Three-dimensional structure of the radial spokes reveals heterogeneity and interactions with dyneins in Chlamydomonas flagella.

Axoneme isolation and cryosample preparation All C. reinhardtii strains used in this study are summarized in Table 1 . Cells were grown in liquid Tris-acetate-phosphate medium ( Gorman and Levine, 1965 ) under a 16-h-light/8-h-dark cycle and harvested by 5-min centrifugation at 1100 × g. After two washes, cells were resuspended in a minimal medium containing 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.4, 1 mM SrCl 2 , 4% sucrose, and 1 mM dithiothreitol. The pH-shock method ( Witman et al. , 1972 ) was used to detach flagella from cell bodies. After deflagellation, 5 mM MgSO 4 , 1 mM ethylene glycol tetraacetic acid (EGTA), 0.1 mM EDTA, and 100 μg/ml aprotinin, pepstatin, and leupeptin were added to the medium, and cell bodies were pelleted by a 10-min spin at 1800 × g at 4°C. The flagella present in the supernatant were purified in two additional centrifugation steps at 2400 × g and 4°C for 10 min using 5 ml of a 20% sucrose solution as a cushion. The flagellar membrane was removed by detergent treatment for 30 min with 0.1% Nonidet P-40 (Sigma-Aldrich, St. Louis, MO) or 1% IGEPAL CA-630 (Sigma-Aldrich) for 30 min at 4°C. The axonemes were collected by centrifugation at 35,000 × g for 1 h at 4°C. The pelleted axonemes were resuspended in 10 mM HEPES, pH 7.4, 25 mM NaCl, 4 mM MgSO 4 , 1 mM EGTA, and 0.1 mM EDTA and processed within 24 h. Quantifoil holey carbon grids (Quantifoil Micro Tools, Jena, Germany) were glow discharged for 30 s at −40 mA before a drop of a 10-nm colloidal gold solution (Sigma-Aldrich) was applied. After drying, grids were washed by briefly dipping them into Milli-Q purified water (Millipore, Billerica, MA), and excess water was removed with a filter paper (Whatman, Piscataway, NJ). After grids dried completely, they were loaded in a home-made plunge-freezing device ( Trachtenberg, 1993 ), and 3 μl of axoneme sample and 1 μl of a 10-fold-concentrated, 10-nm colloidal gold solution were added to the grid. The mixture was blotted with a filter paper (Whatman #1) from the front side for ∼2 s and immediately plunge frozen in liquid ethane to achieve sample vitrification. Grids were stored in liquid nitrogen until examined by EM. Cryo-ET Details of tilt series acquisition and tomogram reconstruction were published earlier ( Heuser et al. , 2009 ; Nicastro, 2009 ). In brief, cryo­samples were transferred into a transmission electron microscope (Tecnai F30; FEI, Hillsboro, OR), which was operated at 300 keV, using a cryoholder (Gatan, Pleasanton, CA). Data acquisition was controlled by the microscope software SerialEM ( Mastronarde, 2005 ). Tilt series were recorded under low-dose conditions by stepwise rotating the sample from −65 to +65° with 1.5–2.5° increments. The cumulative electron dose was limited to ∼100 e/Å 2 . All images were digitally recorded on a 2k × 2k charge-coupled device camera (Gatan) at a magnification of 13,500, resulting in a pixel size of ∼1 nm, at −6 or −8 μm defocus and using an energy filter (Gatan) in zero-loss mode with 20-eV slit width. The IMOD software package ( Kremer et al. , 1996 ) was used for fiducial alignment and cryotomogram reconstruction by weighted backprojection. Only tomograms of noncompressed or mildly compressed and intact axonemes were used for further image processing and data analysis.

Image processing

For subtomogram averaging of the 96-nm-long axonemal repeat along the axonemal DMTs we used the program PEET (Particle Estimation for Electron Tomography; Nicastro et al. , 2006 ). The repeats were computationally cut from the raw tomograms, aligned in three dimensions, and averaged with missing wedge compensation, resulting in an average with increased signal-to-noise ratio and improved resolution ( Nicastro et al. , 2006 ; Nicastro, 2009 ). Two different alignment strategies were tested: in the “global alignment” approach the entire 96-nm repeat unit (volume size, 100 × 100 × 80 nm) was used for 3D alignment, whereas the “local alignment” method optimized the alignment for smaller regions (volume sizes were either 90 × 50 × 32 nm centering on RS1, RS2, and RS3S or 66 × 22 × 32 nm centering on the RS head, respectively). For pf17 the global alignment resulted in the best average and is therefore depicted in all figures of pf17 axonemes. In contrast, the local alignment averages resolved the radial spokes at slightly better detail for WT and pWT and are therefore depicted in all figures of WT/pWT except for Figure 3A and Supplemental Figure S1A. Some of the tomograms were previously used, focusing on other axonemal complexes and using different image processing methods ( Heuser et al. , 2009 ; T. Heuser et al. , unpublished results). The resolution of all analyzed strains was estimated at the stem part of RS2 using the Fourier shell correlation method ( Harauz and Van Heel, 1986 ) and the 0.5 criterion ( Table 1 ). The UCSF Chimera software package ( Pettersen et al. , 2004 ) was used to measure volume sizes and to visualize the axonemal averages by isosurface rendering. Masses of RS parts were estimated by assuming an average protein density of 1.43 g/cm 3 ( Quillin and Matthews, 2000 ) and normalizing the isosurface rendering threshold to the mass of DMT ( Heuser et al. , 2009 ). The placement of boundaries between RS1/RS2/RS3S and the IDA tails is based on comparison of our images with single-particle structures of dynein ( Roberts et al. , 2009 ).