Madin-Darby Canine Kidney type II (MDCK) cells were purchased from the European Collection of Cell Cultures (No 62107) and cultivated as previously described (Muller et al., 2021). MDCK cells with a knockout, overexpression or recovery of TTL are summarized in Table 1. To generate MDCK and MDCK_ΔTTL cells expressing PDX fused with the SNAP-tag (PDX-SNAP), the plasmid pCDNA3.1 (+)-myc-SNAP-GP135 (Addgene, Watertown, United States, (Castillon et al., 2013)) was transfected using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, United States) according to the manufacturer’s instructions. Stable cell clones expressing PDX-SNAP were selected as described previously (Zink et al., 2012) and are listed in Table 1.

MDCK and MDCK_ΔTTL cells were cultured for 6 days, lysed and RNA was purified using the RNeasy Mini Kit (Qiagen, Hilden, Germany). RNA was subsequently converted to cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, United States). Reverse Transcription quantitative PCR (RT-qPCR) was performed using the Stratagene MX3005P system (Agilent Technologies, Santa Clara, United States). To block lysosomal degradation MDCK and MDCK_ΔTTL cells were washed with PBS⁺⁺ and incubated in MEM medium containing 50 mM NH₄Cl for 1 or 2 h or in the absence of NH₄Cl. Cells were then lysed and analyzed by immunoblot.

For preparation of cell lysates, cells were washed with sterile-filtered PBS⁺⁺ (PBS supplemented with 1 mM MgCl₂ and 1 mM CaCl₂), collected in lysis buffer (pH 7.4; 150 mM NaCl; 25 mM HEPES; 5 mM EDTA; 1% NP-40; 0.1% SDS; freshly added protease inhibitor cocktail) and incubated at 4 °C on an overhead shaker for 30 min. Samples were then centrifuged at 17.000 g for 15 min. Protein concentrations in the supernatants were determined principally as described by Lowry et al. (1951) using the DC Protein Assay (Bio-Rad Laboratories, Hercules, United States), separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked for 1 h in 5% skimmed milk powder in PBS or 5% BSA in TBST and incubated with primary antibodies overnight at 4 °C. The used antibodies are listed in Table 2. Detection was performed using horseradish peroxidase (HRP)-conjugated secondary antibodies and enhanced chemiluminescence (ECL) reagent (Thermo Fisher Scientific, Waltham, United States), followed by imaging with the ECL Chemocam Imager (Intas Science Imaging Instruments, Göttingen, Germany). Secondary antibodies, phalloidin conjugates and SNAP-tag substrates are summarized in Table 3. Band densities and molecular weights in Western blots were quantified using LabImage 1D (Kapelan Bio-Imaging, Leipzig, Germany) and the Fiji/ImageJ software package. Band density values were normalized to GAPDH. Fluorescent blots were processed with the Sapphire Biomolecular Imager (Azure Biosystems, Dublin, United States). Statistical analysis was performed using Prism 6 (GraphPad Prism, La Jolla, United States).

Cultivated cells were washed with PBS⁺⁺ and collected in lysis buffer for deglycosylation (pH 7.5; 100 mM NaCl; 25 mM Tris; 1 mM EDTA; 1 mM EGTA; 0.5% Triton X-100; 0.5% NP-40; freshly added protease inhibitor cocktail) by mechanical detachment. The lysates were then incubated at 4 °C for 30 min on an overhead shaker. After centrifugation (17.000 g for 15 min), lysates were treated with PNGase F or the Protein Deglycosylation Mix II according to the manufacturer’s instructions. For immunoprecipitation MDCK cells were washed with PBS⁺⁺, collected in Mg²⁺ lysis buffer (pH 7.4; 100 mM NaCl; 20 mM Tris; 10 mM MgCl₂; 5 mM EDTA; 2 mM DTT; 0.4% NP-40; freshly added protease inhibitor cocktail) and incubated at 4 °C for 30 min. After centrifugation (17.000 g, 15 min), cleared lysates were transferred to new Eppendorf tubes and precleared with non-specific IgG/protein A-sepharose beads for 1 h at 4 °C followed by incubation with anti-PDX antibodies/protein A-sepharose beads for 3 h at 4 °C. Non-specific IgG/protein A-sepharose beads were used as a negative control. Finally, the beads were transferred to Mobicol “Classic” Spin Columns, rinsed three times with Mg²⁺ washing buffer (ph 7.4; 20 mM Tris; 10 mM MgCl₂; 5 mM EDTA; 0.1% Tween 20), followed by three rinses with PBS. For immunoprecipitation of pERM cells were lysed by sonication in Mg²⁺ lysis buffer without detergent. Beads were washed four times with Mg²⁺ lysis buffer without detergent. Subsequently the samples were separated by SDS-PAGE and analyzed by immunoblot.

