Nephrocystin Specifically Localizes to the Transition Zone of Renal and Respiratory Cilia and Photoreceptor Connecting Cilia

Cell Culture

MDCK and HEK293T cells were cultured in DMEM/10% FCS, murine inner medullary collecting duct (mIMCD3) and LLC-PK1 (porcine kidney) were cultured as recommended by American Type Culture Collection (Manassas, VA). All cell culture reagents were from Life Technologies/Invitrogen (Karlsruhe, Germany). Cells were grown on coverslips for 8 d past confluence to allow for epithelial cell polarization and cilia formation, washed with PBS, and subjected to immunofluorescence staining as described below.

Immunoblotting

Protein extracts from Epstein-Barr virus–2-transformed B-lymphocytes, HEK293T, MDCK, and respiratory epithelial cells were prepared by standard procedures using NP-40 or RIPA lysis buffers. Axonemal high-salt protein extracts were obtained from a fresh pig trachea as described previously (25–27). Samples were separated on NuPAGE 4 to 12% bis-tris gels (Invitrogen) and blotted onto polyvinylidene difluoride membranes (Amersham). Blots were processed for ECL plus (Amersham/GE Healthcare, Freiburg, Germany) detection using rabbit anti-nephrocystin (1:2500) and anti-rabbit–horseradish peroxidase (1:2500) antibodies (Santa Cruz, Heidelberg, Germany).

Immunofluorescence Analysis

Respiratory epithelial cells were obtained by transnasal brush biopsy (Cytobrush Plus, Medscand, Malmö, Sweden) and suspended in RPMI 1640 medium without supplements. Cells were spread onto glass slides, air-dried, and stored at −80°C until use. A pig eye was obtained from a local butchery, and the retina was removed carefully using a scalpel. Cryosections (10 μm) were prepared according to standard methods. Samples were treated with 4% paraformaldehyde, 0.2% Triton-X 100, and 5% skim milk (all in PBS) before incubation with primary (at least 2 h) and secondary (30 min) antibodies at room temperature. Slides were washed with PBS after each step. Appropriate controls were performed omitting the primary antibodies. Antibodies were mouse anti–acetylated-α-tubulin and mouse anti–γ-tubulin (Sigma, Taufkirchen, Germany), mouse anti–β-tubulin (Abcam, Cambridge, UK), mouse anti–PACS-1, mouse anti-tensin, and mouse anti-p130^cas (Transduction Laboratories, BD Biosciences, Heidelberg, Germany). Polyclonal rabbit antibodies against DNAH5, nephrocystin, IFT88, and IFT20 as well as mouse anti-nephrocystin antibodies have been described previously (13,23,27–29). Secondary antibodies (Alexa Fluor 488, Alexa Fluor 546) were from Molecular Probes (Invitrogen). DNA was stained with Hoechst 33342 (Sigma). Confocal images were taken on a Zeiss laser scanning microscope (Axiovert 200 LSM510 META) using a 63 × 1.2 numerical aperture water immersion or a 100 × 1.3 numerical aperture oil immersion objective. A four-channel, eight-bit multitracking scan mode was used with a 1024 × 1024 frame size and four-fold average line scan settings. Images were processed with the Zeiss LSM510 software.

High-Speed Video Analysis for Ciliary Beat Assessment

Ciliary beat frequency was assessed with the SAVA system (30). Transnasal brush biopsies were viewed immediately with an Olympus IMT-2 microscope (×40 phase contrast objective) equipped with a Redlake ES-310Turbo monochrome high-speed video camera (Redlake, San Diego, CA) set at 125 frames per second. The ciliary beating pattern was evaluated on slow-motion playbacks.

Electron Microscopy

A mouse eye was cut into slices using a scalpel. Ultrathin retina sections were prepared according to standard methods and subjected to transmission electron microscopy using a Zeiss EM 900.

Ciliogenesis

Respiratory epithelial cells from nasal conchae or polyps were obtained from patients who underwent ear, nose, and throat surgery or after nasal brushing biopsy. Primary cell culture was performed essentially as described previously (31). Briefly, cells were isolated from tissue samples with pronase (Sigma) and grown to confluent monolayers on collagen-coated tissue flasks in F12/DMEM/2% Ultroser G (Pall Life Sciences, Cergy-Saint-Christophe, France). Cell layers were treated with collagenase, cut into pieces, and cultured in Ham′s F12/DMEM/10% NU-serum (BD Biosciences) on a rotary shaker. After 10 d, most of the cells were organized in spheroids covered with motile cilia.

