Localization of RSP3 fusion proteins. (A) Domain organization of RSP3 paralogs predicted by SMART (http://smart.embl-heidelberg.de/). (B) Western blot analyses of ciliary proteins isolated from cells expressing RSP3 paralogs as C-terminal 3-HA tagged proteins under the control of native promoters. A star indicates a band corresponding to the position of fusion proteins, RSP3A-3HA, RSP3B-3HA, and RSP3C-3HA. (C-E) Ciliary localization of the RSP3 paralogs expressed as C-terminally -3HA tagged fusions under the control of the respective native promoters localize along the entire cilia length except for the ciliary tip. (C) RSP3A-3HA, (D) RSP3B-3HA, and (D) RSP3C-3HA labeled with a mix of anti-HA and anti-acetylated tubulin antibodies to visualize the entire cilia. To the right, the magnified fragments of the cells as indicated by the white insets, stained with anti-HA (red), anti-acetylated α-tubulin (green), and merged images (red and green). Below are images showing both channels but with some shifts to better visualize the presence of the RSP3 paralogs along the entire cilia. Note that all RSP3 paralogs localize along the entire cilia length except for a ciliary tip.

Knockout of particular RSP3 paralogs affects cilia beating to different extents. (A-D) WT (A), RSP3A-KO (B), RSP3B-KO (C), and RSP3C-KO (D) trajectories recorded for 3 sec using a high-speed video camera. The cell swimming paths are indicated by the parallel colored lines. Bar = 500 µm. (E) A comparison of the length of trajectories recorded for 3 sec. The red bar position to the right represents the standard deviation. The wild-type cells swam on average 388+/-32 µm/s (n= 100), and the swimming speeds of the RSP3A-KO, RSP3B-KO, and RSP3C- KO mutants were reduced to 272+/-27 µm/s (n=100), 48+/-9 µm/s (n=82), and 217+/-25 µm/s (n=101), respectively. Statistical significance was calculated using student t-test. (F) Drawings showing examples of the most frequently observed subsequent positions of a cilium of WT and RSP3 knockout cells. The ciliary waveform and amplitude in the RSP3B-KO mutant can vary both in the case of neighboring cilia and during subsequent cycles of the same cilium. (G) Graph showing a range of cilia beating frequencies in wild-type and RSP3 knockout mutants. The red bar represents the standard deviation. Statistical significance was calculated using student t-test. (H) Examples of the kymographs used to calculate cilia beating frequency.

Cryo-ET and subtomogram averaging analyses of the ciliary ultrastructure of wild-type cells and RSP3 knockout mutants. (A) Subtomogram averages of WT, RSP3A-KO, RSP3B-KO and RSP3C-KO mutants. RS1: green; RS2: red; RS3: blue. Inner dynein arms (IDAs): light pink, N-DRC: yellow, CCDC96-CCDC113 complex: dark pink. The black arrow in the RSP3B-KO subtomogram points to the place corresponding to the position of IDAc in the WT cilia. (B) Tomographic slices from WT, RSP3A-KO, RSP3B-KO and RSP3C-KO cilia showing the heterogeneity of RS distributions. Scale bar: 50 nm.

Graphical illustration of quantitative proteomics data. (A) Principal component analysis (PCA) score plot from fold-change values of the differentially expressed proteins showing similar grouping of ciliomes identified by mass spectrometry using either LFQ (graphs to the left) or TMT (graphs to the right) strategy based on two independent variables. The analyzed ciliary protein samples were color-coded as follows: WT - red, RSP3A-KO - green, RSP3B-KO - blue, and RSP3C-KO – orange. (B) Volcano plots of wild-type and RS mutant ciliary proteins identified by mass spectrometry using either LFQ (graphs to the left) or TMT (graphs to the right) strategy (N=3) showing changes in mutant ciliomes. FDR=0.05. Proteins Red dots represent proteins that are at a significantly lower level in RSP mutants compared to wild-type cells, while blue dots represent proteins that are more abundant in mutant cilia. Names are provided for known RSP proteins. The statistical significance of the difference in protein enrichment between wild-type and RS mutant ciliomes calculated using Student’s t-test.

