Continued outbreaks of foodborne illness involving dairy products in the United States stress the importance for rapid methods of detection of pathogenic microorganisms in food processing environments. Pathogenic microorganisms, such as Salmonella are widespread, and can be found in a variety of foods, ingredients and in industrial environments. The presence of pathogens in dairy products constitutes great risk for increased exposure, illness and reduces overall quality of the foodstream. As a result, emphasis has been placed on adapting or developing sensitive techniques to rapidly detect notable pathogens, such as Salmonella, Listeria monocytogenes and Escherichia coli O157:H7 in both contaminated foods and industrial environments. Common assays employed in the detection of pathogenic microorganisms, though effective in identification, are time consuming and may require several days for processing. The necessity to quickly screen food products and industrial environments has led to an emphasis to develop rapid, sensitive, automated techniques in food processing operations. Numerous methods of identification and detection have been implemented in food processing environments.

An optimal approach to the rapid detection of microbial pathogens would incorporate several advantages including: 1) improved time-to-result, 2) low-cost, 3) ease of operation and 4) simple interpretation. Such an approach may enable simple and cost-effective sampling of pathogenic microorganisms, which can be used to improve industrial efficiency. As a possible alternative to existing detection efforts, low-cost diagnostic (LCD) tools, particularly paper-based analytical devices (PADs), may be employed for rapid, sensitive and selective detection. PADs are frequently combined with colorimetric detection, in which chromogenic substrates are used to yield a visual representation of detection. Different enzyme-substrate pairs may be employed to accomplish various goals—from simple “presence/absence” to species-specificity. While “presence/absence” is limited, the use of shared enzymes is advantageous during detection and identification of metabolic state. Depending upon environmental factors, bacteria may exist in active or dormant states; reversion of a pathogen from dormancy to a metabolically active state may result in rapid growth and instances of illness.

As the level of enzymatic expression varies between metabolic states, oxidoreductases and alkaline phosphatases (ALP) were investigated as vehicles for colorimetric detection. Oxidoreductases are present in greater amounts in metabolically active bacteria, and are capable of reducing 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-phenyl-2H-tetrazolium chloride (INT) to formazans. Nitrophenyl phosphate (PNPP) is present in dormant bacteria, and cleaves phosphate groups from para-nitrophenyl-phosphate salts, resulting in para-nitrophenol. Combined use of enzymatic substrates, including INT and 5-methylphenazin-5-ium methyl sulfate (PMS) for metabolically active bacteria, and INT and PNPP for dormant bacteria, yielded an improved colorimetric readout visible by eye within 30 min. With detection achieved within 30 min, the two assays, INT-PMS and INT-PNPP, decrease time-to-result, are portable and may be amenable to on-site detection in agricultural, environmental and industrial settings.

While the use of non-specific bacterial enzymes may limit some applications, immobilization of bacteria-specific bacteriophage (P22, T4) onto paper can provide an additional layer of specificity. Bacteriophage are robust, and may be easily absorbed onto paper. In this work, immobilized bacteriophage facilitated specific capture of Salmonella Typhimurium on paper, followed by detection of metabolic state with either the INT-PMS or INT-PNPP assay. This combined approach can be applied to the analysis of mixed cultures, given the generally genera-specific nature of the selected bacteriophages. Moreover, the use of chromogenic substrates simplifies assay design, as color change is easily interpreted by the eye or with basic instrumentation. However, despite these advantages, the requirement for a 48-hour absorption period represents a drawback, lengthening time-to-result.

An alternative to the use of bacteriophage for cell capture are magnetic ionic liquids (MILs). MILs are magnetoactive “molten salt” solvents, containing a paramagnetic component integrated into the cation or anion moiety of the salt. MILs are considered “green” solvents, and are nonvolatile, nonflammable, with tunable physicochemical properties. Due to their hydrophobic and liquid nature, MILs can be quickly be distributed with agitation (stirring or vortexing) throughout aqueous food samples as liquid micro- or nanodispersions. After encountering and binding bacterial cells, cell-MIL complexes can then be collected magnetically or after density-driven sedimentation for further processing. MIL-based capture of bacteria has been previously combined with real-time polymerase chain reaction (qPCR) for the rapid detection of E. coli. While use of qPCR obviates the need for time-consuming steps such as gel electrophoresis, its inherent complexity and cost may prohibit its use in point-of-care or resource-limited settings. Isothermal methods for nucleic acid amplification, such as recombinase polymerase amplification (RPA), may have considerable advantages as alternatives to PCR. RPA results in exponential amplification of nucleic acids and operates at a constant, near-physiological temperature (~40°C), eliminating the need for a thermocycler, generating target-specific amplicons in less than 20 min.