Ca²⁺-Dependent Deciliation

Respiratory epithelial cells from brush biopsies were collected by centrifugation (300 × g, 5 min) and resuspended in deciliation buffer that contained 15.7 mM Tris-Cl (pH 7.5), 79.1 mM NaCl, 1.56 mM EDTA, 0.1% Triton-X 100, 15.8 mM CaCl₂, and protease inhibitors (32). Deciliation (30 to 60 min at room temperature with occasional shaking) was monitored under a microscope. Aliquots were removed, cells were pelleted, and cilia were collected from the supernatant by centrifugation (16,000 × g, 5 min). Samples were spread onto glass slides, air-dried, and used for immunofluorescence staining as described above.

Nephrocystin Associates with the Axonemal Structure at the Transition Zone

We recently demonstrated that nephrocystin localizes to renal monocilia in polarized MDCK cells (2). Here, more detailed analyses by high-resolution immunofluorescence imaging identifies that nephrocystin predominantly localizes to the ciliary base of renal monocilia in MDCK cells (Figure 1A). Similar results were obtained using mIMCD-3 and LLC-PK1 cells (data not shown). Co-staining with γ-tubulin, a marker of the microtubule organizing centers (MTOC), which are located adjacent to the basal bodies, demonstrates that nephrocystin localizes in renal monocilia of mIMCD-3 cells distal to the MTOC within the transition zone (Figure 1B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Nephrocystin localizes to the transition zone of renal monocilia and human respiratory epithelial cilia. High-resolution confocal immunofluorescence imaging of polarized confluent MDCK cells, murine inner medullary collecting duct (mIMCD3) cells, and human ciliated respiratory epithelial cells that were obtained by transnasal brush biopsy. Cells were stained using the indicated antibodies; nuclei were stained with Hoechst 33342. Merged and overlay images are shown on the right. Bars = 10 μm. (A) Co-staining of MDCK cells with antibodies against the cilia marker acetylated α-tubulin (green) and rabbit anti-nephrocystin antibodies (red). Specific nephrocystin staining is observed at the base of each monocilium. (B) Co-staining of γ-tubulin, a component of the microtubule organizing centers (MTOC), and nephrocystin in mIMCD3 cells confirms that nephrocystin localizes to the transition zone but not to the basal bodies. (C) Co-staining of human respiratory epithelial cells with antibodies against acetylated α-tubulin (green) and rabbit anti-nephrocystin antibodies (red). Specific nephrocystin staining is observed only at the ciliary bases. (D) Co-staining with antibodies against the axonemal outer dynein arm heavy chain DNAH5 (red) and mouse anti-nephrocystin antibodies (green). DNAH5 localizes to the entire length of the ciliary axonemes and to the MTOC at the ciliary basal bodies ( 27). Nephrocystin localizes distally to the MTOC at the proximal end of the cilia. The cilia section between the basal bodies and the axoneme is the transition zone. (E and F) Partial co-localization of nephrocystin (red) and its interaction partners p130cas and tensin (green) at the transition zone. (G and H) The intraflagellar transport proteins (IFT) IFT20 and IFT88 (red) localize to the MTOC and, with a speckled pattern, to the ciliary axonemes. Nephrocystin (green) shows a partially overlapping localization with the IFT proteins at the ciliary base but does not co-localize with the IFT proteins at the MTOC and the ciliary axonemes. (I) At the ciliary base, nephrocystin also localizes in close proximity to β-tubulin.

Consistent with these findings, nephrocystin also localizes to the transition zone of respiratory cilia (Figure 1C). To analyze further the role of nephrocystin in respiratory cilia, we tested whether it co-localizes with well-characterized ciliary proteins. Co-staining of the axonemal outer dynein arm motor protein DNAH5, which localizes to the MTOC and to the axonemal doublet structure of cilia (27), confirms that nephrocystin specifically localizes to the transition zone, distally from the MTOC and proximally from the axonemes (Figure 1D).