Cells grown on cover slips or 24-well filter inserts were washed twice with PBS⁺⁺ and fixed with ice-cold methanol for 5 min or 4% paraformaldehyde for 20 min. After fixation, cells were permeabilized with 0.1% Triton X-100 for 20 min and blocked in 5% BSA/PBS⁺⁺ for 1 h. Immunostaining was performed using the indicated primary antibodies in blocking reagent for 2 h or overnight. Secondary antibodies conjugated with the specific Alexa Fluor dyes were applied in PBS⁺⁺ for 1 h. Nuclei were stained with Hoechst 33,342. Following incubation, cells were washed with PBS⁺⁺ and mounted using ProLong Diamond (Thermo Fisher Scientific, Waltham, United States). Confocal images were acquired using a Leica STELLARIS microscope equipped with a ×93 glycerol plan-apochromat objective (Leica Microsystems, Wetzlar, Germany). Immunofluorescence images were quantitatively analyzed (line intensity scan, ROI-measurements) using the Leica Acquire Software X (LasX, Leica Microsystems, Wetzlar, Germany).

For biotinylation MDCK cells were seeded on 24-well filter inserts for 5 days, washed with PBS⁺⁺ and incubated with EZ-Link™ NHS-LC-Biotin (Thermo Fisher Scientific, Waltham, United States) either apically or basolaterally for 30 min at 4 °C. Untreated cells were used as negative controls. After washing with glycine and PBS⁺⁺, cells were lysed as described above. The lysates were incubated with Pierce™ NeutrAvidin™ Agarose beads (Thermo Fisher Scientific, Waltham, United States) for 2 h at 4 °C. Finally, beads were transferred to Mobicol “Classic” Spin Columns and processed as described above. To determine PDX-internalization the cells were incubated with EZ-Link™ Sulfo-NHS-SS-Biotin beads (Thermo Fisher Scientific, Waltham, United States) for 30 min at 4 °C. After washing with glycine and PBS⁺⁺, cells were incubated at 37 °C for 0, 30 or 60 min. Remaining protein-bound biotin was removed by 30 min treatment with 25 mM glutathione in 90 mM NaCl, 1 mM MgCl₂, 0.1 mM CaCl₂, 60 mM NaOH/10% FCS, followed by 10 min incubation with 5 mg/mL Iodoacetamide in PBS⁺⁺ at 4 °C. Cells were then lysed and prepared for immunoblot analysis as described above. Apical surface delivery of MDCK cells expressing PDX-SNAP was measured by cell labelling with SNAP-Surface Alexa 546 for 30 min at 4 °C. The cells were washed and remaining SNAP at the cell surface was blocked with SNAP-Surface Block for 10 min. After washing, newly synthesized material at the cell surface was labeled with SNAP-Surface Alexa 647 for 0, 10, 20 or 30 min at 37 °C as previously described (Stoops et al., 2015). The cells were analyzed by confocal microscopy.

