Nucleus pulposus primary cilia alter their length in response to changes in extracellular osmolarity but do not control TonEBP-mediated osmoregulation

The nucleus pulposus (NP) cells adapt to their physiologically hyperosmotic microenvironment through Tonicity-responsive enhancer binding protein (TonEBP/nuclear factor of activated T-cell5 [NFAT5])-mediated osmoregulation. Primary cilia in different organs serve diverse roles including osmosensing, but its contribution to NP cell osmoadaptive response is unknown. A high percentage of cultured primary NP cells possessed primary cilia that changed length in response to osmotic stimuli. Stable silencing of Intraflagellar Transport 88 (Ift88) or Kinesin Family Member 3 A (Kif3a) to inhibit the formation of primary cilia did not affect hyperosmotic upregulation of TonEBP. While ShKif3a blocked hyperosmotic increase of TonEBP-Transactivation Domain (TAD) activity, overall the knockdown of either gene did not alter the hyperosmotic status of proximal promoter activities and transcription of key TonEBP targets. On the other hand, a small decrease in TonEBP level under hypoosmotic condition was attenuated by Ift88 or Kif3a knockdown. Noteworthy, none of the TonEBP target genes were responsive to hypoosmotic stimulus in control and Ift88 or Kif3a knockdown cells, suggesting the primary role of TonEBP in the hyperosmotic adaptation of NP cells. Similarly, in Kif3a null mouse embryonic fibroblasts (MEFs), the overall TonEBP-dependent hyperosmotic responses were preserved. Unlike NP cells, TonEBP targets were responsive to hypoosmolarity in wild-type MEFs, and these responses remained intact in Kif3a null MEFs. Together, these results suggest that primary cilia are dispensable for TonEBP-dependent osmoadaptive response.


Results
The length of primary cilia in NP cells is responsive to changes in extracellular osmolarity.
Primary cilia were visualized in cultured primary rat NP cells by co-immunostaining acetylated α-tubulin and γ-tubulin, labeling ciliary axoneme and basal bodies, respectively (Fig. 1a,b). Previous studies showed that the length of primary cilia in different types of cells changed in response to extracellular stimuli [39][40][41]

Inhibition of primary cilia formation in NP cells does not affect TonEBP expression in hyperosmotic condition.
The function of primary cilia as an osmosensor has been observed in C. elegans as well as in some types of mammalian cells, including renal tubular epithelial cells, articular chondrocytes, and cholangiocytes [35][36][37][38] . NP cells reside in an osmotically active microenvironment due to high proteoglycan content of the NP matrix and dynamic loading of the spine. We examined if primary cilia of the NP cells play a role in sensing extracellular osmolarity and mediating cellular osmotic response. We inhibited formation of primary cilia in NP cells by performing stable knockdown of Ift88, a component of complex B of the intraflagellar transport (IFT) particles [42][43][44][45] (Fig. 2g). Quantification of the number of cells with primary cilia confirmed this result ( Fig. 2h; p < 0.0001 for all groups). Overall, the lengths of primary cilia that remained after the stable silencing of Ift88 or Kif3a were not significantly different from that of the control cells ( Fig. 2i; p = 0.0334 for ShIft88 #2, all other groups were statistically not significant).
