Supplementary material for Ishikawa et al. (April 3, 2001) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.081543698
Fig. 5.
Simulation studies [Ishikawa, T., Belnap, D., Maurizi, M. R. & Steven, A. C. (2000) Nature (London) 408, 667–668]) to show visualization of small (10–70 kDa) protein "substrates" binding to a Clp ATPase at 30-Å resolution by cryoelectron microscopy. From left: (i) atomic models of ClpYQ [HslUV; Sousa, M. C., Trame, C. B., Tsuruta, H., Wilbanks, S. M., Reddy, V. S. & McKay, D. (2000) Cell 103, 633–643] with and without modeled substrates of masses 10–70 kDa made from the atomic model of individual ClpY subunits or portions thereof; (ii) two-dimensional projection maps at 30-Å resolution, computed from the atomic models; (iii) projection maps convoluted with contrast transfer function (CTF) (2 mm under focus, 7% amplitude contrast, 30-Å resolution); and (iv) difference maps of contrast transfer function convoluted projections between the projections with and without substrates. (a) Control images of ClpYQ without "substrates." (b–e) ClpYQ bound by the 70-kDa substrate (shown in red and orange) at various positions. The difference images in the right columns demonstrate that the substrates are visualized well, independently of their locations with respect to the central axis. To further simulate the experimental conditions, we averaged the four images with the substrate in different axial positions. As we discuss in the text, because the bound substrate binds off-axis, the density appears in various azimuthal positions, resulting in a dilution of the density in averaged images. (f) Left, averaged maps of the four CTF convoluted projections shown in b–e. Right, the difference maps between averaged projection maps and the CTF convoluted projection without substrate. The substrate density is clearly detectable on the apical surface of the ring, even if the density is smeared out. The additional integrated density in f is 90% of that in b, indicative of some underestimation of the mass of associated substrate in averaged maps. At this resolution and without classification to sort images according to substrate position, information about the shape of the substrate and its exact point of attachment on the projected ring surface is not available, but these matters are beyond the scope of the present study. To test the limits of sensitivity at 30-Å resolution, we placed a series of smaller masses in association with ClpYQ and calculated difference maps. (g and h) ClpYQ with a 35-kDa "substrate" (shown in orange). In h, the substrate is rotated 90° from its position in g. The additional density can be seen in the both difference maps. (i) ClpYQ with a 10-kDa substrate. Projection images and the difference map indicate that a 10-kDa substrate can be visualized even with 30-Å resolution.