On the basis of interaction with tensin and p130^cas, it has been speculated that nephrocystin functions at focal adhesions (12,13). In contrast, we find that in human ciliated respiratory epithelial cells, tensin and p130^cas localize at the apical cytoplasmic area, predominantly at the MTOC region, where their signals overlap at the proximal end of the transition zone with nephrocystin (Figure 1, E and F). Because the transition zone is thought to act as the docking site for the intraflagellar transport (IFT) particles, we next tested whether nephrocystin co-localizes with the IFT proteins IFT88 and IFT20. Similarly, as in Chlamydomonas (33), both IFT proteins are localized predominantly at the ciliary base and with a speckled pattern along the ciliary axonemes in respiratory epithelial cells (Figure 1, G and H). Overlapping staining of nephrocystin with IFT proteins is observed only at the proximal part of the transition zone but not within the ciliary axoneme. In addition, antibodies directed against β-tubulin stained the entire axonemes. At the proximal axonemal end, β-tubulin localizes in close vicinity of nephrocystin (Figure 1I).

Because retinitis pigmentosa can be found in patients with NPHP1, we tested whether nephrocystin localizes to the connecting cilium that bridges the outer and inner photoreceptor segments (Figure 2A). Co-staining of retinal cryosections with antibodies against nephrocystin and either acetylated α-tubulin or γ-tubulin demonstrates that nephrocystin exclusively localizes to the connecting cilium in close proximity to the basal bodies (Figure 2, B and C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Nephrocystin localizes to photoreceptor-connecting cilia at the junction of the inner segment (IS) and outer segment (OS). (A) Electron microscopic image of a mouse photoreceptor demonstrating the position of the connecting cilium (cc) and the basal body (bb) between the OS and the IS. Bar = 1 μm. (B) Immunofluorescence staining of a cryosectioned pig retina demonstrates that nephrocystin (red) co-localized with acetylated α-tubulin (green), a marker of the photoreceptor-connecting cilia. (C) Co-staining of a cryosectioned pig retina with antibodies against nephrocystin (red) and γ-tubulin (green) shows that nephrocystin localizes in close proximity to γ-tubulin which marks the positions of the basal body/centrioles. The OS and the IS, the outer nuclear layer (ONL), the outer plexiform layer (OPL), and the inner nuclear layer (INL) are indicated. Nuclei were stained with Hoechst 33342 (blue). Bars = 10 μm in B and C.

To analyze whether nephrocystin is associated tightly with the axonemal structure at the ciliary base, we took advantage of the microtubule severing mechanism by which eukaryotic cells excise cilia and flagella upon treatment with Ca²⁺-containing buffers. The deflagellation mechanism in Chlamydomonas involves the Ca²⁺ binding protein centrin, which is part of the contractible stellate fibers within the transition zone (34). Treatment of human respiratory epithelial cells with deciliation buffer resulted in complete or partial cilia autotomy in most cells (Figure 3). Isolated autotomized cilia showed specific punctual nephrocystin staining at the proximal end (Figure 3A). All partially deciliated cells (Figure 3, B and C) as well as all completely deciliated cell remnants (Figure 3D) showed specific nephrocystin staining at the apical cellular surface. Thus, Ca²⁺-dependent deciliation in respiratory epithelial cells occurs within the region where nephrocystin is localized. This specific staining pattern is observed even after prolonged treatment of the cells with deciliation buffer that contains detergent, which removes the membranes from cell remnants and cilia (25). We therefore conclude that nephrocystin is not a component of the ciliary membrane but associated tightly with the axonemal structure. This also is supported by Western blot analysis, which identified nephrocystin in high-salt protein extracts from purified demembranated respiratory ciliary axonemes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

The nephrocystin-positive transition zone contains the ciliary autotomy site. Human ciliated respiratory cells were subjected to Ca²⁺-dependent deciliation by microtubule severing within the transition zone. (A) Purified autotomized cilia show punctuate nephrocystin staining (red) at the proximal end. Axonemes were stained with antibodies against acetylated α-tubulin. Completely (B) and partially (C) deciliated cell remnants stain positive for nephrocystin (red) at the sites where cilia autotomy occurs. (D) Co-staining of a deciliated cell with antibodies against DNAH5 (red) and nephrocystin (green). DNAH5 staining at the MTOC and nephrocystin staining distally from the MTOC in deciliated cell remnants demonstrate that microtubule severing occurs within the nephrocystin-positive transition zone. Bars = 10 μm in A and B.

