Primary cilia sense glutamine availability and respond via asparagine synthetase

Abstract

Depriving cells of nutrients triggers an energetic crisis, which is resolved by metabolic rewiring and organelle reorganization. Primary cilia are microtubule-based organelles at the cell surface, capable of integrating multiple metabolic and signalling cues, but their precise sensory function is not fully understood. Here we show that primary cilia respond to nutrient availability and adjust their length via glutamine-mediated anaplerosis facilitated by asparagine synthetase (ASNS). Nutrient deprivation causes cilia elongation, mediated by reduced mitochondrial function, ATP availability and AMPK activation independently of mTORC1. Of note, glutamine removal and replenishment is necessary and sufficient to induce ciliary elongation or retraction, respectively, under nutrient stress conditions both in vivo and in vitro by restoring mitochondrial anaplerosis via ASNS-dependent glutamate generation. Ift88-mutant cells lacking cilia show reduced glutamine-dependent mitochondrial anaplerosis during metabolic stress, due to reduced expression and activity of ASNS at the base of cilia. Our data indicate a role for cilia in responding to, and possibly sensing, cellular glutamine levels via ASNS during metabolic stress.

Fig. 1. Primary cilia sense nutrient availability.

a, Representative fluorescence images of MEFs stably expressing ARL13B-GFP, and IF images of hRPE, mIMCD3 and MDCK cells in the indicated culture conditions. Cilia (ARL13B, green), nuclei (DAPI, blue). Scale bar, 5 µm. b, Top: quantification of cilia length in one representative experiment in the indicated cell lines in nutrient-rich (DMEM or DMEM/F12 ± 10% FBS) or deprived (HBSS) medium. n indicates number of cilia whose length was measured in the same representative experiment in the indicated cell line. Bottom: percentage (%) of ciliated cells of one representative experiment in the same conditions. n indicates cilia percentage in three different wells per condition of the same representative experiment. c, Experimental design of NMR spectroscopy on MEF^Ctrl (CT) and MEF^Ift88 (KO) conditioned medium after 24 h culture in DMEM + 0% FBS. d, Hierarchical clustering of extracellular metabolites assessed by NMR spectroscopy as in c. e, Box-and-whisker plots of the levels of glutamine and glucose uptake, glutamate and lactate production in MEF^Ctrl (CT) and MEF^Ift88 (KO) cells assessed by NMR spectroscopy as in c, n = 5 biological replicates. f, Quantification of cilia length in one representative experiment in MEFs and hRPE in HBSS ± d-(+)-glucose (Glc) (20 mM) or l-glutamine (Q) (4 mM). n indicates cilia length measured in the same representative experiment. g, Schematic representation of glutamine utilization for OXPHOS. h, Left: analysis of OCR measurement of one representative experiment in MEFs after 4 h culture in HBSS ± Glc (20 mM) or Q (4 mM) in basal condition and after sequential addition of oligomycin (O), FCCP and antimycin/rotenone (A/R). Right: quantification of ATP-production-coupled respiration as in left. n indicates OCR measured in different wells of the same representative experiment. Box-and-whisker plots show median and minimum to maximum; data in dot and bar plots are mean ± standard deviation. Statistical analysis: Student’s unpaired two-tailed t-test or one-way ANOVA, followed by Tukey’s multiple comparisons test; NS, not significant, ****P < 0.0001. n reported in brackets. Additional replicate experiments in b, f and h are shown in Supplementary Fig. 1. Source data

Fig. 2. Primary cilia respond to glutamine in vitro and in vivo.

