INPP5E Mutations Cause Primary Cilium Signaling Defects, Ciliary Instability and Ciliopathies in Human and Mouse
Total Page:16
File Type:pdf, Size:1020Kb
LETTERS INPP5E mutations cause primary cilium signaling defects, ciliary instability and ciliopathies in human and mouse Monique Jacoby1,11, James J Cox2,11, Ste´phanie Gayral1,11, Daniel J Hampshire3, Mohammed Ayub4, Marianne Blockmans1, Eileen Pernot1, Marina V Kisseleva5, Philippe Compe`re6, Serge N Schiffmann7, Fanni Gergely8, John H Riley9,DavidPe´rez-Morga10, C Geoffrey Woods2,11 & Ste´phane Schurmans1,11 The primary cilium is an antenna-like structure that protrudes embryonic and postnatal death. Analyses confirmed the absence of from the cell surface of quiescent/differentiated cells and Inpp5e protein in mutant cells and tissues (Fig. 1a). Inpp5eD/D mice participates in extracellular signal processing1–3.Here,we presented with bilateral anophthalmos (100%, n ¼ 43) and postaxial report that mice deficient for the lipid 5-phosphatase Inpp5e hexadactyly (62.5%, n ¼ 16; Fig. 1b,c). Histological analyses revealed develop a multiorgan disorder associated with structural that eye development ceased at the optic vesicle stage, just before the defects of the primary cilium. In ciliated mouse embryonic appearance of the optic cup (Fig. 1d). Analysis of kidneys from the fibroblasts, Inpp5e is concentrated in the axoneme of the mice revealed the presence of multiple cysts (100%, n ¼ 10; Fig. 1e). primary cilium. Inpp5e inactivation did not impair ciliary Of the cysts, 84% expressed AQP2 and 14% expressed AQP1, assembly but altered the stability of pre-established cilia after indicating an origin in cortical collecting and connecting ducts serum addition. Blocking phosphoinositide 3-kinase (PI3K) (when AQP2+) as well as proximal tubules and the descending limb activity or ciliary platelet-derived growth factor receptor a of the loop of Henle (when AQP1+)(Supplementary Fig. 2). Only 2% (PDGFRa) restored ciliary stability. In human INPP5E,we of the renal glomeruli were cystic. Inpp5eD/D embryos had skeletal identified a mutation affecting INPP5E ciliary localization abnormalities such as a bifid sternum (50%, n ¼ 6), delayed ossifica- and cilium stability in a family with MORM syndrome, tion of metacarpals and phalanges (100%, n ¼ 5) and cleft palate a condition related to Bardet-Biedl syndrome. Together, (75%, n ¼ 4; Fig. 1f–h). We identified cerebral developmental defects, D/D © All rights reserved. 2009 Inc. Nature America, our results show that INPP5E plays an essential role in the such as anencephaly and exencephaly, in 30% of Inpp5e embryos at primary cilium by controlling ciliary growth factor and PI3K E15.5 (n ¼ 30; Fig. 1i,j). We did not detect liver alterations, laterality signaling and stability, and highlight the consequences of defects or respiratory cilium defects in mutant animals. The tissue INPP5E dysfunction. localization of lesions observed in Inpp5eD/D embryos matched the tissue expression of Inpp5e mRNA during mouse embryogenesis Lipid 5-phosphatases selectively remove the phosphate from position (Supplementary Fig. 3). D-5 of the inositol ring of phosphoinositides and inositolpho- Because most of the defects in Inpp5eD/D mice are features of sphates4,5. To characterize the functions of the 5-phosphatase ciliopathies1–3,9–11, we investigated primary cilia in renal tubular Inpp5e6–8,wegeneratedInpp5eD/+ mice (Supplementary Fig. 1a). cells from these mice (Fig. 2). Primary cilia protruding from renal We obtained no adult Inpp5eD/D mutant mice from intercrosses cells lining Inpp5e+/+ tubules and Inpp5eD/D noncystic tubules were between Inpp5eD/+ mice. However, at embryonic day 13.5 (E13.5) similar in number and morphology; however, in the renal cysts of and E18.5, 16.9% (11/65) and 14.8% (12/81) of embryos were Inpp5eD/D embryos, the number, size and morphology of primary cilia homozygous for the deletion allele, respectively. The mutant mice were markedly altered (Fig. 2a–c). Scanning electron microscopy died soon after birth, indicating that total inactivation of Inpp5e led to (SEM) revealed that cilia were sparse and appeared dilated. Dilation 1Institut de Recherches Interdisciplinaires en Biologie Humaine et Mole´culaire (IRIBHM), Institut de Biologie et de Me´decine Mole´culaires (IBMM), Universite´ Libre de Bruxelles, Gosselies, Belgium. 2Department of Medical Genetics, Cambridge Institute of Medical Research, Wellcome/MRC Building, Addenbrooke’s Hospital, Cambridge, UK. 3Academic Unit of Haematology, Henry Wellcome Laboratories for Medical Research, School of Medicine and Biomedical Sciences, University of Sheffield, Sheffield, UK. 4Psychiatry of Learning Disability, St. Lukes Hospital, Middlesbrough, UK. 5Division of Hematology, Washington University School of Medicine, St. Louis, USA. 6Cellule d’Appui Technologique en Microscopie, Institut de Chimie, Universite´ de Lie`ge, Lie`ge, Belgium. 7Laboratoire de Neurophysiologie, Universite´ Libre de Bruxelles, Bruxelles, Belgium. 8Cancer Research UK, Cambridge Research Institute, Li Ka Shing Centre, University of Cambridge, Cambridge, UK. 9GSK Respiratory Medicines Development Centre, Stockley Park, Uxbridge, UK. 10Laboratoire de Parasitologie Mole´culaire, Institut de Biologie et de Me´decine Mole´culaires (IBMM), Universite´ Libre de Bruxelles, Gosselies, Belgium. 11These authors contributed equally to this work. Correspondence should be addressed to S.S. ([email protected]) or C.G.W. ([email protected]). Received 25 November 2008; accepted 15 June 2009; published online 9 August 2009; doi:10.1038/ng.427 NATURE GENETICS ADVANCE ONLINE PUBLICATION 1 LETTERS ∆ ∆ co-localized with acetylated a-tubulin and a +/+ / b e +/+ ∆/∆ PDGFRa, two markers of the ciliary axoneme 14 Inpp5e in ciliated MEFs (Fig. 2d and Supplemen- +/+ tary Fig. 6c,d). Under the same conditions, a primary cilium with a normal morphology but lacking the 5-phosphatase protein was also present in Inpp5eD/D MEFs (Fig. 2d). Quantitative analyses of serum-starved con- ∆/∆ trol and mutant MEFs revealed that a similar percentage (B70%) of cells assembled a +/+ ∆/∆ cilium, suggesting that Inpp5e is not needed c f g +/+ ∆/∆ for the production of a primary cilium in MEFs (Fig. 3a). Actin Next, we investigated whether stimulating growth factor signaling in MEFs with pre- d +/+ ∆/∆ h +/+ ∆/∆ established cilia might reveal a role for ciliary E12.5 E12.5 Inpp5e in the maintenance of this organelle (Fig. 3a). Treatment of serum-starved ciliated MEFs with 10% serum for 4 h had no effect on the percentage of pre-established cilia in Inpp5e+/+ MEFs but caused a significant ij D/D E18.5 E18.5 percentage of Inpp5e MEFs to lose their cilia. The serum effect was prevented by the PI3K inhibitor LY294002 (Fig. 3a). SEM of Inpp5eD/D MEFs 4 h after serum addition * revealed that the remaining cilia were short with a dilated appearance, as in kidney cysts (Supplementary Fig. 