Rsph4a is essential for the triplet radial spoke head assembly of the mouse motile cilia

Motile cilia/flagella are essential for swimming and generating extracellular fluid flow in eukaryotes. Motile cilia harbor a 9+2 arrangement consisting of nine doublet microtubules with dynein arms at the periphery and a pair of singlet microtubules at the center (central pair). In the central system, the radial spoke has a T-shaped architecture and regulates the motility and motion pattern of cilia. Recent cryoelectron tomography data reveal three types of radial spokes (RS1, RS2, and RS3) in the 96 nm axoneme repeat unit; however, the molecular composition of the third radial spoke, RS3 is unknown. In human pathology, it is well known mutation of the radial spoke head-related genes causes primary ciliary dyskinesia (PCD) including respiratory defect and infertility. Here, we describe the role of the primary ciliary dyskinesia protein Rsph4a in the mouse motile cilia. Cryoelectron tomography reveals that the mouse trachea cilia harbor three types of radial spoke as with the other vertebrates and that all triplet spoke heads are lacking in the trachea cilia of Rsph4a-deficient mice. Furthermore, observation of ciliary movement and immunofluorescence analysis indicates that Rsph4a contributes to the generation of the planar beating of motile cilia by building the distal architecture of radial spokes in the trachea, the ependymal tissues, and the oviduct. Although detailed mechanism of RSs assembly remains unknown, our results suggest Rsph4a is a generic component of radial spoke heads, and could explain the severe phenotype of human PCD patients with RSPH4A mutation.

). How consistently are these proteins reduced in the mutant cells? Could fields that are more representative be shown in these figures? Quantitative western blotting could also potentially address the difference in protein levels that might be difficult to see by immunofluorescence.
Response: Thank you very much for this comment. This is absolutely correct. As pointed, localization of Rsph1 is retained in the ciliated cells in the trachea (Fig.5P) and the brain (Fig.6P) of Rsph4a KO mice in our immunofluorescence (IF) data. It is very difficult to quantify and compare the IF data between wild type and KO mice. Thus, we have carried out western blotting according to your suggestion (Fig.S3). In the trachea, we detect significant difference of protein level of Rsph1 between the wildtype and the Rsph4a KO mice, suggesting that level of Rsph1 is reduced in trachea ciliated cells of Rsph4a KO mice (Fig. S3). Although we cannot explain the exact reason of the difference between IF and western blotting, one possibility is that small amount of Rsph1 protein still remains in Rsph4a KO cells and this may be amplified by antibody reaction. Please note that we have carried out western blotting of Rsph4a using the testis tissue, because our antibody cannot detect clear signal band of Rsph4a protein in the trachea tissue.
In addition, on Rsph23, we have carried out western blotting. In the trachea, we can detect significant difference of protein level of Rsph23 between the wildtype and the Rsph4a KO mice as well as Rsph1 (Fig.S3). Therefore, we conclude that protein both level of Rsph1 and Rsph23 is reduced in the trachea cells of Rsph4a KO mice. This is now mentioned in the text (page 6, "RESULTS" 1 st paragraph, 7th Line & 18th Line) and a new figure (Fig. S3 RSPH3 (Jeanson et al., 2015), which should be referenced. Response: Thank you. This is correct. To prevent misleading, we have deleted the phrase of "and so on" (Page 3, "INTRODUCTION", 1 st paragraph, 5th Line).

