Nommick A et al. (2025), Dual role of Xenopus Odf2 in multiciliated cell...

XB-ART-61200

Dev Biol 2025 Apr 24;520:224-238. doi: 10.1016/j.ydbio.2025.01.014.

Show Gene links Show Anatomy links

Dual role of Xenopus Odf2 in multiciliated cell patterning and differentiation.

Nommick A, Chuyen A, Clément R, Thomé V, Daian F, Rosnet O, Richard F, Brouilly N, Loiseau E, Boutin C, Kodjabachian L.

???displayArticle.abstract???
In developing tissues, the number, position, and differentiation of cells must be coordinately controlled to ensure the emergence of physiological function. The epidermis of the Xenopus embryo contains thousands of uniformly distributed multiciliated cells (MCCs), which grow hundreds of coordinately polarized cilia that beat vigorously to generate superficial water flow. Using this model, we uncovered a dual role for the conserved centriolar component Odf2, in MCC apical organization at the cell level, and in MCC spatial distribution at the tissue level. Like in other species, Xenopus Odf2 localized to the basal foot of basal bodies. Consistently, Odf2 morpholino-mediated knockdown impaired basal foot morphogenesis. Consequently, the rate of microtubule nucleation by Odf2-deficient basal bodies was reduced, leading to cilia disorientation, reduced beating, and ultimately altered flow production across the embryo. Furthermore, we show that Odf2 is required to maintain MCC motility and homotypic repulsion prior to their emergence into the surface layer. Our data suggest that Odf2 promotes MCC spacing via its role in the modulation of cytoplasmic microtubule dynamics. Mathematical simulations confirmed that reduced migration speed alters the spacing order of MCCs. This study provides a striking example of coupling between organizational scales by a unique effector.

???displayArticle.pubmedLink??? 39864486
???displayArticle.link??? Dev Biol

Species referenced: Xenopus laevis
Genes referenced: odf2 pcnt tuba4a tubg1
GO keywords: regulation of microtubule nucleation [+]
???displayArticle.antibodies??? Cetn2 Ab3 GFP Ab20 Odf2 Ab1 RFP Ab5 Tjp1 Ab2 Tuba4b Ab2 Tuba4b Ab4 Tubg Ab8
???displayArticle.morpholinos??? odf2 MO1 odf2 MO2

???attribute.lit??? ???displayArticles.show???

