Fig 2: The astrin-1/CLUH complex is enriched in interphase.(A) Western blots of co-IPs of endogenous CLUH in WT and CLUH KO HeLa cells. Cells have been synchronized as shown in Figure 3—figure supplement 1A. Pan-actin was used as loading control and cyclin D3 as G2 phase marker, pH3-Ser10 as M phase marker and pRPS6-Ser235/236 to assess effective starvation of input samples. (B) Enriched proteins immunoprecipitated with an antibody against endogenous CLUH in G2-synchronized HeLa cells and detected by mass spectrometry (n = 4 independent replicates). Highlighted are all proteins enriched with a log2FC = 3 and q = 0.05. Marked in green are centrosomal proteins; marked in red are previously identified astrin/kinastrin interactors. (C) GO cellular component analysis of enriched proteins highlighted in B analyzed using the EnrichR webtool. (D) Confocal immunofluorescence pictures of HeLa cells overexpressing FLAG-tagged astrin-1 (ATG3 + 4 + 5GGG-SPAG5) stained for CEP43 (green) and FLAG (red) or PCM1 (red) and FLAG (green). DAPI was used to stain nuclei (blue). Scale bar, 10 µm. Small boxes in left corners show a ×4 magnified area of boxed regions. (E) Confocal immunofluorescence pictures of HeLa cells stained for CEP43 (green) and CLUH (red). DAPI was used to stain nuclei (blue). Scale bar, 10 µm. Small boxes in left corners show a ×4 magnified area of boxed regions. Figure 3—source data 1.Uncropped blots for Figure 3A. Figure 3—source data 2.Unedited blots for Figure 3A.

Fig 3: Astrin and kinastrin depletion does not mimic mitochondrial phenotypes seen upon absence of CLUH.(A) Mitochondrial network of WT, SPAG5, and KNSTRN ind-KO HeLa cells grown in basal or galactose media for 16 hr. Mitochondria were stained with an antibody against TOMM20. Scale bar, 10 µm. On the left side, confocal immunofluorescence pictures are shown, on the right side, mitochondrial network is shown after segmentation. Color code for different mitochondrial morphologies: purple: filamentous; green: rod-shaped; orange: fragmented; blue: swollen; black: unclassifiable. (B) Quantification of mitochondrial morphology of experiments as shown in A (n = 3 independent experiments; at least 66 in basal or 44 cells in galactose have been analyzed per genotype per replicate). Bars show the mean of each morphological class with standard error of the mean (SEM). (C) mtDNA levels of WT, SPAG5, and KNSTRN ind-KO HeLa cells grown in basal or galactose media for 16 hr (n = 4 independent experiments). Bars show mean ± SEM and dots represent values of individual replicates. (D) Blue native polyacrylamide gel electrophoresis (BN-PAGE) analysis of respiratory chain supercomplexes and complex I activity staining of isolated mitochondria of WT, CLUH KO and WT, SPAG5, and KNSTRN ind-KO HeLa cells grown in basal or galactose media for 16 hr. Figure 4—source data 1.Uncropped blots for Figure 4D. Figure 4—source data 2.Unedited blots for Figure 4D.

Fig 4: Depletion of CLUH impairs anabolic pathways.Targeted metabolomics of WT and CLUH KO HeLa cells after 8 hr HBSS starvation showing glycolytic intermediates and lactic acid (A), sedoheptulose-7-P (B), pyruvic acid and TCA cycle intermediates (C), amino acid levels (D), ornithine and polyamine levels (E), and nucleotide levels (F). Bars show mean ± standard error of the mean (SEM) and dots represent values of individual replicates. Two-tailed unpaired Student’s t-tests on log converted fold changes were performed with *p = 0.05; **p = 0.01; ***p = 0.001.

Fig 5: The TPR domain facilitates CLUH self-interaction and its interaction with SPAG5. A Western blot analysis of co-IP, between GFP- and 3xHA-tagged CLUH proteins stably expressed in HCT116 cells. The IP is performed using magnetic beads coupled with anti-HA antibodies (IP-HA) with protein extracts treated (+) or not (-) with RNaseA/T1. The proteins are detected using anti-HA and anti-GFP antibodies. GAPDH and ACTIN are used as specificity controls. The loaded samples correspond to 0.5% of the input and 20% of the pulled-down samples. B Ethidium bromide-stained agarose gel loaded with RNase treated (+) or non-treated (-) total protein extracts used for the IP showing the presence of ribosomal RNA. C Scheme showing the human CLUH protein domains identified using Pfam database. The amino acid positions of each domain are indicated. The mutant protein CLUH?TPR has been generated by deleting the TPR domains. D Western blot analysis of co-IP, between 3xHA-tagged wildtype CLUH (in orange) or the CLUH?TPR mutant (in red) with the endogenous CLUH and SPAG5 proteins. A 30-kDa tag corresponding to the BioID2 protein followed by 3xHA peptide is added in N-terminal (Tag-3xHA) of each transgene. A GFP tagged construct is used as a specificity control. The co-IP is performed on total extracts (INPUT) from HCT116 cells stably expressing the different constructs using magnetic beads coupled with anti-HA antibodies (IP-HA). The size of the endogenous CLUH (black), wildtype transgene (orange), and the delta-TPR mutant (red) is indicated. The indicated proteins are revealed using specific antibodies. TUBULIN Is used as a loading control. The loaded samples correspond to 0.5% of the input and 20% of the pulled-down samples