Lee H et al. (2026), DEUP1 functions as a scaffold for basal foot in...

XB-ART-61769

Nat Commun 2026 Mar 12;171:. doi: 10.1038/s41467-026-70661-3.

Show Gene links Show Anatomy links

DEUP1 functions as a scaffold for basal foot integrity and planar polarity in multiciliated cells.

Lee H, Lee J, Shin M, Goutam RS, Kim J, Park S.

???displayArticle.abstract???
Multiciliated cells generate numerous centrioles to support motile cilia essential for fluid flow. This process was thought to rely on the deuterosome, a DEUP1-dependent structure. Here, we show that DEUP1 plays a critical role in the basal foot (BF)-an asymmetric appendage of the basal body required for coordinated ciliary beating. Using high-resolution imaging, we localize DEUP1 to the middle tier of the BF, adjacent to CNTRL. In DEUP1 knockout mice, BF architecture is disrupted, leading to loss of rotational planar polarity, reduced cerebrospinal fluid flow, and ciliary degeneration with age. These defects are recapitulated in Xenopus epidermis, demonstrating evolutionary conservation. Notably, DEUP1 loss alone does not affect basal body or cilia number during early or mature adult stages, with overt degeneration emerging only in aged animals. Together, these findings establish DEUP1 as a key structural regulator of BF integrity that is required for the long-term maintenance of coordinated ciliary motility.

???displayArticle.pubmedLink??? 41820353
???displayArticle.pmcLink??? PMC13125657
???displayArticle.link??? Nat Commun

