Absence of CEP78 causes photoreceptor and sperm flagella impairments in mice and a human individual
Curation statements for this article:-
Curated by eLife
eLife assessment
This paper is of interest to scientists within the cilia and centrosome fields, in particular those studying photoreceptor and sperm development and the diseases associated with their dysfunction. The authors describe the generation and characteristics of Cep78 knockout mice. Consistent with the phenotype observed in patients carrying mutations in CEP78, Cep78 knockout mice show degeneration in photoreceptor cells as well as male infertility associated with multiple morphological abnormalities of the sperm flagella (MMAF). The phenotypic characterisation of Cep78 knockout mice is thorough and convincing, and the Cep78 knockout model will be useful for further elucidating disease mechanism in humans and for potential therapy development. The authors also provide results suggesting that CEP78 directly interacts with IFT20 and TTC21A (IFT139) to form a trimeric complex, but this claim is not justified by the data provided.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Developmental Biology (eLife)
Abstract
Cone-rod dystrophy (CRD) is a genetically inherited retinal disease that can be associated with male infertility, while the specific genetic mechanisms are not well known. Here, we report CEP78 as a causative gene of a particular syndrome including CRD and male infertility with multiple morphological abnormalities of sperm flagella (MMAF) both in human and mouse. Cep78 knockout mice exhibited impaired function and morphology of photoreceptors, typified by reduced ERG amplitudes, disrupted translocation of cone arrestin, attenuated and disorganized photoreceptor outer segments (OS) disks and widen OS bases, as well as interrupted connecting cilia elongation and abnormal structures. Cep78 deletion also caused male infertility and MMAF, with disordered ‘9+2’ structure and triplet microtubules in sperm flagella. Intraflagellar transport (IFT) proteins IFT20 and TTC21A are identified as interacting proteins of CEP78. Furthermore, CEP78 regulated the interaction, stability, and centriolar localization of its interacting protein. Insufficiency of CEP78 or its interacting protein causes abnormal centriole elongation and cilia shortening. Absence of CEP78 protein in human caused similar phenotypes in vision and MMAF as Cep78 −/− mice. Collectively, our study supports the important roles of CEP78 defects in centriole and ciliary dysfunctions and molecular pathogenesis of such multi-system syndrome.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
This manuscript describes the generation and characterization of a mouse knockout model of Cep78, which codes for a centrosomal protein previously implicated in cone-rod dystrophy (CRD) and hearing loss in humans. Previous work in cultured mammalian cells (including patient fibroblasts) also indicated roles for CEP78 in primary cilium assembly and length control, but so far no animal models for CEP78 were described. Here, the authors first use CRISPR/Cas9 to knock out Cep78 in the mouse and convincingly demonstrate loss of CEP78 protein in lysates of retina and testis of Cep78-/- animals. Next, by careful phenotypic analysis, the authors demonstrate significant defects in photoreceptor structure and function in these mutant animals, which become more severe over a 9 (or 18) month period. …
Author Response
Reviewer #1 (Public Review):
This manuscript describes the generation and characterization of a mouse knockout model of Cep78, which codes for a centrosomal protein previously implicated in cone-rod dystrophy (CRD) and hearing loss in humans. Previous work in cultured mammalian cells (including patient fibroblasts) also indicated roles for CEP78 in primary cilium assembly and length control, but so far no animal models for CEP78 were described. Here, the authors first use CRISPR/Cas9 to knock out Cep78 in the mouse and convincingly demonstrate loss of CEP78 protein in lysates of retina and testis of Cep78-/- animals. Next, by careful phenotypic analysis, the authors demonstrate significant defects in photoreceptor structure and function in these mutant animals, which become more severe over a 9 (or 18) month period. Specifically, TEM analysis demonstrates ultrastructural defects of the connecting cilium and photoreceptor outer segments in the Cep78 mutants, which is in line with previously reported roles for CEP78 in CRD and in regulating primary cilia assembly in humans. In addition to a CRD-like phenotype, the authors also convincingly show that male Cep78-/- animals are infertile and exhibit severe defects in spermatogenesis, sperm flagella structure and manchette formation (MMAF phenotype). Furthermore, the authors provide evidence for an MMAF phenotype from a male individual carrying a previously reported CEP78 c.1629-2A>G mutation, substantiating that CEP78 is required for sperm development and function in mammals and supporting previously published work (Ascari et al. 2020).
Finally, to identify the underlying molecular mechanism by which CEP78 loss causes MMAF, the authors perform some biochemical analyses, which suggest that CEP78 physically interacts with IFT20 and TTC21A (an ortholog of Chlamydomonas IFT139) and might regulate their stability. The authors conclude that CEP78 directly binds IFT20 and TTC21A in a trimeric complex and that disruption of this complex underlies the MMAF phenotype observed in Cep78-/- male mice. However, this conclusion is not fully justified by the data provided, and the mechanism by which CEP78 affects spermatogenesis therefore remains to be clarified.
