Skip to main content
ARS Home » Midwest Area » Peoria, Illinois » National Center for Agricultural Utilization Research » Mycotoxin Prevention and Applied Microbiology Research » Research » Research Project #438643

Research Project: Innovative Approaches to Monitor, Predict, and Reduce Fungal Toxins

Location: Mycotoxin Prevention and Applied Microbiology Research

2023 Annual Report

The goals of this project are to reduce exposure to such toxins and to enhance food safety through the development of tools to more effectively monitor for natural toxins. The first goal will be addressed by development of materials capable of being used in the removal of toxins from foods, thereby reducing exposure. The second goal involves surveying one important commodity, oats; improving methods for toxin detection; and, in particular, developing methods to allow for the prediction of toxin contamination. To meet these goals, we have three objectives. Objective 1. Develop materials and methods to maintain commodity value and safety by eliminating mycotoxin contamination. Sub-objective 1.A. Develop antifungal compounds. Sub-objective 1.B. Remediation. Objective 2. Determine the occurrence of mycotoxins in alternative grain commodities, for example the U.S. oat supply. Determine the fungi and oat cultivars associated with reduced mycotoxin levels and/or disease. Sub-objective 2.A. Survey U.S. oats for mycotoxin contamination. Sub-objective 2.B. Survey U.S. oats for fungal contamination. Objective 3. Develop analytical tools to predict and evaluate the presence of natural toxins (mycotoxins) in grain commodities and their related foods. Sub-objective 3.A. Develop tools to predict the presence of toxins in commodities and foods. Sub-objective 3.B. Develop tools to detect emerging toxins, their metabolites and masked forms.

Food crops are commonly infested with fungi, both in the field and in storage. Certain fungi produce toxins (mycotoxins) that can adversely affect human health and the health of domestic animals. Certain toxin-binding materials may also have the potential to be used to remove toxins from foods, and this will be approached through the development and application of novel synthetic materials to reduce exposures. By permitting the timely diversion of contaminated ingredients from the food supply, detection of foodborne toxins can directly improve food safety and the safety of animal feed. Monitoring for the presence of such naturally occurring toxins is widespread and occurs at many of the stages between the producer and the consumer. Increasing the efficiency and improving the accuracy of monitoring results in more appropriate and efficient diversion of contaminated products. The need to monitor for greater numbers of mycotoxins is a trend that will continue, in particular because of recent concern over the so called “masked” mycotoxins. This project will put a particular focus on the food safety of a commodity important to U.S. consumers, oats. A survey of both toxin contamination and fungal contamination in U.S. oats will aid in the assurance of a safe supply of the key ingredient in human diets. Further, this project seeks to address the need for improved toxin detection by developing rapid detection methods leading to the prediction of toxin contamination in a variety of food / feed ingredients.

