Project Description

ABSTRACT

Fungal pathogens pose a significant challenge to the agricultural industry and can cause extensive crop damage affecting quality and total yield. Amongst more detrimental variants of these pathogens are fungi belonging to the Fusarium genus, known for its induction of "Fusarium head blight" (FHB) in cereal crops. Early detection of Fusarium infection is essential to control the fungal spread and to minimize crop losses due to FHB. Current detection methods are lacking as they are time-consuming and require trained laboratory technicians. The team is aiming to develop an early warning system for Fusarium infection using antigen binding-dependent fluorescence protein fusion constructs. We will suspend these protein constructs in a sprayable solvent, which can be quickly applied to crops using commonplace equipment. Once this construct is bound to infected wheat, our construct will undergo fluorescence, releasing red light, which we will detect with our automated drone system. In conjunction, we are also developing an automated drone detection system that would incorporate AI and Robotics.The drone would fly over the field to detect all the fungal presence. Following this, farmers can log into the FungiLink to see a heat map of their infected farmland.

History of Agriculture in Canada

Agriculture is deeply rooted in the development and expansion of Canada, with the industry being a substantial driving force behind the Canadian economy. With over 6.3% of Canadian land devoted solely to farming, $134.9 billion was generated in 2021, accounting for 6.8% of Canadian total gross domestic product (GDP) that year (Overview of Canada's ... n.d.). Multiple countries depend on Canadian exports, and in 2021, Canada shipped over $82.2 billion of food products to over 200 different countries, making us the fifth-largest exporter of agri-food and seafood in the world market (Overview of Canada's ... n.d.).

With an ever-increasing population worldwide, demand for food products will only increase proportionally. With food security being one of the biggest problems humanity is facing, it is essential that our agricultural industry functions nominally to ensure we can meet this ever increasing demand and remain competitive in the global market (FAO et al. 2022). Being based out of Alberta, many of our team members have personally been raised on farms or have close friends in the agricultural industry. Therefore, it was essential for us to utilize synthetic biology to help protect agriculture here in Alberta and benefit the rest of the world.

Significance of Agriculture in Alberta

People often think of oil and natural gas when we discuss major Albertan industries. However, people often gloss over the significant agricultural sector in the province that helps lead the industry in Canada. Alberta is essential to Canadian agriculture, contributing the majority of the total farm revenue earned each year (McComb and St. Pierre 2022). Cattle and grains are Alberta’s most prominent agricultural commodities, with grain crops alone generating $9 billion in 2020 (Overview of Canada's ... n.d.). This revenue totals over 40.4% of Albertan farm earnings annually and is a substantial driving force for the provincial and federal economy (McComb and St. Pierre 2022). As such, cereal crops and grains are a vital source of food and income for Alberta and their yields and quality should be closely monitored and protected.

Crop Infections

Grains such as wheat, barley and oats are some of the world’s biggest cereal crops and are readily consumed worldwide across different cultures. They act not only as an essential food source, but can be used in energy production or as feedstock for cattle and other livestock (Alisaac and Mahlein 2023). However, these crops are susceptible to numerous issues, such as environmental or biological conditions, which can negatively affect crop yield. Fungi and other pathogens present a substantial problem and contribute to global losses of up to 21.5% of total wheat yield (Alisaac and Mahlein 2023). Our project focuses on the fungal genus, Fusarium, which causes massive losses for the agricultural industry here in Alberta, but has also been found throughout the world where multiple epidemics have been reported (Alisaac and Mahlein 2023).

Fusarium Infection

Fusarium is a genus of fungi that affect cereal crops such as wheat and oats and consists of three main species (Korimenko 2018). Out of these candidates, our team chose to focus on Fusarium graminearum as it poses the most substantial threat due to its highly aggressive nature, resulting in a significant loss of grain yield and quality. Fusarium head blight (FHB) is the fungal disease caused by these Fusarium species, characterized by premature bleaching, blighting of the heads of wheat and the appearance of orange, spore-bearing structures called sporodochia at the base of the glumes (Government of Saskatchewan n.d). F. graminearum also secretes mycotoxins which jeopardize livestock feed, baking and milling quality of wheat, biofuel production, and malting and brewing quality of barley (Islam et al. 2022). If these mycotoxins accumulate substantially in these crops, ingestion may become toxic to humans and animals (Islam et al. 2022).

Local Scale / Global Scale

FHB has been reported as early as the 1900s in Eastern Canada and has since spread westward towards the provinces of Alberta and Saskatchewan (Karimento 2018). Yield losses due to FHB can total up to 80%, and upwards of $300 million has been lost annually in Canada since the 1990s (Ali and Calpas 2002). The economic impact resulting from FHB infection can total up to $101 per acre of land and poses a substantial cost to producers and consumers alike. FHB is not centralized to only Canada and poses a problem to countries worldwide (Alisaac and Mahlein 2023). European countries exhibited yield losses of up to 50%, and over 9.9 million hectares of farmland were affected in China due to FHB. Latin America reported seventeen epidemics from 1960 to 2012, with yield losses totalling up to 70%. As well, the United States of America, from 1991 to 1996, had economic losses of $1 billion due to FHB and suffered losses of around 25% for crop yield. The substantial risk that FHB poses to the global agricultural market is what inspired our team to try and solve this problem with the power of synthetic biology.

