Reciprocal regulation of cilia and autophagy via the MTOR and proteasome pathways

Abstract

Primary cilium is an organelle that plays significant roles in a number of cellular functions ranging from cell mechanosensation, proliferation, and differentiation to apoptosis. Autophagy is an evolutionarily conserved cellular function in biology and indispensable for cellular homeostasis. Both cilia and autophagy have been linked to different types of genetic and acquired human diseases. Their interaction has been suggested very recently, but the underlying mechanisms are still not fully understood. We examined autophagy in cells with suppressed cilia and measured cilium length in autophagy-activated or -suppressed cells. It was found that autophagy was repressed in cells with short cilia. Further investigation showed that MTOR activation was enhanced in cilia-suppressed cells and the MTOR inhibitor rapamycin could largely reverse autophagy suppression. In human kidney proximal tubular cells (HK2), autophagy induction was associated with cilium elongation. Conversely, autophagy inhibition by 3-methyladenine (3-MA) and chloroquine (CQ) as well as bafilomycin A1 (Baf) led to short cilia. Cilia were also shorter in cultured atg5-knockout (KO) cells and in atg7-KO kidney proximal tubular cells in mice. MG132, an inhibitor of the proteasome, could significantly restore cilium length in atg5-KO cells, being concomitant with the proteasome activity. Together, the results suggest that cilia and autophagy regulate reciprocally through the MTOR signaling pathway and ubiquitin-proteasome system.

Keywords: 3-MA, 3-methyladenine; 70kDa, polypeptide 1; ANKS6, ankyrin repeat and sterile α motif domain containing 6; ATG/atg, autophagy-related; Ac-TUBA, acetylated-tubulin α; Baf, bafilomycin A1; CF, confluence; CQ, chloroquine; DAPI, 4′-6-diamidino-2-phenylindole; FBS, fetal bovine serum; HK2, human kidney proximal tubular cells; IFT, intraflagellar transport; KAP3, kinesin family-associated protein 3; KD, knockdown; KIF3A/3B, kinesin family member 3A/3B; KO, knockout; LTA, lotus tetragonolobus agglutinin; MAP1LC3B/LC3B, microtubule-associated protein 1 light chain 3 β; MEF, mouse embryonic fibroblast; MTOR; MTOR, mechanistic target of rapamycin; OFD1, oral-ficial-digital syndrome 1; PBS, phosphate-buffered saline; PKD, polycystic kidney disease; RKRB, Krebs-Henseleit saline containing 25 mM NaHCO3; RPS6KB1, ribosomal protein S6 kinase; Rapa, rapamycin; SD, standard deviation; autophagy; cilia; polycystic kidney disease; proteasome.

Figure 1.

Association of cilium length and autophagy from cell growth to differentiation. To determine the association of cilium length and autophagy, HK2 cells were cultured in DMEM-F12 medium containing 10% FBS at subconfluence (Sub-CF), confluence (CF), or for one or 3 d after reaching confluence (Post-CF d 1, 3). (A) Cells were assayed by immunofluorescence staining with the antibody to Ac-TUBA (red) and DAPI (blue) and cell lysates examined by immunoblot analysis of LC3B and ACTB. Scale bar: 5 5 μ. (B) Cilium length was averaged out of over 50 cilia in each group (Sub-CF, n=52; CF, =56; Post-CF d 1, =51; Post-CF d 3, =60). Cilium length: Post-CF d 3>Post-CF d 1>CF>Sub-CF, * P < 0.05. (C) Expression level of KIF3A and IFT88 from cell growth to differentiation. Please note, as previously reported, 2 forms of IFT88 were detected. (D) Positive correlation of cilium length with LC3B-II/-I or LC3B-II/ACTB ratio.

Figure 2.

Shortening of cilia suppresses autophagy through the MTOR pathway. (A, B) Inhibition of autophagy in both IFT88-KD2 and C13 cilia-short cells. To assess the effect of cilium length on autophagy, cilia-suppressed cells and their controls (Control, C4) were incubated with FBS-containing culture medium or serum-free RKRB medium without IFT88 or with 25 μM chloroquine (CQ) for 3 h. Cell lysates were collected for immunoblot analysis of LC3B and ACTB. Protein bands were quantified by densitometry. (C) MTOR hyperphosphorylation in IFT88-KD2 and C13 cells. To define MTOR activity, cilia-short cells were cultured in serum-free DMEM-F12 medium. Cell lysates were collected for immunoblot analysis of p-MTOR, Ser2448; p-RPS6KB1, Thr389; p-AKT1, Ser473, and ACTB. Densitometry was conducted to determine the p-RPS6KB1, Thr389/ACTB ratio. (D) Restoration of autophagy by rapamycin (Rapa) in cilia-suppressed cells. GFP-LC3 was transfected into IFT88-KD2 and C13 cilia-short, and their control cells for the analysis of autophagosome puncta. Fewer puncta per cell were seen in IFT88-KD2 cells (n=36) than control cells (n=33). After 24 h treatment of rapamycin (50 nM), the number of puncta per cell increased in both control (n=55) and IFT88-KD2 cells (n=53), and no difference was detected between them. Similar results were shown for cilia-short C13 cells and their control C4 cells. Fifty-two and 53 positive cells, without rapamycin treatment, were counted from C4 and C13 cells; 52 and 54 positive cells, with rapamycin treatment, were counted from C4 and C13 cells. All the experiments were performed 3 times. P*<0.05, **<0.01, ***<0.001 IFT88-KD2 and C13 vs Control and C4 cells; #<0.05, ##<0.01 +rapamycin vs -rapamycin. Scale bar: 10 μm.

Figure 3.