Specific strengths are weaknesses of the manuscript are listed below.
Strengths:
Overall, the phenotypic characterisation of the Cep78-/- animals appears convincing and provides new evidence supporting that CEP78 plays an important role in the development and function of photoreceptors and sperm cells in vertebrates.
Weaknesses:
- The immunoprecipitation experiments of mouse testis extracts that were used for the mass spectrometry analysis in Table S4 were performed with an antibody against endogenous CEP78 (although antibody details are missing). One caveat with this approach is that the antibody might block binding of CEP78 to some of its interactors, e.g. if the epitope recognized by the antibody is located within one or more interactor binding sites in CEP78. This could explain why the authors did not identify some of the previously identified CEP78 interactors in their IP analysis, such as CEP76 and the EDD-DYRK2-DDB1-VprBP complex (Hossain et al. 2017) as well as CEP350 (Goncalves et al. 2021).
We thank Reviewer #1 (Public Review) for agreeing with us on Cep78 plays an important role in photoreceptors and sperm cells development. We also appreciate Reviewer #1 (Public Review) for pointing out the weaknesses which helped us improve our study.
For the immunoprecipitation experiments of mouse testis extracts, the antigenic sequence of the Cep78 antibody used is p457-741 (NP_932136.2). Cep78 was reported to bind DD-DYRK2-DDB1-VprBP complex, the 1-520aa is responsible for Cep78’s interaction with VprBP, and deletion of p450-497 didn’t affect Cep78’s interaction with VprBP, indicating importance of Cep78 (1-450aa) in interaction with VprBp (Hossain et al. 2017). Our anti-Cep78 antibody is generated using antigen sequence p457-741, the binding of p1-450aa to VprBP is not expected to be blocked by our anti-Cep78 antibody. However, VprBp was not detected by our IP-MS experiment. C-terminal region (395-722aa) of Cep78 overlaps with our Cep78 antibody’s antigenic region (p457-741), and was reported to interact with Cep350 (Goncalves et al. 2021). As a polyclonal antibody, our anti-Cep78 antibody didn’t block the interaction with p457-741, because we still identified Cep350 in our IP-MS. Thus, immunoprecipitation experiments using our Cep78 antibody identified some of the previously known interactors, and the interaction with VprBP may not be blocked by our Cep78 antibody.
The detailed antibody information has now been added to Supplementary Table S7 in our revised supplementary materials.
- Figure 7A-D and page 18-25: based on IPs performed on cell or tissue lysates the authors conclude that CEP78 directly binds IFT20 and TTC21A in a "trimeric complex". However, this conclusion is not justified by the data provided, nor by the previous studies that the authors are referring to (Liu et al. 2019 and Zhang et al. 2016). The reported interactions might just as well be indirect. Indeed, IFT20 is a known component of the IFT-B2 complex (Taschner et al., 2016) whereas TTC21A (IFT139) is part of the IFT-A complex, which suggests that they may interact indirectly. In addition, the IPs shown in Figure 7A-D are lacking negative controls that do not coIP with CEP78/IFT20/TTC21A. It is important to include such controls, especially since IFT20 and CEP78 are rich in coiled coils that tend to interact non-specifically with other proteins.
Thank Reviewer #1 (Public Review) for the comment on protein interaction between Cep78, Ift20, and Ttc21a. As the reviewer pointed out, IFT20 is a known component of the IFT-B2 complex (Taschner et al., 2016) whereas TTC21A (IFT139) is part of the IFT-A complex. Both IFT20 and TTC21a are located at peripheral areas of IFT-B and IFT-A (PMID: 32456460), and are not core components of IFT-A or IFT-B. It is still possible that these two proteins interact with each other. Actually, Liu et al. have revealed interaction between Ift20 and Ttc21a in human sperm (PMID: 30929735). Additionally, to mediate trafficking of ciliary axonemal components, the IFT machinery is recruited to the distal appendages (PMID: 30601682), which is adjacent to the distal end of the (mother) centriole wall, where at the (mother) centriole wall was reported to be located (PMID:35543806). Cep78 may interact with Ift20 and Ttc21a at centriole during cilliogenesis.s
To rule out the nonspecific interaction between Cep78 and Ttc21a or Ift20, we added additional negative controls of Gapdh (Figure 7D) and Ap80-NB-HA (Supplementary Figures S7A-C) in co-IP as the reviewer suggested, and found that the interaction between Cep78 and Ttc21a or Ift20 is specific. To examine if Cep78, Ift20 and Ttc21a formed a complex, we fractionated testicular protein complexes using size exclusion chromatography, and found that Cep78, Ift20 and Ttc21a co-fractioned at the size between158 kDa to 670 kDa (Figure 7E), supporting the formation of a trimeric complex. And our immunofluorescent analysis by SIM also showed co-localization between Cep78 and Ift20 or Ttc21a (Figure 7F). All these data support the interaction among Cep78, Ttc21a and Ift20. In the revised manuscript, we rephrased “direct interaction” as “interaction” at page 18, line 393 in the revised manuscript.