To determine if inhibition of primary cilia formation resulted in dysregulation of cellular osmoresponse, we stably silenced Ift88 or Kif3a and measured the expression of TonEBP/NFAT5, a transcription factor crucial for osmoregulation in NP cells [8][9][10][11] . While there was a trend of increase in TonEBP/Nfat5 transcript levels under hyperosmotic conditions (550 mOsm/kg H 2 O), this increase did not reach a statistical significance in both control and Ift88 knockdown cells ( Fig. 3a; all analyses were statistically not significant). Hyperosmotic increase in TonEBP protein levels was unaffected by Ift88 knockdown (Fig. 3b, Primary cilia do not modulate TonEBP activity and TonEBP target gene expression in NP cells under hyperosmotic conditions. Hyperosmotic stimulus leads to the increased transcriptional activity of TonEBP in the nucleus 9,11,47 . We investigated if primary cilia controlled TonEBP activity without affecting its expression. Changes in proximal promoter activities of well-characterized TonEBP target genes including, AR (aldose reductase), HSP-70 and TauT (taurine transporter) were measured in NP cells transfected with shRNA against Ift88 or Kif3a (Fig. 4a-c). These promoters contain highly conserved TonEBP-binding sites that are active in NP cells 10,15,19,20,47,48 . Hyperosmotic (550 mOsm/kg H 2 O) increase in the AR (Fig. 4a; p < 0.0001) and TauT ( Fig. 4c; p < 0.0001) promoter activities was not affected by either ShIft88 (p = 0.0004 for clone #1, p = 0.0012 for clone #2 for AR-luc; p < 0.0001 for both clones for TauT-luc) or ShKif3a (p = 0.001 for clone #1, p = 0.0003 for clone #2 for AR-luc; p = 0.0009 for clone #1 and p < 0.0001 for clone #2 for TauT-luc). Similarly, the HSP-70 promoter activity increased under hyperosmotic conditions with or without ShIft88 or ShKif3a. While one of the ShIft88 and ShKif3a clone groups did not reach a statistical significance, the trend of hyperosmotic induction of HSP-70 promoter activity was maintained ( Fig. 4b; p = 0.0011 for ShIft88 #2, p = 0.0357 for ShKif3a #2). In addition, we measured the activity of the TonEBP transactivation domain (TAD) in response to the hyperosmotic stimulus (550 mOsm/kg H 2 O) with or without ShIft88 and ShKif3a. While ShIft88 #2 inhibited the hyperosmotic increase of TonEBP-TAD activity, the other clone had no effect ( Fig. 4d; p = 0.0259 for ShIft88 #1). Both clones of ShKif3a prevented a further increase of TonEBP-TAD activity in response to the hyperosmotic stimulus. However, the average level of TonEBP-TAD activity under hyperosmotic conditions was similar or higher in knockdown cells when compared to ShCtr group, suggesting that the overall TonEBP-TAD activity was unaffected by ShIft88 or ShKif3a (Fig. 4d).
To confirm that TonEBP-mediated hyperosmotic response was independent of primary cilia, we measured mRNA levels of TonEBP targets in NP cells with Ift88 or Kif3a knockdown. AR/Akr1b1, TauT/Slc6a6 as well as sodium/myo-inositol co-transporter (SMIT/Slc5a3) are osmolarity-dependent target genes of TonEBP [15][16][17][18] . Ift88 knockdown did not affect the hyperosmotic (550 mOsm/kg H 2 O) increase in the expression of AR/Akr1b1 and SMIT/Slc5a3 genes (Fig. 5a,b; p = 0.0065 for ShCtr, p = 0.0444 for ShIft88 #1, p = 0.0105 for ShIft88 #2 in Fig. 5a; p = 0.0024 for ShCtr, p = 0.0357 for ShIft88 #1, p = 0.0031 for ShIft88 #2 in Fig. 5b). While the increase in TauT/ Slc6a6 levels in response to hyperosmotic stimulus did not reach a statistical significance, the trend of increase  (d-f) qRT-PCR, Western blot, and corresponding densitometry analyses, show significant downregulation of KIF3A after stable knockdown using two different ShKif3a clones (n ≥ 4). (g) Acetylated α-tubulin immunofluorescence staining after lentiviral transduction of ShIft88 or ShKif3a shows inhibition of primary cilia formation in majority of rat NP cells. Scale bar = 75 μm. White arrowheads point to primary cilia. (h,i) Quantitation of percentage of NP cells with primary cilia and primary cilium length after stable silencing of Ift88 or Kif3a (n = 3; at least 150 cells/group). Data are represented as scatter plots (mean ± SEM). ns = not significant. One-way ANOVA or Kruskal-Wallis test with Sidak's, Holm-Sidak's, or Dunn's multiple comparison test was used based on the distribution of the data to determine statistical significance. For statistical comparison of the percentages of NP cells with primary cilia, Fisher's exact test was used. Western blot images were cropped and acquired under same experimental conditions. See Supplementary Fig. S1-1 for un-cropped Western blot images.
