TANGO6 coordinates transcription and translation through nucleocytoplasmic partitioning

Abstract

Coordinated RNA and protein synthesis is fundamental to cell function and survival. Here, we show that TANGO6, a nucleocytoplasmic RNA binding protein, is an integral component of the proteostasis control axis, regulating both transcription and translation. In the nucleus, TANGO6 is recruited to gene promoters alongside RNA polymerase II, where it controls the transcriptional activation of essential genes, including those required for G₁-S cell cycle progression and cell differentiation. Cytoplasmic TANGO6 associates with polysomes to promote protein synthesis. Under stress conditions, the organization of TANGO6 changes; notably, heat stress triggers its translocation to the nucleus, thereby depleting its cytoplasmic pool. Thus, TANGO6 assists in both transcriptional and translation reprogramming to support cell survival. Overall, the nucleocytoplasmic shuttling of TANGO6 is a central mechanism for controlling the cellular proteome, establishing its essential role in cell viability from homeostasis to stress adaptation.

INTRODUCTION

Differentiated cells require precise control of their proteome to maintain cellular physiology and specialization. In addition, cells and organisms face intrinsic and extrinsic stresses that damage DNA, RNA, and proteins, threatening cellular function. To survive, cells rapidly reprogram RNA and protein synthesis, regulating gene expression from transcriptional to posttranslational levels (1–3).

Following the gene expression cascade, proteostasis is initially controlled at the transcriptional level, where RNA polymerase II (Pol II) drives the production of precursor mRNAs. The Pol II holoenzyme is a large complex composed of 12 protein subunits; RPB2 is the second largest subunit and, together with RPB1, structurally build the catalytic site of the polymerization complex (4). A general regulatory mechanism involves the proximal pausing of Pol II at gene promoters, stabilized by NELF and DSIF complexes, and its release upon phosphorylation of the C-terminal domain (CTD) of RPB1 by elongation factors like P-TEFb, enabling rapid response to stimuli (5, 6).

Maintenance of cellular homeostasis persists after transcription through regulatory steps in splicing, translation, and protein stability. Translation initiation is facilitated by several eukaryotic initiation factors (eIFs) and represents a principal regulatory step (7–9). Under stress, translation initiation is rapidly inhibited. This leads to the accumulation of untranslated mRNAs, eIFs, and RNA binding proteins (RBPs) in the cytoplasm, which coalesce to form stress granules (SGs) (10–12).

Here, we characterize TANGO6 as an essential protein regulating protein homeostasis at multiple levels. Originally identified in Drosophila as involved in transport and Golgi organization (13), we show that TANGO6 is a soluble RBP with dynamic localization. In the nucleus, it binds Pol II and controls transcription activation of development and cell cycle related genes, whereas, in the cytoplasm, it associates with ribosomes to promote protein synthesis. Notably, upon heat stress TANGO6 translocates into the nucleus where it increases its association with promoters of the heat stress response (HSR) genes, promoting their induction while contributing to translational down-regulation. Our results identify TANGO6 as a protein that integrates transcription and translation to promote cell survival under stress.

RESULTS

TANGO6 is essential for cell viability

Multiple sequence alignment of TANGO6 orthologs across species confirmed its evolutionary conservation (fig. S1A). In human U2OS cells, small interfering RNA (siRNA)–mediated TANGO6 knockdown reduces cell viability (Fig. 1, A and B). Similarly, yeast TANGO6 ortholog RTP1 is essential for viability (fig. S1, B and C), supporting a fundamental conserved function. RNA sequencing (RNA-seq) of U2OS after 24 hours of TANGO6 depletion (KD-TANGO6-24) revealed deregulation of over 2000 genes, with 1130 significantly down-regulated and 874 up-regulated (Fig. 1C and table S1). Gene Ontology (GO) analysis of the differentially expressed genes showed a substantial down-regulation of genes involved in cell cycle progression (including multiple cyclins, cyclin-dependent kinases, proliferating cell nuclear antigen, and Myc), as well as general DNA metabolism (Fig. 1D). Gene set enrichment analysis (GSEA) highlighted cell cycle as the most down-regulated pathway (Fig. 1E, top), consistent with observed G₁-phase arrest in KD-TANGO6-24 cells, becoming more pronounced after 72 hours of depletion (KD-TANGO6-72) (Fig. 1F). Unexpectedly, GSEA also showed that genes involved in cytoplasmic translation and ribosomal proteins were collectively up-regulated (Fig. 1E, bottom), possibly reflecting a cellular attempt to compensate for impaired gene expression caused by TANGO6 loss. These findings show that TANGO6 is essential for cell proliferation and survival.

Fig. 1. TANGO6 is essential for cell viability.

(A) Immunoblot showing TANGO6 protein depletion 24 and 72 hours after small interfering RNA (siRNA) treatment; calnexin as loading control (means ± SEM, n = 3); one-way analysis of variance (ANOVA), P < 0.001. (B) Live U2OS cell counts every 24 hours post–siRNA transfection; nontargeting siRNA as control; dead cells were excluded by trypan blue (means ± SEM, n = 3); two-way ANOVA, P < 0.001. (C) MA plots of gene expression changes by RNA sequencing (RNA-seq) in KD-TANGO6-24 versus control cells; up-regulated (orange) and down-regulated (blue). (D) Bar plots of enriched Gene Ontology (GO) biological processes and pathways down-regulated in KD-TANGO6-24 cells. (E) Gene set enrichment analysis (GSEA) of KEGG pathways significantly down-regulated (top) or up-regulated (bottom) in KD-TANGO6-24 cells. (F) Cell cycle analysis of U2OS cells 24 and 72 hours post–TANGO6 depletion; serum starvation as G₁ arrest control (n = 3).

TANGO6 is a soluble nucleocytoplasmic protein

It has recently been proposed that TANGO6 is a protein associated with the Golgi membranes, from which it is incorporated into coat protein complex I (COPI) vesicles (14). Nevertheless, our quantitative analysis of endogenous TANGO6 distribution by confocal microscopy showed that it predominantly localizes in the cytoplasm (65%, excluding the Golgi area), with a punctate distribution, in the nucleus (25%), and, to a much lesser extent, around the Golgi region (10%) (Fig. 2A). Cell fractionation confirmed bipartite localization, with nuclear TANGO6 distributed equally between nucleoplasm and chromatin (Fig. 2B). Confocal microscopy–based colocalization analysis revealed that TANGO6 was enriched in the vicinity of the Golgi membranes, exhibiting a heterogeneous pattern (fig. S2A), whereas TANGO6 pools distributed throughout the cytoplasm and nucleus did not show overlap with any of the analyzed markers (fig. S2, B and C). Super-resolution STED microscopy revealed TANGO6 in small peripheral puncta in contact with Golgi membranes marked by GalNT2 and GM130, not colocalized with COPI vesicles (COPA) or endoplasmic reticulum (ER) marker (Hsp47) (Fig. 2C), further supported by Golgi disruption experiments with nocodazole (fig. S3A). Despite predictions of hydrophobic transmembrane stretches, TANGO6 lacks a signal sequence for ER translocation (not shown). Furthermore, after brefeldin A treatment, TANGO6 did not relocalize to the ER as has been described for Golgi transmembrane proteins (15); instead, TANGO6 distribution increased in the cytoplasm, suggesting peripheral rather than integral association with membranes (fig. S3B). Cell fractionation confirmed TANGO6 is a soluble protein, with a minor proportion being peripherally associated to membranes as suggested by its dissociation from them by high salt concentration, urea, or bicarbonate (Fig. 2D and fig. S3C).

Fig. 2. TANGO6 is a nucleocytoplasmic RBP.