- In Figure 7D, the input blots show similar levels of TTC21A and IFT20 in control and Cep78-/- mouse testicular tissue. This is in contrast to panels E-G in the same figure where TTC21A and IFT20 levels look reduced in the mutant. Please explain this discrepancy.
Thank you for pointing this out. Deletion of Cep78 caused down-regulation of Ttc21a and Ift20 proteins. To better reveal the change of interaction between Ttc21a and Ift20, we have to normalize their interaction against expression levels. To achieve this, we increased the amount of total Cep78-/- testicular proteins to ensure that Ttc21a and Ift20 in the input are at similar levels between Cep78+/- and Cep78-/- testes. Using 3 times the amount of the Cep78+/- testicular proteins for Cep78-/- testicular proteins, we detected similar protein levels of Ttc21a and Ift20 between Cep78-/- and Cep78+/- testes, and the interaction between Ttc21a and Ift20 was shown to be down-regulated after Cep78 deletion. Consistently, the analysis of GAPDH as a loading control in input proteins showed more Cep78-/- testicular proteins than Cep78+/- testicular proteins subjected to analysis. To avoid confusion, we have added description of “The amount of Cep78-/- testicular proteins used was 3 times of that of Cep78+/- proteins” in the legend of Figure 7D in the revised version of manuscript.
- The efficiency of the siRNA knockdown shown in 7J-M was only assessed by qPCR (Figure S4), but this does not necessarily mean the corresponding proteins were depleted. Western blot analysis needs to be performed to show depletion at the protein level. Furthermore, it would be desirable with rescue experiments to validate the specificity of the siRNAs used.
Thank the reviewer for the suggestion. To validate the specificity of the siRNAs used, we performed rescue experiments using rescue plasmid with siRNA targeting sequence synonymously mutated (Supplementary Table S6). The efficiency of siRNA knockdown and effects of rescue experiments were evaluated by both qPCR (Supplementary Figures S4.A-C) and Western Blot (Figures 7.J-K, Supplementary Figures S4.D-E, H-I). The results showed that siRNAs significantly reduced the expression of Cep78, Ift20, and Ttc21a at both mRNA (Supplementary Figures S4.A-C) and protein levels (Figures 7.J-K, Supplementary Figure S4.A-C). Meanwhile, with siRNA treatment, the rescue plasmids rescued the expression of Cep78, Ift20, and Ttc21a at both mRNA (Supplementary Figures S4.A-C) and protein levels (Figures 7.J-K, Supplementary Figures S4.D-E, H-I) compared with the control groups.
In the rescue experiments, we further evaluated whether the effects are specific for Cep78, Ift20 and Ttc21siRNAs in the regulation of cilia and centriole lengths. The results showed that suppression of cilia and centriole lengths by Cep78, Ift20 and Ttc21siRNAs could be rescued by overexpression of rescue plasmids of Cep78syn-HA, Ift20syn-Flag and Ttc21asyn-Flag (Figures 7.N-S).
- Figure 7I: the resolution of the IFM is not very high and certainly not sufficient to demonstrate that CEP78, IFT20 and TTC21A co-localize to the same region on the centrosome, which one would have expected if they directly interact.
Thank the reviewer for the constructive comments. To better demonstrate co-localization of CEP78, IFT20 and TTC21A on the centrosome, we overexpressed Cep78-Halo, Ift20-mCherry and Ttc21a-mEmerald in NIH3T3 cells by lentivirus, and acquired super-resolution images with SIM (N-sim, Nikon, Tokyo, Japan). The SIM results showed that Ift20 and Ttc21a co-localized with Cep78 (Figure 7F). Cep78 was previously reported to localize at the centriole (Goncalves et al., 2021). The co-localization of Cep78, Ift20 and Ttc21a indicated possible important roles of Cep78 in the regulation of Ift20 and Ttc21a in centriole. Our interaction analysis revealed that Cep78 interacted with Ift20 and Ttc21a (Figure 7A-C, Supplementary Figure S7), and formed a complex with Ift20 and Ttc21a (Figure 7E). Loss of Cep78 down-regulated the expression of and interaction between Ift20 and Ttc21a (Figures 7D, G-M).
- It is not really clear what information the authors seek to obtain from the global proteomic analysis of elongating spermatids shown in Figure 3N, O and Tables S2 and S3. Also, in Table S2, why are the numbers for CEP78 in columns P, Q and R so high when Cep78 is knocked out in these spermatid lysates? Please clarify.
Thank the reviewer for the comments. Our global proteomic analysis showed that majority of differentially expressed proteins were down-regulated (Figure 3N), and many proteins are centrosome- and cilia-related proteins and important for sperm flagella and acrosome structures (Figure 3O), which provide insights of downstream molecular events in sperm flagella and acrosome defects after Cep78 deletion.