a, Quantification of cilia length in one representative experiment in mIMCD3 after 24 h culture in 0% FBS and HBSS ± Q (0.2, 0.5, 1, 2 and 4 mM). b, Quantification of cilia length in one representative experiment in mIMCD3 cultured for either 8 or 24 h in 0% FBS and for 4, 6, 8 and 24 h in HBSS. c, Quantification of cilia length in one representative experiment in mIMCD3 cultured for 24 h in either DMEM/F12 + 0% FBS or HBSS and replenished with Q for 4, 6 and 24 h. d, Quantification of cilia length in one representative experiment in mIMCD3 after 24 h culture in DMEM/F12 + 0% FBS ± Q (0.5, 1, 2 and 4 mM). e, Experimental design of mice treatment: Crl (fed), fasted (F) for 48 or 24 h, fasted for 24 h + 24 h treatment with 800 mg kg⁻¹ of Q. Sac = Sacrifice. f, Left: distribution of the individual cilia length in kidney sections of one representative experiment of mice treated as in e. Right: average ciliary length calculated from three independent experiments as in e. g, Representative IF images out of three independent experiments of kidney sections of mice Crl (fed), fasted 48 h, and fasted 24 h + Q 24 h. Cilia (ARL13B, green), DBA⁺ tubular cells (DBA, red), nuclei (DAPI, blue). Scale bar, 5 µm. h, Average glycaemic values from three independent experiments in mice treated as in e. i, Left: experimental design of mice treatment: Crl (fed), fasted for 24 h, and fasted for 24 h + treatment with 800 mg kg⁻¹ of Q. Right: average of Q concentration in the serum of mice treated as in left from four independent experiments. j, Average of Q concentration normalized on the kidney weight in mice treated as in e from three independent experiments. Data in dot and bar plots are mean ± standard deviation. Statistical analysis: one-way ANOVA, followed by Tukey’s multiple comparisons test; NS, not significant, ****P < 0.0001. n reported in brackets. Additional replicate experiments in a–d are shown in Supplementary Fig. 3. Source data

Fig. 3. Defective response to glutamine under stress conditions in cilia-ablated cells.

a, Experimental design of LC–MS targeted metabolomics in MEF^Ctrl (CT) and MEF^Ift88 (KO) after 24 h culture in HBSS ± l-glutamine (Q) (4 mM). b, Volcano plot of metabolites in KO versus CT cultured in HBSS for 24 h as assessed by LC–MS targeted metabolomics as in a. FC, fold change. Black dots: P ≥ 0.05 and −0.5 ≤ log₂FC ≤ 0.5; grey dots: P ≥ 0.05 or −0.5 ≤ log₂FC ≤ 0.5; red dots: P < 0.05 and −0.5 > log₂FC > 0.5. c, Box-and-whisker plots of ATP/AMP, GTP/GMP and CTP/CMP ratio in MEF^Ctrl and MEF^Ift88 cells in HBSS for 24 h assessed by LC–MS targeted metabolomics as in a, n = 5 biological replicates. d, Top: Venn diagram showing the metabolites that significantly change in KO versus CT in HBSS and HBSS + Q. Light blue: 49 metabolites that change in KO versus CT only in HBSS. Rose: 14 metabolites that change in KO versus CT only in HBSS + Q. Purple: 37 metabolites that change in KO versus CT in HBSS and HBSS + Q. Bottom: Heat map showing the FC of HBSS + Q versus HBSS, in CT and KO, of the statistically different 51 metabolites. The 14 metabolites that change in KO versus CT in HBSS + Q are expressed in bold. Inf = Infinite. e, Left: analysis of OCR measurement of one representative experiment in MEF^Ctrl (CT) and MEF^Ift88 (KO) after 4 h culture in either HBSS (CT, blue; KO, green) or HBSS + l-glutamine (Q) (4 mM) (CT, red; KO, purple) in basal condition and after sequential addition of oligomycin (O), FCCP and antimycin A/rotenone (A/R). Right: quantification of basal respiration, ATP-production-coupled respiration and maximal respiration as in left. Box-and-whisker plots show median and minimum to maximum; data in bar plots are mean ± standard deviation. Statistical analysis: Student’s unpaired two-tailed t-test or one-way ANOVA, followed by Tukey’s multiple comparisons test; NS, not significant, ****P < 0.0001. n reported in brackets. Additional replicate experiment in e is shown in Supplementary Fig. 5. Source data

Fig. 4. The glutamine response of cilia is mediated by the enzyme ASNS.