(A) Quantitative analysis of TANGO6 subcellular distribution by confocal microscopy in single and summed z-axis views. Raw integrated density of fluoresce measured inside the cell periphery, Golgi area, and nucleus (means ± SEM, n = 30); one-way ANOVA, P < 0.001. Scale bar, 10 μm. ROIs, regions of interest. (B) Immunoblot of TANGO6 in nuclear-enriched and cytoplasmic fractions; glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Histone3 as controls. (C) STED microscopy images of cytoplasmic TANGO6 (green) with different compartment markers GM130, GalNT2, Hsp47, COPA (red). Scale bars, 5 μm. (D) Immunoblot of membrane (pellet) and soluble (supernatant) fractions extracted without detergent; pellet concentrated 4× to visualize TANGO6 in membranes. Tubulin and calnexin as controls. (E) Immunofluorescence of TANGO6 (green) after RNase A treatment in permeabilized cells, GalNT2 staining (red), and DAPI (blue). Bovine serum albumin (BSA)–treated cells as control. Scale bars, 10 μm. (F) Top: Schematic of orthogonal organic phase separation (OOPS) assay using cross-linked Drosophila SL2 spike-in and U2OS cells overexpressing TANGO6-Flag [created in BioRender: T. Guitart (2026), https://BioRender.com/m9d6ij3]. Bottom: Immunoblot showing TANGO6 and control RBPs PDIA6 and Stubarista (Sta). (G) Immunoprecipitation mass spectrometry analysis of TANGO6-Flag; SAINT scores averaged protein enrichment relative to controls (EV-Flag) (n = 3). RBPs are highlighted in green. (H) Interactors with ≥3-fold change analyzed for GO subcellular localization. (I) Overlap of TANGO6 interactome with RBPs (top) and SGs components (bottom); P values by hypergeometric test. MFI, mean fluorescence intensity. (J) Bar plots of enriched GO Pathways for TANGO6 interactome.

Together, our data show that TANGO6 localizes in the nucleus and cytoplasm organized into discrete particles, a subset of which peripherally associates to Golgi membranes.

TANGO6 is an RBP

It has recently been shown that the peripheral Golgi protein GM130 binds to RNA and helps maintain the integrity of the Golgi ribbon together with primarily cytosolic RBPs, such as FXR1 (16). Given TANGO6’s cytoplasmic arrangement near the Golgi membranes, we tested whether its localization depends on RNA. Although ribonuclease (RNase) A treatment of permeabilized cells did not affect transmembrane protein GalNT2 localization, it abolished TANGO6 staining in the cytoplasm and Golgi proximity, likely due to solubilization of higher order RNA-containing structures (Fig. 2E), suggesting that TANGO6 is an RBP (17). To directly test this, we performed orthogonal organic phase separation (OOPS) in U2OS cells, a method to identify RBPs based on differential solubilization of cross-linked RNA:RBP complexes in aqueous:organic partitions (Fig. 2F, top) (18). As expected for an RBP, TANGO6 stayed at the interphase after covalent binding to RNA using ultraviolet (UV) cross-linking, and treatment of such interphase with RNase I released TANGO6 to the organic phase (Fig. 2F, bottom). These results suggest that TANGO6 is a previously unknown RBP that lacks typical RNA binding domains and, hence, is a member of the growing list of noncanonical RBPs (19). We propose that the punctate cellular distribution of TANGO6 reflects its organization into specialized ribonucleoprotein particles (RNPs).

To investigate TANGO6 functions, we immunoprecipitated C-terminally Flag-tagged TANGO6 from U2OS cells and used mass spectrometry to identify interacting proteins. Probabilistic scoring by significance analysis (SAINT) (20) identified 422 TANGO6 candidate interactors with a fold change equal or higher than 3, regardless of Bayesian false discovery rate (BFDR) (Fig. 2G and table S2). This dataset was used as an exploratory resource to generate biologically testable hypotheses based on interactome analysis. GO analysis showed that our hits are primarily cytosolic (35%) and nuclear proteins (29%), consistent with TANGO6 localization (Fig. 2H). Fifty eight percent of interactors with a BFDR ≤0.05 are RBPs, consistent with TANGO6 being RBP, which is supported by network-based RBP prediction approaches (21). Among all identified RBPs, we find SG core components like G3BP1, G3BP2, and CAPRIN1, supporting potential RNA-related functions for TANGO6 (Fig. 2I and table S2) (10). Reactome pathways analysis revealed that TANGO6 interactome was enriched in proteins participating in RNA metabolism and cellular stress responses (Fig. 2J). These findings prompted us to further explore TANGO6’s roles in gene expression and proteostasis control.

Nuclear TANGO6 controls transcriptional activation

Our mass spectrometry data showed a strong interaction between TANGO6 and the Pol II subunit RPB2 (POLR2B), consistent with prior reports in other cell lines (14, 22). We also identified NELFB as a TANGO6 interactor (Fig. 2G). We confirmed these interactions by coimmunoprecipitation (Fig. 3A and fig. S4A). The RPB2 partner, RPB1, was also detected in the precipitated fraction (Fig. 3A). The levels and localization of RPB2 and RBP1 remained unaffected in KD-TANGO6-24 cells (Fig. 3B and fig. S4B). However, Pol II levels were significantly reduced in KD-TANGO6-72 cells, which correlated with reduced viability. Last, de novo RNA synthesis was reduced after 24 hours of TANGO6 knockdown, as revealed by 5-bromo-uridine (BrU) labeling (fig. S4, C and D).

Fig. 3. TANGO6 is recruited to TSSs to promote transcriptional activation.

(A) TANGO6-Flag immunoprecipitation followed by immunoblotting against RPB2 and RPB1 using specific antibodies; anti-Flag blot confirms IP; Mock, EV-Flag. (B) Immunoblot of TANGO6, RPB2, and RPB1 in TANGO6 knockdown versus control cells at 24 and 72 hour; calnexin loading control (means ± SEM, n = 3); two-way ANOVA, P < 0.001. (C) Genomic distribution of TANGO6-binding sites. (D) Metaplot of TANGO6-Flag ChIP-seq signal (blue) ±5 kb around the TSS of 4234 target genes versus input (gray), normalized to sequencing depth. (E) Heatmap of RNA-seq signal (RPKM) on cell cycle–related TANGO6 targets down-regulated in KD-TANGO6-24 cells; genes from Reactome cell cycle pathway (adjusted P = 2.243 × 10⁻⁷). (F) ChIP-seq heatmaps of TANGO6-Flag, RPB1, RPB2, and input ±5 kb around TSS for all genes ranked by TANGO6 intensity. (G) Representative UCSC genome browser snapshots of TANGO6-Flag, RPB2, and RPB1 profiles, normalized to sequencing depth. (H) Metagene plots of RPB2 and RPB1 ChIP-seq for TANGO6 targets versus the rest of the genome; TSS/TES, transcription start/end site, normalized to sequencing depth. (I) Average spike-in normalized ChIP-seq signals at TSSs of RPB2 and RPB1 for TANGO6 target/nontarget genes, in control and KD-TANGO6-24 cells; boxplot lines indicate quartiles; whiskers extend to 10th to 90th percentile. Wilcoxon test, P < 0.0001. (J) Average spike-in normalized ChIP-seq signals of Ser⁵P–Pol II and Ser²P–Pol II at TSSs and gene bodies for TANGO6 target genes, in control and KD-TANGO6-24 cells; boxplot lines indicate quartiles; whiskers extend to 10th to 90th percentile. Wilcoxon test, P < 0.0001. (K) Reverse transcription quantitative polymerase chain reaction (RT-qPCR) quantification of c-FOS and c-JUN mRNAs, in control and KD-TANGO6-24 cells, pre/post–epidermal growth factor (EGF) stimulation (20 ng/ml, 60 min), normalized to GAPDH (means ± SEM, n = 3); two-way ANOVA, P < 0.001.