As to the quantification of CEP78 expression in TMT-based proteomics analysis, the ratio between Cep78-/- and Cep78+/- is relatively high due to the ratio compression effect, a well-known phenomenon in TMT-based proteomics analysis (PMID: 25337643). The actual difference in protein expression is usually higher than the ratio calculated by TMT signals. Actually, our Western blot analysis of CEP78 protein showed absence of expression in Cep78-/- testis. Although TMT labelling has the disadvantage of ratio compression (PMID: 32040177,PMID: 23969891), it is widely used quantitative proteomics analysis, and is demonstrated to be able to identify key pathways and proteins (PMID: 30683861, 33980814).
- Figure 1F and Figure 4K: the data needs to be quantified.
Thank the reviewer for this suggestion. For Figure 4K, we stained Cep78+/- and Cep78-/- spermatids with anti-Centrin 1 to measure the centriole length. The statistical data of centriole length were provided (Figure 4L), showing significantly increased centriole lengths in Cep78-/-spermatids.
For Figure 1F, we quantified the immunofluorescence intensities of cone arrestin of light-adapted retinas of Cep78+/- and Cep78-/- mice at 3-month. The results indicate that immunofluorescence intensity of the cone arrestin was lower in Cep78-/- mice.
- Figure 2A: It is difficult to see a difference in connecting cilium length in control and Cep78-/- mutant retinas based on the images shown here.
Thank you for your suggestion, we have stained retinal cryosections from Cep78+/- and Cep78-/- mice with anti-Nphp1 to visualize connecting cilium, and the data are provided in the revised Figure 2A-B.
Reviewer #2 (Public Review):
In this report, the authors have described the generation and characteristics of Cep78 mutant mice. Consistent with the phenotype observed in patients carrying the mutations in CEP78, Cep78 knock-out mice show degeneration in photoreceptors cells as well as defects in sperm. The author further shows the CEP78 protein can interact with IFT120 and TTC21a. Mutation in CEP78 results in a reduction of protein level of IFT120 and TTC21A and mislocalization of these two proteins, offering mechanistic insights into the sperm defects. Over all the manuscript is well written and easy to follow. Phenotyping is thorough. However, improvement of the background section is needed. In addition, some of the conclusion is not sufficiently supported by the data, warranting further analysis and/or additional experiments. The Cep78 KO mice model established by the author will be a useful model for further elucidating the disease mechanism in human and developing potential therapy.
My comments are the following:
- Introduction. The statement that "CRD usually exists with combination of immotile cilia defects in other systems" is not correct. CRD due to ciliopathy can have cilia-related syndromic defects in other systems but it is a relatively small portion of all CRDs and the most frequently mutated genes are not cilia-related genes, such as ABCA4, GUCY2D, CRX.
Thank the reviewer for the comments. We agree with the reviewer that only a small portion of CRDs are due to cilia defects and can have cilia-related syndromic defects in other systems. We corrected this statement in Line 4, Page 77-79 of the revised version of our manuscript. In our revised version, the statement has been changed to “A small portion of CRDs are due to retina cilia defects, and they may have cilia-related syndromic defects in other systems[1].”
- Introduction: Page 4 CNGB1 encodes channel protein and not a cilia gene. It should be removed since it does not fit.
Thank the reviewer for the comment. According to the reviewer’s suggestion, we removed the description of “mutations in CNGB1 cause CRD and anosmia [3]” at Page 4, Line 81 in the revised manuscript.
- Page 5, given the previous report of CEP78 patients with retina degeneration, hearing loss, and reduced infertility, the statement of "we report CE79 as a NEW causative gene for a distinct syndrome...TWO phenotypes....." Is not accurate.
Thank the reviewer for the comments. We have removed the statement of “NEW” causative gene in Page 5, Line 104 of the revised version of our manuscript. The revised sentence is “In this study, based on results of a male patient carrying CEP78 mutation and Cep78 gene knockout mice, we report CEP78 as a causative gene for CRD and male sterility.”
- Figure 1F, the OS of the cone seems shorter, which might be the reason for weaker arrestin staining in the mutant compared to the heterozygous. Also, it would be better to quantify the staining to substantiate the statement.
Thanks for this suggestion. For Figure 1F, we have quantified the immunofluorescence intensity of cone arrestin in Cep78+/- and Cep78-/- light-adapted retinas at 3-month. The results indicate that immunofluorescence intensity of the cone arrestin was significantly lower in Cep78-/- mice.
- Figure 1K, panel with lower magnification would be useful to get a better sense of the overall structure defect of the retina. Is the defect observed in the cone as well?
Thank the reviewer for the comment. As suggested by the reviewer, we have provided images of lower magnification to show the overall structure by TEM, showing disruption of most outer segment in Cep78-/- retina. It is difficult to distinguish whether the disordered outer segment structure belongs to a cone or a rod cell. The images are now provided as Figure 1L in the revised manuscript.