The combined use of MIL-based extraction and rapid, streamlined pathogen detection using RPA was investigated. The ability of MIL solvents to quickly extract Salmonella Typhimurium was first examined by dispersing MIL into an aqueous suspension, followed by rapid (~30 s) physical enrichment (concentration) and extraction using an applied magnetic field. Following extraction, viable bacteria were desorbed from the MIL extraction phase with exposure to a nutrient-rich broth (Luria Bertani medium), referred here to as a “back-extraction” step. In efforts to improve back-extraction, recovery of the model Gram-negative bacterium Serratia marcescens from the MIL extraction phase was investigated using several back-extraction media varying in ionic strength and nutrient composition. The highest recovery of cells was obtained using a nutrient-rich tryptone medium supplemented with NaCl. This modification of the extraction protocol enabled improvement in MIL-based bacterial concentration, enriching cells by a factor of 5 - 6X within 3–5 min.

The improved MIL assay was then examined in conjunction with RPA for rapid detection of Salmonella Typhimurium. MIL-based sample preparation was compared with use of a commercial sample preparation solution, PrepMan® Ultra Sample Preparation Reagent (PMU), for detection of Salmonella Typhimurium in artificially-contaminated pasteurized foods. PMU is commonly coupled with PCR to eliminate or inactivate PCR inhibitors and uses both heating and centrifugation steps. As an established method for sample preparation, use of PMU served as a benchmark method against which our MIL-based process was compared. In aqueous suspensions of Salmonella Typhimurium, detection was achieved as low as 103 CFU mL-1 using the combined MIL-RPA approach, which is equivalent to the previously investigated MIL-qPCR method, and, in our hands, outperformed the PMU method by an order of magnitude. Visualization of amplified products was achieved using gel electrophoresis or lateral flow readouts. Nucleic acid lateral flow immunoassays (NALFIA) require less than 5 min for amplicon visualization, are portable, require minimal technical expertise during interpretation and are easy to implement outside of laboratory settings. The need for electric-based heating elements for RPA incubation was eliminated through the use of low-cost, portable, supersaturated sodium acetate heat packs. This repurposing of consumer-grade hand warmers for nucleic acid amplification is a novel approach and easily incorporated into the MIL-RPA scheme.

While MILs have been successfully used for capture and concentration of bacteria from foods prior to culture- or nucleic acid-based detection, little is known about their interactions with bacteria—including modes of physical association or potential antimicrobial activities. Further understanding these interactions may facilitate optimization of MIL-based capture in challenging food matrices, as well as modification of downstream procedures to mitigate the impacts of potential bacterial injury during extraction and concentration. To begin this work, a series of multi-strain panels, including seven representative Salmonella DNA subgroups and eight strains of E. coli O157:H7, were exposed to the Ni(II) MIL and plated in parallel on non-selective and selective media. Calculated enrichment factors (EF) were similar between media types, while individual cell counts were nearly identical, suggesting that the Ni(II) MIL, as applied during our capture and concentration assay, does not cause assay-limiting cellular injury in these two pathogens. Observed variability between EF values may result from differences in the extraction efficiency of the MIL, with some strains exhibiting weaker affinity for the MIL compared to other strains tested, which is an area of ongoing research. Importantly, our results demonstrate capture and recovery of strains representative of all seven Salmonella DNA subgroups and all eight strains of E. coli O157:H7 tested, with comparable recovery on non-selective and selective media. This initial and ongoing research on characterization of MIL-bacterial interactions establishes the foundation for further evaluation of new MIL structures for improving the preconcentration and recovery of viable microorganisms from complex food matrices.