Structural organization of the C1a-e-c supercomplex within the ciliary central apparatus

Abstract

Nearly all motile cilia contain a central apparatus (CA) composed of two connected singlet microtubules with attached projections that play crucial roles in regulating ciliary motility. Defects in CA assembly usually result in motility-impaired or paralyzed cilia, which in humans causes disease. Despite their importance, the protein composition and functions of the CA projections are largely unknown. Here, we integrated biochemical and genetic approaches with cryo-electron tomography to compare the CA of wild-type Chlamydomonas with CA mutants. We identified a large (>2 MD) complex, the C1a-e-c supercomplex, that requires the PF16 protein for assembly and contains the CA components FAP76, FAP81, FAP92, and FAP216. We localized these subunits within the supercomplex using nanogold labeling and show that loss of any one of them results in impaired ciliary motility. These data provide insight into the subunit organization and 3D structure of the CA, which is a prerequisite for understanding the molecular mechanisms by which the CA regulates ciliary beating.

Figure 1.

Improved resolution of the averaged CA structure. (A and B) Tomographic slices of the averaged 32-nm repeat of the Chlamydomonas WT CA viewed in cross section to compare data recorded with either a CCD camera (A) or a direct electron detector and Volta Phase Plate (K2/VPP; B). Thin white lines in A and B indicate the locations of the slices shown in E and G, respectively. (C) FSC curves of the CA averages show that the resolution is improved from 3.5 nm for CCD data to 2.3 nm for K2/VPP data (0.5 criterion). (D–G) Tomographic slices to compare the C2 microtubule-associated structures between the CCD (D and E) and the K2/VPP data (F and G) in cross-sectional (D and F) and longitudinal (E and G) views. Note the clear visualizations of (a) filamentous, microtubule-associated proteins between protofilaments 9–12 (white arrowheads in F and G), (b) two distinct domains of the microtubule inner protein (MIP) C2a (black arrows in G), and (c) the tubulin lattice of the microtubule wall (G) in the K2/VPP data, which were not observed or were blurred in the CCD data. (H) Isosurface rendering shows the averaged Chlamydomonas WT CA (K2/VPP data) in cross-sectional view. Naming and coloring of CA projections adopted from Mitchell and Sale (1999) and Carbajal-González et al. (2013). CA protofilaments were not previously numbered; here we assigned protofilament #1 of C1 and C2 to where the C1a and C2a projections attach, respectively. In all figures, cross sections are viewed from proximal (i.e., cell body) to the ciliary tip, and in longitudinal views the proximal side is on the right (except for Fig. 1, E and G; and Fig. S1, A–E). Scale bar in B, 20 nm (valid for A and B); in G, 20 nm (valid for D–G).

Figure 2.

CA projections C1a, e, and c are lost in the pf16 mutant, and the PF16 C-terminus locates to the C1a projection. (A–P) Tomographic slices (columns 1 and 2) and isosurface renderings (columns 3 and 4) of the averaged CA repeats of WT (A–D), 9+2 pf16 (E and F), and the tagged rescue pf16;PF16::BCCP without adding nanogold (I–L, control) and after treatment with streptavidin gold (M–P) in cross-sectional (columns 1 and 3) and longitudinal views (columns 2 and 4). The thin blue line in A indicates the location of the slices shown in column 2. The C1a, e, and c projections (indicated by black brackets in A and C) were missing in the pf16 mutant (white bracket in E). When the C-terminus of PF16 was tagged with BCCP, the additional density of the BCCP-streptavidin-gold label was detected in the C1a projection (blue arrowheads in M–P); this label density was not observed in WT (white arrowheads in A–D) or control samples (white arrowheads in I–L). Scale bar in N, 20 nm (valid for all EM images).

Figure 3.

Motility defects of C1a-e-c mutants, and FAP92 is a C1a protein. (A) Average swimming velocities of WT and mutant cells and rescue strains fap76-1;BCCP::FAP76 (fap76-1res) and fap216;BCCP::FAP216 (fap216res). In the sequence as shown in the histogram n = 30, 30, 30, 30, 30, 30, 32, 40, and 34. *, Significant difference (Student’s t test, P < 0.01) compared with WT. Error bars indicate ± SEM. (B) Each colored line represents the swimming path of one cell recorded for 1 s. The swimming paths of mutants are curving. (C) The percentage of cells displaying photoshock response, i.e., that switched from forward swimming to backward swimming upon light stimulus, compared with WT. (D–K) Comparisons of tomographic slices (D–F and H–J) and isosurface renderings (G and K) between the averaged CA repeats of WT (D–G) and fap92 axonemes (H–K) viewed in cross-sectional (D and H), longitudinal (E, G, I, and K) and top-down (F and J) orientations, showing a sheet-like density that is present in the WT C1a projection (red arrowheads in D–G) but missing in fap92 (white arrowheads in H–K). Thin blue lines indicate the location of the tomographic slices shown in E and F (in D) and I and J (in H). Scale bar in B (WT), 50 µm (valid for all images in B); D, 20 nm (valid for D–F and H–J).

Figure 4.