We observed the abnormality of photopic b-wave amplitudes (Figure 1B, E) and decreased intensity of cone arrestin in light-adapted retinas (Figure 1F, G) in Cep78-/- mice, which indicate that the function of cone cells is damaged.
- Figure 2A, NPHP1 or other markers specifically label CC would be more useful to quantify the length of CC. Also need to provide a notation for the red arrows in Figure 2. In addition, the shape of CC in the mutant seems differ significantly from the control. It seems disorganized and swollen.
Thank the reviewer for the suggestion. According to the reviewer’s suggestion, we have stained anti-Nphp1 in retinal cryosections from Cep78+/- and Cep78-/- mice to visualize connecting cilium, and quantified the length of CC. The results showed that connecting cilia were shorter in Cep78-/- mice. These data are showed in Figure 2A-B.
Besides, we observed that upper parts of connecting cilia were swelled with disorganized microtubules in TEM (Figure 2E-G). The red arrows in Figure 2E-G indicated swelled upper part of connecting cilia and disorganized microtubules of Cep78-/- photphoreceptors, we added this description in the figure legend.
- Evidence provided can only indicate direct interaction among CEP78/IFT20/TTC21A.
Thanks for the comment. To further validate the interaction between Cep78 and Ttc21a or Ift20, we performed reciprocal co-IP between Cep78 and Ttc21a or Ift20 by overexpression (Figure 7A-C), and also added relevant negative control of Gapdh (Figure 7D) and Ap80-NB-HA (Supplementary Figures S7A-C) in co-IP as negative controls to avoid non-specific interaction. Besides, we provided evidence that Cep78, Ift20 and Ttc21a formed a complex, as they all co-fractioned in a testicular protein complex at the size between158 kDa to 670 kDa using size exclusion chromatography (Figure 7E). Additionally, we performed super-resolution analysis of immunofluorescent localizations, and observed co-localization between Cep78 and Ttc21a or Ift20 by SIM. With these data, we think that Cep78 interacts with Ttc21a and Ift20 and they form a complex. We rephrased “direct interaction” as “interaction” in the manuscript.
Reviewer #3 (Public Review):
Authors were aiming to bring a deeper understanding of CEP78 function in the development of cone-rod dystrophy as well as to demonstrate previously not reported phenotype of CEP78 role in male infertility.
It is important to note, that the authors 're-examined' already earlier published human mutation, 10 bp deletion in CEP78 gene (Qing Fu et al., 10.1136/jmedgenet-2016-104166). This should be seen as an advantage since re-visiting an older study has allowed noting the phenotypes that were not reported in the first place, namely impairment of photoreceptor and flagellar structure and function. Authors have generated a new knockout mouse model with deleted Cep78 gene and allowed to convey the in-depth studies of Cep78 function and unleash interacting partners.
The authors master classical histology techniques for tissue analysis, immunostaining, light, confocal microscopy. They also employed high-end technologies such as spectral domain optical coherence tomography system, electron, and scanning electron microscopy. They performed functional studies such as electroretinogram (ERG) to detect visual functions of Cep78-/- mice and quantitative mass spectrometry (MS) on elongating spermatids.
The authors used elegant co-immunoprecipitation techniques to demonstrate trimer complex formation.
Through the manuscript, images are clear and support the intended information and claims. Additionally, where possible, quantifications were provided. Sample number was sufficient and in most cases was n=6 (for mouse specimens).
The authors could provide more details in the materials and methods section on how some experiments were conducted. Here are a few examples. (i) Authors have performed quantitative mass spectrometry (MS) on elongating spermatids lysates, however, did not present specifically how elongating spermatids were extracted. (ii) In the case of co-IPs authors should provide information on what number of cells (6 well-plate, 10 cm dish etc) were transfected and used for co-IPs. Furthermore, authors could more clearly articulate what were the novel discoveries and what confirmed earlier findings.
The authors clearly demonstrate and present sufficient evidence to show CEP78/Cep78 importance for proper photoreceptor and flagellar function. Furthermore, they succeed in identifying trimer complex proteins which help to explain the mechanism of Cep78 function.
The given study provides a rather detailed characterization of human and mouse phenotype in response to the CEP78/Cep78 deletion and possible mechanism causing it. CEP78 was already earlier associated with Cone-rod dystrophy and, this study provides a greater in-depth understanding of the mechanism underlying it. Importantly, scientists have generated a new knock-out mouse model that can be used for further studies or putative treatment-testing.
CEP78/Cep78 deletion association with male infertility is not previously reported and brings additional value to this study. We know, from numerous studies, that-testes express multiple genes, some are unique to testes some are co-expressed in multiple tissues. However, very few genes are well studied and have clinical significance. Studies like this, combining patient and animal model research, allow to identify and assign function to poorly characterized or yet unstudied genes. This enables data to use in basic research, patient diagnostics and treatment choices.
We would like to thank Reviewer #3 (Public Review) for positive comments on our work.