The absence of primary cilia has a minor effect on hypoosmotic downregulation of TonEBP and its targets in NP cells. The tissue osmotic conditions of the NP compartment undergo diurnal changes due to dynamic loading and unloading of the spine. During disc degeneration, the breakdown of the proteoglycan matrix by proteases decreases the tissue osmolarity [49][50][51] . Consequently, NP cells need to maintain their function and integrity under both hyper-and hypo-osmotic conditions 52,53 . We, therefore, assessed the effects of inhibiting  Fig. 6f; p = 0.0139 for ShKif3a #1, p = 0.026 for ShKif3a #2 in Fig. 6h) knockdown showed significant hypoosmotic decrease in TonEBP/Nfat5 transcript, compared to control cells that showed similar decreasing trend (Fig. 6d,i; p = 0.0043 for ShIft88 #1, p = 0.0029 for ShIft88 #2 in Fig. 6d; p = 0.0089 for ShKif3a #2 in Fig. 6i). On the other hand, a small but significant decrease in TonEBP protein levels seen in control cells under the hypoosmotic condition was attenuated following knockdown of Ift88 (Fig. 6b,e; p = 0.0266 for ShCtr, p = 0.0761 for ShIft88 #1) or Kif3a (Fig. 6g,j; p = 0.0515 for ShCtr; Supp. Figs S1 and 2). Although the changes in mRNA and protein levels in response to hypoosmotic stimuli were somewhat www.nature.com/scientificreports www.nature.com/scientificreports/ opposing in cells lacking primary cilia, these results suggested that primary cilia may play a minor role in maintaining TonEBP levels in NP cells under hypoosmotic conditions. We then measured mRNA levels of TonEBP targets to investigate if the small changes in TonEBP levels in cells without primary cilia had functional implications in overall hypoosmotic cellular response. AR/Ark1b1, SMIT/ Slc5a3, and TauT/Slc6a6 were not affected by stable knockdown of Ift88 or Kif3a (Fig. 7a-f; all statistical analyses were not significant in Fig. 7a,e,f; p = 0.0046 for ShIft88 #1 in Fig. 7b; non-significant for ShCtr and ShIft88 #1, p = 0.0139 for ShIft88 #2 in Fig. 7c; non-significant for ShCtr and ShKif3a #2, p = 0.0088 for ShKif3a #1 in Fig. 7d). Interestingly, none of these osmotic targets showed a significant decrease in their expression under hypoosmotic conditions, and this trend was not affected by either Ift88 or Kif3a knockdown ( Fig. 7a-f). Taken together, our results showed that primary cilia may have a small role in modulating hypoosmotic levels of TonEBP, but this did not lead to any appreciable effect on target gene expression.
TonEBP expression and activity are not mediated by primary cilia in mouse embryonic fibroblasts. Similar to NP cells, mouse embryonic fibroblasts (MEFs) induce TonEBP-dependent osmotic response under hyperosmotic conditions 8 . Parallel experiments were performed using Kif3a null MEFs that completely lack primary cilia to further confirm that primary cilia do not control TonEBP-dependent osmotic response 54    www.nature.com/scientificreports www.nature.com/scientificreports/ activity. Hyperosmotic stimulus (600 mOsm/kg H 2 O) significantly increased AR and HSP-70 promoter activities in both wild-type and Kif3a null MEFs (Fig. 9a,b; Fig. 9b). Likewise, the temporal response of TauT promoter to the hyperosmotic stimulus was similar in both wild-type and Kif3a null MEFs, where a robust activation at 16 h post-treatment was observed ( Fig. 9c; p < 0.0001 for 16 h and 24 h in both wild-type and Kif3a www.nature.com/scientificreports www.nature.com/scientificreports/ null MEFs). Interestingly, the AR promoter activity was insensitive to the hypoosmotic stimulus (200 mOsm/kg H 2 O) in wild-type cells, while it was significantly downregulated in Kif3a null MEFs ( Fig. 9d; p < 0.0001). Both wild-type and Kif3a null MEFs showed a hypoosmotic reduction in HSP70 promoter activity ( Fig. 9e; p = 0.0262 for wild-type, p < 0.0001 for Kif3a null MEFs). A gradual and time-dependent decrease in TauT promoter activity was also observed under hypoosmotic conditions ( Fig. 9f; p < 0.0001).
