Predicting herbicide movement across semi-permeable membranes using three phase partitioning

Herbicide efficacy depends on herbicides crossing cell and organelle membranes. We evaluated an artificial membrane system to understand how herbicides cross biological membranes. This understanding aids in predicting herbicide behavior in planta and, consequently, efficacy, mode of action, and whether active transporter-based herbicide resistance mechanisms may be possible. Five herbicides with different log K_ow and pK_a values were assessed: glyphosate, 2,4-D, clopyralid, sulfentrazone and glufosinate. The artificial membrane apparatus included four semipermeable membranes containing buffers with pH 2.7, 5 and/or 7.4, floating in a bath of diethyl ether. These conditions were based on the pH from different cellular compartments and the pK_a for these herbicides. Changes in herbicide concentration due to movement were measured using radioactivity or liquid chromatography mass spectrometry. In general, herbicide behavior followed the pattern predicted by their calculated pK_a and log K_ow. Herbicides added to an acidic phase (pH 2.7) were more mobile than when they were added to the more basic phase (pH 7.4), except when herbicide's pK_a was lower than the pH of the starting phase. Clopyralid, 2,4-D, and sulfentrazone showed significant acid trapping behavior due to their weak acid functional groups. Sulfentrazone and 2,4-D had a high affinity for the nonpolar, diethyl ether bath, especially when they were protonated at low pH. Our findings illustrate the robustness of the system to provide predictions about herbicide behavior at the subcellular level.