Ciliary Imaging pipeline

(A) Primary cilia in the growth plate a of a 6-week old Ift88CTRL mouse. Scalebar: 40 µm. Yellow boxes indicate regions that are enlarged in the resting (B), proliferating (C) and hypertrophic (D) zones. Scalebars: 10 µm. (E) Single cilium viewed from the front. (F) The same cilium viewed from the side. Scalebars: 2 µm. (G) 6-week old Ift88CTRL murine growth plate stained with DAPI and phalloidin enabling confirmation of cytoplasm and hypertrophy. Scalebar: 40 µm. (H-I) Images show appearance in the software. Cell (I), primary cilia (J) and centriole (K) detection. Scalebars: 50 µm. (L) Side view of the cell detection. Scalebar: 20 µm. (M) Primary cilia and centriole detection together. Scalebar: 20 µm. (N) Zoomed in side view of primary cilia detection. Cyan dashed line: cilium pointing sideways; yellow dashed line: cilium pointing backwards. Scalebar: 2 µm.

Cell size and orientation, ciliary prevalence and length, across GP zones and comparing middle and lateral regions.

Cell organisation differs by zone and region in the 6-week old Ift88CTRL growth plate. (A) Safranin-O stain of the growth plate of a 6-week old Ift88CTRL mouse. Boxes indicate regions that were imaged with fluorescence microscopy. Scalebar: 500 µm. (B) Zoom of histology showing zones. Scalebar: 50 µm. (C) Cell volume in the resting, proliferating and hypertrophic zones of the growth plate. Dashed lines represent median (red dashes) and first and third quartiles (black dashes). p**** < 0.0001 (Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 18 mice (2074 to 6359 cells per zone) (D) Major/longitudinal axis (red line) of cells in the proliferating zone of the growth plate. Scalebar: 5 µm. (E) Distribution of θcell, orientation angle with respect to the medial-lateral plane, in the resting, proliferating and hypertrophic zones of the middle and lateral regions of the growth plate. Dashed lines represent median (red dashes) and first and third quartiles (black dashes). p** = 0.0024, p*** = 0.00016, p**** < 0.0001 (Kolmogorov–Smirnov test adjusted for multiple comparisons) n = 18 mice (947 to 2737 cells per zone/region). This panel was created using BioRender.com. (F) Percentage ciliation by growth plate region. Mean ± SEM. p*** = 0.0007 (paired t test). (G) Percentage ciliation in the different zones and regions of the growth plate. Mean ± SEM. p**** < 0.0001. All other comparisons are not statistically significant (two-way ANOVA). n = 18 mice (947 to 3622 cells per zone/region). (H) Primary cilia length in the different zones across the width of the growth plate. Mean ± SEM. p** = 0.0031, p*** = 0.0004 (one-way ANOVA) n = 18 mice (1028 to 3391 cilia per zone). Each data point is the average cilium length in the growth plate zone of one mouse. (I) Primary cilia length in the resting, proliferating and hypertrophic zones of the middle and lateral regions of the growth plate of a 6-week old Ift88CTRL mouse. Mean ± SEM. p* = 0.0198. All other comparisons are not statistically significant (two-way ANOVA). n = 18 mice (466 to 1922 cilia per zone/region). Each data point is the average cilium length in the growth plate zone and region of one mouse. (J) Ki-67 staining in the growth plate of a 6 week-old Ift88CTRL mouse. Scalebar: 50 µm.(K) Percentage of Ki-67 positive cells in the Ift88CTRL Mean ± SEM (unpaired t test) n =3-19 mice.

Cell size, orientation and ciliation in IFT88cKO

(A/B) Cell volume in the resting, proliferating and hypertrophic zones of the Ift88CTRL and Ift88 cKO growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). Black asterisks – p* = 0.0161, p*** = 0.00013, p**** < 0.0001 (Kolmogorov–Smirnov test adjusted for multiple comparisons).n = 14-18 mice (585 to 3622 cells per zone/region). (B) Median cell volumes in the Ift88CTRL and Ift88 cKO growth plates. (C) Distribution of θcell in the resting, proliferating and hypertrophic zones of the middle and lateral regions of the Ift88CTRL and Ift88 cKO growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p* =0.042, p** = 0.0018, p**** < 0.0001. (Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 14-18 mice (585 to 3622 cells per zone/region). This panel was created using BioRender.com. (D) Overall percentage ciliation (unpaired t test). (E) Percentage ciliation by growth plate zone. p* = 0.0291. All other comparisons are not statistically significant (two-way ANOVA). (F) Percentage ciliation by growth plate region. p* =0.0138 (two-way ANOVA). Mean ± SEM. n = 14-18 mice (301 to 1922 cilia per zone/region).