Current Detection Methods and Limitations

Early detection of FHB is essential to prevent the spread of these fungal pathogens and the substantial accumulation of mycotoxins. By detecting the presence of FHB early on, farmers can quickly employ treatment strategies, which may prevent significant loss in crop yields. Numerous methods are currently used to detect and identify the fungal pathogens responsible for FHB.

1. Plate streaking

   

   The simplest and one of the most common methods is to re-isolate the fungus on selective media (Alisaac and Mahlein 2023). Based on the morphological characteristics of the spores and colonies, one can determine the identity of the fungal strain using visual cues. However, this method requires lab technicians well-versed in fungal taxonomy to determine specific fungal species. The process is also quite laborious and time-consuming if screening needs to be done for a multitude of samples.

2. Immunological assays

   

   Immunological assays like enzyme-linked immunosorbent assays (ELISA) take advantage of antibodies that can bind to specific proteins or protein complexes produced by a particular fungus (Alisaac and Mahlein 2023). Yet this method has a substantial drawback it is only genus-specific and will not be able to differentiate between species belonging to the same genus.

3. Polymerase chain reactions (PCRs)

   

   A relatively new molecular biology method that provides both high sensitivity and specificity in plant identification is polymerase chain reactions (PCRs) which allow for the detection of fungal pathogens before any symptoms appear (Alisaac and Mahlein 2023). Unique genomic regions can be used to develop primers for PCR, allowing laboratories to distinguish between species. Yet a substantial drawback for these proposed methods is that they require special lab equipment and trained personnel to perform these tests.

For all the above mentioned tests, seed samples need to be sent to the Canadian Grain Commission (CGC) or to a reputable private testing laboratory, which can be time-consuming and expensive (Ali and Calpas 2002). As such, there is a need for a method that not only allows for the reliable detection of specific fungal species on crops but must have quick results for early detection of FHB. The proposed product must be easy to use and implement so it does not add additional labour hours on top of a farmer's intensive schedule. This is where our team comes in.

Product Summary

Our team has developed the Fungal Early Detection Drone System (FEDDS), a sprayable early warning system for F. graminearum infection in crop fields. FEDDS involves two key components. First is the construction of Fungalescence, a synthetic nanobody construct capable of binding to antigens specific to F. graminearum. Included in our construct are fluorescent proteins that are able to provide a visual signal using fluorescence once bound to the fungus. This nanobody construct will be contained in a sprayable solution where infected wheat will glow red under exposure to light. Second involves Agri-scan, an automated drone detection system detecting fluorescence. The red fluorescence will be detected by the drone where active F. graminearum infection is present. The data will be collected into a heat map which can be accessed by farmers in an easy to use website portal. Not only can we provide a low-cost solution for detecting F. graminearum infection, but can potentially save thousands of dollars for the agriculutral industry in a non-labour intensive manner.

Nanobodies

A major part of our Fungalescence construct is our synthetic nanobody included in our chimera construct. We have designed a synthetic nanobody that is able to recognize an antigen expressed by F. graminearum which will allow our construct to selectively recognize and bind to our fungus of interest with high specificity. Nanobodies (NB, SdAB, VhH) are derivatives of heavy-chain only immunoglobulin G (IgG) antibodies which can bind to a specific antigen. The common conception of antibodies is of a bivalent y-shaped IgG derived from humans or other mammals. However, a wide array of alternative formats exists but certain qualities remain absent (Holliger and Hudson 2005). NBs provide many advantages over regular antibodies and show exciting ways to overcome the limitations normally associated with conventional antibodies.

Nanobodies contain three complementary-determining regions (CDRs) which impart specificity in antigen recognition and these loop domains can be modified to bind to a variety of antigens (Holliger and Hudson 2005; Jin et al 2023; Muyldermans 2013). The size of NBs (12-15 kDa) is much smaller than conventional antibodies and this is less likely to affect the folding of our construct and also allows them to bind to regions of antigens that might have been obstructed otherwise (Holliger and Hudson 2005; Jin et al 2023; Muyldermans 2013). Nanobodies are also easily expressed in high yields in bacteria, mammalian, and plant cells which allows us to potentially generate large quantities of our construct for use (Holliger and Hudson, 2005.; Muyldermans 2013). Nanobodies are highly stable at a wide range of temperatures, being able retain function from storage at -20°C or treatment at 80°C (Holliger and Hudson 2005.; Muyldermans 2013; Zimmermann et al. 2020). Multiple synthetic nanobody libraries have been created and used to find effective binders, thus removing the need for animal immunization associated with standard antibodies (Holliger and Hudson 2005.; Muyldermans 2013; Zimmermann et al. 2020).