Nephrocystin Expression During Ciliogenesis

We next addressed the question of whether the localization of nephrocystin changes during the process of respiratory cell differentiation and in vitro ciliogenesis of primary human respiratory cells. Dissociated cells from nasal polyps were grown to confluence on collagen layers and subsequently were differentiated into ciliated spheroids (cell conglomerates) in suspension cultures (31).

At an early stage of in vitro ciliogenesis, cell morphology of spheroids is characterized by a symmetric cell body and a central nucleus in most spheroids (data not shown). These cells do not express nephrocystin and do not carry cilia as evidenced by the absence of acetylated α-tubulin staining. A few cells within each spheroid then start to express nephrocystin with a diffuse cytoplasmic localization (Figure 4, A and B), which precedes cilia formation (4 to 8 d). In the next period of respiratory cell differentiation (6 to 8 d), most cells in each spheroid show diffuse cytoplasmic nephrocystin expression (Figure 4B) that becomes apically enhanced in those cells that exhibit cell polarization, recognizable by a more prolonged cell body and downward placement of the nucleus. The concentration of nephrocystin at the apical cell region coincides with expression of acetylated α-tubulin that specifically localizes to well-defined spots beneath the apical plasma membrane in these cells, indicating the beginning of axoneme budding and cilia formation (Figure 4, B and C). During later stages of ciliogenesis (8 to 12 d), the punctual localization of nephrocystin at the base of each growing and mature axoneme and proximally to the acetylated α-tubulin remains unchanged (Figure 4, D and E). Staining of nephrocystin at the basolateral side of respiratory cells never was observed.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Nephrocystin translocalizes from the cytoplasm to the ciliary transition zone during respiratory epithelial cell polarization. Confocal immunofluorescence imaging of human respiratory epithelial cells that were subjected to spheroidal growth and in vitro ciliogenesis. Co-staining with the antibodies against cilia-specific acetylated α-tubulin (green) and nephrocystin (red). Nuclei were stained with Hoechst 33342. Bars = 10 μm. (A) In early stages of spheroidal growth, only a few cells express nephrocystin with a diffuse cytoplasmic localization. Cells show a round appearance with no indication of polarization. No specific acetylated α-tubulin staining is detectable. (B) At the beginning of cell polarization, indicated by cone-shaped cells, diffuse cytoplasmic nephrocystin expression becomes locally enriched at the apical cytoplasmic region. In these cells, acetylated α-tubulin is detectable at the apical plasma membrane, indicating the beginning of axoneme budding. (C and D) In later stages of cell polarization, nephrocystin becomes concentrated at the apical cell surface and is tightly associated with the bases of the growing respiratory cilia. (E) In heavily ciliated spheroids, nephrocystin shows a speckled localization pattern at the ciliated apical cell side. Nephrocystin localizes exclusively to the base of cilia, and cytoplasmic localization is no longer detectable.

Nephrocystin Deficiency Does Not Alter Cilia Formation

Because nephrocystin translocalizes from the cytoplasm to the transition zone during respiratory epithelial cell differentiation, we next tested whether absence of nephrocystin affects ciliogenesis or cilia function. We analyzed respiratory epithelial cells from patients with NPHP1 (patients ON-21, ON-23, ON-43, ON-45II1, ON45II2, and ON-50), who had received a diagnosis before for homozygous NPHP1 deletions according to routine molecular genetic testing of juvenile NPHP (11). We first confirmed the deletion of genomic NPHP1 sequences by PCR (data not shown) and the absence of nephrocystin by Western blot in four of these patients using Epstein-Barr virus–transformed B-lymphocytes (Figure 5H).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Nephrocystin-deficient respiratory epithelial cells form normal cilia and basal bodies. Confocal immunofluorescence imaging of ciliated respiratory epithelial cells from healthy probands (control) and patients with nephronophthisis (NPHP) with NPHP1 deletions. Nuclei were stained with Hoechst 33342 (blue). Bars = 10 μm. (A and B) Localization of nephrocystin (red) at the transition zone of respiratory cilia in control cells. Cilia were stained with anti-acetylated α-tubulin antibodies (green). No specific nephrocystin staining (red) is observed in respiratory epithelial cells from patients with NPHP with NPHP1 deletions. In B, microscope settings of the red channel were adjusted to visualize background staining. No red signals are observed when identical settings as in A are used. Nephrocystin deficiency does not impair respiratory epithelial cell polarization and cilia formation. (C and D) Co-localization of phosphoacidic cluster sorting protein-1 (PACS-1; green) and nephrocystin (red) at the transition zone of respiratory cilia in control cells. Normal localization of PACS-1 in nephrocystin-deficient cells indicate normal integrity of the ciliary base. (E and F) Normal localization of γ-tubulin, a com-ponent of the MTOC, and the IFT88 in nephrocystin-deficient cells indicates that the integrity of the ciliary bases and the intraflagellar transport processes per se are not disrupted by the absence of nephrocystin from the ciliary transition zones. (G) In vitro ciliogenesis of respiratory epithelial cells organized in spheroids from a healthy donor (left) and from a patient with NPHP with NPHP1 deletions (right). Nephrocystin deficiency does not inhibit epithelial cell polarization and cilia formation. (H) By Western blot, nephrocystin is detected in crude lysates from tracheal epithelial cells and in high-salt extracts from isolated demembranated respiratory ciliary axonemes, indicating that nephrocystin is associated tightly with axonemal structures. MDCK cells were used as a control (top). Nephrocystin can be readily detected in protein extracts from Epstein-Barr virus–transformed B-lymphocytes from healthy donors (control) but is absent in these cells that were derived from patients with NPHP1 (ON-21, ON-23, ON-43, and ON-50). HEK293T cells were included as a control (bottom).

