A study of Brazilian farmworkers has demonstrated the accuracy and potential of combining exposure assessments, human biomonitoring, high-throughput screening data, and computational toxicology tools developed by NIEHS.

The research tested human samples to establish a link between fungicides used in Brazilian coffee cultivation and decreased levels of the hormones testosterone and androstenedione, as well as other biological effects in men who worked on the plantations. A cooperative effort with Nicole Kleinstreuer, Ph.D., of the NIEHS Division of Translational Toxicology, used the chemical libraries and interpretive models maintained on the Integrated Chemical Environment (ICE) website to replicate the findings of the field study.

“We used laboratory [in vitro] assays, compared them to data drawn from actual people [in vivo], and found an essentially perfect correlation,” Kleinstreuer said. “For the first time, we showed that data from human cell-based assays were highly predictive of effects and exposures and concentrations in people — a true human to human, apples to apples comparison.”

Human volunteers show signs of exposure

The field research, conducted by Luiz Paulo de Aguiar Marciano, Ph.D., of the Federal University of Alfenas, provided a detailed picture of the effects of fungicide use on coffee operations in the Brazilian state of Minas Gerais. His research found high levels of three commonly used chemicals — cyproconazole, epoxiconazole, and triadimenol — in urine samples from a test group of 140 coffee workers and their wives. Test subjects from urban areas of the state exhibited no sign of exposure.

Marciano’s team also collected blood samples from the subjects and found multiple biomarkers in the rural group that indicated effects of exposure.

“They looked at steroid hormone levels, markers of oxidative stress, markers of genotoxicity, markers of liver function through enzymes, and found significant differences in the exposed group versus the urban control,” Kleinstreuer noted.

Most prominent among these differences was the significant reduction of the hormones androstenedione and testosterone in the male farmworkers.

Predictive tools match field findings

Kleinstreuer became involved in the project when Marciano emailed her for assistance. He knew she was overseeing development of computational tools to analyze results from in vitro testing of thousands of chemical substances on human-based cells, and he wanted to test if the high-throughput screening data and computational workflows supported his findings. Presented with the opportunity to test the laboratory predictions against findings from the field, Kleinstreuer arranged a fellowship for Marciano at NIEHS.

That phase of the project centered on the tools available on the ICE site. The research team used a tool called ICE Curve Surfer to assess biological response to various concentrations of the chemicals in the fungicides. They also deployed the in vitro to in vivo extrapolation tool (IVIVE) to determine the equivalent human dosage to the amount of chemicals tested in the laboratory samples.

The research team had expected the findings from the ICE analysis to be compatible with the human testing data but was pleased to learn they matched exactly.

“It was remarkable,” Kleinstreuer said. “Not only were the same targets being perturbed, they were perturbed at the same concentrations as in humans and predicting the same biomarkers measured in humans.”

Potential for the future

Kleinstreuer believes the impact of this study will be felt on multiple fronts. First and foremost, it verifies the human testing results and could benefit the coffee workers and their families.

“It provides evidence that people are at elevated risk of toxicity and that there is a need to develop better risk management strategies, such as setting up more training programs to help them understand the importance of using personal protective equipment and following fungicide usage directions,” she said.

She also anticipated that the success of the materials on the ICE site — especially the in vitro to in vivo extrapolation tool — will expand access to predictive information about the toxicity of substances.

“Previously, you had to be an expert in mathematical modeling and writing code and reading code,” she noted. “You can do all of that now through ICE. It’s really a way of democratizing access to these computational tools.”

Kleinstreuer believes the evidence presented by this study will reduce the need to rely on animal testing. She also noted that the predictive tools could fill the void in cases where human data are not available.

“This gives us some confidence that we can draw on cell-based data from Tox21, apply in vitro to in vivo extrapolation and mathematical modeling, and anticipate adverse effects at certain exposure levels,” she said. 

Results of the study were published Nov. 25 in Science of the Total Environment.