	Graphical abstract
	Fig 1. Odf2 associates with centrioles in multiciliated cells. (A,B) Embryos were injected in ventral animal blastomeres at the 8-cell stage with mRNA encoding Odf2-GFP (A,B) and Centrin-RFP (B), fixed at stage 18 (A), or stage 31 (B), and processed for immunostaining. Maximum Intensity Projections (MIPs) of confocal acquisitions reveal the respective localization of Odf2-GFP (green), Centrin or Centrin-RFP (red, centriole), and -Tubulin (white, basal foot). The bottom row shows higher magnification views of images from the middle row. (C,D) MIPs of confocal acquisitions showing immunostaining against Odf2 (green), and -Tubulin (purple) at stage 18 and 31. Insets in A and C show higher magnification views on few centrioles. The bottom row in B and D shows the top and lateral views of a single BB analysed in 3D with Clear Volume. Scale bars: 5 m (AD), 1 m (insets), 500 nm (bottom row). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 2. Odf2 is necessary for basal body docking and orientation. (A) MIPs of confocal acquisitions of MCCs from uninjected control, odf2 ATG and odf2 Spl morphant embryos at stage 31, immunostained against Centrin (red) to reveal centrioles/BBs. White dotted lines indicate the plane of the Z-projections shown below. (B) Graph displaying the apico-basal (AB) distribution of centrioles/BBs within uninjected control, odf2 ATG, odf2 Spl morphant and rescued (odf2 ATG or Spl MO + odf2 mRNA) MCCs. Each point represents the average distance within a single cell of centrioles/BBs from the most apical centriole-positive confocal slice. The middle horizontal line represents the mean; the top and bottom horizontal lines represent the standard deviation. (C) MIPs of confocal acquisitions of control and morphant MCCs immunostained against Centrin (red) and gamma-Tubulin (white, basal foot). The insets show higher magnification views to best reveal BB orientation. Below each cell, a rose histogram represents the circular distribution of BB orientation angles from 0 to 360degrees. The black line running from the centre of the diagram to the outer edge is the mean angle and the arcs extending to either side represent the confidence limits of the mean fixed at 95% (when the mean angle is not significant, the line is pink). The length of each sector corresponds to the percentage of BBs located in this range of angles. (D) Graph plotting the calculated Circular Standard Deviation of BB orientation angles from control, morphant and rescued MCCs. Each point represents a single cell. The middle horizontal line represents the mean; the top and bottom horizontal lines represent the standard deviation. (E) Graph displaying the percentage of non-polarized MCCs based on the Rayleigh statistical test (MCCs with no significant mean angle of BB orientation) of control, morphant, and rescued MCCs. The 5 um scale bar in (A) applies to all photographs. Scale bar in insets: 1 um. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 3. Odf2 is required for correct basal foot assembly. (A) Transversal Transmission Electron Microscopy (TEM) acquisitions of MCCs from control and odf2 Spl morphant embryos at stage 31 were made to examine BB docking at the apical membrane. Red triangles indicate BBs that failed to dock. (B) TEM acquisitions of control and morphant MCCs from an apical view. Red arrows indicate BB orientation vectors, based on BF direction when it is visible on section. (C) Quantification of the percentage of undocked BBs in control and morphant MCCs from TEM acquisitions. (D) Quantification of BB orientation based on BF direction in control and odf2 morphant MCCs from TEM acquisitions. Each point represents the degrees of deviation of a single BB, from the average angle of all BBs analysed. The middle horizontal line represents the mean; the top and bottom horizontal lines represent the standard deviation. (E) Quantification of BF presence in control and odf2 morphant TEM acquisitions of docked BBs, assuming that the probability of cutting at the BF level is the same in both conditions. (F) Top and lateral views of 3D reconstructed and colorized tomograms of control and morphant BBs. The BB is colorized in green, and microtubules (MTs) in purple. In a normal BB, several MTs emanate from the BF. By contrast, the Odf2-depleted BB has a smaller BF (blue arrow) with no MTs emanating from it. (G) Quantification of the length (from the center of the BB cylinder to the BF tip), the width (at the proximal base of the BF) and the thickness of the BF (along the Z plane) of all control and morphant BBs analysed by 3D tomography. The little TEM inset on the graph shows a BB with the axes considered to measure the length and the width (yellow arrows). Note that one BB entirely lacked the BF and scored 0 for all dimensions. (H) Quantification of the numbers of MTs in contact with BFs of control and morphant BBs analysed by 3D tomography. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 4. Odf2 knockdown impairs apical microtubule network assembly. (A) MIP of confocal acquisitions of a control MCC at stage 31, immunostained against Alpha-Tubulin (green, MTs) and Centrin (red). (B) MIP of confocal acquisitions of an MCC co-injected with odf2 Spl-MO and mRNA encoding Centrin-RFP (red) to serve as injection tracer, and to reveal centrioles. In both (A) and (B) a treatment was applied to remove cilia, so as to properly visualize the cortical MT signal. (A′,B′) Higher magnification views of single apical confocal slices of the MCCs showed in (A,B). The regular MT network that links BBs in control cells is severely reduced in morphant conditions. The bottom row shows Z-projection views of A′ and B’. The 5 μm scale bar in (A) applies to all photographs shown in (A,B). The 1 μm scale bar in (A′) applies to (B′). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 5. Odf2 depletion impairs ciliary beating and fluid flow production. (A) Quantification of ciliary beating frequency (CBF/MCC in Hz) in control and morphant MCCs, based on phase-contrast microscopy and high-speed video-recording. Each dot represents the mean CBF computed over all clearly discernible cilia per MCC. Two experiments are shown, where the mean CBF of control MCCs was different, but systematically decreased upon odf2 knockdown. (B) A pulse of red visible dye was delivered at the anterior tip of control or morphant embryos. The graph displays the percentage of the embryo length reached by the dye front in 12 s. Each bar represents one recorded embryo. Asterisks point the embryos showed in (C). (C) Still frames at 4 time-points showing the antero-posterior progression of the red dye along the flanks of control and morphant tadpoles. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 6. Odf2 is necessary for correct MCC spacing. (A,B) MIPs of confocal acquisitions of control and odf2 morphant embryos at stage 20 (before MCC intercalation, A) and stage 31 (after MCC intercalation, B). MCCs were labelled by fluorescent in situ hybridization (FISH) with the alpha-tubulin riboprobe (red) at stage 20 and by immunostaining against Centrin at stage 31 (red). ZO-1 immunostaining (green) was used to reveal junctions between outer-layer cells. (A,B) Scale bar is 25 μm. (C) Graph displaying the percentage of MCCs found isolated or in contact with one or more MCCs, in control and morphant embryos. (D) Percentage of intercalated MCCs per field, in control and morphant embryos, at stage 20 and stage 31. Intercalation was scored positive when an MCC apical area delimited by ZO-1 staining was clearly discernible. Note that intercalation appeared delayed at stage 20 in morphant embryos compared to control, but was complete at stage 31. (E) Order index of control and morphant embryos at stage 31, calculated from Delaunay tessellation of the centroids of alpha-tubulin-positive MCCs. Each point represents one field of view. (F) Embryos were injected with odf2 Spl or ATG morpholinos, with or without odf2 mRNA, and MCC spacing was scored at st27. Note that MCC spacing was significantly rescued for both odf2 morphant conditions. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Fig. 7. Odf2 is necessary to maintain optimal MCC migration and homotypic repulsion. (A) Still frames of spinning disk confocal video-microscopy from Movie 5. 8-Cell embryos were injected in ventral animal blastomeres with linearized plasmid DNA alpha-tubulin::LifeAct-GFP, to drive expression specifically in MCCs. At stage 14, embryos were processed for live imaging. Colored numbers point different MCCs to appreciate their mobility over time (hh:mm). Note that the respective position of MCCs change over time in control but not in odf2 morphant embryos. White arrows point cellular protrusions, which retract in subsequent frames. Scale bar is 25 μm. (B) Instant speed of control (n = 68) and odf2 morphant (n = 62) MCCs recorded along 4h-long movies. The grey and green areas represent the standard error of all tracked MCCs. The grey and green lines represent the mean instant speed of all tracked MCCs. (C) Graph displaying the cumulative distance covered by each MCC from control and morphant embryos in 4h. (D) Graph displaying the displacement of MCCs (based on their initial and final positions) from control and morphant embryos. (E) Graph displaying the duration of MCC mutual contact in control and morphant embryos. (F) Embryos were injected with mRNA encoding mRFP to reveal the junctions between outer-layer cells. The graph displays the proportion of MCCs that migrate along outer-layer junctions (based on the overlap between GFP and RFP signals) over a period of 4h in control and morphant conditions. (G) The graph displays the order index of MCCs computed in our mathematical model over the fictive time represented by the sequential steps of the Monte-Carlo simulation. Note that when mobility is reduced the order index is stably decreased. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
	Figure S1 Odf2 isoforms. (A) Whole-mount in situ hybridization (WISH) with odf2 riboprobe at stage 20 revealed a salt and pepper pattern typical of MCCs. (B) WISH at stage 25 combined with acetylated-Tubulin IF confirms that odf2 transcripts are particularly enriched in MCCs. (C) Amino-acid (aa) sequences of Xenopus Odf2 isoform 1 (short, 649aa) and Odf2 isoform 2 (long, 827aa). Red boxes indicate the conserved N-terminal aa-sequence similar of the human Cenexin insertion. In blue are shown the C-terminal short and long sequences that differ between the two isoforms. The region expected to be recognized by the human ODF2 antibody used in this study is underlined in blue. (D) Alignment of human CENEXIN1 and Xenopus Odf2 isoforms 1 and 2. In the consensus line, the amino-acids that are identical or similar between the two sequences appear in red, and those that differ appear in blue. Dots point residues that are absent in one or the other protein. Note that isoform 1 was used to generate the Odf2-GFP construct used in this study. (E) MIPs of confocal acquisitions showing Odf2 (green), and γ-Tubulin (purple, centrioles) immunostaining, at centrosomes of non-MCCs at stage 31. (F) MIPs of confocal acquisitions of immunostaining against Centrin (red, centrioles), Odf2-GFP expression (green) and γ-Tubulin (white) at centrosomes of non-MCCs at stage 31. (E,F) Scale bar is 1 μm