Fluorescent Proteins

Funglascence relies on the concept known as Förster or Fluorescence Resonance Energy Transfer (FRET) technology to provide the fluorescence that will be detected by our drone. FRET is a physical phenomenon where a donor fluorophore can be excited and non-radiatively transfer its excitation energy to an acceptor fluorophore nearby, also known as a FRET pair (Bajar et al. 2016). This acceptor fluorophore will be excited by this energy and release its characteristic fluorescence. FRET is highly sensitive to the distance between the two fluorophores and also requires substantial overlap of the donor’s emission and the acceptor’s absorption spectrums. We chose to use fluorescent proteins (FPs) as our FRET pairs due to the ease of fusing these proteins into our constructs via genetic engineering. FPs are also substantially cheaper than conventional fluorescent dyes, which fall outside our scope of funding. Our construct will utilize the mScarlet and ShadowR proteins; mScarlet is amongst the brightest fluorescent proteins (Murakoshi et al. 2019). Red light was chosen as the emitted fluorescence as it contrasts better against a background of plants and farmland with its characteristic greens and browns. ShadowR has no fluorescence, making it an ideal acceptor fluorophore, allowing us to quench mScarlet’s fluorescence in the absence of our fungal antigens.

Nanobody-Fluorescent Protein Chimera Construct

Fungalescence relies on Nanobodies (NBs) for Fusarium recognition and FRET between mScarlet and ShadowR for a visual signal to distinguish between two binding states. Prior to Fusarium exposure, the chimera is in a trapped state where a “fake antigen” is bound to our nanobody. This fake antigen has been engineered to bind to our nanobody with high enough affinity to create a strong bond for FRET interactions, but will dissociate when the actual Fusarium antigen is present. When the nanobody is bound to the fake antigen, a FRET interaction occurs whereby mScarlet fluorescence is quenched using the dark acceptor ShadowR. When native antigens are present, the NB interaction with the fake antigen will be outcompeted with the native antigen and break off the FRET interaction. Hence, this will allow for our mScarlet protein to give off its characteristic emission of red light and thus provide a signal to detect F. graminearum infection.

Application

Our goal is to develop an affordable and reliable detection system. To achieve this, we initiated the development of a prototype using the DJI Matrice 300 RTK drone. This highly robust and versatile drone offers various payload customizations, an extended battery life of up to 2.5 hours, and the capability to execute flight missions, making it an ideal choice for our project. We utilized a multispectral camera attachment to detect the fluorescence of mScarlet, taking advantage of its unique properties.

The flight automation is conducted through DJI's own DJI Pilot software, which allows us to map the flight path over the targeted field. To provide the appropriate excitation for mScarlet, we developed a specialized payload that emits the required wavelength of light. Moreover, the user-friendly controller provided by DJI enables real-time monitoring of the scan through a live feed.

The next step following the development of the first prototype is to automate the detection process using Artificial Intelligence (AI). To achieve this, we are combining a different multispectral camera, the Parrot Sequoia multispectral camera, with the DJI Phantom 4 drone and a Raspberry Pi. By pairing the camera with the Raspberry Pi, we can automate the detection process and record the coordinates of the infection. For this purpose, we are utilizing OpenCV, a programming function library for real-time computer vision. All AI processing will be performed on the Raspberry Pi, with the code written in Python.

The GPS coordinates play a crucial role in generating a heat map and indicating Fusarium infections to the farmers. To obtain these coordinates, we are using the NEO-6M GPS module connected to the Raspberry Pi via a USB connection. Our Python program will read the data from the USB port, which contains the received GPS data, once a fluorescence of the appropriate wavelength is detected. The latitude and longitude will be extracted from this data and saved in a comma-separated values (CSV) file. These coordinates will then be accessed by FungiLink, a web application developed using Django and hosted on Amazon Web Services (AWS). FungiLink is deployed on an Amazon Elastic Compute Cloud instance and utilizes multiple Application Programming Interfaces (APIs). It serves as a centralized platform where users, those performing scans for the farmers, can upload the CSV file, and farmers can log in to view heat maps of the infections in their fields.

Expanding to Other Infectious Microorganism and System Improvements

The adaptable nature of our system lies in its rapid tunability. Due to the nature of our chimera construct, the placement of various restriction enzyme cut sites allow for easy modification and potential optimization. Regulatory sequences such as promoters, ribosome binding sites (RBS) and terminators can all be swapped out with relative ease. This can allow us to further hone and maximize efficiency of our construct while still ensuring fitness of our construct-producing vector. Reducing linker lengths can affect the FRET resonance as the process relies heavily on the distance between the FRET pair fluorophores (Bajar et al. 2016). The fluorescent proteins can also be swapped around where we can try to increase the FRET efficiency and the quantum yield allowing us to maximize the brightness of the emitted fluorescence. We can also change the colour of the emitted fluorescence allowing us to visualize the construct in other scenarios. Through the efficient alteration of our nanobody for antigen specificity, we can derive new fungal targets and theoretically detect other crop pathogens. As such, our project opens a whole realm of possibilities for pathogen detection in the world of agriculture.