A qRT-PCR analysis was performed to measure the expression of select TonEBP target genes in MEFs under different osmotic conditions. AR/Akr1b1 was significantly upregulated in both wild type and Kif3a null MEFs under hyperosmotic conditions (600 mOsm/kg H 2 O) ( Fig. 9g; p = 0.0015 for wild-type, p < 0.0001 for null MEFs). In agreement with its promoter activity, AR/Akr1b1 expression was not changed in wild-type MEFs but was significantly decreased in Kif3a null MEFs in response to the hypoosmotic stimulus (200 mOsm/kg H 2 O) ( Fig. 9g; p = 0.021). SMIT/Slc5a3 expression in both wild-type and Kif3a null MEFs was significantly increased and decreased under hyperosmotic and hypoosmotic conditions, respectively ( Fig. 9h; p < 0.0001 for hyperosmotic condition in wild-type and Kif3a null MEFs, p = 0.0008 for hypoosmotic condition in wild-type MEFs, www.nature.com/scientificreports www.nature.com/scientificreports/ p < 0.0001 for hypoosmotic condition in Kif3a null MEFs). Interestingly, unlike in wild-type MEFs, TauT/Slc6a6 mRNA levels were not induced by hyperosmolarity in Kif3a null MEFs ( Fig. 9i; p = 0.0105 in wild-type MEFs). The hypoosmotic decrease in TauT/Slc6a6 levels was maintained in both wild-type and Kif3a null MEFs ( Fig. 9i; p < 0.0001 in wild-type, p = 0.0001 in null MEFs). Taken together, the osmotic regulation of TonEBP activity in MEFs remained largely unchanged in the absence of primary cilia.

Discussion
Primary cilia serve diverse physiological functions that are well-conserved from C. elegans to vertebrates. Their sensory function in various cell types and organ systems is essential in multiple biological processes such as development and maintenance of tissue homeostasis. There are only a few studies examining the presence of primary cilia in the disc tissues 55 . However, few studies have investigated the role of primary cilia in the intervertebral disc 56 . A few studies have suggested that primary cilia can function as osmosensors in mammalian cells [36][37][38] . In rat cholangiocytes or porcine articular chondrocytes, removal of primary cilia inhibits intracellular Ca 2+ increase  www.nature.com/scientificreports www.nature.com/scientificreports/ in response to hypoosmotic stimuli 36,37 . In addition, renal epithelial cells depleted of primary cilia cannot induce TonEBP target genes, AR/Akr1b3, and Bgt1, under hyperosmotic conditions 38 . These cell types are frequently exposed to dynamic changes in extracellular osmolarity, similar to those experienced by the NP cells. Importantly, since NP tissue is physiologically hyperosmotic, osmoadaptation is a key survival mechanism for the resident cells. We, therefore, investigated whether primary cilia in NP cells have an osmosensory function and if the activity of TonEBP, a key osmoregulatory transcription factor, is dependent on primary cilia. We show, for the first time, that NP cell primary cilia do not have the osmoregulatory function, although they alter their lengths in response to changes in extracellular osmolarity.
In many cellular pathways that are mediated by primary cilia, the localization of signaling molecules to primary cilia is important for activating downstream signaling pathways 22,23,57 . Therefore, the changes in the length of primary cilia are thought to be reflective of their functional modification. Indeed, the length of osteocyte primary cilia has been shown to correlate with their sensitivity to the mechanical stimuli 58 . In addition, several known human ciliopathies present with primary cilia of abnormal lengths 59 . In mouse femoral chondrocytes, primary cilia shorten in response to changes in extracellular osmolarity 40 . Likewise, in kidney epithelial cells as well as bone mesenchymal cells, fluid sheer-mediated bending of primary cilia results in decreased intracellular cAMP, which in turn causes shortening of primary cilia 41 . It is, therefore, possible that the changes in the length of primary cilia in NP cells under different osmotic conditions indicate altered intracellular signaling pathways. However, the pathway mediating the length of cilia in response to osmotic stimuli may be completely independent of the TonEBP-mediated osmotic response.