We first examined the impact of posttranslational microtubule modification on the expression of the apical mucin PDX in polarized MDCK cells, comparing those with and without a TTL gene knockout (MDCK_ΔTTL), cells overexpressing the TTL-GFP fusion protein (MDCK_TTL-GFP), and reconstituted cells (MDCK_{ΔTTL+TTL-GFP}). Five days post-seeding, the cells were lysed, and an immunoblot analysis was performed (Figure 1A). In MDCK_TTL-GFP cells, where TTL-GFP levels were consistently elevated, the amount of detyr-tubulin decreased, while the opposite effect was observed in MDCK_ΔTTL cells. As previously demonstrated, MDCK_ΔTTL cells exhibited depleted levels of tyr-tubulin and increased detyr-tubulin levels (Muller et al., 2021). Notably, TTL-depleted cells showed a significant reduction in PDX-expression down to 21% as compared to the level in MDCK cells (p < 0.0001; n = 8), along with a loss in its molecular weight. Conversely, no changes were observed in the expression of cytosolic ezrin. Furthermore, no significant abnormalities in PDX-expression or size were detected in TTL-overexpressing or -reconstituted cells. We further analyzed these alterations throughout the MDCK cell differentiation process, from day 1–7 after plating (Figure 1B; Supplementary Figure S1A, B). As shown in Figure 1B, PDX levels were consistently reduced in MDCK_ΔTTL cells. The reduced quantities of PDX in MDCK_ΔTTL cells were examined through two scenarios. First, we investigated whether increased lysosomal degradation was responsible for the lower levels of this glycoprotein in MDCK_ΔTTL cells. To test this, we raised the pH of endosomes and lysosomes by adding 50 mM NH₄Cl, which inhibits lysosomal proteases (Straube et al., 2013). As shown in Figure 1C; Supplementary Figure S1C, this treatment did not affect the amount of PDX in either MDCK or MDCK_ΔTTL cells. Therefore, changes in lysosomal degradation were ruled out as the cause of the reduced PDX levels in MDCK_ΔTTL cells. The other possibility was a decrease in PDX biosynthesis, which we analyzed by quantifying PDX mRNA levels in MDCK and MDCK_ΔTTL cells using RT-qPCR (Figure 1D). Our findings revealed that PDX mRNA levels decreased by approximately 50% when TTL was knocked out. Given that PDX is believed to play a role in establishing apical-basal polarity (Meder et al., 2005) and that MDCK_ΔTTL cells initially exhibit a flat morphology before gradually adopting a differentiated columnar epithelial cell shape (Muller et al., 2021), it is significant that proper microtubule architecture seems necessary to keep PDX-expression at a high level in renal epithelial cells.

The other notable alteration of PDX in MDCK_ΔTTL cells was a reduction in molecular weight by approximately 4 kDa. This change was most prominent shortly after plating and converged once the cells became fully polarized, 7 days after seeding (Figure 1E). Since PDX is heavily glycosylated with both N- and O-linked glycans (Meder et al., 2005), we investigated whether this molecular weight shift, observed 1 day after cell seeding, was due to a loss or modification of glycosyl chains. Removal of N-glycans using PNGase F did not eliminate the size difference in PDX between MDCK and MDCK_ΔTTL cell lysates (Figure 1F). This coincides with the observation that an N-glycosylation-deficient PDX mutant barely altered the molecular weight as compared to the wild type molecule (Roman-Fernandez et al., 2023). The size difference was resolved only after removing N- and O-glycosyl chains with a complete mixture of deglycosidases, suggesting that an alteration in the O-glycosylation pattern leads to the molecular weight difference of PDX in MDCK versus MDCK_ΔTTL cells early after seeding.