To further investigate the nuclear roles of TANGO6, we tested whether it binds to chromatin at specific genomic loci by performing chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) in U2OS cells expressing TANGO6-Flag. We identified a total of 4454 regions significantly enriched in TANGO6, corresponding to 4234 target genes (table S3). TANGO6 was mainly bound to promoter regions close to transcription start sites (TSSs) (Fig. 3, C and D). GO enrichment analysis showed that TANGO6 binds genes involved in essential processes, such as transcription and translation initiation (fig. S4E). Notably, a substantial number of TANGO6 targets are key regulators of the cell cycle and become down-regulated upon TANGO6 depletion, which correlates with the observed cell cycle arrest (Fig. 3E). We then performed ChIP-seq experiments using antibodies against the Pol II subunits, RPB2 and RPB1. Results showed a marked overlap between TANGO6 and Pol II binding, particularly at the TSSs (Fig. 3, F and G), with genes enriched in TANGO6 exhibiting higher levels of both RPB2 and RPB1 subunits (Fig. 3H and fig. S4F). KD-TANGO6-24 cells exhibited a moderate yet significant decrease in total Pol II occupancy at both promoters and across gene bodies, with the effect being most pronounced for TANGO6 targets, suggesting that TANGO6 likely regulates a step between Pol II recruitment and productive elongation (Fig. 3I and fig. S4G). Because a similar, albeit less pronounced, effect was observed at genes not classified as TANGO6 targets, we cannot exclude the possibility that TANGO6 is recruited to all Pol II–occupied genes, but that only a subset was detected in our ChIP-seq due to technical limitations.

We then performed independent ChIP-seq experiments using antibodies raised against the serine-5– or serine-2–phosphorylated CTD of Pol II (Ser⁵–Pol II and Ser²–Pol II), which are known to be associated with transcription initiation/promoter-proximal pausing and productive elongation, respectively (23). We found that in KD-TANGO6-24 cells, the remaining Pol II pool showed a specific increase in the Ser⁵–Pol II form at the TSS, concomitant with a reduction of Ser²–Pol II both at the TSS and across gene bodies (Fig. 3J and fig. S4H). Together, this suggests that TANGO6 facilitates the progression of Pol II from early initiation to productive elongation along the transcription unit. However, we cannot formally exclude additional roles of TANGO6 in Pol II recruitment or in regulating specific Pol II–CTD kinases or phosphatases.

Examination of TANGO6 occupancy revealed a strong positive correlation between transcriptional activity and TANGO6 enrichment at TSSs (fig. S4I). To directly assess whether TANGO6 was required for transcription activation, we used a model system in which the addition of epidermal growth factor (EGF) triggers the expression of specific responsive immediate early genes (IEGs), such as c-FOS and c-JUN (24), which we identified as TANGO6 targets (table S3). EGF-mediated induction of these IEGs was significantly impaired in KD-TANGO6-24 cells indicating that TANGO6 is required for efficient transcriptional activation (Fig. 3K). Comparable results were obtained when the nuclear import of TANGO6 was impaired by depleting the leucine-rich repeat–containing protein 59 (LRRC59) (fig. S4, J and K) and in cells previously depleted of NELFB, which leads to proximal-paused Poll II destabilization (fig. S4K) (25). Collectively, these results indicate that TANGO6 is recruited to gene promoters together with Pol II to promote the expression of multiple essential genes, including key regulators of cell cycle progression and cell differentiation.

Cytoplasmic TANGO6 promotes protein synthesis

Besides Pol II components, our IP-MS dataset showed high overlap with translation related proteins including ribosomal and ribosome-associated proteins as well as translation factors representing 53% of the total interactome pool (fig. S5A and table S2). We confirmed interactions with the translation initiation factors eIF3A, eIF3D, and eIF4A1 by coimmunoprecipitation (Fig. 4A). Polysome profiling analysis in different cell lines revealed that a fraction of TANGO6 is associated with polysomes (Fig. 4B and fig. S5B). TANGO6 in polysomes showed a higher size suggesting the presence of posttranslational modifications. Treatment with puromycin led to a shift of TANGO6 toward lighter fractions, indicating genuine polysomal association (Fig. 4C and fig. S5C). TANGO6 distribution in control conditions or upon puromycin treatment was comparable to that of PABPC1, a known translation regulator (Fig. 4D). Comparable results were obtained for overexpressed TANGO6-Flag protein (fig. S5, D and E).

Fig. 4. TANGO6 associates with polysomes and promotes protein synthesis.

(A) TANGO6-Flag immunoprecipitation followed by immunoblotting against eIF3A, eIF3D, and eIF4A1 using specific antibodies; anti-Flag blot confirms IP; Mock, EV-Flag. (B) Polysome profiles of U2OS cells with immunoblotting of selected fractions. RpS6 and PABPC1 are shown as control. (C) TANGO6 shifts to lighter polysomal fractions after puromycin treatment. (D) Quantification of relative levels of PABPC1 and TANGO6 in heavy polysomes (HP > 3 ribosomes) (means ± SEM, n = 3). Unpaired Student’s t test, P < 0.001. (E) Immunoblot and immunofluorescence of TANGO6-FKBP12^F36V after dTAG-13 addition (100 nM) for the indicated time; calnexin loading control. h, hours. (F) Live cell counts of TANGO6-FKBP12^F36V cells every 24 hour post–dTAG-13 treatment versus DMSO; dead cells were excluded by trypan blue staining (means ± SEM, n = 3); two-way ANOVA, P < 0.001. (G) Puromycylation assay after 5 hours of dTAG-13 or DMSO treatment; quantified relative to total protein (Ponceau S) (means ± SEM, n = 3); one-way ANOVA, P < 0.001. (H) OPP incorporation after 5 hours of dTAG-13 or DMSO treatment; microscopy images summed z-axis views; fluorescence intensity normalized by nuclear mask (means ± SEM, n = 9); one-way ANOVA, P < 0.001. (I) BrU incorporation after 5 hours of dTAG-13 or DMSO treatment, n = 3. Unpaired Student’s t test using Welch’s correction, P < 0.001. (J) Puromycylation assay in TANGO6-FKBP^12F36V-KD-LRRC59 cells before and after depleting TANGO6 by 5 hours of dTAG-13 incubation; nontargeting siRNA was used as control. (K) Dot blot of total Mucin-5 AC in HT29-N2 cells in protein-free medium at days 3 and 4 in KD-TANGO6-24 and control cells (left); each dot, replicate; colors show parallel processing. MUC5AC mRNA measure by RT-qPCR, normalized to GAPDH (right) (means ± SEM, n = 3 with nested replicates). One-way ANOVA, P < 0.001.

We found that global translation was reduced (∼20% decrease) in KD-TANGO6-24 cells compared with that in control cells, as assessed by both puromycylation and O-propargyl-puromycin (OPP) incorporation assays (fig. S5, F and G). In addition, to assess the impact of TANGO6 on protein synthesis while minimizing indirect effects resulting from prolonged depletion, we generated a U2OS cell line expressing TANGO6 tagged with the degron tag FKBP12^F36V by CRISPR-Cas9 genome editing. Addition of dTAG-13 leads to recruitment of the VHL-Cullin2 ubiquitin ligase complex to the degron-tagged protein triggering its degradation (26). Tests showed that TANGO6 was depleted to barely detectable levels within 5 hours after addition of dTAG-13 (Fig. 4E). Consistent with previous observations, TANGO6 degradation severely impaired cell viability, confirming the essential nature of this gene (Fig. 4F). In line with prolonged depletion, acute depletion of TANGO6 led to a significant reduction in global protein synthesis (∼40% decrease) (Fig. 4, G and H). Furthermore, analysis of nascent RNA indicated that, even in conditions of acute depletion, ongoing transcription was affected in the absence of TANGO6 (Fig. 4I). To test whether the roles of TANGO6 on translation were secondary to transcription, we blocked its nuclear import by knocking down LRRC59. While this treatment drastically reduced the levels of TANGO6 in the nucleus and impaired transcription activation (see fig. S4, J and K), it did not affect translation, indicating that the effects of TANGO6 in transcription and translation are separable (Fig. 4J). Furthermore, 5 hours of dTAG treatment in these conditions led to inhibition of protein synthesis, indicating that the cytosolic pool of TANGO6 regulates translation (Fig. 4J). To confirm the relevance of TANGO6-mediated translational control, we used HT29-N2 cells that increase mucin production upon differentiation into Goblet cells (27). TANGO6 depletion led to a reduction in Muc5AC synthesis, despite similar MUC5AC mRNA levels, indicating a translational control mechanism (Fig. 4K). These findings support a model where cytoplasmic TANGO6 sustains efficient protein synthesis.