Venn diagrams indicating the overlap of differentially expressed proteins between different radial spoke mutants as estimated by global ciliomes obtained either by LFQ or TMT approaches. (A) Comparison of the total number of proteins identified using LFQ and TMT approaches (B-E). The overlap of proteins that were either reduced (to the left) or elevated (to the right) in different radial spoke mutants as detected using (B) TMT and (C-E) LFQ methods. Images were prepared in Adobe Photoshop based on data sorted using the Excel program.

A schematic summary of the radial spoke protein composition in Tetrahymena cilia prepared based on collected ultrastructural and proteomic data. (A) A summary of the RSP3 paralog-dependent types of RS1 and RS2 spokes. RSP3A-RSP3B heterodimer-containing RS1 and RSP3C-RSP3C-homodimer-containing RS2 are less probable. (B-D) Graphical presentation of the radial spokes composition, (B) RS1 containing either RSP3A or RSP3B homodimers, (C) RS2 containing either RSP3B homodimer or RSP3B-RSP3C heterodimer, (D) RS3. Names of the RS-specific RSPs existing in Tetrahymena as a single ortholog are in light blue; names of the RS- type specific RSP paralogs are in red; new candidate proteins are in orange; RS-associated enzymes are in blue. Note that in the case of RS2-type II (RSP3C-containing), RSPs were reduced in RSP3C-KO cilia and thus likely bind with RSP3C. DYH- dynein heavy chains.

RSP proteins assigned to particular radial spoke types based on bioinformatics, proteomic, and cryo-EM data.

Comparative analyses of the levels of inner dynein arms (IDAs) components in RSP3 mutants

Alterations in RSP3A, RSP3B, and RSP3C loci in engineered Tetrahymena knockout mutants. (A1, B1, C1) Schematic representations of the RSP3A (A), RSP3B (B), and RSP3C (C) loci in the wild-type genome (WT, upper gray bar) and obtained knockout cells (RSP3-KO, lower gray bar) with the size of the fragment of the open reading frame removed in the knockout cells indicated and the position of the annealing of primers (blue and red arrows) used to test changes in the analyzed loci marked. Numbers near the blue arrows indicate the distance of the primer annealing site from the neo4 cassette that replaced the removed fragment of the open reading frame in mutated loci. Note that the primer represented by a red arrow recognizes the nucleotide sequence deleted in mutant cells. Numbers above the gray bars representing the open reading frame indicate the distance between either the start or stop codons and the neo4 cassette insertion site. (A2, B2, C2) A table showing the expected size of the PCR products obtained using genomic DNA purified from wild-type (WT) and knockout cells (KO) and primers as indicated. (A3, B3, C3) PCR analyses of the RSP3 loci RSP3A (A3), RSP3B (B3), and RSP3C (C3) in independently obtained RSP3 mutants using either both primers annealing outside the neo4 cassette or one of the primers recognizing the deleted gene fragment as indicated in A1, B1, and C1, respectively.

Knockout of RSP3B affects cilia beating synchrony. Examples of the kymographs showing cilia beating synchrony.

Consensus subtomogram averages and three-dimensional classification of the radial spokes in Tetrahymena RSP3 knockout mutants. In RSP3A-KO mutant cilia, RS1 is missing in ∼35% of 96-nm axonemal units. The analyses of n=2099 axonemal repeats with RS1 defects revealed either the lack of the entire RS1 (∼30%, n=639 units) or RS1 except for the RS1 base (∼67%, n=1417). Three percent of the recorded repeats were not classified due to the low number of particles. In RSP3B-KO mutant cilia, all analyzed axonemal repeats (n=2092) lacked RS2. Additionally, 77.5% of the axonemal repeats (n=1622) had RS1 spokes defects. The RS1 spokes were either completely missing (51.5%, n=1078) or only their base was well visible (26%, n=544). Among the collected n=2790 axonemal repeats of RSP3C- KO mutant cilia, 173 axonemal repeats (7%) were unclassified. The analyses of the remaining n=2617 axomenal repeats revealed that ∼20% (n=527) lacked the entire RS2. The 3D classification of the remaining 80% of repeats using the RS2 mask covering an RS2 head and stalk showed that the remaining units grouped into two categories: (i) with an intact RS2 structure (∼18%, n=469) and (ii) with a well-visible RS2 base (∼62%, n=1621).

