Microtubule conformational changes examined with SHG. (A) Principle of the measurement. Incident polarization θ dependence of SHG signal reflects the orientation of tubulin subunits in the microtubule. (B) Schematic drawing of the SHG microscope. BPF, bandpass filter; PMT, photomultiplier tube. (C) The ratio of the SHG polarization angle against the concentration of KIF5C motor domain. Parallel GDP microtubule bundles grown from the mitotic asters without Taxol were measured. The ratio of the polarization factor (right vertical axes) reflects the tilt angle (φ) of the dipole moment of the tubulin subunit in the microtubule lattice (left vertical axes). (D) The box-whisker plot of the SHG signal for GDP microtubules without KIF5C and Taxol, with KIF5C, and with both KIF5C and Taxol. The boxes show the first and third quadrants, and the bars in the boxes indicates the median of each dataset. The bottom and top whiskers in each plot indicate minimum and maximum values of a dataset. (E) The effect of apyrase on the SHG signal. Addition of apyrase did not affect the tubulin dipole angle of the GDP-microtubules, suggesting that shift in the SHG signals is induced by the nucleotide free KIF5C. (D and E) P values were calculated using a Steel–Dwass test. *, P < 0.001. (C–E) The numbers of tested samples are shown in the figures.