Autophagy activation increases cilium length. To determine the effect of autophagy activation on cilium length, 3 different approaches for autophagy induction were used. (A) Increase in cilium length during incubation in serum/glucose-free RKRB medium. HK2 cells were incubated for 3 h in full culture medium (medium) or plain RKRB medium, and then subjected to immunofluorescence analysis of Ac-TUBA or immunoblot analysis of the indicated proteins. Forty-five and 52 cilia were measured, respectively, in culture medium and RKRB-incubated HK2 cells. Densitometry was conducted to determine the IFT88/ACTB ratio. (B) Increase in cilium length induced by trehalose. HK2 cells were incubated in the absence (N/A: no addition) or presence of 100 mM trehalose for 24 h, followed by immunofluorescence analysis of Ac-TUBA or immunoblot analysis of the indicated proteins. Seventy-four and 82 cilia were counted for N/A and trehalose-treated cells. Densitometry was conducted to determine the KIF3A/ACTB ratio. (C) Increase in cilium length induced by BECN1 peptide. HK2 cells were incubated with a scrambled sequence or the BECN1 peptide for 4 h, followed by immunofluorescence analysis of Ac-TUBA or immunoblot analysis of the indicated proteins. Fifty one and 66 cilia were counted for control and BECN1 peptide-treated cells. Densitometry was conducted to determine the IFT88/ACTB ratio. In (A, B, and C), autophagy was confirmed by immunoblotting of LC3B and SQSTM1. The results were the summary of 3 separate experiments. P*<0.05, **<0.01, ***<0.001. Scale bar: 5μm ((A)and B) and 10 μm (C).

Figure 4.

Autophagy suppression decreases cilium length. To study the effect of autophagy suppression on cilium length, we used 3 different inhibitors of autophagy. (A) Suppression of cilia by 3-MA. HK2 cells were treated for 2 h with 3-MA and then subjected to immunofluorescence analysis of Ac-TUBA or immunoblot analysis of the indicated proteins. Sixty and 52 cilia were measured for control and 3-MA-treated cells. Densitometry was conducted to determine the KIF3A/ACTB and IFT88/ACTB ratios. (B) Suppression of cilia by chloroquine (CQ). HK2 cells were treated for 2 h with CQ and then subjected to immunofluorescence analysis of Ac-TUBA or immunoblot analysis of the indicated proteins. Fifty-four and 45 cilia were counted for control and CQ-treated cells. Densitometry was conducted to determine the IFT88/ACTB ratio. (C) Suppression of cilia by bafilomycin A₁ (Baf) in a time course. HK2 cells were treated with Baf for up to 24 h in a full culture medium and then subjected to immunofluorescence analysis of Ac-TUBA. Autophagy inhibition by 3-MA and CQ and Baf was confirmed by immunoblotting of LC3B and SQSTM1. The results are the summary of 3 separate experiments. P*<0.05, ***<0.001 vs Control. Scale bar: 5 μm ((A) and B) and 10 μm (C).

Figure 5.

Decrease in cilium length and frequency in proximal tubule-atg7-KO cells. To confirm our in vitro findings, we determined to evaluate cilium length and frequency in proximal tubule-atg7-KO mice. Kidney tissue sections from proximal tubule-atg7-KO mice and wild-type littermate mice were stained for Ac-TUBA and LTA, a proximal tubule marker. One hundred forty seven and 101 cilia were counted for LTA-positive wild-type and atg7-KO cells, respectively. Five different microscope fields of view were used for cilium frequency calculation (cilium number/nucleus number). P*<0.05. Scale bar: 5 μm.

Figure 6.

Proteasome in cilia shortening in autophagy-suppressed cells. (A) Restoration of cilium length by MG132 in 3-MA-treated cells. HK2 cells were treated for 2 h with 3-MA in the absence or presence of 5 μM MG132, followed by immunofluo-rescence analysis of cilia by Ac-TUBA. Thirty six, 35, and 38 cilia were measured for the control, 3-MA-treated, 3-MA+MG132-treated cells respectively. (B) Effect of MG132 on cilium length in atg5-KO cells and induction of KIF3A by MG132. Atg5^+/+ and atg5^−/− cells were incubated without or with MG132 for 2 h, followed by immunofluorescence analysis of Ac-TUBA. Thirty three, 48, 46, and 50 cilia were measured for the wild-type and KO cells at the basal condition or with MG132 treatment. atg5-KO cells had shorter cilia at the basal condition. Cell lysates were collected in (B) for immunoblot analysis. Densitometry was conducted to determine the KIF3A/ACTB ratio. The results are the summary of 3 separate experiments. (C) Proteasome activity increment in autophagy-suppressed cells. Compared to Atg5^+/+ cells, atg5^−/− exhibited enhanced proteasome activity. MG132 significantly suppressed the proteasome activity in HK2 and Atg5^+/+ and atg5^−/− MEF cells. Compared to culture medium, RKRB medium reduced proteasome activity in HK2 cells. Relative fluorescent units (RFU) were used to indicate the proteasome activity in different culture conditions. Results are the summary of 4 separate experiments. P*<0.05, ***<0.001 vs RKRB (A) or Medium (C), atg5^−/− vs Atg5^+/+ (B); #<0.05, ##<0.01, ###<0.001 +MG132 vs -MG132 (A, B, and C). Scale bar (A and B): 5 μm.

Figure 7.

Our working model. Reciprocal interaction between cilia and autophagy. (A) Shortening of cilia activates MTOR, which subsequently inhibits autophagy. (B) Autophagy inhibition or stimulation increases or decreases the proteasome activity, which subsequently affects cilium length regulators. At certain conditions, autophagy may directly regulate cilium- length regulators such as IFT88 in BECN1 peptide-induced cilium elongation. The integrative balance of cilia-suppressing and -elongating factors determines the cilium length.