O-glycans are involved in apical protein sorting (Alfalah et al., 1999; Yeaman et al., 1997). To investigate whether the observed changes in O-glycosylation are related to differences in the polarized distribution of PDX, we conducted an immunofluorescence study and stained PDX for confocal microscopy analysis. For comparison, we also immunostained ezrin, the cytosolic interaction partner downstream of PDX that directly links the glycoprotein to the actin cytoskeleton in kidney cells (Schmieder et al., 2004; Takeda et al., 2001). X/z scans showed that, as previously reported, PDX is localized at the apical membrane, with ezrin concentrated beneath this membrane domain in MDCK cells (Figures 2A,B). Similarly, PDX is found at the apical membrane of MDCK_ΔTTL cells, which exhibit a flat morphology as previously described (Muller et al., 2021). However, in these cells and in stark contrast to MDCK, MDCK_TTL-GFP and MDCK_{ΔTTL+TTL-GFP} cells, ezrin shows a non-polar distribution throughout the cytoplasm (Figures 2A,B; Supplementary Figure 2). This redistribution of ezrin closely aligns with observations made in MDCK cell cysts following PDX depletion (Bryant et al., 2014) and can be attributed to a reduced ezrin-PDX interaction at the apical membrane. We further analyzed this interaction by co-precipitating ezrin with PDX from MDCK cell lysates. The efficiency of this precipitation was notably reduced in MDCK_ΔTTL cells (Figures 2C,D), indicating that alterations in detyr- and tyr-tubulin levels reduce recruitment of ezrin to the apical membrane. Evidence for a depletion in the formation of the PDX/ezrin complex also comes from significantly reduced levels of phosphorylated active ezrin/radixin/moesin (pERM) proteins in MDCK_ΔTTL cells, which is further validated by minor pERM-quantities precipitated by PDX (Figures 1A, 2E,F). PDX and phosphorylation of ezrin are involved in tight junction assembly and regulate expression of claudin 2 (Yasuda et al., 2012). Consequently, we assessed whether the composition of tight junctions was altered in MDCK_ΔTTL cells. Western blot analysis reveals that in contrast to claudin 1, the expression of claudin 2, which is associated with “leaky” nephron segments (Kiuchi-Saishin et al., 2002), is significantly reduced if TTL is knocked out (Figures 3A,B). This nicely corresponds to a previous study showing a depletion in claudin 2-expression in MDCK cells with reduced apical targeting of ezrin (Yasuda et al., 2012). To study the assembly dynamics of tight junctions during early events in monolayer formation, we seeded MDCK cells at high densities and followed tight junctional claudins 1 and 2 by immunofluorescence. Here, we focused on the island and monolayer stages at 0, 12 and 24 h after plating, respectively. To precisely visualize incorporation of claudin 1 and 2 into tight junctions, this membrane region was also immunostained with antibodies directed against zonula occludens-1 (ZO-1), a classical cytoplasmic adaptor protein of tight junctions. Focal planes within and beneath the tight junctional complex were monitored. Figure 3C depicts co-staining of claudin 1 as well as claudin 2 with ZO-1 12 h after plating in MDCK and MDCK_ΔTTL cells, which indicates that the dynamics of tight junction formation is not altered by increased quantities of detyr-tubulin following TTL-depletion. Nevertheless, the claudin composition is obviously altered, which might explain the increased transepithelial resistance peak observed in MDCK_ΔTTL cells early after plating (Muller et al., 2021). In summary, the quality but not the assembly dynamics of the tight junctional complex is altered by TTL-knockout in MDCK cells, while the apical identity promoting PDX/ezrin complex is destabilized.

The apical localization of PDX and ezrin is essential for epithelial cell polarization (Bryant et al., 2014). We thus examined the polarized distribution of PDX by surface biotinylation. In agreement with the fluorescence microscopic images the predominant PDX pool was biotinylated at the apical membrane (Figures 4A,B). However, consistent with data from Stoops et al. a faint fraction of biotinylated PDX (5%) was also found at the basolateral membrane of MDCK cells that traverses this membrane en route to the apical membrane (Stoops et al., 2015). This basolateral fraction could not be detected in MDCK_ΔTTL cells, resulting in an exclusive apical localization of PDX in this cell line. Two scenarios of PDX-trafficking would explain this alteration. First, rapid transcytosis of the small basolateral PDX pool to the apical membrane or, as a second option, post-Golgi transport of PDX exclusively to the apical membrane. To decide between these two scenarios, rapid endocytic PDX uptake was monitored. We therefore labeled PDX with a reducible biotin conjugate and endocytosis was allowed for 0, 30 or 60 min at 37 °C. Thereafter, reduction with glutathione removed the biotin label from polypeptides at the cell surface. Figures 4C,D shows that non-reduced internalized PDX appeared after 30 min of internalization from the apical membrane and after 60 min from the basolateral membrane of MDCK cells. In MDCK_ΔTTL cells no non-reduced PDX internalized from the basolateral membrane even after short time periods could not be detected. This excludes a transcytotic route of PDX in this cell line and supports the idea that in the absence of TTL PDX-trafficking is targeted entirely to the apical membrane.