Centriole position and cilia axoneme orientation

(A) Spherical coordinates of centrioles. (B) Distribution of centriole Φcentriole angles in the resting, proliferating and hypertrophic zones of the middle and growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p***= 0.00024, p**** < 0.0001 (Kolmogorov–Smirnov test adjusted for multiple comparisons).n = 18 mice (1885 to 8204 centrioles per zone/region). (C) Distribution and (D) rose plot of Φcentriole angles across the whole growth plate. n = 18 mice (25,740 centrioles). (E) Distribution of centriole θcentriole angles in the resting, proliferating and hypertrophic zones of the middle and growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p** = 0.0016 (resting middle vs proliferating middle), p** = 0.0023 (hypertrophic middle vs hypertrophic lateral), p**** < 0.0001.(Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 18 mice (1885 to 8204 centrioles per zone/region). (F) Distribution and (G) rose plot of θcentriole angles across the whole growth plate. n = 18 mice (25,740 centrioles). (H) Circular scatter plot of Φcentriole and θcentriole angles across the whole growth plate. n = 18 mice (25,740 centrioles). Primary cilia orientation is consistent throughout the 6-week old Ift88fl/fl mouse growth plate. (I) Spherical coordinates of primary cilia. (J) Distribution of primary cilia Φcilium angles in the resting, proliferating and hypertrophic zones of the middle and lateral regions of the growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p* = 0.022 (resting middle vs hypertrophic middle), p* = 0.016 (proliferating lateral vs hypertrophic lateral), p*** = 0.00082 (proliferating middle vs hypertrophic middle), p*** = 0.00082 (proliferating middle vs proliferating lateral) (Kolmogorov–Smirnov test adjusted for multiple comparisons). N = 18 mice (466 to 1922 cilia per zone/region) (K) Distribution and (L) rose plots and circular scatter plot of Φcilium angles across the whole growth plate. n = 18 mice (5802 cilia). (M) Distribution of primary cilia θcilium angles in the resting, proliferating and hypertrophic zones of the middle and lateral regions of the growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p* = 0.047 (resting middle vs proliferating middle), p* = 0.047 (resting middle vs hypertrophic middle) (Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 18 mice (466 to 1922 cilia per zone/region). (N) Distribution and (O) rose plots and circular scatter plot of θcilium angles across the whole growth plate. n = 18 mice (5802 cilia). (P) Circular scatter plot of Φcilium and θcilium angles across the whole growth plate. n = 18 mice (5802 cilia).

Centriole and primary cilia axoneme anatomical organization.

Centriole (A) and primary cilia axoneme (B) organisation in the resting zone of a 6-week old Ift88CTRL growth plate. For centrioles, the yellow cross denotes the centre of the cell with centriole position relative to this centre. For primary cilia, the yellow cross denotes where the cell would be with cilia pointing out. The geological software Stereonet was used for visualization. This figure was created using BioRender.com.

Transcriptional analysis of naïve and immobilised limb growth plate.

(A) Histology (Safranin O/Fast green staining) before and after tibial growth plate removal. Scalebar: 500 µm. (B) PCA plots for wild type mice, naïve, immobilised and with wheel exercise. (C) Volcano Plot depicts differential gene expression analysis comparing naïve with immobilised. (D) Ridge plot depicts distribution of log2 fold changes for the genes found in corresponding top-most enriched GO terms. (E) Correlation analysis between module genes and sample traits. (F) Over-representation analysis of genes enriched in the black module selected ciliary associated GO terms shown (top terms shown in Table 1.).

The top (lowest p-value) genesets found (uniquely) in each module

Cell volume and orientation upon immobilisation

(A) Cell volume in the resting, proliferating and hypertrophic zones of the control, offloaded and contralateral growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p*** = 0.00024, p**** < 0.0001 (Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 18 mice (2074 to 6349 cells per zone) (naive), n = 4 mice (332 to 1715 cells per zone) (offloaded, contralateral). (B) Median cell volumes in the control, offloaded and contralateral growth plates. (C) Distribution of θcell in the resting, proliferating and hypertrophic zones of the control, offloaded and contralateral growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). p* = 0.019, p**** < 0.0001. (Kolmogorov–Smirnov test adjusted for multiple comparisons). n = 18 mice (2074 to 6349 cells per zone) (naive), n = 4 mice (332 to 1715 cells per zone) (offloaded, contralateral).