Progress Report
In support of Objective 1, “greener” plant-based natural products were investigated for their use as safer antifungal compounds to reduce levels of fungi and associated mycotoxins in post-harvest commodities. U.S. economic losses due to mycotoxins in corn, wheat, and peanuts total $930M annually. Intervention strategies to prevent fungal contamination in harvested commodities are preferable to approaches to remediate commodities upon contamination with mycotoxins. While antifungal compounds can minimize fungal colonization and post-harvest mycotoxin accumulation, these antifungal compounds are generally more toxic than commonly used antibacterial compounds. To examine the potential for safer antifungal treatments, biosynthesized phenolic compounds, including feruloyl glycerol and diferuloyl glycerol complexes, were produced for further study and modelled for properties related to their potential antifungal capabilities. Analyses with the chemical predictive modeling tools included in the “ToxCast” software environment, developed by researchers at the Environmental Protection Agency, indicated that the phenolic compounds are safe alternatives for limiting the growth of mycotoxin producing fungi associated with food decomposition. We modelled 49 antifungal candidates and 10 formulation components for predicted toxicity and identified five antifungal candidates for further testing. Validations of these chemical predictive modelling results are now being completed using antimicrobial assays against Fusarium fungi and other plant pathogens. In further support of Objective 1, commercially available agricultural bioproducts were evaluated for their ability to reduce levels of zearalenone and ochratoxin A from processes related to the utilization of corn for ethanol production and in vitro animal digestion models. Mycotoxins can become concentrated in products of the ethanol production process that are frequently mixed in with animal feed. Zearalenone is an estrogenic mycotoxin that can occur in corn and can harm swine reproductive production. Ochratoxin A can appear in various commodity sources used in food and feed supplies and exposure to ochratoxin A is associated with reduced growth and feed efficiency in animal production. Agriculturally derived biomaterials, such as “biochar” produced from several agricultural waste streams, were evaluated for their ability to sequester ochratoxin A and zearalenone. Charcoals (biochar) are medically used for poison control and marketed as feed additives to sequester and reduce exposure to toxins in animal feed in the U.S. The mycotoxin sequestration capacities of eight different commercially available biochars were evaluated in artificial stomach and ethanol fermentation experiments. In these experiments, coconut shell charcoal exhibited better zearalenone and ochratoxin A sequestration properties compared to the conventionally used montmorillonite clay in conditions related to saccharification during corn ethanol production. In support of Objective 2, oats harvested in crop year 2022 and directed to the U.S. food and feed supply were surveyed for fungal and mycotoxin contamination. A liquid chromatography – tandem mass spectrometry method for determination of several mycotoxins and their masked counterparts was utilized for analysis of representative samples of oat grain obtained from the survey. In response to stakeholder input, a previously developed mycotoxin analytical method for oat samples was modified to include a laboratory scale “de-hulling” step to allow the analyzed samples to more closely reflect the oat commodities utilized in milling and animal feed production. Fusarium species-specific primers were developed and optimized for polymerase chain reaction-based analysis to identify and estimate the quantity of seven different fungal species potentially contaminating the sampled oats. Difficulties persist in demonstration of specificity of some fungal assays in differentiation between closely related Fusarium species responsible for production of trichothecene mycotoxins. Results of this multi-year survey will be used to characterize mycotoxin risks associated with the utilization of oat products in the U.S. food and feed supply. Utilization of these chemical and microbiological tools should allow the characterization of pre-harvest and post-harvest fungal and mycotoxin contamination of oats destined for U.S. processing and prompt diversion of contaminated grain from the food supply. To further support Objective 2 and to develop the ability to sample fungal contamination in developing crops, funding was secured from the National Predictive Modeling Tool Initiative to conduct the project “Quantitative monitoring of mycotoxin - producing fungi in corn production fields” over the years 2023-2024. The project supports the development and implementation of remote sensing devices for the identification and quantitation of air-borne fungal spores in agricultural fields. A goal of the study is to develop tools for the prediction of pre-harvest and post-harvest fungal and mycotoxin contamination of a variety of food/feed grains. During the 2022 crop year the use of in-field spore collection devices in 10 central Illinois locations allowed sampling of air-borne fungal inoculum in corn fields, along with sampling of harvested corn adjacent to spore collection devices. Results of this sampling of air-borne inoculum, along with determination of fungal and mycotoxin levels in harvested grain associated with spore collection sites, should aid in the development and refinement of predictive tools for cereal grain mycotoxin contamination. The focus for Objective 3 was strategies to reduce fungal contamination that benefit from techniques that can be used to rapidly identify the onset of disease or presence of mycotoxins. Attempts were made to determine whether the changes in the composition of maize resulting from fungal contamination could be correlated with mycotoxin content. Samples of maize were “fingerprinted” using ambient ionization mass spectrometry. Machine learning was applied to determine if differences could be observed between contaminated and non-contaminated maize. Predictive models were developed using 253 samples of Illinois maize. Most of the samples did not have detectable mycotoxins. The efficacy of the models was evaluated with a “correctness score”. Scores of greater than 95% are widely regarded as acceptable. The models correctly classified the negative samples as negative, and scores ranging from 85% to 95% were achieved. However, a high proportion of the “positive” samples were also classified as negative. Therefore, despite a high correctness score, the screening method was not acceptable for prediction of the presence of mycotoxins. This result is likely based in the inability of the technique to identify features within the samples that were specific to mycotoxin contaminated maize. In support of Objective 3, we continued research to develop tools for the early detection of growth of mycotoxin-producing fungi in plants. There is currently a lack of convenient methods to observe the development of fungi in crops during the growing season. To facilitate such observations, we have pursued the monitoring of volatile fungal metabolites by mass spectrometry. A modified “direct analysis in real time” ion source connected to a benchtop high-resolution mass spectrometer was used to detect volatiles indicative of growth of specific fungal species on plant substrates. In preliminary experiments, the approach was used to measure volatiles indicating levels of infestation of Fusarium verticillioides in harvested ears of maize from field experiments. The mass spectrometry experiments were conducted with no sample preparation, affording a convenient, rapid and non-destructive method of analysis. Detected levels of characteristic volatile fungal metabolites analysis correlated well with fumonisin contamination of the maize kernels from the monitored field trial ears. Efforts are underway to examine correlation of volatile metabolites to fungal biomass in commodities. Utilization of volatile fungal metabolite analysis provides another potential tool for development of predictive models for commodity contamination. Certain fungal toxins affect the nervous system, causing tremors in livestock and companion animals. Rapid and inexpensive screening tests are desired as a means for keeping such toxins out of the human food and animal feed supply chain. In further support of Objective 3, efforts continued for the development of antibodies for penitrem-A, a potent neurotoxin. Two groups of mice were immunized with two conjugates of penitrem-A with proteins. Sera from the mice were evaluated and found to specifically recognize the toxin-protein conjugates. One was found with a weak response to penitrem-A. Additional toxin-protein conjugates have been synthesized to improve upon this response.