	Figure S2 Characterization of odf2 morpholino antisense oligonucleotides. (A) Schematic representation of Xenopus laevis odf2 (Xenbase entrez-gene 496006). Red bars below exon1 show the position of odf2 ATG and odf2 Spl MOs. In the red dashed box is shown a zoom on exon1-intron1-exon2 region with the position of the primers used to reveal intron1 retention caused by odf2 Spl MO. (B) Validation of odf2 ATG MO by immunostaining. MIPs of confocal acquisitions showing two MCCs, one co-injected with odf2 ATG MO + Centrin-RFP (white asterisk) and the other without MO. (C) PCR on genomic DNA (left gel) extracted from Xenopus embryo and A6 cells amplify a fragment of 260bp containing intron1. PCR on cDNA (middle gel) prepared from total RNA extracted at different embryonic stages reveal a 167bp fragment resulting from intron1 splicing. PCR on Xenopus cDNA from control and odf2 morphant embryos at stage 20 (right gel). The presence of the 260bp band in morphant cDNA confirms intron1 retention. (D,E) For rescue experiments, embryos were sequentially co-injected at the 8-cell stage with odf2 Spl MO, or odf2 ATG MO and centrin-RFP mRNA, and at the 16-cell stage with odf2-GFP mRNA. Micrographs show MIPs of confocal acquisitions of MCCs at stage 31 immunostained for γ-Tubulin (BF, white) and Centrin-RFP (BB, red). Asterisks point MCCs rescued by the presence of Odf2-GFP. The white dashed lines indicate the plane of the Z views shown below. (B,D,E) Scale bar is 25 μm.
	Figure S3 Odf2 is not required for ciliogenesis or rootlet formation. (A) MIPs of confocal acquisitions of odf2 Spl morphant MCCs at stage 31, immunostained for centrioles (Centrin, red), cilia (Acetylated-Tubulin, green) and morpholino tracer (mRFP coinjected with MO, white). Uninjected control and odf2 morphant MCCs are identified with white and pink asterisks, respectively. Scale bar is 50 μm. Dashed boxes delimit the regions magnified in A’ and A’’. In A’’ the top morphant MCC displayed moderate BB clustering and made some cilia, whereas the bottom MCC (pink arrow) displayed severe BB clustering deep in the cytoplasm and made no cilia. The white dashed line indicates the plane of the Z view shown below. Scale bar is 10 μm. (B) MIPs of confocal acquisitions of control and odf2 morphant MCCs immunostained for the rootlet marker Pcnt (green) and Centrin (red). Scale bar is 5 μm. Dashed boxes delimit the region magnified in the views below. Scale bar in the zoomed views is 1 μm. The white dashed line indicates the plane of the Z view shown below. (C) Quantification of Pcnt mean signal intensity. Each dot represents an individual MCC. The middle horizontal line represents the mean, the top and bottom horizontal lines represent the standard deviation from the mean.
	Figure S4 Depletion of Odf2 impairs actin and intermediate filament networks in MCCs. (A,B) MIPs of confocal acquisitions of control and odf2 morphant MCCs at stage 31, stained for F-actin (Sir-actin, white) and Centrin (red, centrioles). Dashed boxes indicate high magnification views in A’, B’. (A’,B’) Single slices of confocal acquisitions showing the apical or subapical (0.6 μm below) actin network in control and morphant MCCs. We observe a strong actin signal reduction in morphant cells. The dashed white line sets the position of the corresponding Z views shown on the bottom row. (C,D) MIPs of confocal acquisitions of control and morphant MCCs at stage 31, immunostained for intermediate filaments (IFs, C-11, green) and Centrin or Centrin-RFP (centrioles, red) coinjected with odf2 Spl-MO. Dashed boxes indicate high magnification views in C’, D’. (C’,D’) High magnification views showing the regular IF network that surrounds BBs in control cells, which appears strongly disorganized in morphant conditions. White dashed lines set the position of the corresponding Z views shown on the bottom row. Scale bars are the same for control and morphant images. (A-D) Scale bar is 5 μm. (A’-D’) Scale bar is 1 μm.
	Figure S5 Cell-autonomous requirement for Odf2 in MCC spacing pattern establishment. (A) MIPs of confocal acquisitions of control and odf2 morphant embryos at stage 20, before MCC intercalation is complete. MCCs were labelled by fluorescent in situ hybridization (FISH) with the alpha-tubulin riboprobe (white). mRFP mRNA was co-injected with odf2 Spl MO. Junctions between outer-layer cells were revealed by ZO-1 immunostaining (green). The dashed line in each view delimits domains where injection targeted primarily the surface cellular layer (right), or the internal cellular layer (left). The dashed boxes 1 and 2 point the areas magnified below. In box 1, MCCs likely received little or no odf2 MO, as mRFP was mostly visible in superficial cells, and spaced properly. In box 2, MCCs likely received significant amounts or odf2 MO, as they were positive for mRFP, and did not space properly. Dashed horizontal white lines indicate the position of the Z-projections shown on the bottom row. Note the superficial position of mRFP in box 1 and its deep position in box 2. Scale bar is 100 μm. Scale bar in the zoomed views is 25 μm.
	Figure S6 MCC spacing depends on proper microtubule dynamics. (A) Control uninjected and odf2 Spl morphant embryos were immunostained at stage 18 to reveal cytoplasmic microtubules (α-Tubulin, green) and BBs in MCCs (Centrin or Centrin-RFP, red). Scale bar is 25 μm. (B) Graph showing the net mean α-Tubulin signal intensity in control and morphant MCCs. (C,D) Embryos were incubated in 2 μM nocodazole from stage 12 to 18 (C) and processed for in situ hybridization, or from stage 12 to 27 and processed for immunostaining (D), to reveal MCCs. Scale bars in (C) and (D) are 50 μm and 25 μm, respectively. (E) Graph displaying the percentage of MCCs found isolated or in contact with one or more MCCs, in control and nocodazole-treated embryos at stage 27.
	Figure S7 Initial and final frames from the simulation Movie 6. The grey meshwork represents outer-layer intercellular junctions, and red circles represent virtual MCCs. The values of the order index (OI) are shown.
	odf2 (outer dense fiber of sperm tails 2) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage 20.