We then focused on the way how PDX is delivered to the apical surface of MDCK cells. Here, PDX is transported to a ring at the base of the primary cilium followed by microtubule-dependent radial movement away from the cilium (Stoops et al., 2015). We thus monitored surface delivery of a modified form of PDX, which contained a SNAP-tag inserted into its extracellular N-terminus (PDX-SNAP) by confocal fluorescence microscopy (Stoops et al., 2015). MDCK and MDCK_ΔTTL cells stably transfected with PDX-SNAP were labeled with cell impermeable SNAP-Surface Alexa Fluor 546 at 4 °C to stain all SNAP fusion proteins yielding a uniform staining pattern of the apical membrane (Figures 5A,B). As a control for staining specificity, cell monolayers were formed by a mixture of transfected and non-transfected cells. Apical delivery of newly synthesized PDX-SNAP was then monitored by first blocking all remaining PDX-SNAP polypeptides on the membrane with non-fluorescent SNAP-Surface Block followed by incubation with SNAP-Surface Alexa Fluor 647 at 37 °C for 0, 10, 20 or 30 min (Figure 5A). The lack of fluorescence at 0 min of Alexa Fluor 647 staining demonstrates that any protein labeling with this fluorophore for longer time periods at 37 °C is attributable to newly delivered PDX-SNAP. After 10 min of labelling Alexa Fluor 647 fluorescence appears at the base of the primary cilium in MDCK cells. This labeling then spreads in the following time intervals across the apical membrane domain, which is in conjunction with data from Stoops et al. (2015). In the absence of TTL we did not observe a focused delivery of newly synthesized PDX-SNAP at the base of the primary cilium. Instead, the whole apical membrane was stained by Alexa Fluor 647 after 10 min of incubation at 37 °C and this staining intensified over time. To check if these alterations in the delivery pattern of PDX-SNAP to the membrane influence kinetics of apical delivery, we measured the appearance of newly synthesized PDX-SNAP labelled with Alexa Fluor 647 on a Sapphire Biomolecular Imager after SDS-PAGE (Figures 5C,D). This biochemical analysis demonstrates that PDX-SNAP appears at the apical membrane of MDCK cells with a similar kinetics as in MDCK_ΔTTL cells. In conclusion, the spatial organization of apical delivery as well as intracellular transport of PDX are altered following TTL-knockout, while the kinetics of appearance is not significantly affected.

The question is how these observations relate to changes in the PTMs of microtubules. To address this point, we studied the distribution of microtubules enriched in acetylated (acetyl-), detyr- and tyr-tubulin in fully polarized MDCK cells by confocal microscopy. Figure 6A, B illustrates that detyr- and tyr-tubulin are predominantly localized at the apical cell pole, with a clear enrichment of acetyl-tubulin in the axoneme of the primary cilium. While detyr-tubulin-enriched microtubules constituted a minor fraction and tyr-tubulin-enriched microtubules represented the major fraction in MDCK cells. This relationship shifted completely in MDCK_ΔTTL cells, with detyr-tubulin-enriched microtubules becoming predominant. The shift resulted in a dramatic accumulation of a detyr-tubulin enriched microtubular network beneath the apical membrane (Figure 6B). A detailed analysis of the distribution of detyr-tubulin within this subapical network was performed by defining circular regions of interest (ROI1) surrounding the base of the primary cilium (Figure 6C). The fluorescence intensity of detyr-tubulin within these ROIs was then compared to the intensity within a second, more peripheral ROI (ROI2). In MDCK cells the quotient of ROI1 and ROI2 was 1.93 ± 0.26 thus indicating subapical accumulation of detyr-tubulin around the base of the primary cilium. This quotient was significantly reduced in MDCK_ΔTTL cells to 1.27 ± 0.06, which demonstrates that in the absence of TTL the subapical network of detyr-tubulin enriched microtubules is spread from a central area towards the cell periphery. These microtubules may serve as tracks for a pan-apical targeting of PDX, and we consequently monitored the association of PDX-containing post-Golgi transport vesicles with detyr- and tyr-tubulin enriched microtubules. For this purpose, we used non-polarized cells. These cells grow with a flat morphology on coverslips, allowing clear visualization of individual microtubules and the allocation of transport vesicles. To trigger post-Golgi transport, newly synthesized material was first accumulated in the Golgi apparatus for 4 h at 20 °C (Simons and Wandinger-Ness, 1990) followed by Golgi-release at 37 °C for 1 h. Staining of PDX-carrying post-Golgi vesicles showed significant overlap with detyr- and tyr-enriched microtubules (Figure 7A). Quantification of microtubule/vesicle-co-staining revealed that in both MDCK and MDCK_ΔTTL cells, a larger fraction of post-Golgi vesicles was associated with detyr-tubulin enriched microtubules than with tyr-tubulin enriched microtubules (Figure 7B). This supports the idea that detyr-tubulin enriched microtubules provide the primary tracks for post-Golgi trafficking of PDX.