TANGO6 changes its distribution upon stress

Because TANGO6 regulates both RNA and protein synthesis, we tested whether these functions couple upon extrinsic stress. We subjected U2OS cells to oxidative stress (SA), heat shock (HS) (28), ER stress (Thaps) (29), UV irradiation, and amino acid starvation (−AA) (30). As expected, all conditions tested induced G3BP1-positive SGs. TANGO6 RNPs coalesced away from their original localization upon SA, HS, and ER stress, and most G3BP1 foci either contained or were in close proximity to TANGO6 granules (fig. S6A). However, most TANGO6 granules did not contain G3BP1. SGs formed upon UV radiation or amino acid starvation, which are known to lack eIF4F and eIF3 (30, 31), also lack TANGO6 (fig. S6A). Unexpectedly, HS triggered a pronounced nuclear localization of TANGO6 (Fig. 5A). This redistribution was confirmed by cell fractionation, revealing a notable increase of TANGO6 in the chromatin fraction (fig. S6B). TANGO6 nuclear import depends on LRRC59, as its knockdown blocked TANGO6 nuclear entry but did not affect cytoplasmic granule association (fig. S6C). Notably, the properties of TANGO6-FKBP12^F36V–tagged protein upon HS were comparable to that of the wild-type (WT) protein in terms of nuclear translocation and granule formation (fig. S6D).

Fig. 5. TANGO6 translocates to the nucleus and coordinates heat shock response.

(A) Confocal microscopy of TANGO6 upon HS or NHS conditions; images show single slides and summed z-axis views. Nuclear fluorescence quantified relative to area (means ± SEM, n = 20 cells per replicate, three replicates); one-way ANOVA, P < 0.001. Scale bars, 10 μm. (B) Overlap of TANGO6 ChIP-seq target genes in NHS and HS. (C) Genomic distribution of TANGO6-binding sites after HS. (D) Metagene plot of TANGO6 ChIP-seq signal in NHS and HS, normalized to sequencing depth. TSS/TES, transcription start/end site. (E) Representative UCSC genome browser view of TANGO6-Flag ChIP-seq in NHS and HS, normalized to sequencing depth. (F and G) Boxplot and heatmap of TANGO6 ChIP-seq signal at HS-responsive gene TSSs, normalized to sequencing depth; peaks ranked by signal intensity. (H) Average ChIP-seq signals of Pol II (RPB1) at TSSs and gene bodies for HS-responsive genes, under NHS and HS. Cells were previously treated with dTAG-13 or DMSO for 5 hours. Data normalized by spike-in; boxplot lines indicate quartiles; whiskers extend to 10th to 90th percentile. Wilcoxon test, P < 0.0001. (I) RT-qPCR of HSR gene mRNAs upon NHS and HS conditions. Cells were previously treated with dTAG-13 or DMSO for 5 hours, normalized to GAPDH (means ± SEM, n = 3); one-way ANOVA, P < 0.001. (J) Immunofluorescence of TANGO6 (green) during recovery (REC; 37°C, 5 hours post-HS) in the presence of cycloheximide (CHX); images show single slides and summed z-axis views. DAPI nuclear staining. Scale bars, 10 μm. (K) Puromycin incorporation assay in NHS and HS conditions and after recovery from stress (REC) upon dTAG-13 or DMSO treatment for 5 hours. (L) Model: TANGO6 shuttles dynamically between the cytoplasm and nucleus in a stress-dependent manner, promoting protein synthesis in the cytoplasm and transcription in the nucleus.

TANGO6 assists cellular HSR

To further investigate the nuclear role of TANGO6 upon HS, we tested whether it binds to chromatin at specific genomic loci. ChIP-seq analysis showed that, upon HS, in concordance with its increased nuclear abundance, TANGO6 was recruited to many additional chromatin sites (11,896 ChIP-seq peaks), corresponding to 9967 target genes, mainly at promoter regions (Fig. 5, B and C). ChIP-seq signal strength was markedly higher compared with non–heat shock (NHS) conditions at 37°C (Fig. 5, D and E; fig. S6E; and table S3). GO terms enriched among de novo TANGO6-bound genes were mainly related to RNA processing, translation, gene expression, and cellular response to heat stress (fig. S6F).

Heat stress typically results in the redistribution of Pol II occupancy across the genome rather than an increase in Pol II levels (32). ChIP-seq analysis of Pol II occupancy revealed a global decrease in Pol II binding upon HS (fig. S7A, top). Additionally, we identified 181 genes with at least a threefold increase in Pol II occupancy that we classified as HSR genes in our system (fig. S7A, bottom). As expected, these genes were enriched for HSR functions and pathways (fig. S7B), including canonical HSR genes such as members of the HSP family of chaperons: HSPA1A, HSPD1, HSPH1, HSP90AA1, etc. Notably, HS triggered a significant increase in TANGO6 recruitment to the promoters of these responsive genes (Fig. 5, F and G).

To test whether TANGO6 is required for proper Pol II recruitment and/or activity under stress conditions, we performed Pol II ChIP-seq experiments in control cells and in cells that had previously been depleted of TANGO6 by dTAG-induced degradation 5 hours before the stress. Notably, TANGO6 depletion impaired the increase in Pol II occupancy at the promoters of heat-sensitive genes and along the gene bodies during HS (Fig. 5H and fig. S7, C and D). Correlating with this, TANGO6 was required for robust expression of a panel of HSR genes as its acute depletion significantly attenuated their rapid activation upon HS (Fig. 5I). Comparable results were obtained in KD-LRRC59 cells (fig. S7E). Last, overexpressing TANGO6-Flag above endogenous levels led to a slight but noticeable increase in the activation of the tested HS genes and an increase in the amount of Pol II binding at their TSS (fig. S7, F and G).

Transcriptional reprogramming upon acute HS is coupled with global inhibition of protein synthesis, while recovery from stress restores normal transcription and translation levels (33). To test whether recovery of translation after stress correlates with relocation of TANGO6 to the cytoplasm, we treated the cells for 5 hours at 37°C after exposure to HS (REC) and monitored TANGO6 localization and global translation. We found that the nuclear translocation of TANGO6 is reversible as returning cells to normal temperature relocates the existing pool of TANGO6 back into the cytoplasm (Fig. 5J). As expected, translation was strongly reduced upon HS, and it was fully recovered to basal levels after recovery (Fig. 5K). Acute depletion of TANGO6 during the recovery period strongly inhibited the restoration of protein synthesis (Fig. 5K). Together, these results suggest that the location-specific roles of TANGO6 described in this work converge during HSR: Increased nuclear TANGO6 is required for transcription activation of heat-responsive genes, while cytoplasmic depletion inhibits translation (Fig. 5L).

DISCUSSION

We provide evidence that TANGO6 is an essential nucleocytoplasmic RBP that shuttles between the two compartments to dynamically adjust gene expression and protein synthesis, thereby ensuring cellular survival under both normal and stress conditions. Although the precise molecular effectors of cell death remain to be defined, our data suggest that TANGO6 depletion triggers a progressive transcription-translation toxic state, ultimately leading to cell cycle arrest, Pol II degradation, and a terminal stress response. As RBPs are increasingly recognized as master regulators of the entire RNA life cycle providing connections between distinct RNA regulatory steps, our model, in which TANGO6 acts at multiple stages of gene expression forming dynamic RNPs that link nuclear events to cytoplasmic output, naturally fits within this emerging paradigm.