1. Measurement of cyclopiazonic acid, a mycotoxin in cheese and maize. Fungi from the genus Penicillium are used in the ripening of two popular types of cheese: soft-ripened cheeses (Brie, Camembert, etc.) and blue-veined cheeses (Roquefort, Stilton, Gorgonzola, DanaBlu, etc.). However, certain strains of these fungi produce toxins (mycotoxins). One such mycotoxin, cyclopiazonic acid, is a neurotoxin that has previously been found in Europe as a common contaminant of soft-ripened cheeses. The toxin has also been found in maize in the United States. Data on the levels of cyclopiazonic acid in cheese and maize were extremely limited. To address this, ARS researchers in Peoria, Illinois, measured cyclopiazonic acid in 254 samples of soft-ripened and blue-veined cheeses. Cyclopiazonic acid was detected in 46% of soft-ripened cheeses and in 24% of blue-veined cheeses, generally at low levels. Higher levels were occasionally found. They also collaborated with researchers at the Complutense University of Madrid, Spain, to develop a novel method to detect it in maize. The method was based upon a ‘mimotope’: a short peptide that mimicked the toxin. This allowed a novel assay format to be used, which resulted in rapid (5 minute) detection of the toxin. This has provided an easy-to-use tool for the rapid, and sensitive, screening of cyclopiazonic acid in maize.

2. Developed portable methods to detect mycotoxins. Measurements of mycotoxins are performed in many locations, from grain elevators to regulatory laboratories. This has led to a diverse range of methods for their measurement. Cost, speed, and the ability to be performed by personnel with minimal training, are among the reasons why immunoassays are frequently used for screening. For regulatory settings, where accuracy is of utmost importance, liquid chromatography with tandem mass spectrometry is often used. In an ideal situation the speed, cost, and portability of immunoassays would be combined with the selectivity and accuracy of mass spectrometry. Producers and consumers alike would benefit from improving the accuracy, speed, and economics of toxin detection. The advent of commercially available portable mass spectrometry instruments has created the opportunity for rapid and accurate testing at locations such as grain elevators where larger laboratory-bound instruments may not be feasible. ARS researchers in Peoria, Illinois, in collaboration with the Illinois Department of Agriculture, have developed methods for testing two groups of mycotoxins in wheat and maize using a portable mass spectrometry instrument. Results suggest that the technology can be applied to settings outside of traditional laboratories.

3. Evaluated AI tools to utilize food databases for safer antifungals. Food databases, such as the USDA FoodData Central, have recently expanded, including information on food composition, flavor molecules, and chemical compound properties. It would be desirable to use these databases to locate specific food components frequently used as natural preservatives, to potentially allow their use to minimize mycotoxin contamination. AI, deep learning and machine learning analyses were evaluated by ARS researchers in Peoria, Illinois, to determine their ability to locate food components in ingredients that have potential properties important to food safety. The models developed from AI tools and food databases serve as cost-effective tools to screen potential antifungal food components for antifungal properties. The results indicate that chemical stability is important for favorable antifungal activities and chemical safety. The combination of increased access to public food databases and AI tools will aid food processors for a variety of food science and food chemistry issues, including food safety and mycotoxin reduction. Results demonstrated their potential application in food safety and antifungal development utilizing approaches aimed at deducing possible pharmacological effects of food components. It is expected that the combination of increased access to public food databases and AI tools will aid food processors for a variety of food science and food chemistry issues, including food safety and mycotoxin reduction.

Review Publications
Tittlemier, S.A., Cramer, B., DeRosa, M.C., Lattanzio, V.M.T., Malone, R., Maragos, C., Stranska, M., Sumarah, M.W. 2023. Developments in mycotoxin analysis: an update for 2021-22. World Mycotoxin Journal. 16(1):3-24.
Maragos, C.M. 2023. Detection of T-2 toxin in wheat and maize with a portable mass spectrometer. Toxins. 15(3). Article 222.
Maragos, C.M., Barnett, K., Morgan, L., Vaughan, M.M., Sieve, K.K. 2022. Measurement of fumonisins in maize using a portable mass spectrometer. Toxins. 14(8). Article 523.
Maragos, C.M., Probyn, C., Proctor, R.H., Sieve, K.K. 2022. Cyclopiazonic acid in soft-ripened and blue cheeses marketed in the USA. Food additives & contaminants. Part B: Surveillance. 16(1):14-23.
Tseng, Y.J., Chuang, P.-J., Appell, M. 2023. When machine learning and deep learning come to the big data in food chemistry. ACS Omega. 8:15854-15864.
Cheng, H.N., Biswas, A., Kim, S., Appell, M., Furtado, R.F., Bastos, M.S.R., Alves, C.R. 2022. Synthesis and analysis of lactose polyurethanes and their semi-interpenetrating polymer networks. International Journal of Polymer Analysis and Characterization. 27(4):266-276.
Cheng, M.H., Maitra, S., Carr Clennon, A.N., Appell, M., Dien, B.S., Singh, V. 2022. The effects of sequential hydrothermal-mechanical refining pretreatment on cellulose structure changes and sugar recoveries. Biomass Conversion and Biorefinery.
Pradanas-Gonzalez, F., Peltomaa, R., Lahtinen, S., Luque-Uria, A., Mas, V., Barderas, R., Maragos, C.M., Canales, A., Soukka, T., Benito-Pena, E., Moreno-Bondi, M.C. 2023. Homogeneous immunoassay for cyclopiazonic acid based upon mimotopes and upconversion-resonance energy transfer. Biosensors and Bioelectronics. 233. Article 115339.