α-Tubulin mutations alter oryzalin affinity and microtubule assembly properties to confer dinitroaniline resistance

Abstract

Plant and protozoan microtubules are selectively sensitive to dinitroanilines, which do not disrupt vertebrate or fungal microtubules. Tetrahymena thermophila is an abundant source of dinitroaniline-sensitive tubulin, and we have modified the single T. thermophila α-tubulin gene to create strains that solely express mutant α-tubulin in functional dimers. Previous research identified multiple α-tubulin mutations that confer dinitroaniline resistance in the human parasite Toxoplasma gondii, and when two of these mutations (L136F and I252L) were introduced into T. thermophila, they conferred resistance in these free-living ciliates. Purified tubulin heterodimers composed of L136F or I252L α-tubulin display decreased affinity for the dinitroaniline oryzalin relative to wild-type T. thermophila tubulin. Moreover, the L136F substitution dramatically reduces the critical concentration for microtubule assembly relative to the properties of wild-type T. thermophila tubulin. Our data provide additional support for the proposed dinitroaniline binding site on α-tubulin and validate the use of T. thermophila for expression of genetically homogeneous populations of mutant tubulins for biochemical characterization.

Fig. 1.

Wild-type T. thermophila cells are dinitroaniline sensitive. (A) Growth of wild-type T. thermophila at 30°C in medium with 0, 7.5, or 15 μM oryzalin. Wild-type T. thermophila cells showed reduced growth at 7.5 μM and complete inhibition at 15 μM. (B and C) Microscopy of T. thermophila cells grown in the absence (B) or presence (C) of 7.5 μM oryzalin indicated that many microtubule structures are disrupted by dinitroaniline treatment. T. thermophila cells were labeled with antitubulin antibody (green) and 4′,6-diamidino-2-phenylindole (DAPI, which labels the nuclei blue). The left panels show merged blue and green fluorescence, the middle panels show the fluorescence image merged with a phase-contrast image, and the right panels show a phase-contrast image alone. (B, left) Normal T. thermophila cell covered with tubulin-containing cilia, but also with underlying microtubules, such as the cortical microtubules (arrows). (B, middle and right) T. thermophila cilia are visible by both tubulin immunofluorescence and phase imaging (arrows). (C, middle) Oryzalin-treated T. thermophila cells lose most or all cilia, making the underlying cortical microtubules (arrows) more visible.

Fig. 2.

Complemented T. thermophila strains show normal growth and express similar levels of tubulin. (A) T. thermophila wild-type (WT) and mutant strains grow with similar kinetics at 30°C in the absence of oryzalin. (B) Immunoblot analysis of cell lysates equivalent to ∼9,750 cells per well indicated that the three T. thermophila strains (the wild-type complemented strain and strains expressing tubulin with the L136F or V252L point mutation) express essentially equivalent concentrations of tubulin per cell (30 pg). This blot was probed with the anti-α-tubulin monoclonal antibody 1-5-2.

Fig. 3.

(A) Growth of complemented strains in 0, 7.5, and 15 μM oryzalin. (A) T. thermophila cells expressing the L136F ATU1 transgene (top panel) and T. thermophila cells expressing the I252L ATU1 transgene (bottom panel). (B to D) Immunofluorescence of T. thermophila cells cultured in oryzalin. Wild-type T. thermophila cells grown in 7.5 μM oryzalin lose most or all cilia but retain underlying cortical microtubules (B), whereas T. thermophila cells expressing the L136F (C) or I252L (D) tubulin mutations retain cilia.

Fig. 4.

Purified tubulin is free of discernible MAPs and shows GTP-dependent assembly. (A) An overloaded Coomassie blue-stained SDS-PAGE gel illustrates the purity of DEAE-purified, cycled L136F tubulin used for biochemistry experiments; the absence of contaminating proteins (particularly any high-molecular-weight microtubule-associated proteins) is typical of all our protein purifications. (B) Assembly of I252L tubulin (as well as the wild-type and L136F samples [data not shown]) is dependent upon the presence of GTP (circles). There was no change in optical density when the assembly reaction mixture contained GDP (squares), indicating that the light scattering reflects assembly rather than aggregation of denatured protein. (C) Electron microscopy of negatively stained samples demonstrated that tubulins polymerize to microtubules and other polymeric forms in the presence of GTP. The wild-type tubulin sample is shown, but all tubulins assembled into similar polymers in this assay. Bar, 200 μm.

Fig. 5.

Binding curves for association of tubulin with oryzalin. Oryzalin binding was modeled using a one-site binding, nonlinear fit. Wild-type T. thermophila tubulin, L136F T. thermophila tubulin, and I252L T. thermophila tubulin all have similar maximal quench values, indicating that oryzalin interacts in the same fashion with each of the three tubulins. The binding curve and maximal quenching are different for vertebrate tubulin. These data indicate that the K_d values for oryzalin are 77 μM for vertebrate tubulin and 0.44 μM for wild-type T. thermophila tubulin, which is reduced to 11 μM for L136F T. thermophila tubulin and 6.7 μM for I252L T. thermophila tubulin.

Fig. 6.

Mutant tubulins show altered assembly properties. (A) Polymerization of 0.7 mg/ml (7 μM) purified T. thermophila tubulin was followed at 351 nm in the presence (gray) or absence (black) of 0 or 5 μM oryzalin. Wild-type T. thermophila tubulin assembles normally in the absence of oryzalin but fails to polymerize in 5 μM oryzalin. L136F T. thermophila tubulin and I252L T. thermophila tubulin both assemble in 5 μM oryzalin. (B) Determination of C_c values for wild-type, L136F, and I252L tubulins. Wild-type T. thermophila tubulin has a C_c of 0.55 ± 0.09 mg/ml. The C_c for L136F tubulin is dramatically decreased to 0.25 ± 0.07 mg/ml, whereas I252L tubulin behaves similarly to wild-type tubulin with a C_c value of 0.56 ± 0.06 mg/ml.