In the nucleus, TANGO6 is recruited to chromatin together with Pol II, predominantly at TSSs of actively transcribing genes influencing expression of genes involved in cell cycle progression among other processes. Regulation of G₁-S transition genes aligns with observed cell cycle arrest upon TANGO6 depletion, highlighting its importance in cell proliferation. Moreover, the requirement of TANGO6 for proper activation of IEGs further underscores its potential relevance in cancer and neurodevelopmental disorders (34).

In the cytoplasm, TANGO6 associates with the translation machinery to promote protein synthesis. Although speculative, given its physical interactions with eIF3 subunits and eIF4A, together with its association with ribosomes and polysomes, an attractive hypothesis is that TANGO6 regulates translation initiation by facilitating efficient 43S-ribosomal complex scanning or by enhancing early translation elongation. Future work will be required to determine whether TANGO6 acts globally or on specific mRNA subsets and whether it modulates distinct mechanisms of translation. Last, we showed that the reversible nucleocytoplasmic shuttling of TANGO6 facilitates transcriptional and translational reprogramming during HS and post–stress recovery.

Recent studies by Feng et al. (14) characterized TANGO6 as a membrane-bound protein of the Golgi complex, proposing that it binds RPB2 at the Golgi membranes, becoming incorporated into COPI vesicles to reach the ER, before translocating into the nucleus (14). However, our findings are inconsistent with this model in terms of TANGO6’s location, topology, and intracellular trafficking. In contrast, our data are consistent with recent high-throughput ChIP–mass spectrometry analysis that found that TANGO6 interacts with nuclear Pol II, specifically with the RPB2, RPB3, and RPB11 subunits (22, 35). Here, we provide a broader previously unknown set of nuclear and cytoplasmic functions for TANGO6, including roles upstream RPB2 trafficking or degradation. Last, given the RBP nature of TANGO6, an intriguing hypothesis is that the Golgi membrane–interacting pool provides a local repository to sense diverse forms of stress; however, further work will be required to test this idea.

Our data revealed distinct effects of TANGO6 depletion that appear to be mechanistically independent. We speculate that the compartmentalized roles of TANGO6 are likely driven by its distinct structural domains, which permit modulation of the activity of specific molecular complexes. Translational inhibition is known to rapidly disrupt the transcriptionally permissive chromatin state and Pol II dynamics via the turnover of short-lived transcription factors that require ongoing translation (36). Therefore, we cannot rule out the possibility that early translation defects triggered by TANGO6 depletion may amplify later TSS Pol II loss. Disentangling these potentially interconnected effects merits future mechanistic investigation.

In summary, TANGO6 emerges as a highly conserved, essential RBP that integrates transcriptional and translational regulation in homeostasis and stress. By dynamically shifting its localization in response to cellular states, TANGO6 enables cells to finely tune protein production. Its central role in directly managing mRNA and protein synthesis underscores its importance for cell viability and positions TANGO6 as a critical node in cellular homeostasis.

MATERIALS AND METHODS

Requests for further information and or reagents may be addressed to corresponding authors, V.M. (vivek.malhotra@crg.eu) and G.B. (gonzalo.bigliani@crg.eu).

Cell culture and transfection

U2OS cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) cell culture medium (Lonza, catalog no. BE12-604F/U1) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; Gibco, catalog no. 10270-106). Cells were kept at 37°C in a humidified incubator supplied with 5% CO₂. To generate stable homozygous U2OS cell line expressing TANGO6 tagged with FKBP12^F36V by CRISPR-Cas genome editing, backbone vector pX330_U6 derived from Zhang Lab was used. Single guide RNAs (sgRNAs) were designed to target right after the C-term of TANGO6. The HDR plasmid has two homology arms of about 1 kb [LHA, 899 base pairs (bp); and RHA, 985 bp], left and right of the insertion site (right before the stop codon). sgRNAs can be found in table S4. For generation of stable doxycycline-inducible cell line expressing TANGO6-3xFlag, lentiviral particles were generated by cotrasnfecting pRSV-REV, pMDLg/pRRE, VSV-G, and the transfer plasmid (doxycycline-inducible) carrying the sequence of the gene to be over expressed with resistance to neomycin into 0.8 × 10⁶ HEK-293 T (American Type Culture Collection, negative for mycoplasma) cells in a 60-mm culture dish with 6 μl of TrasnIT (Mirus Bio, catalog no. 293). Two and 3 days after, the cell medium was collected and filtered through a 0.45-μm membrane. U2OS cells were plated in 60-mm dishes and infected by adding 1 ml of complete medium and 2 ml of medium containing viral particles for 24 hours. Infected cells were selected by supplementing the culture medium with neomycin (Sigma-Aldrich, catalog no. 1142). Induction was performed for 24 hours using doxycycline (Sigma-Aldrich, catalog no. 9891) at a final concentration of 1 μM. Human melanoma Sk-MEL-147 cells were cultured at 37°C in high-glucose DMEM supplemented with pyruvate and GlutaMAX (Thermo Fisher Scientific, no. 31966021), 10% FBS, and 1% penicillin-streptomycin.

RNA interference transfection

RNA interference (RNAi) transfection was performed using Lipofectamine RNAiMAX (Thermo Fisher Scientific, catalog no. 13778075), following the manufacturer’s protocol. Cells were plated so that they were 70 to 80% confluent at the time of transfection. Twenty-four hours after the transfection, either the cells were processed (KD-24) or a second, identical transfection was performed to those cells being silenced for 72 hours since the first transfection (KD-72). RNAi sequences used can be found in table S4.

Protein electrophoresis and Western blot

Cells were washed once with cold phosphate-buffered saline (PBS), and 0.6 ml (six-well plates) of radioimmunoprecipitation assay (RIPA) buffer [previously described in (37)] with 5 μM leupeptin (Focus Biomolecules, catalog no. 10-1346), aprotinin (2 μg/μl; Abcam, catalog no. ab146286), and pepstatin A (2 μg/μl; Panreac Química, catalog no. A2205) was added. The cells were kept in the lysis buffer on ice for 5 min before being transferred to 1.5-ml tubes and placed on a rotating rack at 4°C for 30 min. Samples were centrifuged at 16,000g at 4°C for 15 min. The supernatant was recovered in a new tube and 1:6 in volume of 6× loading buffer was added (lysate). Lysates with loading buffer were boiled at 95°C for 5 min. Twenty to 40 μl of the lysate were loaded into a 10% polyacrylamide gel for SDS–polyacrylamide gel electrophoresis (PAGE) electrophoretic (Bio-Rad, Mini-PROTEAN) separation of proteins. Proteins were transferred to a 0.20-μm nitrocellulose membrane (Amersham, catalog no. 10600001) and blocked with a solution of 5% (w/w) of nonfat dry milk in TBS-T for 30 min at room temperature (RT) on a laboratory rocker. Primary antibodies were diluted in TBS-T supplemented with 2.5% bovine serum albumin (BSA), and membranes were incubated with this solution overnight at 4°C. Membranes were washed three times, 5 min each, before being incubated with the conjugated secondary antibody diluted in 2.5% milk in TBS-T for 60 min at RT in a laboratory rocker. Membranes were then washed three times for 5 min each with TBS-T. Enhanced chemiluminescence (ECL) signal was generated by incubating the membranes with Immobilon Forte Western HRP Substrate (Millipore, catalog no. WBLUF0100). ECL emissions were digitalized using the iBright imaging system (Thermo Fisher Scientific) equipped with a high-resolution (9.1 megapixels) charge-coupled device camera. Information on all antibodies used can be found in table S4.

