Field evaluation of microfluidic paper-based analytical devices for microbial source tracking

Foodborne pathogens on fresh produce can lead to serious health issues. The source of these pathogens is often wild animals or animal feeding operations. Current methods for assessing the risk of animal-source contamination on fresh produce have limitations because they are costly, time-consuming, and lab-based. This project will develop a novel growers’ risk assessment biomarker investigative tool that can be used in the field and provide results of animal-source contamination within an hour (instead of waiting for days with traditional methods). This novel tool will be based on methods for detecting DNA from feces of animals and it will distinguish between different animal hosts (swine, poultry, ruminants). The tool builds on paper-based devices (similar to a pregnancy test) that are simple-to-use, low-cost, and portable. To validate this tool, it will be tested in a variety of conditions: controlled lab setting, around animal feeding operations, during the growing season in the field, and during the harvesting season in the field. As a result of this project, specialty crop operations will have a new method of assessing the risk of contamination in their product. Managing risk from animal sources will ultimately lead to safer food and fewer foodborne illnesses.

Animal-specialty crop interface poses risk factors that can lead to the contamination of fresh produce with foodborne pathogens. Current practices include collecting samples from the crop and sending them to a lab for testing. These methods, typically based on culturing or quantitative polymerase chain reaction, are slow and can require several days to provide a result. Thus, it is challenging to obtain detailed spatial information (for example, a heat map of proximity zones from animal feeding operations to plan growing strategies) or to obtain rapid results (for example, on an hourly basis during harvest to determine when the tools should be cleaned). In this proposal, our overall goal is to build and test a field-deployable microfluidic paper-based analytical device to determine contamination events. We will determine the levels of fecal and pathogenic contamination that are naturally present in fields, on harvest tools, and around animal feeding operations to establish a baseline (Objective 1). We will also optimize our microfluidic paper-based analytical device by testing samples in the lab and the field (Objective 2). We will use Bacteroidetes as indicator organisms for fecal contamination because of differences between the types of Bacteroidetes found in different hosts (humans, poultry, swine, and ruminants). These bacteria are currently used for tracking the source of fecal contamination in water and this process is known as microbial source tracking. To detect these bacteria, we will target specific sequences of DNA within their genome using an isothermal amplification method call loop-mediated isothermal amplification. Since this method only requires heating at a constant temperature (instead of cycling of temperature as in polymerase chain reaction), it is easier to design a portable device to conduct DNA amplification. Additionally, storing the reagents required for loop-mediated isothermal amplification on microfluidic paper-based analytical devices makes the assay user-friendly because the user does not need to handle multiple reagents. Using the same method, we will also detect Shiga toxin-producing Escherichia coli and Salmonella to enable a tool for making decisions during harvest. For objective 1, we will collect almost 2000 samples in the form of culture swabs from romaine lettuce, harvest tools, and animal feeding operations. We will characterize the fecal and pathogenic bacterial levels in these samples using an established method (quantitative polymerase chain reaction) and compare it to our novel method (quantitative loop-mediated isothermal amplification). With these data, we will be able to provide insight into methods for characterizing proximity zones around animal feeding operations and animal intrusion. For objective 2, we will test 150 samples in the lab and 200 samples in the field after optimizing the microfluidic paper-based analytical devices and portable detection system. We will compare the results of our portable system to a commercially available lab-based system and aim to provide similar results in under an hour. The anticipated outcome of this project is a validated field-deployable growers’ risk assessment biomarkers investigative tool that enables science-based approaches to measure and manage risk at the interface of animals and specialty crops.

1. Establish background levels of fecal and pathogenic contamination in the field to determine the limits of detection that are needed for a field-based assay. 

2. Design and test a portable microfluidic paper-based analytical device (µPAD) that can detect contamination and provide results within an hour in the field.

The following were the most important findings: 

1. We were the first to report baseline studies surveying the background concentration of Bacteroidales as a fecal contamination indicator in fresh produce fields and also adjacent to animal feeding operations. As expected, the concentrations of Bacteroidales in the sampled commercial fields were very low (0 – 2.00 copies/cm2). On the other hand, fields adjacent to animal feeding operations have Bacteroidales concentrations over 104 copies/cm2. 

2. The harvester samples showed that Bacteroidales concentrations on harvesters were higher (0 – 100 copies/cm2) compared to those observed in the fields. However, this concentration should still be considered as being within acceptable safety limits. 

3. Host-specific Bacteroidales constitute a small proportion of the total Bacteroidales population and are thus not always present when fecal contamination occurs, particularly at low levels of contamination. When choosing a host-specific marker, there is a trade-off between specificity and sensitivity; where a decreased cross-reaction rate between various hosts might also affect the sensitivity of the assay. Therefore, for studies attempting to detect fecal contamination, we strongly recommend using not only host-specific Bacteroidales markers but also a universal Bacteroidales marker to reduce the incidence of false negatives. 

4. We developed a fully-integrated LAMP testing platform which included components such as heating, imaging, fluid delivery, and paper-based LAMP assay, and deployed it on a commercial lettuce farm. The unit, operating in the back (trunk) of a car, was powered by a portable power station. Our platform enabled in-situ identification of fecal contamination within 60 min of sampling. We are the first to implement a portable paper-based LAMP testing platform in fresh produce farms. It serves as an enabler for establishing future nucleic acid amplification tests (NAATs) as part of standard growing and harvesting practices during fresh produce production.