Effect of immobilization on primary cilia.

Differences in growth plate length measured from thresholded microCT images of the naïve(A), Offloaded (B) and contralateral (C) Ift88CTRL growth plates. (D) Thresholded growth plate lengths ofnaïve, offloaded and contralateral limbs. Mean ± SEM. p* = 0.0168, p** = 0.0023 (male) (one-way ANOVA) p = 0.005988 (female) (unpaired t test corrected for multiplicity).n = 4-18 mice. Primary cilia imaging in the middle of the growth plates offloaded (E) and contralateral (F) limbs. Scalebar: 30 μm. Percentage ciliation in the CTRL, offloaded and contralateral growth plates. (G) Overall percentage ciliation (one-way ANOVA). (H) Percentage ciliation by growth plate zone. All comparisons were not statistically significant (two-way ANOVA). Mean ± SEM. n = 14-18 mice (1660 to 6349 cells per zone) (Ift88CTRL, n = 4 mice (332 to 1715 cells per zone) (offloaded, contralateral). (I) Cilia length Mean ± SEM. p* = 0.0240 (proliferating vs hypertrophic Ift88CTRL contralateral), p** = 0.0031 (resting vs hypertrophic Ift88CTRL), p*** = 0.0004 (proliferating vs hypertrophic Ift88CTRL). All other comparisons were not statistically significant (two-way ANOVA). n = 14-18 mice (778 to 3299 cilia per zone), n = 4 mice (128 to 901 cilia per zone) (offloaded, contralateral). (J) Differences in primary cilia orientation between the 6-week old CTRL, offloaded and contralateral growth plates. Distribution of primary cilia Φcilium angles in the resting, proliferating and hypertrophic zones across the growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). n = 14-18 mice (778 to 3299 cilia per zone) (Ift88CTRL), n = 4 mice (128 to 901 cilia per zone) (offloaded, contralateral). (K) Distribution of primary cilia θcilium angles in the resting, proliferating and hypertrophic zones across the growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes) n = 14-18 mice (778 to 3299 cilia per zone) (ctrl), n = 4 mice (128 to 901 cilia per zone) (offloaded, contralateral). (L) Distribution of centriole Φcentriole angles in the resting, proliferating and hypertrophic zones across the growth plates. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes). n = 14-18 mice (2518 to 14,587 centrioles per zone) (Ift88CTRL), n = 4 mice (950 to 4611 centrioles per zone) (offloaded, contralateral). (M) Distribution of centriole θcentriole angles in the resting, proliferating and hypertrophic zones across the growth plate. Dashed lines represent median (longer dashes) and first and third quartiles (shorter dashes) n = 14-18 mice (2518 to 14,587 centrioles per zone) (Ift88CTRL), n = 4 mice (950 to 4611 centrioles per zone) (offloaded, contralateral).

Schematic for conceptualization of ciliary axoneme orientation to avoid single X/Y/Z signal ‘blindness’.

(A). Physical compressive forces are in the proximal-distal (cranial-caudal) Z direction, cells within GP may experience signals or gradients in medial-lateral or ventral-dorsal (Y or X) direction and Z direction. (B). Cilia are never orientated vertically or medial-lateral horizontal, an orientation that is not parallel to any single axis, enables integration of multiple signals and avoids blindness in any axis. This figure was created using BioRender.com.

Spherical coordinates.

(A) Spherical coordinate system, (B) with anatomical directions and angle values. (C) Example point A has spherical coordinates Φ = 10° & θ = 45°. Spherical coordinate system position to measure primary cilium axoneme orientation (D) and centriole position (E) spherical coordinates, and cell angle θcell (F). This figure was created using BioRender.com.

(A) Safranin Orange/Fast green, Alcian blue (B) and Toluidine blue (C) stained sections of tibial epiphysis. uCT image (D) growth plate length measurements (E) Growth plate lengths at different ages. Mean ± SEM. n = 2-5 mice.

Flowchart describing the steps of the analysis pipeline run in R to calculate cell, primary cilia and centriole characteristics.

Safranin-O staining

(A-D) and primary cilia imaging (E-H) of murine growth plates at different ages. E, G and H were acquired with the SR8Y Airyscan mode and F with the Co8Y mode (lower resolution, faster). Yellow boxes are representative of the regions of the growth plate imaged for fluorescence.

Full description of WGCNA Modules.

(A) depicts the dendrogram as modules are split out. (B) Heatmap of module eigengene expression. (C) Top GO BP terms for each module, fuller descriptor in Supplementary Table 1.