13][14][15][16]. Since the isothermal amplification method the thermophilic form of HDA (tHDA). The mHDA reacts at a medium temperature of 37 • C using UvrD helicase/Exo-Klenow polymerase; the tHDA reacts at a higher temperature (60-65 • C) using a thermostable Tte-UvrD helicase/Bst DNA polymerase [39]. Compared to mHDA, tHDA has a higher sensitivity and efficiency, and simplifies the reaction because it does not need MutL and SSB proteins to stabilize the DNA sequence [11]. The magnesium ion, which serves as a cofactor for the helicase and polymerase, is also used in HDA to increase the enzyme activity, thereby making the enzymes compatible with structurally modified primers [9,40]. Subsequently, amplification products are detected using gel electrophoresis, colorimetry assay, LFA, and real-time monitoring [11].
NASBA targets 16s rRNA genes or mRNA transcripts for bacterial detection, enabling the analysis of bacterial viability [16,41]. NASBA amplifies single-stranded RNA using two primers and three enzymes (avian myeloblastosis virus reverse transcriptase (AMV-RT), RNase H, and T7 DNA-dependent RNA polymerase (DdRp)) [42]. While AVM-RT generates complementary complementary DNA by extending primers, dsDNA is formed through RNase H. However, T7 DdRP recognizes the exposed T7 promoter of the dsDNA and initiates transcription to initiate the reaction [42]. Since NASBA enzymes are heat labile, amplifications can be performed at a relatively low temperature, with optimal conditions of 41 • C for 1.5-2 h [43]. The low reaction temperature of NASBA, however, can produce to result in false-positive results because of nonspecific primer interactions. Nevertheless, adding DMSO and betaine can prevent this limitation [44,45]. Subsequently, amplification products are detected using gel electrophoresis, enzyme-linked immunosorbent assay (ELISA), enzyme-linked gel assay, electrochemiluminescent (ECL), and real-time monitoring with molecular beacons [9,46]. When the clustered regularly interspaced short palindromic repeats (CRISPR)/Cas system and NASBA are combined, RNA can be detected with high sensitivity through the NASBACC (NASBA-CRISPR Cleavage) system [47].
RCA has high specificity and sensitivity targeting RNA and DNA [48]. The RCA reaction is conducted using DNA/RNA polymerase (phi29 DNA polymerase or T7 RNA polymerase), short DNA or RNA linear single-stranded primer, circle template, and ligase [49]. DNA/RNA polymerase produces a long single-stranded RCA product (RCAP) complementary to a circular template. Unique oligonucleotide padlock probes (PLP) and T4 DNA ligase or special ssDNA ligase synthesize bacterial single-or double-stranded RNA/DNA templates into single-stranded circular DNA [9]. The various modified RCA systems have been developed for efficient amplification: a linear-RCA (LRCA) or an exponential-RCA (ERCA), such as multiply-primed RCA (MPRCA), hyperbranched RCA (HRCA), and primer-generation RCA (PG-RCA) [42]. The saltatory RCA (SRCA), a simplified form of RCA, has been developed, requiring ligase and PLP for cyclization [14,50]. RCA generally reacts for 1-1.5 h at 30-65 • C depending on the reaction system, and amplification products are detected using gel electrophoresis, colorimetry, and real-time monitoring [42].
MDA randomly and massively amplifies single-cell genomic DNA and is compatible with whole genome amplification (WGA) [51,52]. MDA is conducted using modified random hexamer primers, phi29 DNA polymerases (strand-displacing DNA polymerase from bacteriophage Ø29), denatured template DNA, and dNTPs [53]. The modified random hexamer primers eliminate the need to design target-specific primers are designed to anneal to random areas on each strand of the target DNA, thereby forming hyperbranched intermediates and dsDNA amplicons after exponential amplification [54]. The phi29 DNA polymerase enables amplification at a relatively low temperature (typically 30 • C) because of its high strand displacement activity; it has a higher replication fidelity and lower error rate than Taq DNA polymerases and Bst DNA polymerases [13,55]. Furthermore, the addition of PEG to the reaction for high-efficiency MDA causes molecular crowding, which enables sensitive allele detection in multiplex short tandem repeat genotyping [56].