Our Ift88 and Kif3a knockdown studies showed that the expression of TonEBP and its target genes was unaffected by the absence of primary cilia in NP cells. Under hyperosmotic conditions, TonEBP inhibits excessive sodium influx by increasing the levels of intracellular non-ionic osmolytes including taurine, sodium/ myo-inositol, and betaine by inducing transcription of TauT/Slc6a6, Bgt1, SMIT/Slc5a3, and AR/Akr1b1 [15][16][17][18] . Therefore, disruption of TonEBP activity under hyperosmotic condition can compromise cell survival. The fact that both Ift88 and Kif3a knockdown did not affect the hyperosmotic induction of TonEBP and its target genes suggests that this was the result of the absence of primary cilia rather than a specific effect of the knockdown of either gene. Interestingly, despite the decrease in TonEBP protein levels, the expression of its transcriptional targets, TauT/Slc6a6, SMIT/Slc5a3, and AR/Akr1b1 was not significantly affected under hypoosmotic conditions. This may be due to the fact that the osmoregulatory role of TonEBP is more important under hyperosmotic rather than hypoosmotic condition. Nonetheless, neither Ift88 nor Kif3a silencing significantly affected how these genes were regulated in response to hypoosmotic stimuli, strongly suggesting the lack of primary cilia's involvement in this process. In addition, the proximal promoter activities of select TonEBP targets, as well as TonEBP-TAD activity, were mostly unaffected by ShIft88 or ShKif3a regardless of the osmotic status, suggesting that primary cilia play a minimal role, if any, in osmotic response of NP cells.
Our parallel studies using MEFs further confirmed that primary cilia do not control TonEBP-dependent osmotic response. The use of Kif3a null MEFs was advantageous mainly because they completely lack primary cilia, and because MEFs have TonEBP-dependent osmoregulatory pathways similar to that of NP cells 8,54 . Interestingly, unlike NP cells, Kif3a null MEFs were unable to upregulate TauT mRNA in response to the hyperosmotic stimulus. This was intriguing since TauT promoter activity was not affected by the absence of primary cilia in Kif3a null MEFs. This is perhaps due to the high sensitivity of luciferase reporter assay and the fact that it measures the activities of shorter promoter fragments. We cannot exclude the possibility of involvement of either a regulatory sequence outside the analyzed promoter fragment or a primary cilia-associated post-transcriptional regulation. Nonetheless, other osmoresponsive TonEBP targets, SMIT/Slc5a3 and AR/Akr1b1, were upregulated under the hyperosmotic condition in Kif3a null MEFs, indicating that the overall cellular osmotic response to hyperosmotic stimuli is intact. Furthermore, we did not observe any obvious increase in Kif3a null MEF cell death in response to hyper-or hypo-osmotic stimuli, indicating that these cells were able to maintain their cellular homeostasis without primary cilia. The results of our Kif3a null MEFs are particularly interesting because the importance of primary cilia in modulating cell signaling during embryonic development is well-known.
Our data clearly demonstrate that TonEBP-mediated osmoregulation is not controlled by primary cilia in both NP cells and MEFs. The previously reported osmoregulatory and osmosensing function of primary cilia in cholangiocytes 36 and kidney epithelial cells 38 are therefore likely cell-type specific. Although the in vitro culture system is a simplification of the complex in vivo extracellular environment, the osmoregulatory role of primary cilia in other cell types was shown to be preserved in in vitro settings, suggesting that primary cilia are indispensable in these cell types. On the other hand, the results of our studies indicate that NP cells are capable of initiating TonEBP-mediated cellular osmoresponse without primary cilia. A recent study showed that mouse NP tissue cultured in an ex vivo setting decreased their TRPV4 expression in response to cyclic hyperosmolarity 60 . Future studies examining the localization and function of ion channels such as TRPV4, TRM3, or PC-1 and PC-2 that have been associated with cilia osmosensing would be helpful in better understanding how NP cells sense osmotic stimuli.