a, Box-and-whisker plots of the levels of intracellular glutamine and asparagine in MEF^Ctrl and MEF^Ift88 cells in HBSS ± l-glutamine (Q) assessed by LC–MS targeted metabolomics as in Fig. 3a, n = 5 biological replicates. b, Western blot for ASNS and pS6RP^S235/236 of total cell lysates from mIMCD3 transiently knocked down for Asns (siAsns) compared with control (Scr) after 24 h in either 0% FBS or HBSS ± Q (4 mM). c, Average cilia length values from six independent experiments (in each counting >50 cilia per condition) in siAsns mIMCD3 compared with Scr after 24 h culture in either DMEM/F12 + 0% FBS or HBSS ± Q. d, Quantification of cilia length of one representative experiment in mIMCD3 after 24 h in either 0% FBS ± Asnase (5 U ml⁻¹), HBSS ± Q, or Q and Asnase, or asparagine (N) (0.1 mM). e, Left: analysis of OCR measurement of one representative experiment in siAsns mIMCD3 compared with Scr after 4 h culture in HBSS (Scr, blue; siAsns, red) followed by acute injection (+) of Q (Scr, green; siAsns, purple), in basal condition and after sequential addition of oligomycin (O), FCCP and antimycin A/rotenone (A/R). Right: quantification of acute response as in left. f, Scheme of ¹⁵N₂-glutamine usage by ASNS. g, Levels of intracellular glutamine m + 2 and asparagine m + 1 release assessed by ¹⁵N₂-glutamine labelled LC–MS targeted metabolomics in MEF^Ctrl (CT) and MEF^Ift88 (KO) at 4 and 24 h. n = 5 biological replicates. h, Top: scheme of cilia enrichment by subcellular fractionation. Bottom: representative western blot for TSC1, IFT88, ASNS and S6RP of fraction 1 (Fr1: cytoplasmic), fraction 2 (Fr2: organelles), fraction 3 (Fr3: cilia) and total extract (TE) from mIMCD3 as in top. i, Fluorescence live imaging of eGFP-ASNS transiently transfected mIMCD3 cells in 0% FBS. eGFP-ASNS (green, arrows), nuclei (Hoechst, blue). Scale bar, 10 µm. Insets show magnification. Scale bar, 5 µm. j, Representative IF images of eGFP-ASNS transiently transfected mIMCD3 cells in 0% FBS. eGFP-ASNS (green), cilia (ARL13B, red), nuclei (DAPI, blue). Scale bar, 10 µm. Arrows indicate eGFP-ASNS at the base of cilia. Insets are magnifications (merged and single channels). Scale bar, 5 µm. k, Representative IF images of mNeon-ASNS stable mIMCD3 lines in HBSS (8 h). mNeon-ASNS (green), centrosomes (γ-tubulin, red), cilia (ARL13B, white), nuclei (DAPI, blue). Scale bar, 10 µm. Arrows indicate co-localization with centrosomes. Insets are magnifications (merged and single channels). Scale bar, 5 µm. Box-and-whisker plots show median and minimum to maximum; data in dot and bar plots are mean ± standard deviation. Average plot is mean ± standard error of the mean. Statistical analysis: one-way ANOVA, followed by Tukey’s (a, d, e, g) or Bonferroni’s (c) multiple comparisons test; NS, not significant, ****P < 0.0001. n reported in brackets. Additional replicate experiments in d, e and h are shown in Supplementary Fig. 7. Source data

Extended Data Fig. 1. Referred to Fig. 1.

a, Left: western blot for IFT88 of total cell lysates from MEF^Ift88 and MEF^Ctrl generated by CRISPR/Cas9 technology. Right: Representative IF images of MEF^Ift88 and MEF^Ctrl. Cilia (ARL13B, green), nuclei (DAPI, blue). Scale bar: 10 µm. Arrows indicate cells in the insets. b, Growth curves in MEF^Ift88 and MEF^Ctrl cultured in DMEM + 10% FBS for 8, 24, 48 hours. c, Analysis of OCR measurement in one representative experiment in MEF^Ift88 and MEF^Ctrl cultured in DMEM + 10% FBS in basal condition and after sequential addition of oligomycin (O), FCCP, and antimycin A/rotenone (A/R). d, Left: western blot for IFT88 of total cell lysates from mIMCD3^Ift88 and mIMCD3^Ctrl generated by CRISPR/Cas9 technology. Right: Representative IF images of mIMCD3^Ift88 and mIMCD3^Ctrl. Cilia (ARL13B, green), nuclei (DAPI, blue). Scale bar: 10 µm. Arrows indicate cells in the insets. e, Analysis of OCR measurement in one representative experiment in mIMCD3^Ift88 and mIMCD3^Ctrl cultured in DMEM/F12 + 10% FBS in basal condition and after sequential addition of O, FCCP, and A/R. f, Left: analysis of OCR measurement in one representative experiment in MEF^Ift88 and MEF^Ctrl cultured in DMEM + 0% FBS in basal condition and after sequential addition of O, FCCP, and A/R. Right: analysis of OCR measurement in one representative experiment in mIMCD3^Ift88 and mIMCD3^Ctrl cultured in DMEM/F12 + 0% FBS in basal condition and after sequential addition of O, FCCP, and A/R. g, Principal Component analysis (PCA) of extracellular metabolites assessed by NMR spectroscopy as in Fig. 1c. h, Hierarchical clustering of extracellular metabolites assessed by NMR spectroscopy as in Fig. 1c, including metabolite concentrations in unconditioned medium (MO) and media conditioned by MEF^Ift88 (KO) and MEF^Ctrl (CT) cells. Data in dot and bar plots are mean ± SD. Source data