Schematic diagram and summary of isothermal amplification techniques are shown in Figure 1 and Table 1. molecular crowding, which enables sensitive allele detection in multiplex short tandem repeat genotyping [56].
Schematic diagram and summary of isothermal amplification techniques are shown in Figure 1 and Table 1.   High [57] Abbreviations used in the table: LAMP, loop-mediated isothermal amplification; RPA, recombinase polymerase amplification; HDA, helicase-dependent amplification; mHDA, mesophilic form of HDA; tHDA, thermophilic form of HDA; NASBA, nucleic acid sequence-based amplification; RCA, rolling circle amplification; MDA, multiple strand displacement amplification.

Inhibitors That Originate from Food during the Isothermal Amplification Process
Isothermal amplification methods, with a high detection efficiency, sensitivity, and specificity using geometrically amplifying DNA/RNA, have been widely used to detect food-borne pathogens ( Table 2) [58]. However, complex food matrices can inhibit the detection of these food-borne pathogens. Therfore, food-origin inhibitors possess inhibitors that make the DNA/RNA extraction challenging, denature DNA or polymerases, or bind magnesium ions to interfere with amplification [22,24,25]. Furthermore, it has been reported that these inhibitors are complex polysaccharides, phenolic compounds, and calcium ions [19,26].
Polysaccharides in vegetables, fruits, and seafood can inhibit nucleic acid amplification [19]. Cationic polysaccharides can therefore interfere with DNA amplification initiation to inhibit DNA polymerase activity, thereby inducing competitive binding at the primer's binding site owing to their binding properties to anionic particles, such as the template DNA of bacteria [27]. Interference is due to the electrostatic interaction of the anionic phosphate and carboxyl groups in the lipopolysaccharide (LPS) of the extracellular membrane with the cationic molecule, particularly in gram-negative bacterial cells [27,59]. Similar to this inhibition mechanism, chitosan, contained in crab and shrimp, inhibited LAMP by aggregating negatively charged template DNA molecules [27]. A study also reported that polysaccharides in rice interfered with LAMP by inhibiting activity of Bst DNA polymerase [60]. Additionally, polysaccharides can inhibit nucleic acid amplification by forming high-viscous DNA solutions through mechanical spatial entanglements with DNA [61]. This inhibition works because DNA can be denatured by viscous formation processes. High-viscous DNA solutions can inhibit subsequent DNA quantification and amplification [62]. In a previous study, the column/membrane of the DNA extraction kit was blocked because of the highly viscous DNA solution formed by soybean polysaccharides, thereby leading to low DNA yield [63]. Recently, commercial DNA extraction kits have increased the DNA extraction yield by suppressing the formation of viscous DNA [64,65]. However, since this method controls the high-viscosity of a DNA solution formed after the inhibition mechanism of polysaccharides, it is impossible to fundamentally remove the inhibition before forming the high-viscosity DNA solution that interferes with amplification. Therefore, future research should alleviate the inhibitory mechanism of polysaccharides in food in advance.  Phenolic compounds contain secondary metabolites, such as polyphenols, tannic acids, caffeic acids, and chlorogenic acids, found in strawberries, citrus fruits, and potatoes [19,101]. These secondary metabolites can react with primary reactive residues of enzymes/proteins during amplification to change their intrinsic properties [102]. Consequently, the extraction of high-purity DNA is complex, and amplification can be inhibited [19,26]. Tannins are readily oxidized to bind proteins and nucleic acids irreversibly, thereby forming high molecular weight complexes, which makes DNA extraction difficult [103,104]. It is possible to prevent oxidation of polyphenols by adding polyvinylpyrrolidone and sodium acetate, and by adding soluble PEG to prevent the formation of free radicals, high-purity DNA can be obtained [105,106]. However, besides this, phenolic compounds can inhibit amplification by interfering with the activity of DNA polymerases, DNA restriction enzymes, and RNA-dependent reverse transcriptases [19,57].