Species referenced: Xenopus laevis
Genes referenced: deup1

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

	Fig. 1. DEUP1 localizes to BBs in multiciliated ependymal cells of the mouse brain.a Schematic representation of LW tissue in immature and mature states, illustrating abundant cell type composition. Created in BioRender. Park, S. (2026) https://BioRender.com/ei6atfy. b Immunostaining of LWs from heterozygous mice at P17 and P48 showing DEUP1 expression (green), type B1 cells (red), and cell boundaries. Schematics highlight mutually exclusive expression of DEUP1 and VCAM1. Quantification shows a significant reduction in VCAM1+ cells over time (P17: N = 6 fields, n = 178 VCAM1+ cells; P48: N = 6, n = 39). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. c As in panel B, with immunostaining for CEP164 to mark multi-ciliated cells. Quantification: P17: N = 3 fields, n = 50 CEP164+ cells; P48: N = 3, n = 91. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. d Immunostaining of LWs from WT and DEUP1 KO mice at P50, showing loss of DEUP1 in KO cells while FOP remains detectable. Regions outlined by dashed boxes are shown at higher magnification in the subsequent panels. e Schematic of primary ependymal cell culture preparation from neonatal mouse brain. Created in BioRender. Park, S. (2026) https://BioRender.com/bajfnvq. f In vitro–differentiated ependymal cells (day 14), transfected with EGFP-DEUP1, were stained for BB markers. A representative BB, indicated by a white arrowhead, is shown at higher magnification to provide a clearer view of DEUP1 localization. Arrows indicate the vector from the BB center to the EGFP-DEUP1 signal. A defined region of the 3D-SIM image in the x–y plane (indicated by a dashed box) was reconstructed into longitudinal (x–z) section views using Imaris software. These reconstructions were used to quantify the longitudinal distances between EGFP–DEUP1 and FOP or CEP164. Quantification: CEP164, n = 30; FOP, n = 26. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. g Schematic of in utero electroporation of EGFP-DEUP1 into the LW at E14.5. Created in BioRender. Park, S. (2026) https://BioRender.com/bajfnvq. h Immunostaining of LWs at P30 confirms DEUP1 localization at BBs in vivo. A representative BB (white arrowhead) is shown at higher magnification to highlight DEUP1 localization. Arrows indicate orientation from the BB center to the EGFP-DEUP1 signal. A dashed-boxed region was selected to generate longitudinal section views.
	Fig. 2. DEUP1 is a BF component aligned parallel to CNTRL.a Schematic of experimental design: AAV2-EGFP-DEUP1 was injected into the LV to visualize DEUP1 in relation to established BF components in multi-ciliated ependymal cells. Created in BioRender. Park, S. (2026) https://BioRender.com/srlnnvg. b The LW of the LV has a distinct three-dimensional curvature rather than a uniformly flat orientation relative to the objective lens. By collecting and comparing BB images from various orientations, we inferred three representative views—top, side, and front—of CNTRL and DEUP1 within the BF. The “top-view” BBs are used for measuring the radial distances of BF components. c Immunostaining and 3D-SIM imaging of LWs revealed the spatial relationship between DEUP1 and CNTRL relative to FOP-labeled BBs. Directional vectors from the center of the FOP ring to CNTRL or EGFP-DEUP1 are illustrated; a dotted cone denotes the putative BF structure. d–f As in panel B, using antibodies against ODF2 (c), NIN (d), or γ-TUB (e) to map additional BF components. Large GFP-positive aggregates (marked with an X in panel (e) were excluded from our analysis. g Quantification of radial distances from the BB center to each BF protein is shown as box-and-whisker plots. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. Sample sizes: ODF2, n = 97; CNTRL, n = 77; EGFP-DEUP1, n = 76; NIN, n = 96; γ-TUB, n = 82 from at least three independent P25 mice. h Schematic model of BF architecture incorporating positional data from panel (f). i Immunostaining combined with 3D-SIM imaging of LWs from P25 mice illustrating the spatial localization of CEP152 or CEP63 relative to CNTRL.
	Fig. 3. DEUP1 is required for BF architectural integrity in ependymal cells.a Schematic illustrating the region of interest (ROI) in the LW used to assess BF organization in DEUP1 KO mice. Ad, adhesion area. Created in BioRender. Park, S. (2026) https://BioRender.com/3r4xcqo. b 3D-SIM images of LWs from 5-month-old WT and DEUP1 KO mice, immunostained for BF components. Insets (yellow boxes) are schematized to show BF protein localization relative to the BB center, with corresponding intensity-distance profiles (y-axis: fluorescence intensity; x-axis: distance from BB center). c Quantification of BF protein positioning from panel (b). Distances were measured from the BB center to the signal peak of each BF marker. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT ODF2, N = 8 cells, n = 215 BBs; DEUP1 KO ODF2, N = 9, n = 315; WT CNTRL, N = 7, n = 191; DEUP1 KO CNTRL, N = 10, n = 232; WT NIN, N = 9, n = 213; DEUP1 KO NIN, N = 10, n = 241; WT γ-TUB, N = 8, n = 257; DEUP1 KO γ-TUB, N = 10, n = 248 from WT, 3 mice; KO, 3 mice. d, e Same analysis as in panels B and C, performed on 21-month-old mice. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT ODF2, N = 7 cells, n = 202 BBs; DEUP1 KO ODF2, N = 6, n = 94; WT CNTRL, N = 7, n = 153; DEUP1 KO CNTRL, N = 3, n = 58; WT NIN, N = 9, n = 219; DEUP1 KO NIN, N = 10, n = 227; WT γ-TUB, N = 8, n = 136; DEUP1 KO γ-TUB, N = 9, n = 152 from WT, 3 mice; KO, 3 mice. f, h Transmission electron microscopy (TEM) analysis of BF ultrastructure in 5-month-old (f) and 21-month-old (h) mice. Serial 50 nm sections were collected on a one-hole grid and reconstructed to estimate BF volume. Yellow asterisks indicate sections with relatively intact BF. g, i Quantification of BF volume from TEM datasets (left panels). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. For 5 months of age, WT, n = 19 BBs; DEUP1 KO, n = 18 BBs from 2 mice; for 21 months of age, WT, n = 17 BBs; DEUP1 KO, n = 31 BBs from 2 mice. In addition, the distance from the BB center to the most distal visible point of the BF was measured in each section (right panels). For 5 months of age, WT, n = 25 BBs; DEUP1 KO, n = 27 BBs from 2 mice. For 21 months of age, WT, n = 24 BBs; DEUP1 KO, n = 30 BBs from 2 mice.