Extended Data Fig. 2. Referred to Fig. 1.

a, Quantification of cilia length in one representative experiment in mIMCD3 and MDCK after 24 hours culture in HBSS ± L-Glutamine (Q) (4 mM). Shared control (HBSS) with Fig. 1b. b, Left: analysis of OCR measurement in one representative experiment in MEF after 16 hours culture in either HBSS (blue) or HBSS + D-(+)-Glucose (Glc) (20 mM) (green) or Q (4 mM) (red) in basal condition and after sequential addition of O, FCCP, and A/R. Right: quantification of basal respiration, ATP-production coupled respiration, and maximal respiration as in Left. c, Left: analysis of OCR measurement in one representative experiment in MEF after 4 hours culture in HBSS (blue) followed by acute injection (+) of either Glc (green) or Q (red) and after sequential addition of O, FCCP and A/R. Right: quantification of basal respiration, acute response, ATP-production coupled respiration, and maximal respiration as in Left. d, Left: analysis of ECAR measurement referred to Fig. 1h. Right: analysis of ECAR measurement referred to b. Data in dot and bar plots are mean ± SD. Statistical analysis: Student’s unpaired two-tailed t-test or one-way ANOVA, followed by Tukey’s multiple comparisons test; ns: not significant, ****p < 0.0001. n reported in brackets. Source data

Extended Data Fig. 3. Referred to Fig. 1.

a, Scheme of Mitochondrial Respiratory Chain Complexes targets of rotenone, antimycin, and oligomycin. b, Quantification of cilia length in one representative experiment in MEF and hRPE after 24 hours culture in DMEM or DMEM/F12 + 0% FBS ± either A/R or O. c, Western blot for pAMPK^T172 of total cell lysates from MEF cultured for 1, 2, 4, 6 hours in HBSS or for 4 hours in either DMEM + 10% FBS ± AICAR (1 mM) or Metformin (2 mM). Vinculin is used as loading control. d, Representative IF images of one representative experiment after 24 hours culture in either DMEM/F12 + 10% FBS ± vehicle (veh) or AICAR. Cilia (ARL13B, green), nuclei (DAPI, blue). Scale bar: 5 µm. e, Left: Quantification of cilia length in one representative experiment in hRPE after 24 hours culture in the indicated conditions. Right: % of ciliated cells in one representative experiment in hRPE after 24 hours culture in the indicated conditions. f, Left: Quantification of cilia length in one representative experiment in mIMCD3 after 24 hours culture in the indicated conditions. Right: % of ciliated cells in one representative experiment in mIMCD3 after 24 hours culture in the indicated conditions. Data in dot and bar plots are mean ± SD. Statistical analysis: Student’s unpaired two-tailed t-test or one-way ANOVA, followed by Tukey’s multiple comparisons test; ns: not significant, ****p < 0.0001. n reported in brackets. Source data

Extended Data Fig. 4. Referred to Fig. 2.

a, Left: experimental design of mice treatment: Crl (Fed), Fasted for 24 hours. Center: quantification of cilia length in one representative experiment in kidney sections of mice treated as in Left. Right: average from 3 independent experiments of glycaemia values in mice treated as in Left. b, Left: experimental design of mice treatment: Crl (Fed), Fasted for 48 hours. Center: quantification of cilia length in one representative experiment in kidney sections of mice treated as in Left. Right: average from 3 independent experiments of glycaemia values in mice treated as in Left. c, Left: experimental design of mice treatment: Crl (Fed), Fasted for 48 hours. Right: average from 3 independent experiments of Q concentration in the serum of mice treated as in Left. d, Scheme of renal tubular segment with Cre-recombinase driven inactivation of Opa1 gene. e, Left: Transmission Electron Microscopy (TEM) images of kidney sections of control (Crl) and Opa1 KO mice at P2. Scale bar: 1 µm. Dashed Dots indicate mitochondria. Right: Enzymatic activity staining for Succinate Dehydrogenase (SDH) and Cytochrome c Oxidase (COX) of kidney sections of Crl and Opa1 KO mice at P30 with zoom in. Dashed dots indicate medulla edge. Scale bar: 500 µm. f, Left: representative IF images of kidney sections of Crl and Opa1 KO mice at P30. Right: quantification of cilia length in one representative experiment of DBA + tubular cells in kidney sections of Crl and Opa1 KO mice at P30. Cilia (ARL13B, green), DBA + tubular cells (DBA, red) nuclei (DAPI, blue). Scale bar: 5 µm. Arrows indicate cilia in the insets. Data in dot and bar plots are mean ± SD. Statistical analysis: Student’s unpaired two-tailed t-test; ns: not significant, ****p < 0.0001. n reported in brackets. Source data

