Proteomic and structural comparison between cilia from primary ciliary dyskinesia patients with a DNAH5 defect

Abstract

Introduction: Primary ciliary dyskinesia (PCD) is a genetic disorder affecting motile cilia across various organs, leading to recurrent respiratory infections, subfertility, and laterality defects. While several diagnostic tools exist-such as high-speed video microscopy, immunofluorescence staining, electron microscopy, and genetic screening-the relationship between different pathogenic variants within a single PCD gene and their effects on ciliary composition, structure, and clinical phenotype remains poorly understood.

Methods: To investigate this, we analyzed cilia from PCD patients with different mutations in axonemal dynein heavy chain dnah5 using mass spectrometry and cryo-electron tomography. These methods allowed us to examine both the protein composition and ultrastructural organization of motile cilia in affected individuals.

Results: Though all analyzed patients present similarly in traditional diagnostic methods, we observed differences in axonemal composition among patients carrying different dnah5 mutations. Specific reductions in ciliary components varied between individuals, indicating a mutation-specific impact. Notably, proteins such as VWA3B, KIAA1430/CFAP97, and DTHD1-not previously identified as components of human respiratory motile cilia-were detected in wild type cilia, but not in patient cilia. Lastly, we confirmed some changes in protein abundance in the 96-nm repeated unit of the axoneme between wild-type and PCD samples.

Discussion: These findings suggest that mutations in dnah5 result in varied and specific alterations in axonemal composition, reflecting the heterogeneity of the disease at the molecular level. The discovery of novel ciliary proteins and mutation-specific differences enhances our understanding of the complexity of PCD pathogenesis and may inform future diagnostic and therapeutic strategies.

Keywords: DNAH5; axoneme; cryo-electron tomography; dynein; mass spectrometry; primary ciliary dyskinesia.

FIGURE 1

Schematic overview of motile cilia structure. (a). Cross section view of the 9 + 2 arrangement of the axoneme. (b). Longitudinal view of a single microtubule doublet, demonstrating periodic arrangement of the outer dynein arms (yellow) at a 24-nm interval, inner dynein arms (pink, lilac), nexin-dynein regulatory complex (grey), and radial spokes (green) at a 96-nm periodicity.

FIGURE 2

Traditional diagnostic procedures for all patients demonstrate similar phenotypes in TEM and IF. (a). Schematic of the genetic variants in dnah5 of patients included in this study. (b). Representative images of the cross section of wild type and patient cilia as imaged by transmission electron microscopy. Extensive loss of ODA can be seen in all patient samples. Pink arrows indicate examples of clearly visible ODA in the healthy sample. (c). Representative immunofluorescence images of wild type and PCD patient cilia. All patients show a loss of DNAH5 (red) from the axoneme. Acetylated tubulin is stained in green and DNA (DAPI) in blue.

FIGURE 3

Loss of DNAH5 from the axoneme is associated with loss of other axonemal components. (a). Volcano plot of the significantly less (blue) and more (pink) abundant ciliary proteins in PD441 compared to WT. ODA components are highlighted by their gene names. (b). Volcano plot of the significantly less (blue) and more (pink) abundant ciliary proteins in PD480 compared to WT. ODA components are highlighted by their gene names. (c). Bar charts comparing the intensities of ODA (left) and IDA components (right) between wild type and PD441 and PD480 isolated cilia. All samples were normalized to their respective sample TUBB4B intensity and expressed as a percentage of wild type abundance.

FIGURE 4

DNAH5-defective patient cilia carrying different dnah5 mutations show distinctive protein composition patterns. (a) Bar chart comparing the intensities of newly identified ciliary components between wild type and PD441 and PD480 isolated cilia. All samples were normalized to their respective sample TUBB4B intensity and expressed as a percentage of wild type abundance. (b–d). Protein interaction networks for the newly identified ciliary proteins based on reported STRING interactions. (e) Volcano plot comparing the significantly less (blue) and more (pink) abundant ciliary proteins in PD441 than in PD480. GSTA2 and several IFT proteins are highlighted by their gene name.

FIGURE 5

Structural comparison of wild-type and patient outer dynein arms (ODAs) by cryo-electron tomography and subtomogram averaging. (a). Isosurface rendering of a wild-type microtubule doublet (MTD) average. The ODAs are highlighted on the bottom of the doublet. (b). Tomographic slices of the healthy MTD (left) and the double-headed ODA (right). The orientation of the ODA slice is indicated by the dotted green line. (c). Isosurface rendering of wild type MTD with 24-nm periodicity with ODA view. (d). Isosurface rendering comparison of wild type and patient MTD with 24-nm periodicity with ODA view. Star denotes an example of the additional globular density identified in PCD patient axonemes. (e). Fitting of the ODA docking complex chains from atomic model RRO7 (Gui et al., 2021) into the tomographic map of PD480.

FIGURE 6

Overlay of 96-nm periodic structure of the wild-type (grey, transparent) and PD441 (orange) microtubule doublet. Zoomed in view of several components that are demonstrating reduced density in PD441 compared to WT: (a). IDAe (DNAH7). (b). IDAf heavy chain alpha (DNAH10) (both shown by black arrows). (c). N-DRC. Tomographic slices of the reduced (d). IDAf, and (e). N-DRC structures, generated with 3dmod. Pink arrows in d and e denote the location of the IDAf HC and the N-DRC, respectively.