As to the suggestions to provide some details in the materials and methods by the reviewer, we added the description of STA-PUT method for spermatids purification at Page 34, Line 729-741 in the revised manuscript, the amount of cells used for co-IPs “10 cm dish HEK293T were transfected (Vazyme, Nanjing, China) wit 5μg plasmid for each experimental group.” at Page 36, Line 783-784 in the revised manuscript.
We also highlighted our new discovery and ensured that all previous published findings are accompanied by references, we added “We further explored whether c.1629-2A>G mutation in this previously visited patient would disturb CEP78 protein expression and male fertility. Blood sample was collected from this patient and an unaffected control for protein extraction.” at Page 17, Line 335. We also added “The major findings of our study are as follows: we found CEP78 as the causal gene of CRD with male infertility and multiple morphological abnormalities of the sperm flagella using Cep78-/- mice. A male patient carrying CEP78 c.1629-2A>G mutation, whom we previously reported to have CRD [8], was found to have male infertility and MMAF in this study. Cep78 formed a trimer with sperm flagella formation enssential proteins IFT20 and TTC21A (Figure 8), which are essential for sperm flagella formation[16, 18]. Cep78 played an important role in the interaction and stability of the trimer proteins, which regulate flagella formation and centriole length in spermiogenesis. ” at the first paragraph of discussion, which is Page 21, Line 447-456 of our revised manuscript.
-
eLife assessment
This paper is of interest to scientists within the cilia and centrosome fields, in particular those studying photoreceptor and sperm development and the diseases associated with their dysfunction. The authors describe the generation and characteristics of Cep78 knockout mice. Consistent with the phenotype observed in patients carrying mutations in CEP78, Cep78 knockout mice show degeneration in photoreceptor cells as well as male infertility associated with multiple morphological abnormalities of the sperm flagella (MMAF). The phenotypic characterisation of Cep78 knockout mice is thorough and convincing, and the Cep78 knockout model will be useful for further elucidating disease mechanism in humans and for potential therapy development. The authors also provide results suggesting that CEP78 directly interacts with IFT20 …
eLife assessment
This paper is of interest to scientists within the cilia and centrosome fields, in particular those studying photoreceptor and sperm development and the diseases associated with their dysfunction. The authors describe the generation and characteristics of Cep78 knockout mice. Consistent with the phenotype observed in patients carrying mutations in CEP78, Cep78 knockout mice show degeneration in photoreceptor cells as well as male infertility associated with multiple morphological abnormalities of the sperm flagella (MMAF). The phenotypic characterisation of Cep78 knockout mice is thorough and convincing, and the Cep78 knockout model will be useful for further elucidating disease mechanism in humans and for potential therapy development. The authors also provide results suggesting that CEP78 directly interacts with IFT20 and TTC21A (IFT139) to form a trimeric complex, but this claim is not justified by the data provided.
-
Reviewer #1 (Public Review):
This manuscript describes the generation and characterisation of a mouse knockout model of Cep78, which codes for a centrosomal protein previously implicated in cone-rod dystrophy (CRD) and hearing loss in humans. Previous work in cultured mammalian cells (including patient fibroblasts) also indicated roles for CEP78 in primary cilium assembly and length control, but so far no animal models for CEP78 were described. Here, the authors first use CRISPR/Cas9 to knock out Cep78 in the mouse and convincingly demonstrate loss of CEP78 protein in lysates of retina and testis of Cep78-/- animals. Next, by careful phenotypic analysis, the authors demonstrate significant defects in photoreceptor structure and function in these mutant animals, which become more severe over a 9 (or 18) month period. Specifically, TEM …
Reviewer #1 (Public Review):
This manuscript describes the generation and characterisation of a mouse knockout model of Cep78, which codes for a centrosomal protein previously implicated in cone-rod dystrophy (CRD) and hearing loss in humans. Previous work in cultured mammalian cells (including patient fibroblasts) also indicated roles for CEP78 in primary cilium assembly and length control, but so far no animal models for CEP78 were described. Here, the authors first use CRISPR/Cas9 to knock out Cep78 in the mouse and convincingly demonstrate loss of CEP78 protein in lysates of retina and testis of Cep78-/- animals. Next, by careful phenotypic analysis, the authors demonstrate significant defects in photoreceptor structure and function in these mutant animals, which become more severe over a 9 (or 18) month period. Specifically, TEM analysis demonstrates ultrastructural defects of the connecting cilium and photoreceptor outer segments in the Cep78 mutants, which is in line with previously reported roles for CEP78 in CRD and in regulating primary cilia assembly in humans. In addition to a CRD-like phenotype, the authors also convincingly show that male Cep78-/- animals are infertile and exhibit severe defects in spermatogenesis, sperm flagella structure and manchette formation (MMAF phenotype). Furthermore, the authors provide evidence for an MMAF phenotype from a male individual carrying a previously reported CEP78 c.1629-2A>G mutation, substantiating that CEP78 is required for sperm development and function in mammals and supporting previously published work (Ascari et al. 2020).
