Abstract
Commercial pesticide formulations contain co-formulants, which are generally considered as having no toxic effects in mammals. This study aims to compare the toxicity of 8 major herbicide active ingredients-namely glyphosate, dicamba, 2,4-D, fluroxypyr, quizalofop-p-ethyl, pendimethalin, propyzamide and metazachlor-with a typical commercial formulation of each active ingredient. Cytotoxicity and oxidative stress capability was assessed in human hepatoma HepG2 cells. Using an MTT assay, formulations of glyphosate (Roundup Probio), fluroxypyr (Hurler), quizalofop-p-ethyl (Targa Super) and dicamba (Hunter) were more toxic than the active ingredient alone. Metazachlor and its formulation Sultan had similar cytotoxicity profiles. Cytotoxicity profiles were comparable in immortalised human fibroblasts. Toxilight necrosis assays showed the formulation of metazachlor (Sultan50C) resulted in significant membrane disruption compared to the active ingredient. Generation of reactive oxygen species was detected for glyphosate, fluroxypyr, pendimethalin, quizalofop-p-ethyl, the formulation of 2,4-D (Anti-Liserons), and dicamba and its formulation Hunter. Further testing of quizalofop-p-ethyl and its formulation Targa Super in the ToxTracker assay system revealed that both products induced oxidative stress and an unfolded protein response. In conclusion, these results show that most herbicide formulations tested in this study are more toxic than their active ingredients in human tissue culture cell model systems. The results add to a growing body of evidence, which implies that commercial herbicide formulations and not just their active ingredients should be evaluated in regulatory risk assessment of pesticides.