Materials and Methods
Reagents and plasmids. Lentiviral ShIft88 #1 (TRCN0000178064), ShIft88 #2 (TRCN0000182620), ShKif3a #1 (TRCN0000339512), ShKif3a #2 (TRCN0000339514), and control shRNA plasmids were purchased from Sigma. psPAX2 (12260) and pMD2.G (12259) developed by Dr. Didier Trono were obtained from Addgene. Luciferase reporter plasmids were kindly provided by Dr. Takashi Ito, Osaka University (wild type and TonEmutant TauT-Luc) 15  www.nature.com/scientificreports www.nature.com/scientificreports/ Cell culture and treatments. All procedures regarding collection of animal tissues was performed as per approved protocols by Institutional Animal Care and Use Committee (IACUC) of the Thomas Jefferson University, in accordance with the IACUC's relevant guidelines and regulations. Rat NP cells were isolated using a method described by Risbud et al. 61 . After isolation, cells were maintained in Dulbecco's Modified Eagles Medium (DMEM) (Corning, 10-013-CV) with 10% fetal bovine serum (FBS) (Sigma-Aldrich, F6178) supplemented with antibiotics in flask in normoxia (20.9% pO 2 ). NP cells isolated from two rats were pulled into one T25 flask (P0) and cultured till confluency before passaging (P1). From this point, cells were either plated directly into experiment-specific plates (P2) or passaged. Approximately total of 16  Real-time RT-PCR analysis. Total RNA was extracted from NP cells using RNAeasy mini columns (Qiagen). The purified, DNA-free RNA was converted to cDNA using RNA to cDNA EcoDry TM Premix (Clontech). Template cDNA and gene specific primers were added to SYBR green master mix (Applied Biosystems) and mRNA expression was quantified using the Step One Plus Real-Time PCR System (Applied Biosystems). Hprt1 was used to normalize gene expression. All the primers used were synthesized by Integrated DNA Technologies, Inc. (Coralville, IA).
Transfections and dual luciferase assay. Cells were plated on 48-well plates (2 × 10 4 cells/well) one day before transfection. Cells were transfected with 250 ng of wild type TauT, mutant TauT, wild type HSP70, mutant HSP70, or AR reporter plasmid and 250 ng of pRL-TK plasmid. For measuring TonEBP-TAD activity, cells were transfected with 200 ng of pFR-Luc, 200 ng of GAL4dbd-548-1531 (Ton-TAD) and 100 ng of pRL-TK plasmids. For each transfection, plasmids were premixed with the transfection reagent Lipofectamine 2000 (Invitrogen). The treatments were done so that the cells were lysed 48 h after the transfection. Dual-Luciferase TM reporter assay system (Promega) was used for sequential measurements of firefly and Renilla luciferase activities using TECAN Infinite200 Pro microplate reader (TECAN). At least three independent transfections were performed and all analyses were carried out in triplicate.