Coimmunoprecipitation

The cells were washed once with cold PBS, and 1.2 ml of lysis buffer [50 mM tris, 150 mM NaCl (pH 7.3), 1% Triton X-100, 5 μM leupeptin (Focus Biomolecules, catalog no. 10-1346), aprotinin (2 μg/μl; Abcam, catalog no. ab146286), and pepstatin A (2 μg/μl; Panreac Química, catalog no. A2205)] was added. The cells were kept in the lysis buffer on ice for 5 min before being transferred to 1.5-ml tubes and placed on a rotating rack at 4°C for 30 min. Samples were centrifuged at 16,000g at 4°C for 15 min. The supernatant (0.1 ml) was recovered as the input fraction in a new tube and 20 μl of 6× loading buffer [375 mM tris-HCl; 12% SDS; 60% glycerol, 600 mM dithiothreitol (DTT), and 0.6% bromophenol blue] was added. The supernatant (1 ml) was recovered in a new tube containing 30 μl of ANTI-FLAG M2 affinity gel (Milipore, catalog no. A2220). Samples were incubated in a rotating wheel for 60 min at 4°C. The affinity resin was pelleted by centrifugation at 2500g for 5 min and washed four times with 500 μl lysis buffer. The IP fractions were eluted in 45 μl of 1.2× loading buffer.

Immunofluorescence

The cells were washed once with PBS and fixed for 20 min with a solution of paraformaldehyde (PFA) 4% (w/v) (Sigma-Aldrich, catalog no. P6148) in PBS, prewarmed at 37°C. Cells were then washed three times with PBS, 5 min each before being permeabilized by incubation for 6 min with 0.1% Triton X-100 (Sigma-Aldrich, catalog no. T9284) (w/v) at RT. Nonspecific binding sites were blocked with a solution of BSA 1% (w/v) in tris-buffered saline (blocking solution) for 30 min at RT. Primary antibody was diluted in blocking solution, and cells were incubated with this solution overnight at 4°C. After primary antibody incubation, cells were washed three times with PBS, 5 min each. All secondary antibodies were diluted to 2 μg/μl in blocking solution. Cells were incubated with the secondary antibodies’ solution for 1 hour at RT. Cells’ nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) at 0.5 μg/μl. Last, cells were washed three times with PBS, rinsed in Milli-Q water once, and mounted on a microscopy slide with FluorSave reagent (CalBiochem, catalog no. 345789). For RNase washout experiments, live cells were permeabilized for 20 min at RT using Tween 20 (Sigma-Aldrich, catalog no. T1379) 0.6% (w/v) and treated with RNase A (Thermo Scientific, catalog no. EN0531) solution (1 mg/ml) for 30 min, as described in (17). After the treatment, cells were washed three times in BSA 1% and three times with ice cold PBS, 5 min each. Then, cells were fixed, and immunofluorescence was performed as previously described. All antibodies used can be found in table S4.

Confocal microscopy

High-resolution images were acquired in an inverted Leica STELLARIS confocal microscope equipped with photomultipliers and hybrid detectors. A Plan-Apo 63×, 1.4 numerical aperture, oil-immersion objective was used for specimen imaging. Pixel size and z-step sizes were set to fulfil Nyquist criterion. In all cases, laser and detection spectral bands were chosen to maximize signal recovery while avoiding signal bleed-through. Scanning speed was set to 600 Hz and bidirectional. Between one and three lines were averaged to generate the final image. Bit-depth was set to 8 bits. Images were processed using ImageJ software.

STED microscopy

Super-resolution images were acquired in Leica confocal STED 3X FALCON system with white light laser (WLL) and three depletion wavelengths (WLL: 470 to 670 nm, 592 nm CW, 660 nm CW, 775 nm pulsed). Secondary antibodies used can be found in table S4.

Images quantification

ImageJ/Fiji (38) was used for the analysis of all digitalized images from Western blots experiments. To quantify ECL or fluorescence intensity, the background was subtracted from the images. We generated an intensity profile for the membrane and measured the area of the peaks of interest.

Nuclear fraction isolation

Cells were tripsinized, centrifuged 5 min at 150g, and washed once with cold PBS. Cells pellet (total) was resuspended in 10 volumes of hypotonic buffer [10 mM Hepes-K (pH 7.9), 10 mM KCl, 1.5 mM MgCl₂, 0.34 M sucrose, 10% glycerol, 1 mM DTT, 10 mM b-GliP, 1 mM SOV, 1× PIC, and 0.10% Triton X-100] and incubated for 5 min on ice followed by low-speed centrifugation: 1300g for 5 min at 4°C. Supernatant was collected in a separate tube, and pellet (nuclei) was washed and resuspended in 1 volume of buffer. Ten volumes of 2× SDS lysis buffer [50 mM tris-Cl (pH 7.5), 2% SDS, and 2 mM EDTA] were added and incubated for 30 min on ice followed by centrifugation 1700g for 5 min at 4°C. Supernatant was collected and clarified by centrifugation 15,000g for 15 min at 4°C (nucleoplasm). Pellet was washed and resuspended in 1× SDS buffer, boiled at 98°C for 10 min, and sonicated until it clarified.

Membrane fractionation

Cells were tripsinized, centrifuged at 150g for 5 min, and washed once with cold PBS. Cells pellet (total) was resuspended in 10 volumes of hypotonic buffer without detergent and pass-through syringe (G22 needle) 20 times to mechanically break the cells. Lysate was low-speed centrifugated (5 min at 150g), and supernatant was transferred to a new tube for ultracentrifugation 1 hour at 100,000g in Tabletop Max-XP centrifuge. Supernatant fraction was transferred, and pellet was resuspended in 0.25 volumes of lysis buffer with 1% Triton X-100.

Pull-down and mass spectrometry analysis

The anti-FLAG beads used in the immunoprecipitation were cleaned three times with 500 μl of 200 mM ammonium bicarbonate, after which 60 μl of a solution containing 6 M urea and 200 mM ammonium bicarbonate was added. The samples were then reduced with dithiothreitol (30 nmol at 37°C for 60 min), alkylated in the dark with iodoacetamide (60 nmol at 25°C for 30 min), and diluted to 1 M urea with 200 mM ammonium bicarbonate for trypsin digestion (1 μg at 37°C for 8 hours; Promega, catalog no. V5113). After digestion, the peptide mixture was acidified with formic acid and desalted using a MicroSpin C18 column (The Nest Group Inc.) before liquid chromatography–tandem mass spectrometry analysis. The samples were analyzed using an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA), which was coupled to an EASY-nLC 1200 [Thermo Fisher Scientific (Proxeon), Odense, Denmark]. Acquisition was performed in data-dependent acquisition mode, with full mass spectrometry scans acquired at a resolution of 120,000 over a mass range of 350 to 1400 mass/charge ratio, using the Orbitrap mass analyzer for detection. Digested BSA (New England Biolabs, catalog no. P8108S) was analyzed between each sample to avoid carryover and ensure instrument stability. The acquired spectra were analyzed using the Proteome Discoverer software suite (v2.5, Thermo Fisher Scientific) and the Mascot search engine (v2.6, Matrix Science). Data were searched against a Swiss-Prot human database (as of May 2022, containing 20,401 entries) and a list of common contaminants, along with all corresponding decoy entries. The false discovery rate for peptide identification was set to a maximum of 5%. The SAINTexpress algorithm was used to score protein-protein interactions.