DISCUSSION, 2nd paragraph: The authors state that "the stability of the neck/arch is most likely different among species." This is an important point, but is it possible that assembly of the neck/arch could be different as well?
Response: Thank you very much for this comment. This is absolutely important point.
Our data suggests that the neck/arch is disrupted in the trachea cilia of Rsph4a KO mice.
To address this issue, we should compare our cryo-ET structure with cryo-ET structure of trachea cilia in Rsph4a-deficient human patients or the other vertebrate.
Unfortunately, we have not obtained any cryo-ET data of human or the other vertebrate yet, we have better change the tone of the sentence at this time. We have changed the phrase of "is most likely" to "may be" in the sentence (Page 7, "Discussion" 2 nd paragraph, 10th Line) Response: Thank you very much for this comment. This is absolutely important. At first, we have submitted again images of acetylated tubulin with shorter exposure time accordingly (Fig.5-7). Next, we have tried to quantify intensity of immunofluorescence (IF) data of Rsph9 and Rsph1 and compare the wildtype with the Rsph4a KO. However, it is very difficult for us to evaluate because there is no linear relationship between intensity of IF images and amount of protein. Thus, alternatively, we have carried out western blotting using the trachea and the testis tissue. In result, we have observed significant differences of Rsph1 between wild type (Band intensity = 131,000) and

METHODS, Generation of
Rsph4a KO mouse (Band intensity = 5,180) in the trachea tissue (Fig. S3). The IF data of the trachea show some remaining intensity of Rsph1 in cilia of Rsph4a KO mice (Fig.5P). Although we cannot explain the exact reason of the difference between immunofluorescence and western blotting, one possibility is that small amount of Rsph1 protein still remains in Rsph4a KO cells and this may be amplified by antibody. Please note that we have carried out western blotting of Rsph4a using the testis tissue, because our antibody cannot detect clear signal band of Rsph4a protein in the trachea tissue.
On Rsph9, we believe that an apparent difference of Rsph9 is observed between wildtype and Rsph4a KO in our IF data (Fig.5G&J, Fig.6G&J, Fig.7G&J). These revisions are now mentioned in the text (Page 6, "RESULTS" 1 st paragraph, 5th line) and Figures (Fig.5-7 & a new figure (Fig. S3)). We are trying to determine cryoEM structure of mouse oviduct cilia. Unfortunately, however, number of isolated cilia is not sufficient to obtain 3D structure (bottom Fig), thus this must be our future work. On the ependymal tissue, we have not succeeded to isolate cilia. This is technically tough at this time.

CryoET of mouse oviduct cilia (Preliminary) c) How do you explain differences in localization of especially Rsph1 and Rsph23 in trachea, ependymal cells, oviduct? Especially the normal appearing Rsph1 localization in trachea and ependyma of Rsph4 mutant mice?
Response: Thank you for this comment. As you point out, our IF data show Rsph1 localization seems normal in the trachea and the ependymal cells of Rsph4a KO mice.
To clarify protein level of Rsph1 and Rsph23, we have carried out western blotting using the trachea tissue. In result, we have observed significant difference of level of both Rsph1 and Rsph23 between the wild type and the Rsph4a KO mouse (Fig.S3).
This suggests that Rsph1 and Rsph23 decrease in the Rsph4a KO cells in the trachea.
We conclude that the spoke head proteins (Rsph1, 9) and the neck protein (Rsph23) are affected in the trachea cells of Rsph4a KO mice. This is now mentioned in the text (Page 6, "RESULTS" 1 st paragraph, 5th line& 17th line) and a new figure (Fig. S3). Figure 7 appears to be basal body staining, please comment Response: Thank you for this comment. In Fig.7L, Rsph9 seems to stay at the basal body in the oviduct of Rsph4a KO mice. But, we are not sure this is indeed basal body localization since we do not see basal body marker. Response: Thank you for this comment. This is absolutely correct. We have motion data of sperm in Rsph4a KO mice, but the number of example is not sufficient (N = 2 cells).

d) Rsph9 staining in
We have deleted information on the sperm in M&M.

Supplementary videos 1,5,6 appear to be different format and cannot be assessed.
Response: Thank you. We have revised and unified the format of Video1, Video5, and Video6: (the number of frames and the speed of movie).

Please use proper nomenclature for proteins in mouse
Response: Thank you. We have corrected "Rsph4a KO mice" to italic "Rsph4a KO mice.

Reviewer #3
The