Isolates of aflatoxin-producing strains of Aspergillus can grow on Glycine max beans. Furthermore, both mixed aflatoxins and ariatoxin BI inhibit the elongation of both attached and excised G. max roots. In addition, the toxin inhibits the capacity of the excised roots to take up [14C]leucine. This suggests that the toxin may inhibit the activity of an uptake system responsible for uptake of low-molecular-weight (MW) compounds by possibly altering membrane structure and/or function. In this connection, incubation of excised roots in media containing either mixed ariatoxins or ariatoxin B a results in a diminution of acid-insoluble protein, but neither sterol nor lipid phosphorus levels, of an 80,000 • g pellet (purported "crude" plasmalemma fraction). To provide preliminary evidence that mixed aflatoxins can decrease the amount of a specific plasmalemma protein (which might regulate the uptake of low-MW compounds), we incubated roots with and without mixed aflatoxins and then gel-filtrated integral proteins which were released by detergents from 80,000 X g pellets that had been obtained by differential centrifugation of Miracloth filtrates. The released proteins were gelfiltrated on Sephadex G-100 columns. Sodium dodecyl sulfate, Triton X-IO0 and Tween 20 each solubilized greater than 85% of the pellet's protein. Gel fdtration yielded 280 nm absorbing void volume and retarded peaks for substances which were either precipitated by trichloroacetic acid (total protein) or solubilized by detergents (integral protein) from pellets that were derived from ariatoxin-treated and nontreated roots. The amounts of protein which were recovered within column void volumes following gel fdtration of 80,000 X g, acid-precipitable protein were not significantly different for treated and nontreated roots, suggesting that incubation of roots with aflatoxins does not reduce the pellet's content of an acidprecipitable, high-MW (100,000 or more) protein. However, this conclusion is highly tentative as less than 15% of the acid-insoluble protein which was layered onto the column was recovered. The amplitude of the void volume peaks for detergent-released proteins from treated roots was consistently less than that for nontreated roots, suggesting that treatment of roots with aflatoxins diminishes the "crude" plasmalemma fraction's content of high-MW protein(s). This suggestion was supported through calculation of the amounts of protein which were recovered within both the void volume and retarded peaks and comparing these to the total protein levels that were recovered from the columns. The amount of protein which was found within the void volume peak following gel filtration of either Triton X-IOO or Tween 20-released proteins from pellets that were derived from treated roots was less than that for nontreated roots.