Extended Data Fig. 5. Referred to Fig. 3.

a, Heat-map showing the fold-change of HBSS + L-Glutamine (Q) vs HBSS, in MEF^Ift88 and MEF^Ctrl, of the 49 metabolites that change in a statistically significant way in the comparison between KO vs CT in HBSS. b, Left: analysis of OCR measurement in one representative experiment in MEF^Ift88 and MEF^Ctrl after 24 hours culture in either HBSS (CT: blue, KO: green) or HBSS + Q (4 mM) (CT: red, KO: purple) in basal condition and after sequential addition of oligomycin (O), FCCP, and antimycin A/rotenone (A/R). Right: quantification of basal respiration, ATP-production coupled respiration, and maximal respiration as in Left. c, Left: analysis of OCR measurement in one representative experiment in mIMCD3^Ift88 and mIMCD3^Ctrl cells after 4 hours culture in either HBSS (CT: blue, KO: green) or HBSS + Q (4 mM) (CT: red, KO: purple), in basal condition and after sequential addition of O, FCCP, and A/R. Right: quantification of basal respiration, ATP-production coupled respiration, and maximal respiration as in Left Data bar plots are mean ± SD. Statistical analysis: one-way ANOVA, followed by Tukey’s multiple comparisons test; ns: not significant, ****p < 0.0001. n reported in brackets. Source data

Extended Data Fig. 6. Referred to Fig. 4.

a, Western blot for pS6RP^S235/236 of total cell lysates from mIMCD3 cells after 24 hours culture in either DMEM/F12 ± 10% FBS or HBSS ± L-Glutamine (Q) (4 mM). Vinculin is used as loading control. b, Quantification of cilia length in one representative experiment in mIMCD3 cells after 24 hours culture in either HBSS ± Q or HBSS + Q + Rapamycin (100 nM). c, Western blot for pS6RP^S235/236 of total cell lysates from mIMCD3 cells after 24 hours culture in either HBSS ± Q or L-Leucine (L) (0.5 mM). d, Quantification of cilia length in one representative experiment in mIMCD3 cells after 24 hours culture in either HBSS ± D-(+)-Glucose (Glc) (20 mM) or Q (4 mM) or L (5 mM). e, qRT-PCR analysis of Asns expression in one representative experiment in mIMCD3 transiently knocked down for Asns (siAsns) relative to control (Scr). f, Quantification of cilia length in one representative experiment in mIMCD3 after 8 hours culture in HBSS or HBSS + Q (4 mM). g, Quantification of cilia length in one representative experiment in mIMCD3 after 24 hours culture in either DMEM/F12 + 0% FBS or HBSS or HBSS + Q (0.5 or 4 mM) or HBSS + L-Asparagine (N) (0.1 or 1 mM). h, Left: analysis of OCR measurement in one representative experiment in siAsns MEF cells compared to Scr after 4 hours in HBSS (Scr: blue, siAsns: red) followed by acute injection (+) of Q (Scr: green, siAsns: purple), in basal condition and after sequential addition oligomycin (O), FCCP, and antimycin A/rotenone (A/R). Right: quantification of acute response as in Left. i, qRT-PCR analysis of Asns expression in one representative experiment in MEF^Ift88 and mIMCD3^Ift88 relative to MEF^Ctrl and mIMCD3^Ctrl. Data in dot and bar plots are mean ± SD. Statistical analysis: Student’s unpaired two-tailed t-test or one-way ANOVA, followed by Tukey’s multiple comparisons test; ns: not significant, ****p < 0.0001. n reported in brackets. Source data

Extended Data Fig. 7. Proposed model.

a) Schematic representation of how primary cilia sense nutrient availability and facilitate glutamine utilization through ASNS affecting the TCA cycle fuelling and, as a consequence, energy production.