Finally, to identify the underlying molecular mechanism by which CEP78 loss causes MMAF, the authors perform some biochemical analyses, which suggest that CEP78 physically interacts with IFT20 and TTC21A (an ortholog of Chlamydomonas IFT139) and might regulate their stability. The authors conclude that CEP78 directly binds IFT20 and TTC21A in a trimeric complex and that disruption of this complex underlies the MMAF phenotype observed in Cep78-/- male mice. However, this conclusion is not fully justified by the data provided, and the mechanism by which CEP78 affects spermatogenesis therefore remains to be clarified.
Specific strengths are weaknesses of the manuscript are listed below.
Strengths:
Overall, the phenotypic characterisation of the Cep78-/- animals appears convincing and provides new evidence supporting that CEP78 plays an important role in the development and function of photoreceptors and sperm cells in vertebrates.
Weaknesses:
The immunoprecipitation experiments of mouse testis extracts that were used for the mass spectrometry analysis in Table S4 were performed with an antibody against endogenous CEP78 (although antibody details are missing). One caveat with this approach is that the antibody might block binding of CEP78 to some of its interactors, e.g. if the epitope recognized by the antibody is located within one or more interactor binding sites in CEP78. This could explain why the authors did not identify some of the previously identified CEP78 interactors in their IP analysis, such as CEP76 and the EDD-DYRK2-DDB1-VprBP complex (Hossain et al. 2017) as well as CEP350 (Goncalves et al. 2021).
Figure 7A-D and page 18-25: based on IPs performed on cell or tissue lysates the authors conclude that CEP78 directly binds IFT20 and TTC21A in a "trimeric complex". However, this conclusion is not justified by the data provided, nor by the previous studies that the authors are referring to (Liu et al. 2019 and Zhang et al. 2016). The reported interactions might just as well be indirect. Indeed, IFT20 is a known component of the IFT-B2 complex (Taschner et al., 2016) whereas TTC21A (IFT139) is part of the IFT-A complex, which suggests that they may interact indirectly. In addition, the IPs shown in Figure 7A-D are lacking negative controls that do not coIP with CEP78/IFT20/TTC21A. It is important to include such controls, especially since IFT20 and CEP78 are rich in coiled coils that tend to interact non-specifically with other proteins.
In Figure 7D, the input blots show similar levels of TTC21A and IFT20 in control and Cep78-/- mouse testicular tissue. This is in contrast to panels E-G in the same figure where TTC21A and IFT20 levels look reduced in the mutant. Please explain this discrepancy.
The efficiency of the siRNA knockdown shown in 7J-M was only assessed by qPCR (Figure S4), but this does not necessarily mean the corresponding proteins were depleted. Western blot analysis needs to be performed to show depletion at the protein level. Furthermore, it would be desirable with rescue experiments to validate the specificity of the siRNAs used.
Figure 7I: the resolution of the IFM is not very high and certainly not sufficient to demonstrate that CEP78, IFT20 and TTC21A co-localize to the same region on the centrosome, which one would have expected if they directly interact.
It is not really clear what information the authors seek to obtain from the global proteomic analysis of elongating spermatids shown in Figure 3N, O and Tables S2 and S3. Also, in Table S2, why are the numbers for CEP78 in columns P, Q and R so high when Cep78 is knocked out in these spermatid lysates? Please clarify.
Figure 1F and Figure 4K: the data needs to be quantified.
Figure 2A: It is difficult to see a difference in connecting cilium length in control and Cep78-/- mutant retinas based on the images shown here.
-
Reviewer #2 (Public Review):
In this report, the authors have described the generation and characteristics of Cep78 mutant mice. Consistent with the phenotype observed in patients carrying the mutations in CEP78, Cep78 knock-out mice show degeneration in photoreceptors cells as well as defects in sperm. The author further shows the CEP78 protein can interact with IFT120 and TTC21a. Mutation in CEP78 results in a reduction of protein level of IFT120 and TTC21A and mislocalization of these two proteins, offering mechanistic insights into the sperm defects. Overall the manuscript is well written and easy to follow. Phenotyping is thorough. However, improvement of the background section is needed. In addition, some of the conclusion is not sufficiently supported by the data, warranting further analysis and/or additional experiments. The …
Reviewer #2 (Public Review):
In this report, the authors have described the generation and characteristics of Cep78 mutant mice. Consistent with the phenotype observed in patients carrying the mutations in CEP78, Cep78 knock-out mice show degeneration in photoreceptors cells as well as defects in sperm. The author further shows the CEP78 protein can interact with IFT120 and TTC21a. Mutation in CEP78 results in a reduction of protein level of IFT120 and TTC21A and mislocalization of these two proteins, offering mechanistic insights into the sperm defects. Overall the manuscript is well written and easy to follow. Phenotyping is thorough. However, improvement of the background section is needed. In addition, some of the conclusion is not sufficiently supported by the data, warranting further analysis and/or additional experiments. The Cep78 KO mice model established by the author will be a useful model for further elucidating the disease mechanism in human and developing potential therapy.