	Fig. 4. DEUP1 is essential for maintaining rotational planar polarity of the BF and ciliary coordination.a Schematic illustrating the region of interest (ROI) along the CSF flow pathway used to assess BF polarity. Created in BioRender. Park, S. (2026) https://BioRender.com/kfvphq2. b En face TEM images of the apical surface from 5-month-old mice of WT and DEUP1 KO ependymal cells. c Quantification of BF angular orientation from panel B shown as histograms. d–f High-speed time-lapse imaging of fluorescent bead movement over the LW surface of live mice at P45 or P52. Individual beads are tracked over time (d, e); merged trajectories are shown in (f). g Quantification of bead velocity. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT, n = 56 beads from 3 mice; DEUP1 KO, n = 84 beads from 4 mice; P < 0.001. h, i Immunostaining for FOP in the ventral region of the AD pathway to quantify BB numbers in 5-month-old (h) and 21-month-old (i) mice. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. For 5 months of age, WT, n = 15 fields; DEUP1 KO, n = 16 fields from 2 mice; for 21 months of age, WT, n = 24 fields; DEUP1 KO, n = 27 fields from 2 mice. j Scanning electron microscopy (SEM) of the LW surface in 21-month-old mice showing the age-dependent ciliary phenotype. k μMRI-based volumetric analyses were conducted to quantitatively compare brain ventricular volumes between aged DEUP1 KO mice and their wild-type WT littermates. Color-coded images display brain ventricles along the anterior–posterior (A–P) axis. Red indicates anterior regions, while blue indicates posterior regions of the ventricular system. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT from 8 mice; DEUP1 KO from 6 mice of 21-month-old.
	Fig. 5. DEUP1 plays a conserved role in BF organization and planar polarity in Xenopus epidermal MCCs.a Schematic comparison of BB organization in mouse and Xenopus MCCs, highlighting the putative conserved role of Deup1 in Xenopus. Created in BioRender. Park, S. (2026) https://BioRender.com/qsil2ki. b EGFP-DEUP1 (30 pg) was injected into Xenopus embryos. At stage 30, embryos were immunostained and analyzed by 3D-SIM. The dashed box shows a magnified region; the arrows indicate the direction from the BB center to EGFP-DEUP1 puncta. c Morpholino (MO)-injected embryos were analyzed at stage 34 by 3D-SIM; arrows point from the center of the Fop-positive BB ring to Fop-negative BF regions. d Quantification of BF orientation angles. e–g BB organization in control vs. deup1 MO embryos at stage 34. e Fop staining; f BB patch size (control MO n = 29 cells from 6 embryos; deup1 MO n = 36 cells from 5 embryos); g BB number per cell (control MO n = 49 cells from 6 embryos; deup1 MO n = 84 cells from 6 embryos). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. h Radial distances from the BB center to each BF protein signal were measured: Cntrl, n = 42 BBs from 2 cells; EGFP-DEUP1, n = 42 BBs from 2 cells; γ-Tub, n = 44 BBs from 3 cells. i Spatial positioning of Cntrl and EGFP-DEUP1 in stage 34 embryos. j, m Cntrl and γ-Tub localization was assessed in control and deup1 MO embryos. k, n Quantification of distances from the BB center for Cntrl (k) and γ-Tub (n) (Control MO Cntrl, n = 76 BBs from 11 cells; deup1 MO Cntrl, n = 75 BBs from 9 cells; Control MO γ-Tub, n = 174 BBs from 17 cells; deup1 MO γ-Tub, n = 159 BBs from 14 cells). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. l, o Quantification of fluorescence intensities for Cntrl (l) and γ-Tub (o) in control vs. deup1 MO cells (Control MO Cntrl, n = 24 BB patches; deup1 MO Cntrl, n = 28; Control MO γ-Tub, n = 29; deup1 MO γ-Tub, n = 36). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values.
	Fig. 6. Dual ablation of CNTRL and DEUP1 results in additive BF architectural disruptions in ependymal cells.a Schematic illustrating the induction of CNTRL KO in the LW using the AAV2-CRISPR/CAS9 system. Created in BioRender. Park, S. (2026) https://BioRender.com/bu5i6ed. b 3D-SIM images of LWs from 3-month-old mice immunostained for BF components. A representative BB (a white arrowhead) is enlarged in an inset to illustrate the localization of NIN relative to the BB center, accompanied by corresponding fluorescence intensity–distance profiles. The region indicated by the yellow box is shown at higher magnification to highlight BB rotational polarity, with arrows connecting the BB center to the FOP-negative region corresponding to the BF extension. c Quantification of BB numbers from panel (B). The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT, N = 22 cells; CNTRL1 KO, N = 17; DEUP1 KO, N = 16; double KO, N = 19 from 3 LWs. d Quantification of NIN protein positioning from panel B. Distances were measured from the BB center to the signal peak of each BF marker. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT, N = 11 cells, n = 83 BBs; CNTRL1 KO, N = 11, n = 69; DEUP1 KO, N = 12, n = 79; double KO, N = 11, n = 71 from 3 LWs. e Quantification of BB rotational orientation within individual cells based on the data shown in panel B. Rotational polarity was calculated by analyzing at least 20 BBs per cell with clearly identifiable BF, and expressed as the mean vector length. The boxes represent the interquartile range (IQR) with the line inside indicating the median. Whiskers (error bars) extend to the minimum and maximum values. WT, N = 14 cells; CNTRL1 KO, N = 11; DEUP1 KO, N = 16; double KO, N = 21 from 3 LWs.
	Fig. 7. Model of BF development and dynamic remodeling in mouse ependymal cells during maturation and aging.Created in BioRender. Park, S. (2026) https://BioRender.com/whcm9c4. In neonatal brains, newly formed BBs dock at the apical surface of immature ependymal cells, initiating ciliogenesis. At this early stage, precursors of the BF emerge laterally from BBs, although BF structures remain small and randomly oriented. As ependymal cells mature, CSF flow strengthens. In response, BF structures enlarge, and their orientation becomes increasingly aligned with the direction of CSF flow, establishing robust rotational planar polarity that reinforces coordinated ciliary beating. This alignment is essential not only for generating a strong CSF flow but also for minimizing shear stress exerted on the ependymal surface by the continuous movement of CSF. During aging, both CSF flow and BF organization progressively decline. In DEUP1-deficient ependyma, BFs remain diminutive and disorganized, failing to adapt to mechanical cues from CSF flow. This mismatch results in chronic shear stress, ultimately leading to BB and cilia disintegration over time.