High concentrations of calcium ions in milk can inhibit nucleic acid amplification through competitive binding to DNA polymerases with magnesium ions, a cofactor of DNA polymerase and helicase [104]. Milk-based amplification inhibition occurred during MDA for detecting S. Typhimurium and Cronobacter sakazakii [107]. Additionally, MDA was amplified in the presence of 14-16% milk, which was even lower than the quantity which RPA was inhibited (15-25%), thereby confirming the weak resistance to food matrices. Additionally, Santiago-Felipe et al. [57] reported that this inhibition could be because of calcium ions in milk.

Inhibitors Originating from Isothermal Amplification Processes
Various inhibitors repress the isothermal amplification technique for detecting bacteria, including food-borne pathogens. These inhibitors originate from amplification or detection processes and the food metrix (Table 3). Bst DNA polymerase and helicase that enables isothermal amplification require a relatively high concentration of magnesium ions of about 4-8 mM compared to Taq DNA polymerase [38]. However, magnesium ions can inhibit molecular amplification depending on the concentration in the reaction mixture [110]. Murakami et al. [110] reported that by in- creasing the background signal amplification in RCA, the concentration of magnesium ions inhibited the signal during target gene amplification. Nb.BsmI, a nicking enzyme, needs magnesium ions to enhance amplification efficiency. Therefore, optimizing magnesium ion concentrations (4-6 mM MgSO 4 ) and decreasing dNTPs and DNA polymerase concentrations can decrease background signal amplifications. Doseeva et al. [109] conducted a study to alleviate magnesium ion-dependent inhibition because the concentration of magnesium ions affected the amplification efficiency of tHDA. Results indicated that as the concentration of magnesium ions and that of the dATPs improved, the signal-to-noise ratio increased 1.5-2.0 times. It was also observed that the optimal concentrations of MgSO 4 , dATP, and dNTP were 4, 3, and 0.4 mmol/L, respectively. Additionally, betaine, DMSO, and sorbitol, which help the combined effect of magnesium ions and dATP during molecular amplification, were added to increase the efficiency and specificity of amplification in tHDA.
The inappropriate concentration of primers can inhibit molecular amplification processes. Thus, SSB proteins either inhibit the strand exchange activity of recombinase T4 UvsX or compete with recombinase proteins in RPA [118]. Additionally, primers for one target can inhibit the amplification of another target [32]. Therefore, it is essential to optimize primer concentration [32,36].
DNA/RNA templates as amplification targets can also inhibit molecular amplification. RPA reactions are more sensitive to inhibitors when the DNA concentration of the reaction mixture for amplification is close to that of the detection limit or when the background DNA concentration is high [36]. Additionally, in the initial denaturation step for denaturing the secondary structure of mRNAs before NASBA amplification, short or partially digested DNA molecules can be denatured and inhibited, thereby acting as a substrate for NASBA [118]. Simpkins et al. [84] overcame the induction of mRNA degradation using RNase enzymes in cells through pasteurization and a 15 min incubation before the nucleic acid extraction. The combination of RNase A, treatment, and molecular labeling based real-time NASBA assay successfully detected viable L. monocytogenes cells in meat and salmon products without an initial denaturation step that can cause inhibition. Additionally, Sidoti et al. [112] suggested that adding 60% DMSO to NASBA primers and beacon mixtures, although not food samples, improved results through oligonucleotide stabilization maximization.
The colorimetric method, which is widely used to identify amplicons after amplification, can be inhibited using some compounds, such as intercalating dyes or cationic polymers [22]. SYBR Green I is a fluorescent dye widely used for colorimetric isothermal amplification detection because of its high detection sensitivity of amplified nucleic acids [15,114,119]. Therefore, it has been reported that if such a fluorescent dye was added directly to the reaction mixture before LAMP and reacted during amplification, the amplification process will be inhibited. However, if a fluorescent dye was added after amplification, the amplification result can vary owing to cross-contamination and aerosol contamination [114,120]. This inhibition was due to the strong bonding of the dye to ssDNA, and the ability to bind LAMP and loop primers in real-time to prevent primer annealing or interference with the activity of Bst DNA polymerase [121]. Hence, to prevent this inhibition, studies have allowed these fluorescent dyes to react after amplification in a closed tube system [114,122]. Wax capsules containing fluorescent dye have been reported too. They preserve the fluorescent dyes before amplification and allow the fluorescent dyes to react with amplicons when the wax melts after amplification [114]. Additionally, a study was also reported in which the fluorescent dye was applied to the inner part of the tube's cap. Then the tube was rotated after amplification to allow the fluorescent dye to react with the amplicon [121]. Similarly, nonmutagenic noncytotoxic fluorescent intercalator (EvaGreen) had less amplification inhibition capacities compared to SYBR Green I because of its characteristic of fluorescing only when bound to dsDNA. Thus, it was suggested that it could replace SYBR Green I [11,123].