FAP76 is a C1c protein with multiple connections to neighboring structures. (A–H) The comparison of tomographic slices (A, B, E, and F) and isosurface renderings (C, D, G, and H) between the averaged CA repeats of WT (A–D) and fap76-1 axonemes (E–H) viewed in cross-sectional (A, C, E, and G) and longitudinal (B, D, F, and H) orientations shows a triskelion-shaped structure that is present in the WT C1c projection (magenta brackets and arrowheads in A–D) but missing in fap76-1 (white brackets and arrowheads in E–H). The thin blue line in A indicates the location of the tomographic slices shown in B and F. (I and J) The isosurface rendering of a difference map between the averaged CA of WT and fap76-1 shows the triskelion-shaped FAP76 densities (magenta) in longitudinal view superimposed on the averaged CA of fap76-1 (gray) in overview (I) and zoomed-in (J, location indicated by box in I). Note the three connections (magenta arrowhead and #1–3) of FAP76 with neighboring structures, i.e., #1 with the C1c projection close to C1e (I), and #2 and #3 with C1d densities (yellow in J). Scale bar in F, 20 nm (valid for A, B, E, and F).

Figure 5.

The FAP76 N-terminus is located at the interface between C1c and C1e projections. (A) Immunofluorescence light microscopy images of axonemes isolated from fap76-1 (left) and fap76-1;BCCP::FAP76 rescue (right) probed by anti-acetylated-tubulin antibody (red) and fluorescently tagged streptavidin (green). The streptavidin signal indicates that the BCCP-tagged FAP76 was assembled into the axoneme of the rescued strain. (B) SDS–polyacrylamide gel stained by a silver enhancement kit (top) to show that a specific band of appropriate relative mobility could be detected in the fap76-1;BCCP::FAP76 axonemes treated with streptavidin-Au, but not in the control (–Au). Coomassie brilliant blue (CBB) staining (bottom) shows the tubulin bands as loading controls. (C) A classification analysis of the C1c psu1 (black arrows in E and F) in WT and fap76-1, and fap76-1;BCCP::FAP76 rescue axonemes (white arrows in H, I, K, and L). The particle numbers (n) included in the averages for WT, fap76-1, fap76-1;BCCP::FAP76 (+Au), and fap76-1;BCCP::FAP76 (–Au) were 664, 927, 1,089, and 935 (see Table S3). (D–L) Comparison of tomographic slices (D, E, G, H, J, and K) and isosurface renderings (F, I, and L) of the averaged CA repeats of WT (D–F) versus fap76-1;BCCP::FAP76 rescue strain either without (G–I) or treated with (J–L) streptavidin gold, viewed in cross-sectional (D, G, and J) and longitudinal (E, F, H, I, K, and L) orientations. The additional density of the streptavidin-gold label at the interface between the C1c and C1e projections in the gold-treated rescue strain (blue arrowheads in J–L) is not observed in WT or control CA (white arrowheads in D–I). Thin blue line in D indicates the location for the tomographic slices shown in E, H, and K. Scale bar in A, 5 µm (valid for all fluorescence images); K, 20 nm (valid for all EM images in D–L).

Figure 6.

FAP81 is required for the stable assembly of the C1e-c subcomplex. (A–F) Comparison of tomographic slices (A and D) and isosurface renderings (B, C, E, and F) between the averaged CA repeats of WT (A–C) and fap81 (D–F), viewed in cross-sectional (A and D) and longitudinal (B, C, E, and F) orientations, shows that the C1e and C1c projections (orange/yellow brackets in A) are present in WT but missing in fap81 (white bracket in D). (G) Schematic drawing of the WT CA structure in longitudinal view to outline the densities of the C1a (red) and C1e (orange) projections and their interactions. Note the transition from 16-nm periodicity (C1a) to 32-nm periodicity (C1e). Blue dots indicate the locations of the FAP76 N-terminus. (H) An overlay of the averaged CA repeats of fap76-1 (transparent gray) and fap81 (colored) shows the interaction (cyan arrowheads) between a rod-shaped C1d density and the gray C1c density, which remains present in the fap76-1 CA and consists of at least the FAP81 protein. Scale bar in A, 20 nm (valid for A and D).

Figure 7.

FAP216, a C1c protein that may bridge the C1 microtubule and peripheral C1c densities. (A–H) Comparisons of tomographic slices (A, B, E, and F) and isosurface renderings (C, D, G, and H) between averaged CA repeats of WT (A–D) and fap216 (E–H) viewed in cross-sectional (A, C, E, and G) and longitudinal (B, D, F, and H) orientations shows that the fap216 CA lacks a small C1c density (white arrowheads in E–H) that in WT connects between C1 microtubule protofilament 2 and peripheral C1c densities (yellow arrowheads in A–D). In addition, part of the C1e projection (orange arrowheads in A and C) that bridges between the C1e-c subcomplex and the C1a projection is reduced in fap216 (light orange arrowhead in E and G). Classification analysis confirmed that 57% of the fap216 CA repeats lack this C1e density (see Fig. S4). Scale bar in F, 20 nm (valid for all EM images).

Figure 8.

Models of the C1a-e-c supercomplex. (A) The isosurface rendering shows the 3D structure of the averaged Chlamydomonas WT CA repeat viewed in longitudinal orientation. (B and C) A schematic drawing of the 32-nm repeat of the Chlamydomonas WT CA in longitudinal orientation (B; same orientation as in A) and cross-sectional view (C) to highlight the protein compositions of the C1a-e-c supercomplex within the ciliary CA; this includes densities containing the subunits PF16 and FAP92 (dark red), FAP76 (magenta), FAP81 (yellow), and FAP216 (green). Blue dots indicate the C-terminus of PF16 and N-terminus of FAP76 proteins. (D–H) Schematic drawings of the 32-nm CA repeat of various C1a-e-c mutants in cross-sectional view to show the observed structural defects.