Orthogonal organic phase separation

We followed OOPS protocol described in (18) with some modifications. One 100-mm culture dish with TANGO-Flag overexpressing U2OS cells at 80 to 90% confluency (UV cross-linked at 0.15 J/cm² or not cross-linked) was mixed with a spike-in control consisting of 2 million to 3 million Drosophila SL2 cross-linked cells. Acidic guanidinium-thiocyanate-phenol-chloroform (AGPC) phase separation was performed by addition of 1 ml of TRIzol (Invitrogen, 15596018) and 200 μl of chloroform (VWR, 1024451000) and centrifugation at 12,000g for 15 min at 4°C. While aqueous and organic phases were discarded, the interface was subjected to two additional phase separation rounds. The resulting interphase was precipitated with methanol, and, after partial resuspension in water, pellets were digested with 1000 U of RNase T1 (Thermo Fisher Scientific, EN0541) and 10 μg of RNase A (Roche, 10109142001) for 15 min at 4°C. Samples were further solubilized by adding 1% SDS and 100 mM TEAB (Fluka, 17902) followed by six sonication cycles (30 s on/off) in a Bioruptor Pico (Diagenode). Samples were incubated at 95°C for 5 min, cooled, and digested with 1000 U of RNase T1 and 10 μg of RNase A o/n at 37°C upon shaking. A final AGPC phase separation round was performed, and the interface and the organic phase (containing the RNA-released proteins) were precipitated by one or two rounds of methanol precipitation (respectively). Resulting samples were analyzed by Western blot.

Reverse transcription quantitative polymerase chain reaction

RNA was isolated from cultured U2OS cells with the RNeasy Plus Mini Kit (QIAGEN), and cDNA was generated from 1 μg of RNA with the qScript cDNA Synthesis Kit (Quantabio), both following the manufacturer’s instructions. Real-time polymerase chain reaction (PCR) reactions were performed using SYBR Green I PCR Master Mix (Roche) and the Roche LightCycler 480. All primers used are listed in table S4.

BrU incorporation assay

Cells were plated at a density that did not exceed 60 to 70% confluency and allowed to recover overnight. Pretreated cells were pulse labeled with 0.5 mM BrU in complete medium for 30 min. Cells were then trypsinized, collected in 15-ml tubes, and neutralized with medium containing 10% FBS. Fixation was performed by adding PFA to a final concentration of 4% and incubating for 10 min at RT on a tube rotator. Glycine was subsequently added to a final concentration of 0.125 M, and samples were incubated for an additional 5 min under the same conditions. All centrifugation steps were performed at 700g for 5 min. Following fixation, cell pellets were washed twice with 10 ml of PBS. For permeabilization, pellets were resuspended in 1 ml of PBT buffer and incubated for 30 min at RT with rotation. Cells were then washed with 1 ml of PBS containing 1% BSA and incubated with anti-BrU primary antibody diluted in PBS/1% BSA for 1 hour at RT. After washing, cells were incubated with an Alexa Fluor–conjugated secondary antibody, followed by two additional washes with PBS/1% BSA. Cells were lastly resuspended in PBS and analyzed on a Cytek Aurora spectral flow cytometer. More than 100,000 events were acquired per sample. Forward scatter (FSC) and side scatter (SSC) parameters were used to define the main cell population based on size and internal complexity, respectively. Cell doublets and aggregates were excluded by gating on FSC-A (area) versus FSC-H (height), and only singlet cells were included in downstream analyses. Autofluorescence was separated from specific fluorescence signals using the unmixing algorithm implemented in SpectroFlo software. Mean fluorescence intensity, defined as the geometric mean fluorescence of the indicated parameter, as well as cell frequencies, were quantified using FlowJo software. Statistical analyses were performed in GraphPad Prism 10 using Student’s t test.

Polysome profiling

U2OS cells were seeded in 150-mm culture dishes, grown up to 80 to 90% confluency, and preincubated at 37°C for 5 min in medium containing cycloheximide (CHX; 0.1 mg/ml) or for 10 min in medium containing puromycin (100 μg/ml). Cells were washed in ice-cold PBS with CHX (0.1 mg/ml) and scrapped in ice-cold lysis buffer [15 mM tris-HCl (pH 7.4), 5 mM MgCl₂, 100 mM NaCl, 0.5% IGEPAL, 0.2 M sucrose, CHX (0.1 mg/ml), RNasin (200 U/ml; Promega, no. N2111)] and 1 mM DTT. Extract was incubated 3 min on ice and cleared for 5 min at 1300g. Between 5.5 and 12 A₂₆₀ units were loaded onto linear 10 to 50% sucrose gradients and centrifuged for 3 hours at 35,000 rpm at 4°C in a Beckman SW41 rotor. Gradients were prepared by placing 50% sucrose dissolved in polysome buffer [15 mM tris-HCl (pH 7.4), 5 mM MgCl₂, 100 mM NaCl, 1 mM DTT, and CHX (0.1 mg/ml)] underneath a 10% sucrose solution and mixed using a Gradient Master (BioComp). Absorbance was measured at 260 nm, and fractions were collected using a Gradient Station (BioComp). Eluted fractions were analyzed by Western blotting after ethanol precipitation. For TANGO-Flag expressing cells, eluted fractions corresponding to 80S, light polysomes (two to three ribosomes) and heavy polysomes (more than three ribosomes) were precipitated, pooled, and loaded on SDS-PAGE for immunoblot analysis. For SK-MEL-147 cells, a protocol with small modifications was used. Puromycin treatment was performed for 5 min at 37°C. Cells were washed twice in ice-cold PBS with CHX (0.1 mg/ml), scrapped, and collected by centrifugation at 300g for 7 min at 4°C. Pellets were resuspended in ice-cold lysis buffer [10 mM tris-HCl (pH 7.4), 10 mM MgCl₂, 100 mM KCl, 0.5% Triton X-100, CHX (0.1 mg/ml), SuperaseIn (1 U/ml; Ambion, no. AM2694), and protease inhibitors (Roche, no. 11873580001)]. The cell suspension was vortexed three times for 10 s in the presence of glass beads, with 5-s intervals on ice, incubated 15 min on ice, and vortexed again. The lysate was cleared for 5 min at 8000 rpm. Twelve A₂₆₀ units were loaded onto linear 10 to 50% sucrose gradients prepared in polysome buffer [15 mM tris-HCl (pH 7.4), 5 mM MgCl₂, 80 mM KCl, and CHX (0.1 mg/ml)].

TANGO6 and PABPC1 levels were quantified with ImageJ/Fiji. The levels in heavy polysomes (more than three ribosomes) were divided to total signal (sum of fractions from 40S onward). For each protein, values obtained in the presence of puromycin were shown relative to those measured in its absence. In the case of TANGO6-Flag, total signal included the sum of 80S, light and heavy polysomal pools.

Puromycin incorporation assay

Identical numbers of cells were grown making sure to avoid exceeding 60 to 70% confluence at the end of the experiment. Cells were incubated with puromycin (10 μg/ml; Gibco, catalog no. A11138-03) for 30 min at 37°C, 5% CO₂. Control cells were incubated with 100 μM CHX (Sigma-Aldrich, catalog no. C7698) since 5 min prior puromycin treatment. Puromycin was added to the cells without changing the cell culture medium. The cells were then immediately transferred to ice, and the medium was discarded to stop puromycin incorporation. The cells were washed three times with ice-cold PBS. The whole-cell lysates for each condition were prepared using RIPA lysis buffer. Samples were run in SDS-PAGE, and the levels of puromycin incorporation were determined by Western blot analysis using an anti-puromycin antibody.

OPP incorporation assay

Plate cells at 60 to 70% confluency and allow cells to recover overnight. Treat the cells with dTAG (1 μM) or dimethyl sulfoxide (DMSO) for 5 hours. Pretreat the negative controls with CHX for 5 min. Add to all conditions 20 μM Click-iT OPP final concentration of working solution following the manufacturer’s indications (Invitrogen, catalog no. C10458). Incubate cells for 30 min. After incubation, remove the medium containing Click-iT OPP and wash the cells once with PBS. Fix the cells with 4% formaldehyde in PBS. Incubate for 20 min at RT. Remove fixative. Add 0.1% Triton X-100 in PBS and incubate for 6 min at RT. Wash cells twice with PBS. Add Click-iT Plus OPP reaction cocktail (Invitrogen, catalog no. C10458) to each well and mix well. Incubate for 30 min at RT, protected from light. Remove the reaction cocktail and wash once with Click-iT Reaction Rinse Buffer (Invitrogen, catalog no. C10458). Add 1× HCS NuclearMask Blue Stain (Invitrogen, catalog no. C10458) working solution and incubate for 30 min at RT, protected from light. Wash twice with PBS. Remove the wash solution and proceed to imaging and analysis. For microscopy, coverslips were mounted using FluorSave (Merck, catalog no. 345789). For fluorescence quantification, 96 well plates were scanned using TECAN M200-B.