DNA-binding dyes can also inhibit detection after nucleic acid amplification [122]. When polyethyleneimine (PEI), a cationic polymer widely used as a nucleic acid precipitating agent for nucleic acid purification, is added to the amplicon to precipitate the insoluble LAMP amplicon-PEI complex, it enables the sequence-specific visual detection of trace quantities of nucleic acids [124]. However, since the insoluble DNA-PEI complex neutralizes the negative charge of DNAs and consequently inhibits LAMP, it should be added after amplification [115]. Metal ion indicators, such as MnCl 2 and calcein, can also be added to the reaction mixture before amplification, thereby enabling amplification and detection to be conducted in a closed tube system. Wastling et al. [115] compared the sensitivity of LAMP using a metal ion indicator and calcein supplemented with MnCl 2 . Results indicated a lower LAMP sensitivity than calcein alone.
During electrochemical monitoring developed to digitize amplicon detection, amplification inhibition is proposed to occur due to partial defects in redox active compounds [22]. Methylene blue (MB) has a lower LAMP amplification efficiency than other redox molecules with binding activity (10 4 -10 5 M −1 ) to dsDNA, a Hoechst 33258 redox molecule, which is unsuitable for real-time monitoring of LAMP, as it limits DNA amplification and detection in the solution phase, and strongly inhibits enzyme activity of polymerase for endpoint DNA detection [116]. Replacing DNA dyes can improve this LAMP inhibition. Osmium redox showed high binding activity and stability during isothermal amplification. In contrast, real-time electrochemical detection without inhibition detected ruthenium hexamine (RuHex) because it lacked an intercalating ligand and bound electrostatically to the anionic dsDNA backbone [125,126]. Ahmed et al. [127] used RuHex to successfully detect without inhibition because it can be used as an indicator to monitor LAMP amplicons instead of immobilizing the probe to the electrode's surface. Additionally, the voltammetry mode, developed based on the electrochemical reaction of free dGTP molecules in the carbon nanotube array electrode, can be used without inhibition because no secondary indicator exists. Polydopamine-doped paper disks eliminated LAMP inhibitors by reacting with milk components and calcium ions, thereby enabling the simultaneous multiplex detection of food-borne pathogens without inhibition [43,117,120]. Alternatively, the electrochemical monitoring of HDA uses an electroactive intercalator instead of a fluorescent probe. Here the redox probe inhibits the amplification of the specific sequence of the E. coli plasmid, so that the amplification rate of electrochemical monitoring would be slower than that of fluorescent real-time amplification [11].

Conclusions
Inhibitors originating from food matrices, DNA extraction reactions and nucleic acid amplification reactions can adversely affect isothermal amplification techniques by inhibiting amplification to detect food-borne pathogens. Nevertheless, due to the complex nature of the food matrix, food-origin inhibitors, including inhibition mechanisms for molecular analysis, have not been fully characterized yet. Additionally, studies on the inhibitory mechanisms of inhibitors obtained during isothermal amplification processes are still insufficient and have not been clearly established. The above-mentioned removal strategies for inhibiting nucleic acid amplification are rather limited and uneconomical because they only target specific inhibitors, and have not been proven to be applicable to various inhibitors. Therefore, further studies on nucleic acid amplification inhibitors and inhibitor removal strategies are needed to detect food-borne pathogens in food using isothermal amplification technologies without inhibition of nucleic acid amplification inhibitors. In the future, these studies will lead to the manufacture of a ready-to-use kit that simultaneously purifies and removes inhibitors. In this case, efficiency is guaranteed for accurate detection of food-borne pathogens in complex food matrices and for the POCT application of these isothermal amplification technologies.

Conflicts of Interest:
The authors declare no conflict of interest.