Immunofluorescence microscopy. NP cells were plated on poly-L-lysine-coated glass coverslips. After treatments, cells were fixed and permeabilized with 4% paraformaldehyde at room temperature for 15 minutes, washed with PBS and then blocked with 5% normal goat serum in PBS with 0.3% Triton X-100 (Sigma Aldrich, T8787) for 1 h at room temperature. Cells on coverslip were then incubated with anti-acetyl-α-tubulin (Lys40) (D20G3) XP ® antibody (Cell Signaling Technology, #5335, 1:700), or anti-acetylated-α-tubulin (Sigma, T6793, 1:300) and anti-γ-tubulin antibody (Abcam, ab11317, 1:1000) for co-staining in blocking buffer at 4 °C overnight, washed with PBS, and then incubated with Alexa Fluor ® -594 conjugated anti-rabbit, or Alexa Fluor ® -594 conjugated anti-mouse, or Alexa Fluor ® -488 conjugated anti-rabbit secondary antibody (Jackson ImmunoResearch Lab, Inc.), at a dilution of 1:700 for 1 h at room temperature in dark. Then the coverslips were washed with PBS and mounted with ProLong ® Gold Antifade Mountant with DAPI (Thermo Fisher Scientific, P36934). Mounted slides were visualized using a Zeiss AxioImager A2 (Carl Zeiss, Germany), or Zeiss LSM510 confocal microscope (Carl Zeiss, Germany). Three independent experiments were performed for quantitative analysis of cilia length and incidence using ImageJ software (http://rsb.info.nih.gov/ij/). Briefly, each cilium was traced using Segmented Lines of the Line Selection Tools, and then the length of the tracing was measured using Measure function. The lengths of at least 150 primary cilia per treatment group were measured. Protein extraction and western blotting. Following treatment, cells were immediately placed on ice and washed with ice-cold PBS. All the wash buffers and the final cell lysis/re-suspension buffers included 1X complete TM Mini Protease Inhibitor Cocktail (Roche, 11836153001), NaF (5 mM) (Sigma Aldrich, 201154) and Na 3 VO 4 (200 μM) (Sigma Aldrich, S6508). Total cell proteins were resolved by electrophoresis on 8-12% SDS-polyacrylamide gels and transferred by electroblotting to PVDF membranes (EMD Millipore, IPVH00010). Dual-color Lane marker (Bio-Rad, 161-0394) was used as a protein ladder to guide identification of the band sizes. The membranes were blocked with 5% non-fat dry milk in TBST (1% Tween 20 in TBS) and incubated overnight at 4 °C in 5% non-fat dry milk in TBST with the antibodies against IFT88 (Proteintech, #13967-1-AP, 1:500), KIF3A (D7G3) (Cell Signaling Technology, #8507, 1:1000), TonEBP/NFAT5 (Novus Biologicals, NB120-3446, 1:1000), or GAPDH (Novus Biologicals, NB300-221, 1:3000). Immunolabeling was detected using the Amersham TM ECL TM Prime Western Blotting Detection Reagent (Thermo Fisher Scientific, 45-002-401). All Western blot experiments were performed at least three independent times. Lentiviral particle production and viral transduction. HEK 293 T cells (ATCC, CRL-3216) were plated in 10 cm plates (5 × 10 6 cells/plate) in DMEM with 10% heat-inactivated FBS one day before transfection. Cells were transfected with 9 μg of ShCtr, ShIft88, or ShKif3a plasmids along with 6 μg psPAX2 and 3 μg pMD2.G using Lipofectamine 2000 (Invitrogen). After 6 h, transfection medium was replaced with DMEM with 10% heat-inactivated FBS and penicillin-streptomycin. Lentiviral medium was harvested at 48 to 60 h (2019) 9:15469 | https://doi.org/10.1038/s41598-019-51939-7 www.nature.com/scientificreports www.nature.com/scientificreports/ post-transfection, and mixed with 7% PEG 6000 (Sigma Aldrich, 81253) solution and incubated overnight at 4 °C to precipitate virus particles. PEG solution was removed from virus medium before transduction by centrifugation at 1,500 × g for 30 min to pellet virus particles. NP cells were plated in DMEM with 10% heat-inactivated FBS one day before transduction. Cells were transduced with fresh DMEM with 10% heat-inactivated FBS containing viral particles along with 8 μg/ml polybrene (Sigma Aldrich, H9268). After 16 h, the medium was removed and replaced with DMEM with 10% FBS. Cells were harvested for protein extraction 4-5 days after transduction to ensure maximum knockdown efficiency without affecting cell viability. At least three independent experiments were performed. Statistical analysis. All experiments were performed at least three independent times. Data are presented as the mean ± SE. Differences between multiple groups were assessed by one-way ANOVA or Kruskal-Wallis test depending on the distribution of the data with Sidak's, Holm-Sidak's, or Dunn's multiple comparison test for post-hoc analyses using Prism7 (GraphPad Software). Some data were log-transformed before statistical analyses to account for non-Gaussian distribution. P < 0.05 was considered statistically significant.