Data collection and analysis parameters for cryo-ET and subtomogram averaging.

(A-I) Immunofluorescence confocal images of Tetrahymena cells expressing RSP proteins as C-terminal –HA-BirA* fusions under the control of respective native promoters detected using anti-HA antibodies. (A) RSP3A-HA-BirA*, (B) RSP3B-HA-BirA*, (C) RSP3C-HA-BirA*, (D) RSP4A-HA-BirA*, (E) RSP4B-HA-BirA*, (F) RSP4C-HA-BirA*, (G) CFAP206-HA-BirA*, (H) CFAP61-HA-BirA*, (I) CFAP91-HA-BirA*. (J-M) Western blot analyses of the ciliary proteins isolated from Tetrahymena cells expressing RSP proteins as –HA- BirA* fusions under the control of respective native promoters. A star indicates a band corresponding to the position of fusion proteins. (J) Western blot-based identification of the RSP- HA-BirA* fusion proteins. (K-M) Western blot-based analyses of the biotinylated ciliary proteins in either WT cells or cells expressing RSP-HA-BirA* under the control of the respective native promoter, all grown for 4 hrs in a 10 mM Tris-HCl buffer (pH 7.4) supplemented with biotin, detected using HRP-conjugated streptavidin.

Two-dimensional analyses of the RSP isoforms. The radial spokes proteins were expressed as C-terminal 3HA fusions under the control of the respective native proteins. The isoelectric focusing of the ciliary proteins (30 µg) was performed using 7 cm 4-7 or 3-10 ready strips. The theoretical pI and Mw values were calculated using https://web.expasy.org/compute_pi/. The RSP isoforms were detected using anti-HA antibodies.

Identification of Tetrahymena RSP orthologs

Figure 6-figure supplement 3-Table 1A. List of Tetrahymena thermophila radial spoke proteins identified in Tetrahymena Genome Database (TGD, https://tet.ciliate.org) using Chlamydomonas RSP proteins as baits. In the case of multiple hits (see details in Table S1B), it was analyzed if identified Tetrahymena RSP orthologs are present in the Tetrahymena ciliomes (Figure 4-Source Data 2’), co-immunoprecipitate with known RSPs (Figure 6-Source Data 1’), or are biotinylated in cilia of cells expressing RSP proteins fused with mutated BirA* ligase (Figure 6-Source Data 2’). The identification of Tetrahymena orthologous proteins was verified by a reverse blast search against Chlamydomonas reinhardtii proteins (https://blast.ncbi.nlm.nih.gov). Domains were identified using SMART (http://smart.embl-heidelberg.de/) and InterProScan (https://www.ebi.ac.uk/interpro) programs.

Figure 6-figure supplement 3-Table 1B. Tetrahymena RSP orthologs identified with multiple hits in TGD. In cases when numerous proteins were identified with similar scores in blastp search of Tetrahymena total proteome (TGD) using Chlamydomonas RSP as a bait, it was analyzed which of those proteins are present in Tetrahymena ciliome (Tables S3 and S4) and either co- immumoprecipitate (co-IP, Table S5) with known RSPs or are biotinylated (BioID, Table S6) in cells expressing known RS proteins as fusions with mutated BirA* ligase. Arrows indicate if there was a significant change (↑ increase or ↓ decrease) in the protein level in studied knock-out mutants (based on TMT analyses, Table S4). ns – change not significant, nd – protein not detected in the Tetrahymena ciliome.

Figure 6-figure supplement 3-Table 2. Comparative analyses of the levels of the radial spoke proteins in wild-type cells and radial spoke mutants

Figure 6-figure supplement 3-Table 3. Comparative analyses of the levels of potential radial spoke-associated proteins in wild-type cells and radial spoke mutants

Figure 6-figure supplement 3-Table 4. Comparative analyses of the levels of proteins with an enzymatic activities in radial spoke mutants