We then confirmed by immunofluorescence imaging that respiratory epithelial cells from all six patients are deficient for nephrocystin (Figure 5, A and B). Because cell and cilia morphology were indistinguishable from controls, nephrocystin deficiency does not change the overall integrity of respiratory cilia.

To test whether nephrocystin localization also is altered in patients with other forms of NPHP, we analyzed respiratory epithelial cells from two patients who had adolescent NPHP and carried compound heterozygous NPHP3 mutations (patients F23II1 and F23II2) and one patient who had NPHP type 5 and carried compound heterozygous NPHP5 mutations (patient A19), which were reported previously (7,8). In these patients, cell morphology and nephrocystin localization were normal (data not shown), demonstrating that NPHP3 and NPHP5 mutations in the analyzed patients do not alter nephrocystin localization. In addition, we found normal nephrocystin localization in two patients with BBS, five patients with autosomal recessive polycystic kidney disease, and three patients with NPHP (for which NPHP1 deletions have been excluded) and unknown mutations (data not shown). These results demonstrate that absence of nephrocystin from the transition zone is a specific finding in patients who have NPHP with NPHP1 deletions rather than a common alteration in cystic kidney diseases.

Next, we asked whether nephrocystin is essential for correct targeting of ciliary/basal body proteins. However, staining of nephrocystin-deficient respiratory cells with antibodies against PACS-1 (Figure 5, C and D); γ-tubulin, a component of the pericentriolar material around the basal bodies; and IFT88 (Figure 5, E and F) revealed normal localization of all three proteins at the ciliary base. These results demonstrate that nephrocystin is not required for targeting of these proteins to the ciliary base and that the structural integrity of the ciliary base remains unaffected in nephrocystin-deficient cells. Furthermore, because the localization IFT88 along the ciliary axonemes also remains unaffected in nephrocystin-deficient cells, intraflagellar transport per se obviously is not impaired by the absence of nephrocystin from the transition zone.

We also tested whether the loss of nephrocystin impairs the function of respiratory cilia because three of the patients with NPHP1 deletions (ON-21, ON-23, and ON-50) reported mild respiratory symptoms with chronic sinusitis and rhinitis suggestive of a cilia dysmotility defect. High-speed video-microscopic analyses of respiratory epithelial cells from these patients showed normal ciliary beat frequencies (5 to 9 Hz at room temperature) and normal beat amplitudes. However, evaluation of slow-motion playbacks revealed that ciliary motility is slightly irregular (Supplementary Videos 1 and 2).

To examine whether the observed dysmotility is caused by secondary effects, we grew respiratory epithelial cells that were obtained by transnasal brush biopsies from patient ON-50 to heavily ciliated spheroids in vitro, which bypasses secondary ciliary dyskinesia (35). Nephrocystin-deficient spheroids (Figure 5G) had normal morphology and normal ciliogenesis when compared with control cells (Figure 4) as well as normal ciliary beat frequencies and amplitudes. Although we noted a slightly irregular beating pattern, the degree of the observed abnormality was not as severe as usually observed in primary ciliary dyskinesia (Supplementary Videos 3 and 4).