My comments are the following:
1. Introduction. The statement that "CRD usually exists with combination of immotile cilia defects in other systems" is not correct. CRD due to ciliopathy can have cilia-related syndromic defects in other systems but it is a relatively small portion of all CRDs and the most frequently mutated genes are not cilia-related genes, such as ABCA4, GUCY2D, CRX.
2. Introduction: Page 4 CNGB1 encodes channel protein and not a cilia gene. It should be removed since it does not fit.
3. Page 5, given the previous report of CEP78 patients with retina degeneration, hearing loss, and reduced infertility, the statement of "we report CE79 as a NEW causative gene for a distinct syndrome...TWO phenotypes....." Is not accurate.
4. Figure 1F, the OS of the cone seems shorter, which might be the reason for weaker arrestin staining in the mutant compared to the heterozygous. Also, it would be better to quantify the staining to substantiate the statement.
5. Figure 1K, panel with lower magnification would be useful to get a better sense of the overall structure defect of the retina. Is the defect observed in the cone as well?
6. Figure 2A, NPHP1 or other markers specifically label CC would be more useful to quantify the length of CC. Also need to provide a notation for the red arrows in Figure 2. In addition, the shape of CC in the mutant seems differ significantly from the control. It seems disorganized and swollen.
7. Evidence provided can only indicate direct interaction among CEP78/IFT20/TTC21A. -
Reviewer #3 (Public Review):
Authors were aiming to bring a deeper understanding of CEP78 function in the development of cone-rod dystrophy as well as to demonstrate previously not reported phenotype of CEP78 role in male infertility.
It is important to note, that the authors 're-examined' already earlier published human mutation, 10 bp deletion in CEP78 gene (Qing Fu et al., 10.1136/jmedgenet-2016-104166). This should be seen as an advantage since re-visiting an older study has allowed noting the phenotypes that were not reported in the first place, namely impairment of photoreceptor and flagellar structure and function. Authors have generated a new knockout mouse model with deleted Cep78 gene and allowed to convey the in-depth studies of Cep78 function and unleash interacting partners.
The authors master classical histology techniques …
Reviewer #3 (Public Review):
Authors were aiming to bring a deeper understanding of CEP78 function in the development of cone-rod dystrophy as well as to demonstrate previously not reported phenotype of CEP78 role in male infertility.
It is important to note, that the authors 're-examined' already earlier published human mutation, 10 bp deletion in CEP78 gene (Qing Fu et al., 10.1136/jmedgenet-2016-104166). This should be seen as an advantage since re-visiting an older study has allowed noting the phenotypes that were not reported in the first place, namely impairment of photoreceptor and flagellar structure and function. Authors have generated a new knockout mouse model with deleted Cep78 gene and allowed to convey the in-depth studies of Cep78 function and unleash interacting partners.
The authors master classical histology techniques for tissue analysis, immunostaining, light, confocal microscopy. They also employed high-end technologies such as spectral domain optical coherence tomography system, electron and scanning electron microscopy. They performed functional studies such as electroretinogram (ERG) to detect visual functions of Cep78-/- mice and quantitative mass spectrometry (MS) on elongating spermatids.
The authors used elegant co-immunoprecipitation techniques to demonstrate trimer complex formation.
Through the manuscript, images are clear and support the intended information and claims. Additionally, where possible, quantifications were provided. Sample number was sufficient and in most cases was n=6 (for mouse specimens).
The authors could provide more details in the materials and methods section on how some experiments were conducted. Here are a few examples. (i) Authors have performed quantitative mass spectrometry (MS) on elongating spermatids lysates however, did not present specifically how elongating spermatids were extracted. (ii) In the case of co-IPs authors should provide information on what number of cells (6 well-plate, 10 cm dish etc) were transfected and used for co-IPs. Furthermore, authors could more clearly articulate what were the novel discoveries and what confirmed earlier findings.
The authors clearly demonstrate and present sufficient evidence to show CEP78/Cep78 importance for proper photoreceptor and flagellar function. Furthermore, they succeed in identifying trimer complex proteins which help to explain the mechanism of Cep78 function.
The given study provides a rather detailed characterization of human and mouse phenotype in response to the CEP78/Cep78 deletion and possible mechanism causing it. CEP78 was already earlier associated with Cone-rod dystrophy and, this study provides a greater in-depth understanding of the mechanism underlying it. Importantly, scientists have generated a new knock-out mouse model that can be used for further studies or putative treatment testing.
CEP78/Cep78 deletion association with male infertility is not previously reported and brings additional value to this study. We know, from numerous studies, that testes express multiple genes, some are unique to testes some are co-expressed in multiple tissues. However, very few genes are well studied and have clinical significance. Studies like this, combining patient and animal model research, allow to identify and assign function to poorly characterized or yet unstudied genes. This enables data to use in basic research, patient diagnostics and treatment choices.
-