7). These results indi- D/D Figure 1 Phenotypic characterization of Inpp5e mice. (a) Protein blot analysis on E13.5 brain cate that the MEF model mimics the struc- protein extracts with a rabbit polyclonal antibody directed against the N-terminal end of the mouse D/D tural primary cilium defects observed in Inpp5e protein. No Inpp5e is detected in brain from Inpp5e mice. Actin served as a loading control. D/D Similar conclusions were drawn from protein blot analysis on protein extracts from ciliated Inpp5eD/D Inpp5e kidney, and that loss of ciliary MEFs (data not shown) and from immunofluorescence analysis of ciliated Inpp5eD/D MEFs (Fig. 2d). Inpp5e sensitizes quiescent MEFs to resorb (b) General appearance and anophthalmia in E18.5 Inpp5eD/D embryos. (c) Postaxial hexadactyly in pre-established cilia in response to one or E18.5 Inpp5eD/D embryo. (d) Eye development in E12.5 and E18.5 Inpp5e+/+ (left) and Inpp5eD/D more factors in serum that signal via the (right) embryos. Mutant embryos show developmental arrest at the optic vesicle stage before PI3K pathway. Factors known to activate D/D appearance of the optic cup. No eye is detected in E18.5 Inpp5e embryos (asterisk indicates the the PI3K pathway, such as IGF-1 (10 nM), normal location of the eye). (e) Multiple cysts on kidney section of E18.5 Inpp5eD/D embryos. (f,g) Bifid sternum and delayed ossification of sternum, metacarpals and phalanges in E18.5 Inpp5eD/D embryos. insulin (200 nM), EGF (50 ng/ml), FGF © All rights reserved. 2009 Inc. Nature America, Cartilage is stained in blue (alcian blue) and bone in red (alizarin red). The ossified surface is (100 ng/ml), PDGF-AA (50 ng/ml) and significantly decreased on Inpp5eD/D sternum (f) and metacarpals and phalanges (g) as compared with PDGF-BB (50 ng/ml), individually had no Inpp5e+/+ embryos (64–86% of control embryos, P o 0.005). (h) Frontal section of head showing cleft significant effect on the stability of the cilium palate in E18.5 Inpp5eD/D embryos. (i,j) Example of anencephaly and exencephaly in E13.5-E15.5 (P 4 0.15). However, when testing different D/D Inpp5e embryos. Scale bars, 10 mm. growth factor combinations, we found that PDGF-AA, which signals through its alpha receptor (PDGFRa) localized on the cilium frequently occurred at the cilia tip but was also seen at the base of the membrane in ciliated MEFs (Fig. 2d and ref. 14), was necessary to cilium or along the ciliary shaft (Fig. 2b,right,andSupplementary induce ciliary instability (Fig. 3a). Stimulation of Inpp5eD/D MEFs Fig. 4a). Transmission electron microscopy (TEM) confirmed the by PDGF-AA and either IGF-1, EGF or FGF rapidly destabilized presence of bulges at the distal ends of the cilia and along the ciliary the cilium, whereas stimulation by IGF-1 + EGF + FGF did not shafts (Fig. 2c). As previously reported in mice and in humans with (Fig. 3a). Blocking ciliary PDGFRa signaling with a neutralizing kidney cysts due to primary cilium defects12,13, we identified apoptotic antibody significantly prevented ciliary instability in serum- or cells in the cysts (Supplementary Fig. 5). These ciliary defects were PDGF-AA + EGF-stimulated Inpp5eD/D MEFs (Fig. 3a). This indi- not specific for cystic tubules, as abnormal cilia and apoptotic cells cates that ciliary instability in Inpp5eD/D MEFs requires the com- were also observed in Bowman’s capsule from Inpp5eD/D cystic bined activation of ciliary PDGFRa and other growth factor glomeruli (Supplementary Fig.