Chromatin immunoprecipitation

Cells were cross-linked in 1% formaldehyde for 10 min at RT, followed by quenching in 0.125 M glycine for 5 min at RT. The cross-linked cells were washed with ice-cold PBS, harvested by gently scrapping the plates (on ice) in PBS and protease inhibitors, and centrifuged at 3400g at 4°C for 5 min. Pellets were frozen at −80°C until use. For TANGO6-Flag ChIPs, a double cross-linking was used: first with ChIP Crosslink Gold (Diagenode, C01019027) according to the manufacturer’s instructions, followed by the regular formaldehyde cross-linking. Pol II ChIPs upon HS conditions were performed in U2OS-TANGO6-FKBP12^F36V cell line, and depletion of TANGO6 was induced for 5 hours of incubation with dTAG prior the stress.

For chromatin preparation, the pellets were resuspended in ice cold swelling buffer [25 mM Hepes-KOH (pH 7.9), 1.5 mM MgCl₂, 10 mM KCl, and 0.1% NP-40] and proteinase/phosphatase inhibitors and incubated 10 min on ice. Cells were disrupted using a Dounce homogenizer and centrifuged at 3000g for 5 min at 4°C. Nuclei pellets were lysed in sonication buffer [50 mM Hepes-KOH (pH 7.9), 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Na-deoxycholate, and 0.1% SDS] and proteinase/phosphatase inhibitors, sonicated in a Bioruptor Pico (Diagenode) to generate 100- to 500-bp chromatin fragments, and centrifuged at 16,000g at 4°C for 20 min to remove the debris.

ChIP experiments were performed using 60 μg of chromatin (DNA) and 0.25% of sonicated Drosophila melanogaster chromatin (prepared in-house) as “spiked-in.” Aliquots (1%) were removed as input material. Chromatins were incubated with 10 μg of the specific antibodies and 1 μg of Drosophila spike-in antibody (Active Motif, no. 104597) overnight at 4°C on rotation. The day after, 60 μl of Dynabeads protein A (for rabbit antibodies) or A+G (for mouse antibodies) were added, and samples incubated for another 3 hours at 4°C. After incubation, beads were successively washed with 1 ml of sonication buffer, buffer A [50 mM Hepes-KOH (pH 7.9), 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Na-deoxycholate, and 0.1% SDS], buffer B [20 mM tris (pH 8.0), 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, and 0.5% Na-deoxycholate], and TE [10 mM tris (pH 8.0) and 1 mM EDTA]. ChIPed material was eluted in 200 μl of freshly prepared elution buffer (1% SDS and 100 mM NaHCO₃) at 65°C in a shaker (1000 rpm) for 1 hour. Eluted chromatin and input samples were de–cross-linked overnight at 65°C with the addition of 200 mM NaCl, followed by protein digestion with proteinase K (final concentration, 0.1 mg/ml) for 1 hour at 45°C. ChIPed DNA and inputs were purified using the QIAquick PCR purification kit (QIAGEN) and eluted in 60 μl. The following antibodies were used in the ChIP experiments: anti-Flag M2 (Sigma-Aldrich, no. 1804), Rpb1 NTD (Cell Signaling Technology, no. 14958), RNA PolII-S5p (Abcam, no. ab5131), and POLR2B (Santa Cruz Biotechnology, no. sc166803). Library preparation for ChIP-seq experiments was performed at the UPF/CRG Genomics Unit. Libraries were sequenced using Illumina HiSeq2000 sequencer.

PolyA–RNA-seq analysis

The RNA-seq samples were mapped against the hg38 human genome assembly using TopHat (39) with the option –g 1 to discard those reads that could not be uniquely mapped in just one region. DESeq2 (40) was run to quantify the expression of every annotated transcript using the RefSeq catalog of exons and to identify each set of differentially expressed genes. Expression values shown in the boxplots and heatmaps correspond to the average RPKMs across the three replicates in each condition. GSEA of the preranked lists of genes by DESeq2 stat value was performed with the GSEA software (41). Human genes were ranked by the ratio between TANGO6 KD and WT RNA-seq expression.

ChIP-seq analysis

ChIP-seq samples were mapped against the hg38 human genome assembly using Bowtie with the option –m 1 to discard those reads that could not be uniquely mapped to just one region (42). ChIP-seq samples normalized by spike-in were mapped against a synthetic genome constituted by the human and the fruit fly chromosomes (hg38 + dm3) using Bowtie with the option –m 1 to discard reads that did not map uniquely to one region. MACS was run with the default parameters but with the shift size adjusted to 100 bp to perform the peak calling against the corresponding control sample (43). The genome distribution of each set of peaks was calculated with SeqCode (44) by counting the number of peaks fitted on each class of region according to RefSeq annotations (45). Promoter is the region between 2.5 kilo–base pairs (kbp) upstream and 2.5 kbp downstream of TSS. Genic regions correspond to the rest of the gene (the part that is not classified as promoter), and the rest of the genome is considered to be intergenic. Peaks that overlapped with more than one genomic feature were proportionally counted the same number of times. Reports of functional enrichments of GO and other genomic libraries were generated using the EnrichR tool (46). The aggregated plots showing the average distribution of ChIP-seq reads around the TSS of each target gene were generated with SeqCode (44) by counting the number of reads for each region according to RefSeq and then averaging the values for the total number of mapped reads of each sample and the total number of genes in the particular gene set. To generate with SeqCode (44), the aggregated plots showing the distribution of ChIP-seq reads along the body of a metagene derived from a target gene set, each gene was converted into a uniform region of 100 positions to count the number of reads along this region, calculating the mean at each point of the resulting metagene profile afterward. This graphical representation was integrated into the neighboring genomic region, calculated as described above. The heatmaps displaying the density of ChIP-seq reads around the TSS of each target gene set were generated by counting the number of reads in this region for each individual gene and normalizing this value with the total number of mapped reads of the sample. Genes on each ChIP heatmap were ranked by the logarithm of the average number of reads in the same genomic region. Boxplots showing the ChIP-seq level distribution for a particular ChIP experiment on a set of genomic regions were calculated by determining the maximum value on this region at this sample, which was assigned afterward to the corresponding region. Each point on the scatterplots of ChIP-seq intensities between two conditions were calculated by determining the maximum value of the sample inside each peak at each condition. The UCSC genome browser was used to generate the screenshots of the profiles produced by SeqCode (44) for each group of experiments along the manuscript (47). In all cases, the values of the samples including spike-in were corrected by the number of fly reads mapped in the sequencing experiment.

Statistical analysis

All statistical analysis and graphical representation of the data were performed with GraphPad Prism version 10 software for Windows (Boston, MA). When the experimental design required a single pairwise comparison, Student t tests were applied. When the experimental design required the comparison of multiple conditions, an analysis of variance (ANOVA) with orthogonal contrasts was applied. In the cases where we needed to do multiple comparisons, we applied an ANOVA test followed by a Tukey’s post hoc test or Wilcoxon test. The exact value of n, what it represents, and center and dispersion measures are described in the figures and figure legends. The exact P value of the statistical test is written in the figures or figure legends. Shapiro-Wilk tests were done to control for normal distribution of and equal variance of the samples.

Supplementary Materials

The PDF file includes:

Figs. S1 to S7

Legends for tables S1 to S3

Table S4

Other Supplementary Material for this manuscript includes the following:

Tables S1 to S3